fy13 lanl apr la ur 14 22168 b2bf0

Summary

FY13 ANNUAL PROGRESS REPORT Laboratory Directed Research and Development Los Alamos National Laboratory This page intentionally left blank. About the Cover ChemCam is the Chemistry and Camera instrument, one of 11 science instruments onboard the NASA Mars Science Laboratory rover, Curiosity. As the name implies, ChemCam is actually two different instruments combined as one: a Laser-Induced Break- down Spectrometer (LIBS) and a Remote Micro Imager (RMI). The purpose of the LIBS instrument is to p…

Page 1

FY13 ANNUAL PROGRESS REPORT Laboratory Directed Research and Development Los Alamos National Laboratory

Page 2

This page intentionally left blank.

Page 3

About the Cover ChemCam is the Chemistry and Camera instrument, one of 11 science instruments onboard the NASA Mars Science Laboratory rover, Curiosity. As the name implies, ChemCam is actually two different instruments combined as one: a Laser-Induced Break- down Spectrometer (LIBS) and a Remote Micro Imager (RMI). The purpose of the LIBS instrument is to provide elemental composi- tions of rock and soil, while the RMI will give ChemCam scientists high-resolution images of the sampling areas of the rocks and soil that LIBS targets. The LIBS technology was developed at Los Alamos National Laboratory with early support from the LDRD program. The research focused on LIBS technique for detection work in hazardous environments. A 2010 LDRD project explored direct application of LIBS on the ChemCam instrument, including LIBS detection limitations in Mars and Venus environments. Today LDRD is exploring the applications of this technology for finding and analyzing nuclear materials. Disclaimer The Los Alamos National Laboratory strongly supports academic freedom and a researcher’s right to publish; therefore, the Laboratory as an institution does not endorse the viewpoint of a publication or guarantee its technical correctness. With respect to documents available from this server, neither the United States Government nor the Los Alamos National Security, LLC., nor any of their employees, makes any warranty, express or implied, including the warranties of merchantability and fitness for a particular purpose, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial products, process, or service by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or the Los Alamos National Security, LLC. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or the Los Alamos National Security, LLC., and shall not be used for advertising or product endorsement purposes. Unless otherwise indicated, this information has been authored by an employee or employees of the Los Alamos National Security, LLC. (LANS), operator of the Los Alamos National Laboratory under Contract No. DE-AC52-06NA25396 with the U.S. Department of Energy. The U.S. Government has rights to use, reproduce, and distribute this information. The public may copy and use this information without charge, provided that this Notice and any statement of authorship are reproduced on all copies. Neither the Government nor LANS makes any warranty, express or implied, or assumes any liability or responsibility for the use of this information. Issued March 2014 LA-UR-14-22168

Page 4

Structure of this Report In accordance with U.S. Department of Energy Order (DOE) 413.2B, the Laboratory Directed Research and Development (LDRD) annual report for fiscal year 2013 (FY13) provides summaries of each LDRD-funded project for the fiscal year, as well as full final reports on completed projects. The report is organized as follows: Overview: An introduction to the LDRD Program at Los Alamos National Laboratory (LANL), the program’s structure and strategic value, the LDRD portfolio management process, and highlights of outstanding accomplishments by LDRD researchers. Project Summaries: The project summaries are organized first by science and technology categories: Physics, Chemistry and Material Sciences, Environmental and Biological Sciences, Information Science and Technology, and Technology. Within each category, summaries are organized by LDRD component in the following order: Directed Research (DR, Exploratory Research (ER), Early Career Research (ECR), and Postdoctoral Research and Development (PRD). Full final reports are included at the end of each section. Projects are listed in numerical order according to their project identification number, which consists of three parts. The first is the fiscal year in which the project began; the second is a unique numerical identifier; and the third identifies the project component. Acknowledgements Technical Review William Priedhorsky Jeanne Robinson Lennett Rendon Publication Editor and Designer Andrea Maestas Team Contributors Stephen Schultz Lisa Lujan Susan Whittington Debbie Martinez Peter Haase

Page 5

Table of Contents 14 Overview 82 Next Generation Ionic Liquids for Plutonium Science, Separation, and Production George S. Goff Chemistry and Material Sciences 88 Functional Materials by Design: Explosives with 30 Energy Storage Tailored Optical Response Properties Fernando H. Garzon Robert J. Scharff 33 Organic Electronic Materials: Designing and 92 Alternative Explosives Signatures for Detection Creating Functional Interfaces David S. Moore Sergei Tretiak 95 Room Temperature Oxidation and Corrosion of 37 Plutonium-242: A National Resource for the Plutonium Fundamental Understanding of 5f Electrons Alison L. Pugmire Eric D. Bauer 97 Novel Mesoscale Modeling Approach for 40 Modern Challenges in Actinide Science Investigating Energetically Driven Nanoscale Albert Migliori Defect/Interface Interactions Abigail Hunter 46 Accelerating Materials Certification Through Co- Design 100 Magnetic Field Effects on Convection-Modified Jack S. Shlachter Solid-Liquid Interfaces Amy J. Clarke 48 Fighting Carbon with Carbon: All-Carbon Nanomaterial Photovoltaics 102 Understanding The Catalytic Conversion Stephen K. Doorn of Oligosaccharides to Fuels and Chemical Feedstocks 51 Design Principles for Materials with Magnetic Andrew Sutton Functionality Joe D. Thompson 103 Effects of Joining Processes on Bimetal Interface Content and Radiation Damage Resistance 53 Non-Precious Metal Electrocatalysts for Clean John S. Carpenter Energy Piotr Zelenay 106 Probing Interface Reactions of Calcite Nanocrystals at Elevated Temperatures and 56 Phase Stability of Multi-Component Pressures Nanocomposites under Irradiation Katharine L. Page Blas P. Uberuaga 109 Exploiting Non-Innocent Ligands in Catalysis: New 59 Radioparagenesis: Robust Nuclear Waste Form Base Metal Catalysts for the Reduction of Carbon Design and Novel Materials Discovery Dioxide Christopher R. Stanek Susan K. Hanson 64 Hydrogen Effects in Delta-Stabilized Pu Alloys: 112 Advanced Materials for Exploring New Electronic Fundamental Thermodynamics and Interactions and Optical Behav-iors in Single-Walled Carbon at Reduced Dimensionality Nanotube Systems Daniel S. Schwartz Juan Duque 69 First Reactions: Simple Molecule Chemistry 115 Valence, Coordination and Reactivity at Interfaces Behind the Shock Front Christopher D. Taylor Dana M. Dattelbaum 123 Plasmas of Magnetic Monopoles in Artificial 76 Innovative and Validated Sub-micron to Meso- Spin Ice: Designing Frustration, Engineering scale Modeling of the Evolution of Interface Magnetricity Structure and Properties under Extreme Strains Cristiano Nisoli Irene J. Beyerlein

Page 6

129 Valence-Fluctuation-Mediated Superconductivity 173 Very Low Temperature Scanning Point Contact in Heavy-Fermion Materials Spectroscopy Investigation of Inhomogeneous Marc Janoschek States on the Nano-scale Roman Movshovich 135 Understanding Earth’s Deep Water Cycle: Neutron Diffraction and Calorimetric Studies of Hydrous 175 Excited State Quantum Interactions in Carbon Minerals Nanotubes Hongwu Xu Stephen K. Doorn 138 Developing Potentials for Atomistic Modeling of 178 “Upscaling” Nanoscale Thermoelectrics: The Defect Phenomena at Metal-Ceramic Interfaces Meso-macroscale Design Challenge for Real- Steven M. Valone World Energy Needs Jennifer A. Hollingsworth 141 From Waste to Fuels and Feedstocks: Reduction of CO Using Main-Group Catalysts 181 Threat Reduction via Nanomaterials: Engineering 2 Andrew Sutton a Novel SERS Plat-form for Chemical Detection Nathan H. Mack 143 First Principles Many-Body Approaches to Strongly Correlated Actinide Metals 184 Accurate Interfacial Structures for Atomistic Jianxin Zhu Simulations: Minimizing the Grand-Canonical Free Energy 146 Lanthanum Bromide Glass Ceramics for Gamma- Danny Perez Ray Spectroscopy Markus P. Hehlen 187 Understanding and Controlling Magneto-Electric Coupling in Multiferroic Materials 149 Are Nanoscale Foams Radiations Resistant? Dmitry A. Yarotski Jose A. Caro 190 Phase Transitions at Extremes: Emergence of 151 A Novel Mo-99 Separation Process Designed for Topological Defects Next Generation Medical Isotope Production Vivien Zapf Iain May 193 Redox active Catalysts for C-C Coupling Reactions 153 Ultrafast Spectro-microscopy for Nanoscale Relevant to Renewable Energy Magnetic Domain Imaging John C. Gordon Richard L. Sandberg 195 Novel Chemical Architectures for Supercapacitor 159 Plasmon-Exciton Interactions in Single-Wall Electrolytes: Comparing In Situ Scattering Carbon Nanotube – Metal Nanostructure Measurements to Theory and Simulation Complexes Cynthia F. Welch Han Htoon 198 Superconducting Vortices in Magnetic Media 162 Novel Inverted Nanoshells for Multimodal Boris A. Maiorov Diagnostic Imaging and Cancer Therapy Jennifer A. Hollingsworth 202 Novel Anti-Perovskite Electrolytes for Superionic Lithium Transport 165 Forensic Archaeology of a Manhattan Project Era Luc L. Daemen Nuclear Site Warren J. Oldham 205 Quasiparticle Scattering for Multiscale Modeling of Electronic Materials 168 Exploiting Metal/Organic Interfaces as Potential Sergei Tretiak Bulk Heterojuntions: Unlocking the Efficiency of Organic Photovoltaics 209 Solving the Active Site Conundrum in Oxygen Hsing-Lin Wang Reduction Catalysis Piotr Zelenay 171 3-Dimensional Characterization of Nuclear Fuels: Microstructural Evolution under Representative 217 Innovative Process for Making Ultra-thin Temperature and Thermal Gradients Dielectrics Donald W. Brown Quinn R. Marksteiner

Page 7

219 Embedding Plasmonic Nanostructures at 265 Controlled Synthesis and SERS Responses Semiconductor Interfaces for Enhancing of Metal Nanostructures with Complex Photovoltaic Efficiency Morphologies Todd L. Williamson Hsing-Lin Wang 222 Bio-analogue Catalysts: Evolved Aptazymes 270 Developing a Materials Strategy for d-electron for the Hydrolysis of Organophosphorous Heavy Fermion Systems Compounds Filip Ronning Robert F. Williams 272 Multiferroic Magnetoelectric Hybrid Inorganic- 226 Developing a Mild Catalytic Route for the Organic Frameworks Reduction of N to NH Michael R. Fitzsimmons 2 3 John C. Gordon 274 Catalysts Containing Earth Abundant Metals for 228 Determination of Fluid Properties at Carbonate Energy Applications Interfaces - An Integrated Experimental and John C. Gordon Theoretical Approach Donald D. Hickmott 276 Direct Tracking of Charge Carriers in Heterostructured Nanowires 233 Photo-triggerable Immolative Polymers: A New Rohit P. Prasankumar Modality for Radiation Dosimetry Robert D. Gilbertson 279 Fabrication of an All-Carbon Solar Cell Andrew M. Dattelbaum 236 Stable Uranium Benzyne Complex Jaqueline L. Kiplinger 281 Chemically Modifying the Uranyl Ion Jaqueline L. Kiplinger 239 Characterization of Products from Hydrolysis of UF6 283 Designing and Probing Novel Materials by Marianne P. Wilkerson Pressure Tuning of Nanocrystals Hongwu Xu 242 Re-invigorating LANL Quantum Computing Research 286 Functional Soft Materials by Assembling Metallo- Antoinette J. Taylor Biopolymers Reginaldo C. Rocha 245 Attosecond Probing of Dynamical States in Solids - Graphene 289 nTuned Optical Properties of Low-Dimensional George Rodriguez Carbon Nanomaterials for Energy Harvesting Stephen K. Doorn 250 Mechanistic Investigation of C-C Bond Forming Reactions for the Production of Long Chain 291 Enhanced Structural Robustness in Metal/ Hydrocarbons Nonmetal Nanocomposites using Bio-inspired Enrique R. Batista Structure Design Amit Misra 253 Exploiting the Fermi Surface to Understand Quantum Criticality in Complex Electronic 293 Nanoscopic Recrystalization Dynamics and Phase Materials Behavior in Liquid Crystals and Polymers Eric D. Bauer Juan Duque 257 Non-aqueous Organic Materials for Bio-defense 295 The Fate of the Three Dimensional Electron Far James M. Boncella Beyond the Quantum Limit Ross D. Mcdonald 261 Metal Catalysts for Oxidation of Lignocellulose and Reduction of Carbon Dioxide 295 Benefit to National Security Missions; 297 Susan K. Hanson Access to Industrially Important Chemical Compounds by Direct Hydrogenation of Biomass Derived Molecules Containing Carbonyl Fragments John C. Gordon

Page 8

299 Stimuli Responsive, Functional Biopolymers: 352 Discovery Science of Hydraulic-fracturing: Quinic Acid-Based Polymers and Their Assemblies Developing Innovative Concepts for Novel Hsing-Lin Wang Working Fluids and Their Interactions with Rocks, Fractures, and High Value Hydrocarbons 301 NMR Study of Quantum States of Matter Hari S. Viswanathan Joe D. Thompson 357 Defeating Bacterial Multi-Drug Resistance: An 303 Efficient Carbon Nanotube Growth on Graphene- Experimentally Driven Multi-Scale Study of Efflux Metal Surfaces Machinery Stephen K. Doorn Sandrasegaram Gnanakaran 305 Hybrid Nanostructures for Photoreduction of CO 361 Identifying, Creating, and Controlling Functional 2 to Hydrocarbons Hotspots in DNA Hongwu Xu Boian Alexandrov 307 Alternating Positive-Negative Charge Systems: 364 Construction and Functionalization of Novel New Compounds and Synthetic Routes Biomimetic Human Liver David E. Chavez Jun Gao 308 Graphene Quantum Dots for Carrier- 366 The World’s First Drought and Insect Caused Multiplication-Enhanced Solar Cells Global Tree Mortality Monitoring System Victor I. Klimov Chonggang Xu; Erin H. Lay 310 Microstructured Biohybrid Synthesis of 370 Mechanistic Studies Toward an Improved Photosynthetic Assemblies Carbonic Anhydrase for Biofuel Production Gabriel A. Montano Suzanne Z. Fisher 312 Photoactive Energetic Materials for Quantum 372 Low-Frequency Acoustic Interferometry for Optical Control Probing the Stratosphere Sergei Tretiak Stephen J. Arrowsmith 313 Poly-Triphenylamine-Functionalized Graphene 405 Grayscale Flow Cytometry - Multicoil NMR Quantum Dots for Flexible High-Performance Sensors for Portable Flow Cytometry Polymeric Memory Devices Pulak Nath Hsing-Lin Wang 411 Biofuel from Magnetic Algae 315 Understanding and Controlling Magnetism in Pulak Nath Multiferroics with THz Pulses Rohit P. Prasankumar 416 Understanding Thunderstorm Effects on the Ionosphere: a New Approach to Investigate Environmental and Biological Sciences Possible Convective and Electrical Coupling Mechanisms 332 Terrestrial Vegetation, CO Emissions, and Climate Xuan-Min Shao 2 Dynamics 421 High Density Neuronal Recording Using Nanowire Nathan G. Mcdowell Capacitor Sensors 334 Advanced Metagenomic Analysis to Understand John S. George Dynamics of Soil Microbial Community under 427 Restoring Neurological Function with Self- Conditions of Climate Change assembled Lipid Bilayers Shunsheng Han Srinivas Iyer 339 Multi-Scale Science Framework for Climate Treaty 431 Understanding the Mechanisms of Biological Verification: Attributing and Tracking Greenhouse Transport for Nano-technology Applications Gas Fluxes Using Co-emitted Signatures Sandrasegaram Gnanakaran Manvendra K. Dubey 345 Multi-scale Dynamics of Biological Systems Robert E. Ecke

Page 9

436 Nonlinear Resonances between DNA and THz 518 Exact Renormalization Method for Frustrated Radiation Systems and Optimization Boian Alexandrov Cristian D. Batista 439 Biomarkers for the ‘Signature Wound’ of Iraq and 522 Computational Modeling of Topo-Taxis: Afghanistan: Traumatic Brain Injury Directing the Motion of Bacteria and Cells with Harshini Mukundan Microfabricated Topologies Charles Reichhardt 451 Analysis of Protein Structure-Function Relations in Antibiotic Biosynthesis and Signal Transducing 526 Algorithmic Co-Design: Paradigms for Receptors Unstructured Problems on Accelerated Louis A. Silks III Architectures Sunil Thulasidasan; 20110195ER 455 Quantitative Modeling of Cellular Noise Michael E. Wall 532 Extreme Quantum Simulation: Co-Design from Desktop to Exa-Scale 459 Modeling the Surface Mass Balance and Nicolas Bock Freshwater Runoff from the Greenland Ice Sheet in a Changing Climate 537 Smart Grid Control and Optimization Matthew W. Hecht Michael Chertkov 461 Novel Laboratory and Field Observations to 539 Development of a Novel Algorithm to Generate Model Climate Warming by Aged Absorbing User-Defined Databases of Genomic Signatures Aerosols Patrick S. Chain Manvendra K. Dubey 541 Multi-Phenomenology Explosion Monitoring Information Science and Technology (MultiPEM) Dale Anderson 466 Optimization Principles for Co-Design applied to 543 Climate Impacts Research through Large Remote- Molecular Dynamics Sensing Datasets Stephan J. Eidenbenz Timothy M. Kelley 469 CoCoMANS: Computational Co-design for Multi- 548 Cyber Security of the Smart Grid scale Applications in the Natural Sciences Russell W. Bent Dana A. Knoll 472 Hierarchical Sparse Models for Robust Analysis of 550 Edge Traffic Monitoring in Computer Networks Video Data Joshua C. Neil; Eli Ben-Naim Steven P. Brumby Physics 476 Empowering the Expert: Machine Learning with User Intelligence Reid B. Porter 555 TeV Jets: Nature’s Particle Accelerators Brenda L. Dingus 479 Multi-Perspective Network-Scale Modeling & Detection for Cyber Systems 558 IMPACT: Integrated Modeling of Perturbations Michael E. Fisk in Atmospheres for Conjunction Tracking - A New Orbital Dynamics and Drag Model to Avoid 483 Correlations and Dynamics in Information Science Collisions in Space Robert E. Ecke Alexei V. Klimenko 489 Gating and Emission Enhancement of Diamond 561 Hydrodynamical Mix Studies at the National Field-Emitter Arrays Ignition Facility (U) Heather L. Andrews Michael J. Steinkamp 515 Rare Category Detection 563 Advancing the Fundamental Understanding of James P. Theiler Fission (U) Morgan C. White

Page 10

566 Physics Beyond the Standard Model with the 637 Petascale Kinetic Plasma Simulation of the Long-Baseline Neutrino Experiment Interaction Among Laser Speckles in Nonlinear Christopher M. Mauger Optical Systems Lin Yin 568 Disruptive Innovation in Numerical Hydrodynamics 639 Multiscale Spacecraft Charging Simulations in Jacob I. Waltz Support of New Space Missions Gian L. Delzanno 570 Illuminating the Origin of the Nucleon Spin Ivan M. Vitev 642 Laboratory Study of Cosmically-Relevant Collisionless Shocks 573 Peta-scale Studies of Cosmic Explosions and Scott C. Hsu Supernova Shock Breakout with Palomar Transient Factory 644 Nuclear Quadrupole Resonance: From no Field to Przemyslaw R. Wozniak Ultra-Low Field Michelle A. Espy 576 High Performance Atom-Based Sensors for Fields and Rotations 647 Laser-Driven Relativistic Mechanics, Radiation, Malcolm G. Boshier and Ion Acceleration James A. Cobble 579 Quantum Chemistry, Information, Materials and Metrology 649 Monte Carlo for Quantum Transport Robert E. Ecke James E. Gubernatis 583 Non-Equilibrium Phenomena in Materials, Fluids, 652 Nonequilibrium Spin Noise in Semiconductors: and Climate Physics and Applications Robert E. Ecke Nikolai Sinitsyn 586 CLEAN Detection and Identification of Dark 654 Flipping the (Light) Switch on Nano-magnetism: Matter Emergent Photomagnetization in Quantum- Andrew Hime confined Semiconductors Scott A. Crooker 591 Multi-Messenger Signals from Low-Mass Supernovae 656 A Novel Exploration of Nature’s Largest Christopher L. Fryer Explosions W T. Vestrand 597 Network-Centric Quantum Communications Richard J. Hughes 658 Transport by Thermodynamic Cross-terms in ICF Capsule Plasmas 603 First Direct Measurement of Particulate Evolution Xianzhu Tang in Isochorically Heated Dense Plasma (U) Brian J. Albright 660 Co-Evolution of Protoplanets and Transitional Protoplanetary Disks: Pathway to Giant Exoplanet 625 The Dynamic Environment of Iron Core Formation Formation and Collapse Hui Li Casey A. Meakin 662 A Computationally Efficient Model for Warm 629 Foundational Methods and Experiments in Dense Mixtures Ultracold Molecular Physics Didier Saumon Michael D. Di Rosa 665 A New Hypothesis to Explain the Variability of the 632 Next Generation Earth Models Outer Radiation Belt: Can we Predict Post-storm Monica Maceira Fluxes of Energetic Electrons Based only on Pre- storm Fluxes of the Lower-energy Population? 634 Non-Equilibrium Fluctuation-Induced Interactions Gregory S. Cunningham Diego A. Dalvit 667 A New Approach to Multiscale Plasma Physics Simulations Gian L. Delzanno

Page 11

670 Enhancing Thermoelectric Properties of 721 “Listening” to the Noise of a Single Electron Topological Insulators through Nanostructuring Spin: Methods for Fast, Ultra-sensitive, Non- Nikolai Sinitsyn perturbative Noise Spectroscopy Scott A. Crooker 672 Giving Cold Atoms Weight: creating Heavy Fermions in Optical Lattices 723 The Terahertz Quantum Hall Effect Cristian D. Batista Rohit P. Prasankumar 674 Topology in Superposition: Quantum 728 A Compact, Brilliant, Coherent X-ray Source Based Decoherence in Many-body Systems on a Dense Relativistic Electron Mirror Wojciech H. Zurek Juan C. Fernandez 677 Ultra-Bright Electron Beam Acceleration in 733 Majorana Neutrinos Dielectric Wake Accelerators Steven R. Elliott Evgenya I. Simakov 735 Quantum Hydrodynamics Approach to Non- 680 Wide Field-of-View Plasma Spectrometer equilibrium Modeling of Compressed Plasmas Ruth M. Skoug Michael S. Murillo 683 Magnetic Nanomarker Detection and Imaging 736 Toward Laser Manipulation and Quantum Control with SQUIDs of Th-229 Nuclei Embedded in a Solid Andrei N. Matlashov Xinxin Zhao 686 Beyond the Standard Halo 738 Materials Study for Small Self-Regulating Fission Michael S. Warren Reactor System for Space Applications Patrick R. McClure 689 Coherent Diffractive Imaging of Ultrafast Ejecta Processes 743 pRad Measurements of Meteorites Cynthia A. Bolme Chad T. Olinger 691 In Search of Light WIMPs 791 Physics of Cosmic Ray Shocks and the High Energy Alexander Friedland Universe Hui Li 693 Electron Capture Spectroscopy for Neutrino Mass: Isotopes, Science, and Technology Development 794 Non-Condon and State Interaction Effects in Michael W. Rabin Carbon Nanotubes Stephen K. Doorn 695 Jet Probes of New Physics at RHIC and at the LHC Ivan M. Vitev 798 A New Regime of Carrier Multiplication Using Intraband Re-Excitation of Nanocrystals 700 Daylight Imaging with Seismic Noise Victor I. Klimov Xiaoning Yang 803 Heavy quarks in Cold Nuclear Matter and in the 705 Balanced and Unbalanced Turbulent Cascades Quark Gluon Plasma Balasubramanya T. Nadiga Ming X. Liu 709 Time-Dependent Quantum Molecular Dynamic 807 The Formation and Evolution of Black Holes in the Simulations of Dense Plasmas Supporting Universe Thomson X-ray Scattering Experiments Christopher L. Fryer Jerome O. Daligault 811 Auger-Recombination-Free Nanocrystals by 715 Precision Measurement of Atomic Parity Violation Rational Design of Confinement Potential in Trapped Yb+ Victor I. Klimov Martin M. Schauer 816 Universal Physics with Ultracold Atoms 718 The Largest Cosmic Implosions: Formation of Joseph A. Carlson Supermassive Black Holes Hui Li

Page 12

818 A New Drift Shell Integration Technique for Inner 860 Emittance-Reduction System for Future Magnetospheric Space Weather Models Accelerator Solutions Josef Koller Kip A. Bishofberger 822 “Listening” to the Noise of a Single Electron Spin: Technology Ultra-sensitive, Non-perturbative Spin Noise Spectroscopy 863 Harnessing Nonlinearity for Transformative Scott A. Crooker Metamaterial Technology Houtong Chen 824 Determining the Origin of the Highest Energy Cosmic Rays with TeV Gamma-Ray Observations 871 Exploiting Hamiltonian Properties of Beams Brenda L. Dingus to Revolutionize X-Ray Free-Electron Laser Architectures 828 Probing the Nucleon Structure via Spin Bruce E. Carlsten Observables Ivan M. Vitev 876 Development of an Interface-Dislocation Dynamics Model to Incorporate the Physics 832 Data Mining for the M-Sigma Relation of Interfaces in Predicting the Macroscopic Aimee L. Hungerford Mechanical Properties of Nanoscale Composites Jian Wang 836 Modeling X-ray Bursts Christopher J. Fontes 880 Exploration of Megawatt Heat Pipe Reactor Concepts 838 Experimental and Computational Studies of Patrick R. Mcclure Engineered Nanoparticle Aggregation in Human Tissue Models 885 Full-Frame Programmable Spectral Imagers Based Charles Reichhardt on Micro-Mirror Arrays Steven P. Love 840 Quantum and Correlation Effects in High-Energy Density Plasmas 892 Novel Broadband Tera Hertz Sources for Remote Jerome O. Daligault Sensing, Security and Spectroscopic Applications John Singleton 842 Genetically Encoded Materials: Libraries of Stimuli-responsive Polymers 898 Pyroelectric Heat Engines: Highly Efficient, Jennifer Martinez Environmentally Friendly Cooling Markus P. Hehlen 845 Using Microreactors for Efficient Plutonium Separations (U) 904 Spatial and Wavelength-diversity Lasercomm Stephen L. Yarbro System Anatoly V. Efimov 847 Battlefield MRI Michelle A. Espy 908 Hyperspectral Intensity Correlation Interferometry 850 A Time Reversal Non Contact Acoustical Source David C. Thompson for NEWS (Nonlinear Elastic Wave Spectroscopy) Pierre-Yves Le Bas 912 Microfluidic Uranium Enrichment through Electrochemically Driven Solvent Extraction 853 Measurement of Pressure and Temperature in Kirk R. Weisbrod Thermal Explosions Markus P. Hehlen 915 Utilization of Conducting Polymers for Embedded Electrical Circuits (U) 855 High-Energy Segregated Fuel-Oxidizer Solid Kevin M. Hubbard Rocket Propulsion System Bryce C. Tappan 919 Toward Development of a Uniquely Specific and Sensitive Detection Technology 858 High Performance Thin Film Super Capacitors Antonietta M. Lillo Aditya Mohite

Page 13

924 Neutron Sensor based on Integrated Scintillator and Photodiode Kiril D. Ianakiev 925 Time Reversal Acoustic Communications for Rescue Operations Brian E. Anderson 930 Remote Raman-LIBS Spectroscopy: Preliminary Tests for the Next Mars Rover Samuel M. Clegg 932 Optimization of Heterogeneous Sensing Systems for Risk-Minimized Decision Making Charles R. Farrar

Page 14

Laboratory Directed Research & Development: Making every mission possible by William Priedhorsky, Los Alamos LDRD Program Manager R&D is one of the best investments a country can make. In his 2013 remarks to the National Academy of Sciences, President Obama said, “With the pace of technological innovation today, we can’t afford to stand still for a year or two years or three years. We’ve got to seize every opportunity we have to stay ahead. And we can’t let other countries win the race for ideas and technology of the future.” The national laboratories have been entrusted with the role of serving as incubators for in- novation — places where novel solutions to resolve future national security challenges are cultivated, directed, and sustained. The Laboratory Directed Research and Develop- ment program at Los Alamos National Laboratory has a fundamental role in maintain- ing intellectual vitality and innovation in our enduring capabilities. It is also critical in anticipating and addressing technical challenges that would otherwise compromise the Laboratory’s ability to execute its missions to • ensure the safety and reliability of U.S. nuclear deterrent, • reduce the threat of weapons of mass destruction, proliferation, and terrorism, and • solve national problems in defense, energy, environment, and infrastructure. These missions demand that we anticipate, innovate, and deliver to solve the unsolved and know the unknown. We seek innovation that can drive breakthroughs in mission performance, whether that mission demands prediction of weapons aging, the reme- diation of legacy contamination, or the mitigation of transnational terrorism. At Los Alamos, LDRD is an open market for ideas, inspired by mission. We succeed be- cause of the diversity of our investments, which mix applied and fundamental research. Although no single project can guarantee results, the aggregate results of LDRD invest- ment drive the Laboratory forward. The LDRD program is extremely important in the recruitment and retention of technical staff, as well as for our postdoc pipeline and collaborations with academia. Many of the innovations and technological advances that put Los Alamos on the map are built upon the quality and success of the LDRD program. Our missions require breadth and depth in science and technology. Their complexity de- mands a span of excellence that crosses nearly every discipline in science. For example, the core nuclear weapons program demands prediction and control of materials in extreme conditions, computing at the petascale and soon the exascale, the understand- ing of turbulent fluid flow, miniaturized smart sensors, and advances in nuclear and high-energy density physics. By investing in LDRD, Los Alamos builds its vision: to be the National Security Science Laboratory of choice, ready to take on whatever complex issue the nation asks of us. 14

Page 15

Laboratory Directed Research and Development is the that are needed to address Los Alamos missions. In FY13, most prestigious source of research and development the LDRD Program Office engaged thought leaders across funding at the Los Alamos National Laboratory. It follows the Laboratory to restructure the DR component for the a strategic guidance derived from the missions of the U.S. upcoming fiscal year. The new scheme consists of five Department of Energy, the National Nuclear Security Ad- Focus Areas that map directly to the Los Alamos science ministration, and the Laboratory. To execute that strategy, pillars, plus two additional multi-competencies not cap- the Los Alamos LDRD program creates a free market for tured by the pillars. The new structure applies to the FY14 ideas that draws upon the bottom-up creativity of the LDRD program plan. Laboratory’s best and brightest researchers. The combina- tion of strategic guidance and free-market competition Exploratory Research provides a continual stream of capabilities that position The ER component is focused on developing and maintain- the Laboratory to accomplish its missions. ing technical staff competencies in key strategic disciplines that form the foundation of the Laboratory’s readiness The LDRD program provides the Laboratory Director with for future national missions. Largely focused on a single the opportunity to strategically invest in forward-thinking, discipline, ER projects explore highly innovative ideas that potentially high-payoff research that strengthens the underpin Laboratory programs. In FY13, LDRD funded 135 Laboratory’s capabilities for national problems. Funded in ER projects, which represents approximately 34% of the FY13 with approximately 6.9% of the Laboratory’s overall program’s research funds. Exploratory Research projects budget, the LDRD program makes it possible for research- are typically funded up to a maximum of $390Kperyearerstopursuecutting-edgeresearchanddevelopment.Thisforthreeyears.inturnenablestheLaboratorytoanticipate,innovate,anddeliverworld-classscience,technology,andengineering.UnlikeDRproposals,divisionendorsementsarenotrequiredforERproposals;instead,thiscomponentoftheProgramStructureLDRDprogramisoperatedasanopenandcompetitivepathforeverystaffmembertopursuefundingforhis/hergreatidea.TheERcomponentisacriticalchannelforTheLosAlamosLDRDprogramisorganizedintofourpurelybottom-upcreativityattheLaboratory.Nonethe-programcomponentswithdistinctinstitutionalobjectives:less,itisstronglydrivenbymissionneedsviathedefinitionDirectedResearch(DR),flagshipinvestmentsinmissionofthe12ERresearchcategories,andtheassignmentofsolutions;ExploratoryResearch(ER),smallerprojectsthatinvestmentbetweenthem.investinpeopleandskillsthatunderpinkeyLaboratorycapabilities;EarlyCareerResearch(ECR),supportingtheEarlyCareerResearchdevelopmentofearly-careerresearchers;andPostdoc-TheECRcomponentoftheLDRDprogramisdesignedtoralResearchandDevelopment(PRD),recruitingbright,tostrengthentheLaboratory’sscientificworkforcebyqualified,early-careerscientistsandengineers.InFY13,providingsupporttoexceptionalstaffmembersduringtheLDRDprogramfunded300projectswithtotalcoststheircrucialearlycareeryears.Theintentistoaidinof$145.7 million. These projects were selected through a the sometimes challenging transition from postdoc to rigorous and highly competitive peer-review process and full-time staff member, and to stimulate research in are reviewed formally and informally throughout the fiscal disciplines supported by the LDRD program. In FY13, the year. LDRD program funded 28 ECR projects, which represents approximately 3% of the programs research funds. Early Directed Research Career Research projects are funded up to $225KperyearTheDRcomponentmakeslong-rangeinvestmentsinfortwoyears,andonlyupto60multidisciplinaryscientificprojectsinkeycompetencyorcanbefromtheLDRDprogram.technology-developmentareasvitaltoLDRD’slong-termabilitytoexecuteLaboratorymissions.InFY13,LDRDPostdocResearchandDevelopmentfunded54DRprojects,whichrepresentsapproximatelyThePRDcomponentensuresthevitalityoftheLaboratory56byrecruitingoutstandingresearchers.Throughthisinvest-projectsaretypicallyfundeduptoamaximumof$1.8M ment, the LDRD program funds postdoctoral fellows to per year for three years. work under the mentorship of PIs on high-quality projects. The primary criterion for selection of LDRD-supported The DR component is guided by the LDRD Strategic Invest- postdocs is the raw scientific and technical talent of the ment Plan, which is in turn guided by eight Grand Chal- candidate, with the exact specialty of the candidate a lenges that define the advances in science and technology 15

Page 16

Directed Research Grand Challenges Mission Impact Sensitive instrumentation and tools to manipulate massive Beyond the Standard Model data volumes, in support of national security missions Energy sources, efficiency and storage; sensing for threat Materials: Discovery Science to Strategic Applications reduction; materials underpinnings of stockpile security Complex Biological Systems Energy, national security, health and the environment Information Science and Technology Overarching capability supporting all Laboratory missions Earth and Energy Systems Energy and climate security Nuclear Performance Stockpile safety, surety and reliability Nuclear weapons of mass destruction, space situational awareness, Sensing and Measurement Science for Global Security global environmental treaty monitoring, and emerging threats Engineered Systems Systems-level solutions for all missions Exploratory Research Technical Categories Laboratory Capability Biological, Biochemical, and Cognitive Sciences Biosciences Chemistry and Chemical Sciences Chemistry Information and knowledge sciences, computer and Computational and Numerical Methods computational sciences High-performance computing, data analysis, and data-driven Computer Science, Mathematics, and Data Science science Theoretical, computation and modeling, and experimental Defects and Interfaces in Materials methods to understand defects and interfaces in materials Earth and Environmental Sciences and Space Physics Earth and space sciences Weapons science and engineering, advanced manufacturing, Engineering Applications sensors, and remote sensing Theory, computation and modelling, and experimental Emergent Phenomena in Materials Functionality methods to understand behavior of materials High-energy Density, Plasma, and Fluid Physics High-energy density plasmas and fluids and beams Measurement Science, Instrumentation, and Diagnostics Measurement methods that enable new scientific discovery Nuclear and Particle Physics, Astrophysics, and Cosmology Nuclear physics, astrophysics, and cosmology Fundamental interactions and excitations in atomic, optical, Quantum and Optical Science and molecular systems secondary factor. In FY13, LDRD funded 82 PRD projects, nize their role as future science and technology leaders, which represents 6% of the program’s research funds. these appointments are named after some of the greatest These postdocs are supported full-time for two years. leaders of the Laboratory’s past. In addition to approximately 47 Director’s Postdocs, the More postdocs are hired through DR and ER projects than LDRD program supported 17 distinguished postdoctoral directly through PRD appointments. Counting both ave- fellows at a higher salary and for a three-year term. Dis- nues, the LDRD program supported 61% of the 596 post- tinguished postdoctoral fellow candidates typically show docs at the Laboratory in FY13. evidence of solving a major problem or providing a new approach or insight to a major problem and show evidence of having a major impact in their research field. To recog- 16

Page 17

Postdoctoral Research and Development Projects First nuclear material detection by short-pulse-laser-driven neutron source Los Alamos researchers successfully demonstrated for the first time that laser-generated neutrons can be enlisted as a useful tool for national security. The international research team used the short-pulse laser at Los Alamos’ TRIDENT facility to generate a neutron beam An intense burst of laser energy slams into an with novel characteristics that interrogated a closed container extremely thin foil target to produce neutrons at Los to confirm the presence and quantity of nuclear material Alamos National Laboratory’s TRIDENT laser facility inside. The successful experiment paves the way for creation during a recent experiment, which proved that laser- of a table-top-sized or truck-mounted neutron generator that driven neutrons can be used to detect and interdict could be installed at strategic locations worldwide to thwart smuggled nuclear materials. smugglers trafficking in nuclear materials. See this scientific achievement in motion on the Laboratory’s You Tube page. 17

Page 18

Program Reserve Reserve Project Delivers Results Numerous space probes have taken advantage of radioiso- tope thermoelectric generators (RTGs) powered by pluto- Most LDRD investments are selected in a rigorous, multi- nium. However, the end of the Cold War brought about a step peer-review process during the nine months preced- shortage of plutonium. In collaboration with NASA Glenn ing the new fiscal year. However, in a fast-changing world, Research Center and National Security Technologies, a there are needs that cannot wait until the next cycle of team of LDRD researchers developed an alternative type competition. The LDRD Program Office holds a reserve of nuclear reactor, one that uses plentiful uranium as its each year for these needs. This reserve is small on a pro- fuel source. Known as KiloPower, the reactor has six major gram basis, but allows the opportunity to make modest components: a reactor core, core reflector, a rod to start investments that address new opportunities. In FY13, the the reactor, heat pipes to move energy, radiation shielding reserve budget was approximately $3.2M. and Stirling engines to provide power. Measuring just 1 ft, the fast reactor uses neutrons with a high-energy range, Reserve decisions are guided by strategic priorities articu- in excess of 0.2 meV, to produce fission and employs heat lated by Division Leaders, Associate Directors, and more pipes to remove heat from the core. The heat pipes trans- senior management. They are also informed by the just- fer the heat to a Stirling-engine power-conversion system. concluded competitive cycle, in which selection panels The Stirling engine then produces 500 to 1,500 W of elec- sometimes identify ideas that are exciting but just not tricity over 15 to 30 years. mature enough for a full-scale investment. Reserve is held foremost to exploit opportunity, rather than fix problems, With KiloPower, it is possible for NASA and other govern- and the expectation is that reserve projects will be the first mental and industrial organizations to continue developing step in a chain that leads to bigger things. probes and spacecraft for deep-space exploration. Researchers test novel power system for space travel A team of Los Alamos researchers developed a small-space reactor known as Kilo- Power that can provide long-term power—approximately 15 to 30 years—to a deep-space probe or satellite. To produce electricity, KiloPower uses a nuclear fission system as a heat source that transfers heat via a heat pipe to a small Stirling-engine- based power convertor to produce electricity. KiloPower uses plentiful uranium instead of scarce plutonium; generates 500 to 1,500 watts of electricity; and minimizes hazards and guarantees performance as a result of its safe and simple design. With KiloPower, it is possible for NASA and other government and industrial organizations to continue developing probes and space- craft for the exploration of deep space. Other applications include providing power on the surfaces of planets, mobile power for forward-operating bases (of interest to the Department of Defense), and power in remote loca- tions of interest to intelligence agencies. Los Alamos research on the project was funded by the LDRD program. NASA Glenn and NSTec also used internal support to fund their contributions to the experiment. “Perhaps one of the more important aspects of this experiment is that it was taken from concept to completion in 6 months for less than a million dollars,” said Los Alamos engineer David Dixon. “We wanted to show that with a tightly- knit and focused team, it is possible to successfully perform practical reactor testing.” See this scientific achievement in motion on the Laboratory’s You Tube page. 18

Page 19

Project Budgets Senior management reviews project sizes annually with a goal of keeping projects as small as possible, while still accomplishing our strategic objectives. Typical project sizes for each LDRD component are identified above. Protecting ports, borders, and critical infrastructure The Multi-Mode Passive Detection System (MMPDS) combines technology invented by scientists at Los Alamos with private sector investment and expertise from Decision Sciences, to deliver totally safe, effective and reliable automated scanning to speedily detect both shielded and unshielded nuclear and radiological threats. Additional modality enables explosive and contraband detection. Harnessing the natural occurrence of muons in the atmosphere, MMPDS tracks muons through even heavily shielded materials and computes a 3-D image of what is being scanned. In combination with a gamma radiation detector, MMPDS can scan a typical 40-foot shipping container in 45 seconds, on average, providing accurate and safe scanning while facilitating the flow of commerce. The MMPDS technology produces no ionizing radiation, meaning it is completely safe for people, animals, plants and food. MMPDS modular construction enables the system to be scaled up or down to scan any type of vehicle, rail cars and cargo containers. A key driver behind the success of MMPDS is a technology called muon tomography. Muons are naturally occurring cosmic ray-induced particles that continuously rain down from the Earth’s upper atmosphere, harmlessly penetrating everything they touch. MMPDS tracks these muons, detecting and recording their deflection signatures with advanced proprietary software and algorithms as they pass through an object within the MMPDS. LDRD investment in muon tomography dates back to 2001. The Laboratory granted Decision Sciences Corporation exclusive commercial license for muon tomography, and the tech- nology won a 2013 R&D 100 Award. Watch a Nova segment about MMPDS on the Decision Sciences website. 19 DRP RCE RE RD

Page 20

The Los Alamos LDRD portfolio reflects diversity at all levels, from innovative projects that span the DOE missions, to a diverse pool of researchers that range from early career to senior scientist. LDRD Researchers Across Divisions and Directorates 20 stnapictiraP Organization LDRD Supports Researchers at All Career Stages srotagtisevnI lapicnirP Years since highest degree

Page 21

Project Selection sible, the appraisal is held as part of a broader workshop hosted by the Laboratory. The LDRD program is the vehicle by which the Labora- Written appraisals, held in the LDRD archives, address: tory harvests the ideas of some of our best and brightest (1) Brief summary of accomplishments; (2) Assessment scientists and engineers to execute DOE/NNSA missions. of quality of science and technology, relevance to Labo- This bottom-up approach is balanced by a program man- ratory and national missions, progress toward goals and agement strategy in which Senior Laboratory leadership milestones, project leadership, and the degree to which sets science and technology priorities, then opens an LDRD the project may establish or sustain a position of scientific competition for ideas across the breadth of the Laboratory. leadership for the Laboratory; and (3) Recommendations Panels formed from the Laboratory’s intellectual leaders by the committee for changes in the scope or approach rigorously review proposals. Conflict of interest is carefully of the project. The criteria for the most important point regulated, and evaluation criteria include innovation and – number (2) above – are derived from criteria developed creativity, potential scientific impact, viability of the re- by the National Academy of Science to assess all federally search approach, qualifications of the team and leadership, sponsored research. and potential impact on Laboratory missions. The selection processes are modeled on best practices established by Continuing ER and ECR projects are appraised in their first the National Science Foundation (NSF) and National Insti- and second years. The LDRD Deputy Program Director col- tutes of Health (NIH). laborates with the technical divisions to conduct project appraisals. Like DRs, the projects are appraised according To guarantee fairness and transparency, and to ensure that to the Federal criteria of quality, performance, leadership, the strongest proposals are funded, the selection panels and relevance. include managers and technical staff drawn from the full range of technical divisions. Serving on an LDRD selection LDRD service provides benefit to all panel is often a starting point on the path to leadership roles in the scientific community. Past LDRD panelists The mission of the Laboratory is to solve the nation’s most have gone on to be Laboratory Fellows, division leaders, difficult national security problems. By their nature, these program directors, association Fellows, and chief scientists, problems lack a well-defined path to solution. In fact, the while others have become leaders in academia. path is often completely unknown. It is rare that such creative work is done alone; the ideas and results from many colleagues are needed, often drawn out in confer- Independent project appraisals ences, hallway conversations, journals, and seminars. LDRD In FY13, the LDRD Program Office conducted an appraisal is an internal arena in which Laboratory staff serve as peer of every ongoing project it intended to fund in the next reviewers and play a key role of interaction in the scientific fiscal year. The primary objective is to assess progress and process. Proposal selection panelists are chosen for their provide peer input to help PIs maintain the highest quality subject-matter expertise, and the discussions in which of work. The appraisals also help the LDRD Program Office they engage are not only critical to the LDRD process, but monitor and manage the program portfolio. In addition to they also provide an opportunity for panelists to educate formal project appraisals, which are conducted annually, themselves on the latest results and practices, and expose the LDRD Program Director and Deputy Program Director themselves to opportunities for collaboration. As noted in meet informally with PIs in their labs at least once a year an evaluation of peer review conducted by the UK House to discuss their projects. The purpose of these one-on-one of Commons, “Peer review is regarded as an integral part meetings is to give PIs individualized assistance and to de- of a researcher’s professional activity; it helps them be- termine what the LDRD Program Office can do to positively come part of the research community.” impact the success of the project. Every DR project has also been assigned a Program Development Mentor to as- sist the transition of LDRD successes to mission. Any weak- nesses are actively addressed, and the project leader is asked to respond to the report with a revised project plan. Continuing DR projects are appraised every year of the life of the project, with at least one of the reviews including external reviewers. The internal-external review is open to all Laboratory staff and leaders. Four project appraisers – two internal and two external – are nominated by the PI and approved by the LDRD Program Director. When pos- 21

Page 22

DR Project Appraisal Highlights Plutonium-242: A National Resource for the Dynamic Earthquake Triggering, Granular Physics Fundamental Understanding of 5f Electrons and Earthquake Forecasting (20120007DR) (20120024DR) “The project is well conceived in that it consists of “The project team has established new experimental laboratory, simulation, and observational efforts that are and theoretical capabilities to study plutonium lead by experts and employ students and early career materials that are clearly aligned with the missions scientists. Further, the team of scientists from diverse of the Laboratory and key parts of the Plutonium backgrounds and numerous, including international, Science and Research Strategy Implementation institutions have done an admirable job coordinating Plan. This project contributes important work to efforts. For example, modeling efforts are focused on maintaining Los Alamos as the NNSA Plutonium understanding experimental results through accurate Center of Excellence.” reproduction of the experimental setup at other institutions. Scientists often want to solve the most interesting problem to them, personally, and it is the exception rather than the norm when this level of cooperation is found. “ Analysis of FY13 DR project appraisals showed an average score of 4.3, which is between excellent and outstanding. This average reflects 21 appraisals. All Carbon Solar Cells: Fighting Carbon with Carbon (20130026DR) “The team already has 17 publications to their credit (7 published, 3 submitted and 7 in preparation) and most of them have appeared in journals of high impact, including ACS-Nano, NanoLett, Phys. Rev. Lett., and J. Am. Chem. Soc. An essential breakthrough was the development of an aqueous two phase (ATP) method for isolating single chirality nanotubes.” Genetically Encoded Materials: Libraries of Stimuli-Responsive Polymers (20120029DR) “The results of this project have the potential to have a significant impact within the technical field. In particular, the development of combinatorial methods producing libraries of mono-disperse polymers of known structure is unprecedented, and could be a transformational technology in combinatorial synthesis. The results…also have potential to impact a broad range of other technological areas including electronics and sensors. Internally, the significance of the results from this project have been recognized by the LANL Technology Transfer Department as they have fast tracked this work for patent protection.” 22

Page 23

Mission Relevance Mission relevance is one of the most important criteria in scientific discovery and innovation. LDRD work also ben- the evaluation of a potential LDRD project; it is carefully efits the national security missions of the Department of considered in project selection, and it is tracked annually Homeland Security, the Department of Defense, and Other through the data sheet process. Many of the technologies Federal Agencies. As a result, the scientific advances and that put Los Alamos on the map have deep roots in LDRD technology innovations from LDRD provide multiple ben- and are valuable to DOE/NNSA mission areas of nuclear efits to all Los Alamos stakeholders, consistent with Con- security, energy security, environmental remediation, and gressional intent and the Laboratory’s scientific strategy. FY13 LDRD Investment in DOE Missions ($M)Thechartaboveshowsthemissionrelevanceofthe$145.7M FY13 LDRD portfolio. The dollar amounts total more than 100 percent of the total FY13 budget because many LDRD projects impact and have relevance to more than one mission area. The Los Alamos LDRD program plan includes a list of every project to be funded in the upcoming fiscal year and indicates which mission areas the project will impact. 23

Page 24

The NASA Mars Science Laboratory rover Curiosity identify samples that would be of greatest interest to sci- launched November 25, 2011, and landed on Mars in Gale entists for analysis by other instruments onboard Curiosity. Crater on August 5, 2012. Since its landing, Curiosity has Determining the composition of rocks on the Martian been assessing whether the Martian environment ever surface is usually a laborious, time-consuming task, even was, or is, capable of supporting microbial life. Like its for advanced spacecraft such as the Mars Exploration predecessors, the Curiosity rover will rely on proven tech- Rovers Spirit and Opportunity. Most rocks on the surface of nology, as well as new, innovative technologies to carry out Mars are covered with a layer of dust. Many rocks are also its mission. Curiosity represents the next big step in the covered with a layer of material that has been altered by exploration of Mars because it will: wind and possibly water. To determine the true composi- tion of a rock, spacecraft must first clear away the dust • demonstrate the ability to land a very large, heavy and the altered layer of rock. This involves the spacecraft rover to the surface of Mars (which could be used for a rolling up to the rock and using a tool to clear away the future Mars Sample Return mission that would collect unwanted layers. Dust can be easily removed but altered rocks and soils and send them back to Earth for labora- layers of rock usually need to be removed by grinding tory analysis) away the material. Another obstacle is the rock grinder. • demonstrate the ability to land more precisely in a They wear down. In fact the rock grinders on the previous 20-kilometer (12.4-mile) landing circle and MER rovers wore down long ago. When conducting a Mars mission, this seemingly menial task may require at least • demonstrate long-range mobility on the surface of the one day’s worth of operations, a long time by mission red planet (5-20 kilometers or about 3 to 12 miles) for standards. ChemCam, with its unique laser system, will be the collection of more diverse samples and studies. able to perform the same task in a fraction of the time and Technology development makes missions possible. Each without having to be in contact with the rock. Mars mission is part of a continuing chain of innovation. Each relies on past missions for proven technologies and LDRD Investment Made ChemCam Possible contributes its own innovations to future missions. This The LIBS technology was developed at Los Alamos National chain allows NASA to push the boundaries of what is cur- Laboratory with early support from the LDRD program. The rently possible, while still relying on proven technologies. research focused on LIBS technique for detection work in hazardous environments. A 2010 LDRD project explored direct application of LIBS on the ChemCam instrument, ChemCam including LIBS detection limitations in Mars and Venus ChemCam is the Chemistry and Camera instrument, one environments. Today LDRD is exploring the applications of of 11 science instruments onboard Curiosity. As the name this technology for finding and analyzing nuclear materials. implies, ChemCam is actually two different instruments The multi-mission nature of this LDRD work is representa- combined as one: a Laser-Induced Breakdown Spectrom- tive of the large majority of the LDRD portfolio, which eter (LIBS) and a Remote Micro Imager (RMI). The purpose boasts numerous projects with results that impact several of the LIBS instrument is to provide elemental composi- mission areas. tions of rock and soil, while the RMI will give ChemCam scientists high-resolution images of the sampling areas of the rocks and soil that LIBS targets. Simply stated, ChemCam will tell us what the rocks are made of in the Curiosity rover’s landing region. The primary objectives of ChemCam are to rapidly analyze rocks and soil to determine their compositions and to 24

Page 25

Mission Impact An LDRD research project titled Hierarchical Sparse Models for Robust Analysis of Video Data brings together world- LDRD delivers capabilities and explores solutions for all Los class researchers in statistical modeling, computer vision, Alamos missions: Nuclear Security, Global Security, Energy and high performance computing. Their Video Analysis Security, and Scientific Discovery. In some cases these lead Search Technology (VAST) is being developed for integra- to clearly identified follow-on projects. In part, this can be tion into an intelligence system for the DoD and other attributed to the program development mentors who help communities. transition these projects from LDRD to external funding. Early LDRD investment in VAST exercised and strengthened Putting a quantitative metric on mission impact is difficult. Laboratory capabilities in image analysis, biosecurity, and The obvious metric (follow-on funding) omits other ben- data analysis and pattern recognition, and serves critical efits like new ideas and approaches that feed into ongoing Federal and industry sponsors. work. Counting dollars is also imperfect because, although LDRD is part of what enables new mission solutions, the Laboratory brings other assets to bear, including facilities, Uniquely combining Magnetic ideas, and personnel from a range of programs. Nonethe- Resonance Imaging (MRI) and less, admittedly rough numbers show that projects active X-ray technology, LDRD scien- in FY13 played an important part on the path to $53M of tists unlocked new threat externally funded R&D. This number will grow with time as technology that could make follow-on funding is often acquired after projects end. airport travel faster and safer. This advanced technology, The path from LDRD to mission is clearly demonstrated by called MagRay, builds on follow-on projects from sponsors that range, inside DOE, previous Lab technology, cap- from the nuclear weapons program to nonproliferation tures measurements of MRI, programs, applied energy programs, and the Office of Sci- which is better suited for screening liquids than traditional ence, and more broadly, across a range of sponsors that baggage scanners, alongside x-ray density, and a new pa- include DoD (DARPA, ONR, and DTRA), DHS and DNDO, NIH rameter—proton content—that is not available in either and CDC, classified agencies, and a suite of industrial part- x-ray or MRI alone. ners. The capabilities sustained by these external sponsors feed back once again to enable the Laboratory’s execution Traditional baggage scanners, which use x-rays, have lim- of NNSA missions. ited sensitivity for liquid discrimination, whereas MagRay quickly and accurately distinguishes between liquids that Here we highlight two successful projects that acquired visually appear identical. For example, what appears to be external funding in FY13: a bottle of white wine could potentially be nitromethane, a clear liquid that can be used to make an explosive. The Vision is one of operator interface is straightforward: A red light signals a the most challeng- hazardous liquid; a green light signals a benign liquid. The ing and intrigu- technology could eliminate the need for the 3.4-oz con- ing problems in tainer limitation (the 3-1-1 rule) for travelers. artificial intelligence and computer science. Advances in the field have hinged on three fundamental challenges: (1) Early LDRD investment in the MRI technology used in how to create mathematical models of natural video se- MagRay supported the Laboratory’s global security mission quences that can generate and interpret the visual scene; and Science of Signatures pillar. The Department of Home- (2) how to learn image feature representations in such land Security is supporting the development of MagRay for models across many spatial and temporal scales and for use in airports across the nation. many object categories; and, (3) how to exploit the sheer size and richness of large-scale video datasets now becom- ing available. These three problems not only need to be solved together but are now within reach as a result of new theoretical, experimental and computational advanc- es. Los Alamos researchers developed novel video analysis models and algorithms that are expected to approach hu- man analyst performance and will surpass human speed. 25

Page 26

Awards and Recognitions The LDRD program supports some of the Laboratory’s most accomplished researchers, as well as many of its most prom- ising young scientists and engineers. In the past year, LDRD researchers received many awards and recognitions, includ- ing R&D 100 Awards; fellowships from professional associations; poster, publication, and research awards; and many more. Here we highlight only a few. dination chemists – devoted to understanding and promot- Zelenay receives Research Award from the ing the richness of the chemistry of the elements. Electrochemical Society The Energy Technology Division of Currently an LDRD PI, Kiplinger is internationally recog- The Electrochemical Society pre- nized as a scientific leader in actinide and lanthanide sented the Research Award to Piotr chemistry. She is a Fellow of the American Association Zelenay. The award recognizes his for the Advancement of Science and the Royal Society of “outstanding and original contribu- Chemistry, and she has received two R&D 100 Awards, a tions to the science and technology Los Alamos Fellows Prize for Research, and a Los Alamos of energy-related research areas Distinguished Mentor Award. She has served on the that include scientific and techno- Editorial Board for the American Chemical Society journal logical aspects of fossil fuels and Organometallics and as an Alternate Councilor for the ACS alternative energy sources, energy Division of Inorganic Chemistry. Kiplinger has published management and environmental consequences of energy nearly 80 journal articles and has received more than 2100 utilization.” citations. Currently PI on an LDRD DR project, Zelenay’s research José Olivares elected to Board of Directors of the focuses on polymer electrolyte fuel cells (PEFCs), includ- Algae Biomass Organization ing the direct-methanol fuel cell (DMFC), precious-metal/ The Algae Biomass Organization non-precious metal electrocatalysis, membrane and (ABO) chose José Olivares to serve membrane-electrode assembly development, and perfor- on its Board of Directors. The ABO mance optimization of fuel cell materials and components. is a non-profit organization whose He is the recipient of a 2003 Laboratory Distinguished mission is to promote the develop- Performance Award and a 10-time recipient of the Labora- ment of viable commercial markets tory’s Patent & Licensing Award. He garnered a 2010 DOE for renewable and sustainable com- Hydrogen Program R&D Award in recognition of outstand- modities derived from algae. ABO’s ing contributions to fuel cell technologies, and 2001 DOE board guides the organization to educate the general “Energy 100” and “Energy @23” awards for fuel cells for public, policy makers, and industry about the benefits transportation. He has authored more than 100 scientific of algae to address energy security, food production and papers, holds six patents, and serves on the editorial board stainability, and to advocate for policies that can acceler- of the Springer journal Electrocatalysis. ate the development of commercial markets for products made from algae. Jaqueline Kiplinger selected as Subdivision Chair in the American Chemical Society Olivares’ research focus has been primarily in instrumenta- Jaqueline L. Kiplinger was selected tion for mass spectrometry and biological and chemical to serve as the chair-elect for the sensor development for national security and health Organometallic Subdivision of the related applications. He is the Division Leader for the American Chemical Society (ACS) Bioscience Division and serves on the DR Strategy Team. Division of Inorganic Chemistry. He is the Executive Director of the National Alliance for She will act as a liaison between Advanced Biofuels and Bioproducts, a consortium of 40 the Division and the organometallic institutions funded by the DOE - Energy Efficiency and community, proposing symposia Renewable Energy Office of Biomass Programs. Olivares is for future ACS meetings and com- an adjunct member of The Danforth Plant Science Center, municating opportunities for Division support for organo- a non-profit institution dedicated to plant research for metallic conferences. The Division of Inorganic Chemistry food and energy, and co-Editor-in-Chief of the scientific represents a diverse body of scientists – organometallic, journal Algal Research. bioinorganic, solid-state, materials, nanoscience and coor- 26

Page 27

Performance Metrics Patents and Disclosures Another indication of the cutting-edge nature of the research funded by LDRD is the contribution the program The LDRD program is a key resource for addressing the makes to the intellectual property of the Laboratory. In long-term science and technology goals of the Laboratory, FY13, LDRD-supported research resulted in 32 patents, as well as for enhancing the scientific capabilities of Labo- 42% of the Laboratory’s total, and 30 disclosures, 27% of ratory staff. Through careful investment of LDRD funds, the Laboratory’s total. the Laboratory builds its reputation, recruits and retains excellent scientists and engineers, and prepares to meet Patents Granted evolving national needs. The impacts of the LDRD program are particularly evident in the number of publications and FY10 FY11 FY12 FY13 citations resulting from LDRD-funded research, the number of postdoctoral candidates supported by the program, and LANL Patents 61 59 72 77 the number of awards LDRD researchers received. LDRD Supported 15 15 11 32 Publications % due to LDRD 24% 25% 15% 42% The numerous publications made possible with LDRD fund- ing help the Laboratory maintain a strong presence and scientific reputation in the broader scientific community. Disclosures In FY13, LDRD researchers generated 534 peer-reviewed FY10 FY11 FY12 FY13 publications, accounting for 24% of the Laboratory’s total. The quality of these publications is evidenced by the num- LANL Disclosures 116 129 103 110 ber of times they were cited. LDRD publications published in FY13 were cited 481 times, accounting for 38% of the LDRD Supported 16 28 34 30 Laboratory’s citations, as well as one of the Laboratory’s top 10 cited publications. With increased efforts to collect % due to LDRD 21% 22% 33% 27% post-project performance metrics, LDRD publication and citation counts increase year-to-year. Recent analysis of Postdoctoral Support citations since the inception of the LDRD program reveals 13 of the top 50 high cited publications are linked to LDRD. LDRD remains an important vehicle for recruiting the brightest researchers to the Laboratory, where they Peer-Reviewed Publications become innovators and scientific leaders. In FY13, LDRD supported 367 postdocs, accounting for 61% of the Labora- FY10 FY11 FY12 FY13 tory’s total. LANL Pubs 1741 2079 2119 2263 Postdoc Support LDRD Supported 408 452 458 534 FY10 FY11 FY12 FY13 % due to LDRD 23% 22% 22% 24% LANL Postdocs 547 580 581 596 LDRD Supported 325 330 349 367 Citations % due to LDRD 59% 57% 60% 61% FY10 FY11 FY12 FY13 LANL Citations 14416 6899 3373 1278 Postdoc Conversions LDRD Supported 3944 1866 1096 481 FY10 FY11 FY12 FY13 % due to LDRD 27% 27% 32% 38% LANL Conversions 53 44 41 57 Top 10 Most Highly Cited Publications LDRD Supported 33 17 21 34 LDRD Supported 4 4 1 2 % due to LDRD 62% 39% 51% 59% 27

Page 28

External collaborations are often critical to a successful LDRD project. Formal collaborations between LDRD and researchers at other national laboratories, academia, and industry enable access to world-leading facilities and knowledge. Such collaborations enable LDRD researchers to be active and prominent members of the broad scientific community. There were more than 1,000 external collaborators associated with FY13 LDRD projects, many of which involved no exchange of funds. New global HIV vaccine design shows promise in pre-clinical trials The considerable diversity of HIV worldwide represents a critical challenge for designing an effective HIV vaccine. Now, it appears that a vaccine bioinfor- matically optimized for immunologic coverage of global HIV diversity, called a mosaic vaccine and designed by Los Alamos researcher Bette Korber and her team may confer protection from infection. Bette Korber developed a component of These vaccines are specifically designed to present the most common forms a new vaccine against HIV that is now of parts of the virus that can be recognized by the immune system. This being tested in monkeys. new insight regarding a mosaic vaccine’s ability to protect from infection is the result of work by a scientific team led by Beth Israel Deaconess Medical Center (BIDMC), including Los Alamos LDRD researchers. The study, which was conducted in monkeys, was published in the journal Cell. “These data suggest a path forward for the development of a global HIV vaccine and give us hope that such a vaccine might indeed be possible,” says lead author Dan H. Barouch, MD, PhD, the director of the Center for Virology and Vaccine Research at BIDMC and professor of medicine at Harvard Medical School. “We are planning to advance this HIV vaccine candidate into clinical trials next year,” he adds. 28

Page 29

Chemistry and Material Sciences

Page 30

Chemistry and Material Sciences Directed Research Continuing Project Energy Storage Fernando H. Garzon 20120003DR Introduction Progress The storage of energy from intermittent renewable A number of major accomplishments were achieved in energy sources to supply the irreducible base needed FY2013. The high pressure experimental research capa- by electrical utilities is a critical limitation to widespread bility of the Laboratory was enhanced by the completion of renewable energy. Ammonia has been proposed of a high pressure reaction chemistry facility. The facility, by many as a carbon-free energy carrier and storage located in MPA- Division, includes a custom designed medium (NH holds 14 MJ/liter, 10-100 times more than high pressure systems (200 atm) to perform gas adsorp- 3 the best battery). However, current methods of ammo- tion, infrared spectroscopy of surfaces and high pres- nia production have high capital costs, low to moderate sure electrochemical experimentation. The laboratory efficiency, and are not scalable to utility levels. Our main now can safely perform ammonia synthesis, character- goals are to: 1) Improve our theoretical understanding of ize ammonia adsorption in nonaqueous electrolytes, nitrogen reduction on surfaces, multi-electron transfer determine the reaction species and intermediates on to adsorbed molecules, and proton transport in anhy- electrochemical catalyst surfaces and measure funda- drous electrolytes. 2) Develop, characterize and optimize mental thermodynamic properties of ionic liquid electro- new materials for electrolytes and electrocatalysts for lyte ammonia mixtures. The team also developed a high ammonia electrosynthesis. 3) Demonstrate improved ef- temperature electrochemical experimental capability for ficiency of scalable ammonia electrosynthesis concepts. characterization of solid proton conductor based electro- chemical reactors and an RF magnetron in situ thin film Benefit to National Security Missions sputter deposition capability suitable for the creation of nanoscale catalytic electrode materials relevant to the Los Alamos National Laboratory has a stated mission in needs of the project. the area of Energy Security. This project directly sup- ports the Laboratory’s grand challenges in Energy and Important milestones were achieved with the newly Earth Systems and Materials, coupling to the themes developed experimental capabilities. The deposition of energy storage, interfaces, emergent properties, systems yielded high surface area molybdenum nitride and nanotechnology. Success will lead us to the design thin films on a variety of substrates. The crystal structure principles for materials to be used in the next genera- morphology and surface chemistry of these nanoscale tion of electrochemical energy storage systems with electrode materials were characterized by electron substantially improved efficiency. The work is expected microscopy, X-ray diffraction, X-ray photoemission spec- to lead to follow-on proposals in BES in the areas of troscopy and gas adsorption measurements. New indium new materials, interfaces, emergent properties, energy tin pyrophosphate and strontium cerate proton conduct- conversion, and energy storage. In the applied area, such ing membranes were synthesized and characterized. as ARPA-E, DOE-EERE, and DOE-OE, this work will posi- The proton conduction mechanism of tin pyrophosphate tion LANL to lead or be a strong team member with ANL is currently under debate within the scientific commu- for the development of improved transportation, local, nity. Experiments and theoretical modeling conducted and distributed-grid-level electrical storage systems. The by LANL determined that the protons moved by a grain low cost synthesis of ammonia from renewable energy is boundary transport mechanism and yielded several pre- also of importance to the USDA. 30

Page 31

sentations and publications. A large number of indium tin ammonia synthesis conversion rate will be evaluated by DC pyrophosphate materials with varying metal to phosphate polarization measurements and optimized by alterations of ratios were synthesized using solution precipitation meth- the electrochemical cell design. ods developed and patented by LANL. The crystal struc- tures, surface areas and ionic conductivities as a function As the Los Alamos Neutron Scattering Facility is now opera- of temperature, and gas composition were determined by tional this year, supporting studies will include structural electrochemical transport measurements. Our data clearly neutron diffraction studies of the molybdenum nitride showed that no structural changes in the crystalline phase electrode materials and solid electrolytes. Neutron vibra- were occurring as the sample composition was varied from tion spectroscopy will characterize the proton dynamics stoichiometric to excess phosphate compositions. However is the pyrophosphate material. IR spectroscopy and NMR the presence of a phosphate-rich grain boundary phase spectroscopy of the ionic liquid -ammonia systems will was detected and the measured proton conductivity was a provide molecular information about the interactions of strong function of the grain boundary phase. ammonia with the protonic ionic liquids. The strontium cerate proton conducting materials with Theoretical studies will include Density Functional Theory electro-catalytic electrodes were incorporated into small based calculations of the nitrogen reduction reaction on scale electrochemical reactors for measurements of am- metal nitride surfaces and studies to elucidate the ionic monia electrochemical synthesis from nitrogen and hydro- transport mechanisms in both solid and liquid electrolyte gen feedstock materials. The reactors successfully electro- systems. Pathways for homogenous nitrogen catalysis on synthesized ammonia. Protonic ionic liquids were prepared metal centered organic catalysts will evaluated using mo- and purified for low temperature liquid electrolyte cells. lecular dynamics. Ionic conductivity of the protic ionic liquids was character- ized using AC impedance spectroscopy and the thermal Conclusion stability was determined. This project will investigate the conversion of electrical energy into liquid fuel, greatly increasing our ability to Neutron vibrational spectroscopy is a powerful tool for store GWhs of energy onsite. The project will increase our investigating the surface reactions occurring on the am- fundamental understanding of nitrogen reduction chem- monia synthesis catalysts. The technique requires relatively istry, decrease the cost of plant fertilizer from renewable large homogenous samples, thus we developed nanosyn- resources, and enable the more efficient utilization of thesis methods to yield large quantities of well-charac- intermittent renewable energy sources. The project will terized molybdenum nitride catalysts for our upcoming expand our theoretical knowledge of a meaningful electro- LANSCE run cycle. The nanosynthesis and characterization synthetic process, charge transfer theory, and emergent was presented and published in FY13. ion transport properties at nanoscale interfaces. Future Work Publications In the following fiscal year we will continue our electro- Kreller, C. R., M. S. Wilson, R. Mukundan, E. L. Brosha, and catalyst materials development and electrode optimiza- F. H. Garzon. Stability and Conductivity of In3+-Doped tion for ammonia synthesis. We will test our high surface SnP2O7 with Varying Phosphorous to Metal Ratios. area thin film molybdenum nitride thin films deposited on 2013. ECS ELECTROCHEMISTRY LETTERS. 2 (9): F61. conductive support materials in electrochemical cells con- Kreller, C. R., M. S. Wilson, and F. H. Garzon. Enhancing taining ionic liquid electrolytes and on solid oxide proton the Protonic Conductivity of Tin Pyrophosphates by conductors. Continued electrolyte studies will include ionic Increasing Phosphate Content . 2013. Electrochemical transport measurements on protic ionic liquids particularly Society Transactions. 57 (11): 1009. ammonia -ionic liquid interaction studies at high pressures in our new test facility. We will measure the solubility of Kreller, C. R., R. Mukundan, and F. H. Garzon. ammonia in protic ionic liquids, and characterize how the Electrosynthesis of Ammonia using Solid-State the binary mixture properties change with composition. Proton Conductors. To appear in 224 Meeting of the Solid state electrochemical cells will be fabricated us- Electrochemical Society. (San Francisco, 27 Oct.-4 Nov.). ing thin ceramic and polymer reinforced solid electrolyte membranes. Liquid cells will be fabricated for pressurized Ogumi, Z., H. Matsuoka, F. Garzon, H. Kim, L. J. Wan, K. studies of the nitrogen reduction electrochemistry of Tamao, and M. Nakamura. Electrochemistry 80th nitrogen. Cyclic voltammetry and AC impedance spectros- Anniversary Special Issue. 2013. ELECTROCHEMISTRY. copy will characterize ionic transport in these systems. The 81 (3): 140. 31

Page 32

Sansinena, J. M., J. Chilstunoff, N. C. Tomson, J. M. Boncella, and F. H. Garzon. Ionic Liquids for Ammonia Electrosynthesis and Energy Storage . Presented at 224th Meeting of the Electrochemical Society. (San Francisco, 27 Oct- 4 Nov. 2013). 32

Page 33

Chemistry and Material Sciences Directed Research Continuing Project Organic Electronic Materials: Designing and Creating Functional Interfaces Sergei Tretiak 20120019DR Introduction and (iii) by discovering emergent phenomena in complex The description of disordered organic materials and the systems. The project also strongly supports the Energy interfaces between these materials presents an excit- and Earth Systems challenges by providing design strate- ing challenge that requires the establishment of new gies for molecular materials suitable for clean energy approaches lying between those of condensed matter production. (which typically relies on crystal periodicity) and mo- lecular chemistry (which typically relies on small system Progress size). We have developed new theoretical tools and During the second year of the project our effort focused mastered sophisticated experimental techniques, which on several research directions, including synthesis, together will enable a quantitative jump in our under- device fabrication and characterization, and theoretical standing of physical and dynamical processes in disor- modeling. Our synthetic efforts result in synthesized and dered organic electronic materials. By manipulating both characterized series of thiophene oligomer derivatives: the nanostructure of the molecular building blocks and terthiophene, quatertiophene and pentathiophene de- the structure of the interfaces between them through rivatives with thiolated and syloxinated end group. The synthetic and fabrication processes, we aim to engineer latter allow for covalent bonding to the Au/Ag and glass materials and fabricate devices with emergent proper- substrate surfaces, respectively. In particular, Scanning ties and unprecedented efficiencies. The proposed inter- Tunneling Microscope (STM) characterization of self-as- disciplinary research harnesses the group’s unique depth sembled monolayer (SAM) thin films on single crystalline of expertise in theory (spanning all length scales from gold surfaces, reveal a surprising fact that only specific molecular electronic structure simulations to mesoscale oligomers from the family are able to organize into device modeling), time- and spatially resolved spectro- ordered monolayers, as opposite to the globular struc- scopic and transport measurements, organic synthesis, tures. These properties make clear differences on the device fabrication and characterization. device performance (as described below) due to chang- ing the underlying interface structure. Benefit to National Security Missions The project objectives will be achieved through integra- Ongoing quantum-chemical simulations are able to ratio- tion of LANL strengths in theory, spectroscopy, chemis- nalize the observed phenomena. Our device fabrication try, high-performance computing, and materials science and spectroscopic efforts focused on four fundamental on multiple length-/time-scales, thereby realizing the processes that determine the performance of typical vision of “co-design” and directly aligning with the Ma- organic photovoltaic devices. Typically exciton dissocia- RIE goal to “couple theory, experiment, and simulation tion rate is faster than it needs to be by several orders of tools.” The outcome will favorably position the Lab for magnitude, and charge transfer state recombination is future investments in the strategic field of Organic Func- also too fast. We inserted a LiF insulating layer between tional Materials, with application to a range of sensing the donor and acceptor layer in a prototypical system missions including, potentially, nuclear nonproliferation. (tetracene(Tc)/fullerene bi-layer) and find that there is We primarily address the Materials: Discovery Science an optimum thickness of the insulator layer for improv- to Strategic Applications challenge by (i) a thorough ing device performance. We applied ultrafast pump- understanding of dynamical processes at interfaces, (ii) probe spectroscopy to follow the dynamics of singlet their application to control of the material functionality and triplet excitons induced by ultrashort laser pulse 33

Page 34

absorption in thin Tc/fullerene films. These studies reveal Finally, to evaluate the influence of interfaces on the de- a complex dynamics of multiple excited states and present vice performance, we have applied a physics-based one- a first spectroscopic evidence of timescales and pathways dimensional model to assess the effect of incorporating a related to the exciplex formation in this system. We further thin tunnel barrier between the electron and hole trans- study donor/acceptor bi-layers (such as the MTDATA/ port layers of organic photovoltaic devices. The results Bphen and P3HT/C60) by photoconductivity and PL/TRPL evidence that by optimally choosing the insulating layer experiments (Thrust 2). thickness, the organic solar cell power efficiency can be significantly improved. The results of our first and second These experiments are performed on films and prototype year work have been summarized in 25 articles published/ devices to understand the effect of electric field on ener- submitted to the international journals. Several student getic and dynamic of charge-separated (exciplex) states. and postdoc hires were also made. Our studies reveal a clear presence of luminescent inter- facial exciplex state. Spectroscopic studies in the presence Future Work of magnetic field brought unique information on emission In the third year of the project we will continue joint dependence on the applied magnetic field evidencing experimental/theoretical studies of well-ordered materi- impact of hyperfine interactions in the dynamics of spin als to evaluate the impact of aggregation on the electronic states. Concurrent electronic structure studies interpret properties (Thrust 1). In particular, we will construct a self- the underlying phenomena in terms of ‘hot’ and ‘cold’ assembled monolayer of thiophene oligomers at air-water exciplex states. interface and on various substrates and conduct various optical and structural characterizations of self-assemblies. We further study the effects of inserting an insulating By incorporation synthesized self-assemblies into device layer at the interface of the P3HT/C60 pair. By incorporat- architectures as built in interfaces, we will further study ing a thin insulating LiF layer we are able to decrease the impacts of these interfaces on the device performances exciplex recombination rate and to decrease the exciplex and will obtain valuable insights, establishes correlation binding energy that results in a dramatic increase in pho- between interfacial structure and charge transfer dynam- tocurrent by two orders of magnitude. Insertion of well- ics. ordered monolayers of thiophene oligomers (see above), increase photocurrent even further. This already suggests Spectroscopic measurement (e.g., photoluminescence (PL) new synthetic strategies targeting optimization of energy and time-resolved PL (TRPL) study) will provide under- conversion efficiency and minimizing losses in organic standing of the effect of electric field on charge separation semiconductor devices. Our theoretical modeling span and recombination processes. Our supplemental theoreti- multiple time and length timescales. cal modeling using the LANL-developed NA-ESMD code and multiscale transport models will provide details of Using our LANL-developed Non-Adiabatic Excited State structure/property relationships and will help understand- Dynamics (NA-ESMD) code we investigate energy trans- ing of underlying electronic dynamics. fer processes in several polymeric systems and discover multiple channels in energy transport, which emphasizes We will further study donor/acceptor bi-layers (such as the complexity of this process going beyond simple Foerser- MTDATA/Bphen) by photoconductivity and PL/TRPL experi- like phenomenological approaches. We are finalizing ments (Thrust 2). These experiments will be performed not development of a computational platform to simulate the only on films but also on prototype devices to understand electronic (e.g., photoexcitation) and vibrational dynamics the effect of electric field on energetic and dynamic of of large organic molecules within a complex environment. charge-separated (exciplex) states. This Quantum Mechanics/Classical Mechanics (QM/MM) We will also attempt to tune the lifetimes of exciplexes computational framework combines available codes for by incorporating a tunnel barriers in the interface. The classical molecular dynamics of large organic systems with concurrent theoretical modeling will use the organic accurate simulations of excited state dynamics using the interface device model developed in FY12 to interpret NA-ESMD code. One the larger scale, our developed mul- current-voltage and capacitance-voltage measurements tiscale macroscopic charge transport models ‘bridging the on bilayer organic devices. Finally, we will finalize develop- gap’ between local properties of the polaron and exciton ment of theoretical tools (Quantum Mechanics/Molecular (nanoscale) and mesoscopic charge mobilities and exciton Mechanics (computational platform for modeling excited (micron-scale) are currently being extended to disordered state dynamics beyond the Born-Oppenheimer approxima- inhomogeneous systems. tion and extend organic interface device model to describe 34

Page 35

bulk heterojunction organic solar cells) and experimental Liu, F. L., B. K. Crone, P. P. Ruden, and D. L. Smith. Control capabilities (a time-resolved Optical Parametric Oscillator of interface microscopic processes in organic bilayer system at NHMFL). structures and their effect on device performance. 2013. Journal of Applied Physics. 113: 044516. Conclusion Liu, F., P. Paul. Ruden, I. H. Campbell, and D. Smith. Our effort will focus on the fundamental and applied sci- Electrostatic capacitance in single and double layer ence of organic π-conjugated materials including films organic diodes. 2012. APPLIED PHYSICS LETTERS. 101 and blends of conducting polymers, small molecules, (2): 023501. fullerenes, and graphene, all of technological importance, Liu, F., P. Paul. Ruden, I. H. Campbell, and D. Smith. Device present-day or anticipated. If successful, this will signifi- model for electronic processes at organic/organic cantly advance the knowledge and development of mul- interfaces. 2012. JOURNAL OF APPLIED PHYSICS. 111 tifunctional organic π-conjugated materials towards their (9): 094507. use in a variety of additional applications. Moreover, the theoretical tools/experimental techniques that we develop Nayyar, I. H., E. R. Batista, S. Tretiak, A. Saxena, D. L. Smith, and experience we gain will be readily applicable to a mul- and R. L. Martin. Effect of trans- and cis-isomeric titude of soft materials in the future focus of LANL mission defects on the localization of the charged excitations in -conjugated organic polymers. 2013. JOURNAL OF impact. POLYMER SCIENCE PART B-POLYMER PHYSICS. 51 (12): 935. Publications Alberti, S. F., V. D. Kleiman, T. Nelson, A. E. Roitberg, and Nayyar, I., E. Batista, S. Tretiak, A. Saxena, D. L. Smith, S. Tretiak. Shishiodoshi unidirectional energy transfer and R. L. Martin. Role of Geometric Distortion and mechanism in phenylene ethynylene dendrimers. Polarization in Localizing Electronic Excitations in 2012. Journal of Chemical Physics. 137: 22A526. Conjugated Polymers. 2013. Journal of Chemical Theory and Computations. 9: 1144. Campbell, I. H., and B. K. Crone. Improving an organic photodiode by incorporating a tunnel barrier between Nelson, T., S. Fernandez-Alberti, A. E. Roitberg, and the donor and acceptor layers. 2012. APPLIED PHYSICS S. Tretiak. Nonadiabatic excited-state molecular LETTERS. 101 (2): 023301. dynamics: Treatment of electronic decoherence. 2013. JOURNAL OF CHEMICAL PHYSICS. 138 (22): -. Clark, J., T. Nelson, S. Tretiak, G. Cirmi, and G. Lanzani. Femtosecond torsional relaxation. 2012. Nature Nelson, T., S. Fernandez-Alberti, A. E. Roitberg, and S. Physics. 8 (3): 225. Tretiak. Conformational disorder in energy transfer: beyond Forster theory. 2013. PHYSICAL CHEMISTRY Fernandez-Alberti, S., A. Roitberg, T. Nelson, and S. Tretiak. CHEMICAL PHYSICS. 15 (23): 9245. Identification of unavoided crossings in nonadiabatic photoexcited dynamics involving multiple electronic Nelson, T., S. Fernandez-Alberti, V. Chernyak, A. Roitberg, states in polyatomic conjugated molecules. 2012. and S. Tretiak. Nonadiabatic excited-state molecular Journal of Chemical Physics. 137 (1): 014512 (10 pp.). dynamics: Numerical tests of convergence and parameters. 2012. Journal of Chemical Physics. 136 (5): Furmanchuk, A., J. Leszczynski, S. Tretiak, and S. Kilina. 054108. Morphology and optical response of carbon nanotubes functionalized by conjugated polymers. 2012. Journal Park, Y. I., C. Y. Kuo, J. S. Martinez, Y. S. Park, O. Postupna, of Physical Chemistry C. 116 (12): 6831. A. Zhugayevych, S. Kim, J. Park, S. Tretiak, and H. L. Wang. Tailored Electronic Structure and Optical Gorshkov, V. N., S. Tretiak, and D. Mozyrsky. Semiclassical Properties of Conjugated Systems through Aggregates Monte-Carlo approach for modelling non-adiabatic and Dipole-Dipole Interactions. 2013. ACS APPLIED dynamics in extended molecules. 2013. NATURE MATERIALS & INTERFACES. 5 (11): 4685. COMMUNICATIONS. 4: -. Sheng, C. X., S. Singh, A. Gambetta, T. Drori, M. Tong, Kilina, S., E. Batista, A. Saxena, D. L. Smith, R. L. Martin, S. Tretiak, and Z. V. Vardeny. Ultrafast intersystem- and S. Tretiak. Effect of Packing on Formation of Deep crossing in platinum containing pi-conjugated polymers Carrier Traps in Amorphous Conjugated Polymers. with tunable spin-orbit coupling. 2013. SCIENTIFIC 2013. Journal of Physical Chemistry Letters. 4: 1453. REPORTS. 3: -. Kilina, S., J. Ramirez, and S. Tretiak. Brightening of the Soler, M. A., A. E. Roitberg, T. Nelson, S. Tretiak, and S. F. lowest exciton in carbon nanotubes via chemical Alberti. Analysis of state-specific vibrations coupled functionalization. 2012. Nano Letters. 12 (5): 2306. 35

Page 36

to the unidirectional energy transfer in conjugated dendrimers. 2012. Journal of Physical Chemistry A. 116: 9802. Zhugayevych, A., O. Postupna, R. Bakus , G. Welch, G. Bazan, and S. Tretiak. Ab-initio study of a molecular crystal for photovoltaics: light absorption, exciton and charge carrier transport. 4920. Journal of Physical Chemistry C. 117: 2013. 36

Page 37

Chemistry and Material Sciences Directed Research Continuing Project Plutonium-242: A National Resource for the Fundamental Understanding of 5f Electrons Eric D. Bauer 20120024DR Introduction materials science, directly addresses the Materials Grand Plutonium is arguably the most complex element in the Challenge with an emphasis on the emergent phenom- periodic table. Plutonium’s complex electronic structure ena arising from Pu’s 5f electrons. A Plutonium Science is due to its position in the actinide series where 5f elec- Strategy has recently been developed as part of the Plu- trons move from being tied to the atoms (localization) to tonium Center of Excellence. The cutting edge research freely moving around the metal (itineracy). Understand- contained within this project for a fundamental under- ing the behavior of the 5f electrons at this crossover standing of the 5f electrons of plutonium specifically poses a significant challenge to the condensed matter address many aspects of the Plutonium Science Strategy. community. As a major step to address the challenge of Our program will assert LANL leadership in Pu science understanding the 5f electrons of Pu, we will synthesize and stretch the Lab’s capability base in new directions high-purity single crystals of delta-242Pu and 242Pu- through the development of a robust, long-term delta- based compounds to, for the first time: 1) use advanced Pu single crystal grain growth capability and determina- photon spectroscopies to determine the valence of Pu’s tion of Pu’s electronic structure using the most sophisti- 5f electrons, which will significantly constrain current cated experimental and theoretical techniques available. theories of Pu electronic states; 2) apply novel photon Plutonium science also supports nuclear nonprolifera- spectroscopy and neutron scattering techniques to tion, nuclear energy, and environmental remediation. determine the magnetic configuration of delta-Pu’s 5f electrons and their charge/spin fluctuations, resolving a Progress decades-long debate; and 3) perform seminal measure- This year, we have performed seminal measurements of ments of the electronic band structure of delta-Pu using the magnetic fluctuation spectrum in large polycrystal- our unique angle-resolved photoemission capability to line samples of delta-242^Pu (22 g) and 242^PuCoGa 5 significantly constrain equation of state models. The (15 g) at the Spallation Neutron Source at Oak Ridge fundamental and comprehensive understanding of the National Laboratory. In particular, we found evidence for 5f electrons of Pu obtained through this integrated effort magnetic scattering centered at 90 meV and derived a in experiment and theory will provide the scientific basis magnetic form factor—for the first time ever. This result for Pu’s equation of state, phase stability, and phase implies that the Kondo effect (i.e., a fluctuating magnetic transformation kinetics under extreme conditions.This moment) is the origin of the missing magnetism in delta- national program will stretch the Lab’s expertise in char- Pu, resolving a debate that has persisted for more than acterizing and modeling complex materials, as well as re- 20 years. Experiments on 242^PuCoGa were successful 5 vitalize plutonium science in this country and make it an and preliminary analysis of the data suggests additional attractive field of study for the next generation of LANL scattering at very low energy (1 meV) deep in the super- scientists; it is a crucial step towards fulfilling LANL’s role conducting state. Further analysis is ongoing. as the Plutonium Center of Excellence and the Plutonium Science Strategy. We were successful in our attempts to grow large single grains of delta-Pu up to 1.3 mm in diameter. Starting Benefit to National Security Missions with fine-grained Ga-stabilized delta-Pu, we strained The fundamental understanding of plutonium obtained the material between 2-4%, then annealed for 100 from this project will provide the scientific underpinning hours at about 400 C. This produced a number of large for the Lab’s core missions of stockpile sustainability and single grains, a couple of which are large enough to 37

Page 38

carry out band structure measurements of delta-Pu via In a three-pronged effort, the theoretical work for the angle-resolved photoemission (ARPES) for the first time. project spans the range from providing exact solutions in We have begun to harvest two of the large grains and will the extreme itinerant limit (local density approximation), collect our initial ARPES before the end of the fiscal year. to the localized limit (hybrid functionals), and in between In parallel, we performed two de Haas van Alphen (dHvA) these limits where the Coulomb repulsion and kinetic en- measurements to probe the Fermi surface on very small ergies of the 5f electrons are comparable dynamical mean delta-Pu grains (100 microns) at the National High Mag- field theory (DMFT). In a tour de force effort, we recently netic Field Laboratory using a novel technique involving completed calculations on alpha-Pu, for the first time, a micro-silicon cantilever. While these first two attempts within the framework of density functional theory (DFT) were not successful, we plan to measure the Fermi surface and the generalized gradient approximation (GGA) to- with this technique in the new modulation coils, which gether with dynamical mean-field theory (DMFT). In short, improves the sensitivity of the technique by more than an alpha Pu appears to have the capacity to simultaneously order of magnitude. have multiple degrees of electron localization/delocaliza- tion of Pu 5f electrons within this elemental phase of Pu. We carried out a number of measurements on the PuMIn 5 and PuMGa (M=Co, Rh) superconductors to investigate Future Work 5 the effects of the 30% volume collapse from PuRhIn5 to For the next fiscal year, we will focus on the following PuCoGa , which is similar to the volume difference be- goals: 5 tween alpha-Pu and delta-Pu. By performing resonant ultrasound measurements on a single crystal of PuCoGa ,

1. Harvest large single grains of delta-Pu prepared by the 5 which is most similar to delta-Pu, we extracted all six strain-anneal technique for first-ever measurement of elastic constants of this tetragonal system. Analysis of delta-Pu band structure via angle resolved photoemis- the superconducting and normal states reveal that only sion experiments. the compressional modes show an anomaly below Tc and further indicate an unusual softening of the lattice below 2. Perform measurements of the angular momentum 50 K. These results suggest that valence fluctuations may components of of delta-Pu at the Advanced Photon mediate the superconductivity in PuCoGa . Extensive mea- Source to determine magnetic configuration its the 5f 5 surements on the PuRh(In1-xCdx)5 single crystals reveal electrons. that long-range antiferromagnetic (AFM) order is found
2. Perform de Haas van Alphen measurements on small, for x>0.07 Cd with a narrow region of coexistent supercon- readily available single grains of delta-Pu using a new ductivity and AFM order. This is the first example of such modulation coil for improved sensitivity at the National interplay between superconductivity and magnetism in a High Magnetic Field Laboratory. plutonium compound. To further explore the nature of the 5f electrons of plutonium in these materials, we performed 4. Carry out measurements of the phonon dispersions angle-resolved photoemission (ARPES) and x-ray pho- on single grains of delta-Pu at the Advanced Photon toemission spectroscopy (XPS) measurements on single Source to determine the softening of the specific pho- crystals of PuCoIn , PuPt In , and PuCoGa . These experi- non mode related to the delta-alpha phase transitions. 5 2 7 5 ments reveal the degree of localization of the Pu 5f elec-
3. Investigate origin of multiconfigurational ground state trons. The ARPES measurements also reveal the electronic recently observed by our team in alpha-Pu and delta- band structure, which when compared to dynamical mean Pu through combination of dynamical mean field field theory, delineate the degree of hybridization of the 5f theory and detailed x-ray absorption spectroscopy and conduction electron states that characterize the strong measurements. electronic correlations in these materials.
4. Perform inelastic neutron scattering measurements on We continue to search for the direct Pu-239 nuclear mag- delta-242^Pu and 242^PuCoGa5 and various neutron netic resonance (NMR) signal in a number of Pu interme- facilities (Spallation Neutron Source, Neutron Center tallic compounds. Preliminary results for observing the for Research, and again at LANL). Pu-239 signal in the superconductor PuCoGa look promis- 5 ing for observing this fingerprint of plutonium’s nucleus 7. Search for the Pu-239 direct NMR signal in various will allow us to study its 5f electrons, which play a decisive Pu-based superconductors and magnets to reveal the role in controlling the degree of localization/itineracy of its nature of the 5f electrons in these materials. 5f electrons. 38

Page 39

Conclusion Study of the Entangled Crystal, Magnetic, and Electronic Structures of PuGa3 . 2013. The complex behavior of Pu arising from its unique posi- arXiv:1307.0762arXiv:1307.0762 . tion in the periodic table remains an enigma to this day and poses one of the most significant challenges to DOE’s Wang, C. C., M. D. Jones, and J. Zhu. Fermi surface national security missions. Sustaining the aging nuclear topology and de Haas–van Alphen orbits in PuIn_\{3\} stockpile at significantly reduced levels demands a funda- and PuSn_\{3\} compounds. 2013. Physical Review mental understanding of Pu’s 5f electrons beyond what is BPhysical Review B. 88 (12): 125106. currently known. Our approach consists of performing ex- Yasuoka, H., G. Koutroulakis, H. Chudo, S. Richmond, D. periments on single crystals of delta-Pu and Pu compounds K. Veirs, A. I. Smith, E. D. Bauer, J. D. Thompson, G. containing the isotope Pu-242. Our work will provide a D. Jarvinen, and D. L. Clark. Observation of Pu-239 fundamental and comprehensive understanding of plutoni- Nuclear Magnetic Resonance. 2012. ScienceScience. um necessary for the continued success of DOE’s stockpile 336 (6083): 901. sustainability and other national security missions. Yasuoka, H., G. Koutroulakis, H. Chudo, S. Richmond, D. Publications K. Veirs, A. I. Smith, E. D. Bauer, J. D. Thompson, G. D. Jarvinen, and D. L. Clark. Observation of Pu-239 Bauer, E. D., P. H. Tobash, J. N. Mitchell, and J. L. Sarrao. Nuclear Magnetic Resonance. 2012. ScienceScience. Single crystal growth of plutonium compounds 336 (6083): 901. from molten metal fluxes. 2012. Philosophical MagazinePhilosophical Magazine. 92 (19-21): 2466. Zhu, J., R. C. Albers, K. Haule, G. Kotliar, and J. M. Wills. Site-selective electronic correlation in α-plutonium Bauer, E. D., P. H. Tobash, J. N. Mitchell, and J. L. Sarrao. metal. 2013. Nat CommunNat Commun. (in press). Single crystal growth of plutonium compounds from molten metal fluxes. 2012. Philosophical MagazinePhilosophical Magazine. 92 (19-21): 2466. Booth, C. H., Y. Jiang, D. L. Wang, J. N. Mitchell, P. H. Tobash, E. D. Bauer, M. A. Wall, P. G. Allen, D. Sokaras, D. Nordlund, T. C. Weng, M. A. Torrez, and J. L. Sarrao. Multiconfigurational nature of 5f orbitals in uranium and plutonium intermetallics. 2012. Proceedings of the National Academy of Sciences of the United States of AmericaProceedings of the National Academy of Sciences of the United States of America. 109 (26): 10205. Haga, Y., E. D. Bauer, P. H. Tobash, J. N. Mitchell, O. Ayala- Valenzuela, R. D. McDonald, C. H. Mielke, and Z. Fisk. Shubnikov-de Haas oscillation in PuIn3. 2013. Journal of the Korean Physical SocietyJournal of the Korean Physical Society. 63 (3): 380. II, M. F. Beaux, J. J. Joyce, T. Durakiewicz, K. S. Graham, E. D. Bauer, P. H. Tobash, and S. Richmond. Electronic Structure, Localization and 5f Occupancy in Pu Materials. Mate. Jian-Xin, Z., P. H. Tobash, E. D. Bauer, F. Ronning, B. L. Scott, K. Haule, G. Kotliar, R. C. Albers, and J. M. Wills. Electronic structure and correlation effects in PuColn 5 as compared to PuCoGa 5. 2012. Europhysics LettersEurophysics Letters. 97 (5): 57001 (5 pp.). Mitchell, T. E., J. P. Hirth, D. S. Schwartz, and J. N. Mitchell. The beta-to-alpha phase transformation in plutonium. 2013. Acta MaterialiaActa Materialia. 61: 2895. Rudin, S. P.. Density Functional Theory Calculation 39

Page 40

Chemistry and Material Sciences Directed Research Continuing Project Modern Challenges in Actinide Science Albert Migliori 20120167DR Introduction nuclear forensics, nuclear weapons, nonproliferation, The purpose of Modern Challenges in Actinide Science and the associated agencies by studying the important is to support cutting-edge research for targeted tasks at modern experimental, engineering, technical, and theo- the forefront of actinide science today that directly sup- retical questions that govern the properties of actinides. port LANL missions. Because each of the mission-relevant areas has as an irreducible central component of the actinide elements, Broadly defined, this project supports experimental research on modern scientific challenges in actinides is and theoretical research that underlies LANL missions a required component of each of the relevant areas. It in fundamental actinide science (all aspects), nuclear is expected that the selected sub-tasks will each lead to weapons, nuclear energy and energy security, nuclear knowledge and capabilities that will attract internal and safeguards and detection of actinide materials, nuclear external funding where new science merits it and where nonproliferation and forensics, fate and transport of ac- that science is instrumental in resolving technical and tinides in the environment, interactions of actinides with engineering issues. biological systems, and nuclear waste management—all deemed strategic for the Laboratory. Some of these will Progress be in areas that will enhance the “plutonium center of • The projects led by postdoctoral fellows in the last excellence” designation for Los Alamos. Other areas will year have resulted in 28 publications. support MaRIE and the “LANL-LLNL Plutonium Science • Resonant Ultrasound Studies of plutonium elas- and Research Strategy.” Noting that “actinides” are more tic moduli versus temperature and time required than just plutonium, and that the “other” actinides are advances in the analysis of huge amounts of data. exceptionally important for forensics, environmental-re- These advances enabled studies of the temperature lated missions, biological and waste management issues dependence of the elastic moduli of superconduc- and more, a balance will be maintained among the many tors with unprecedented precision. The result is actinide research areas chosen for this project. the discovery of a new thermodynamic phase in high temperature superconductors, “Bounding For definiteness, the project will support research in the pseudogap with a line of phase transitions in sub-tasks as described above, identified and worked on YBa2Cu3O6+δ,” Nature 498, 75-77 (2013). by LANL postdoctoral fellows. Each sub-task, and the postdoctoral fellow proposing it, will be evaluated by • Work to date has indicated that installation of a third a selection team that will judge the proposed work for U=E multiple bond may lead to highly reactive spe- scientific merit, relevance to LANL mission, and quality cies, necessitating more research into the groups of of the applicant, with precedence among equal-quality atoms used to occupy the remainder of the metal’s proposals given to areas important to LANL missions coordination sphere. identified above that are underrepresented. • A comprehensive study of the materials PuCoIn5 and PuIn3 has been completed and the relevant Benefit to National Security Missions manuscripts will be submitted to Journal of Physics: “Modern Challenges in Actinide Science” will support Condensed Matter within a couple of months. scientific discovery and innovation, environmental re- • Discovery of the direct plutonium-239 Nuclear mediation, nuclear waste management, nuclear energy, Magnetic Resonance (NMR) signal, published in 40

Page 41

the jounal Science: First observation of 239Pu NMR ing U(VI) tris(imido) complexes (imido = [R-N=]2–; in any material, reported by Yasuoka et al., Science, R = alkyl, aryl), for the purpose of discovering a 2012, 336 (6083): 901-904. The use of NMR to direct general route into these species. Protonolysis re- observe plutonium-239 in the solid state and perhaps actions of the U(VI) bis(imido) carbene species solution opens up a powerful new window to observe U(NAr)2(dppmS2)(t-Bu2bpy) (Ar = aryl groups; dppmS2 the chemical and electronic behavior of plutonium for = 1,1-bis(thiodiphenylphosphino)methane dianion; many applications. t-Bu2bpy = 4,4’-di-tert-butyl-2,2’-bipyridyl) with acidic amines liberates free H2dppmS2, but we have been • Our photoluminescent measurements on Cs2NpO- unable to positively identify the uranium-containing 2Cl4 doped into a Cs2UO2Cl4 matrix (referred to as product from this reaction. Attempts at hydrogen- Cs2U(Np)O2Cl4) revealed energy transfer between the atom abstraction reactions from low-valent uranium Cs2UO2Cl4 host matrix and the Cs2NpO2Cl4 analyte. amides have resulted in either electron transfer from The energy transfer results in a large increase in the a U(III) pentakis-arylamide or unidentified products near-IR luminance intensity from Cs2NpO2Cl4. To in the cases of U(IV) imido amides. Further work will study this effect in a quantitative way we produced a continue in this area to generate suitable starting series of Cs2U(Np)O2Cl4 crystals containing varying materials for investigating hydrogen-atom abstraction amounts of Cs2NpO2Cl4. This allowed us to gain an chemistry. We have also investigated the formation of understanding of the mechanisms which are involved bis(imido)-oxo or dioxo-imido U(VI) complexes using in the transfer of energy between actinide-containing nitrite salts as O1– sources, but the only products iso- molecules and to develop a model for the energy lated from these reactions are uranyl (UO22+) deriva- transfer in this system. We have recently produced tives, suggesting that undesired chemistry is occurring the analogous plutonium-containing Cs2U(Pu)O2Cl4 with the imido group. system and are preparing to studying energy transfer in this system as well. • Development and programming of analytical methods needed to conduct project-related measurements • Synthesized a new type of carboxyl-functionalized pyrazolium ionic liquid that may have useful properties • Preliminary actinide measurements involving selected for dissolving and separating actinide metal ions and test materials. To date, such measurements have con- characterized the new ionic liquid with spectroscopic sistently shown the proposed sources to be superior methods. Studied the uranyl(VI) coordination chemis- to traditional approaches, with as much as a 2400% try with a carboxyl-functionalized phosphonium ionic enhancement in sample utilization being recorded at liquid using various spectroscopic methods. Three trace loading levels (e.g. picograms). papers were published on this work. • We have observed the first ever signal from ultracold Future Work neutron-generated fission in actinides. • Study the complexation of actinides in new ionic liq- • Published a paper describing the electronic structure uids with X-ray diffraction analysis and spectroscopic of several novel uranium compounds in the Journal of methods such as Raman, FT-IR, UV-vis-NIR and NMR. the American Chemical Society. • Continue focus on developing synthetic methods for • Developed capabilities to collect solution-phase sulfur generating small molecule analogs to UO3. Expand to K-edge X-ray absorption spectroscopic measurements. the study of imido complexes of uranium that could All measurements to date have been with solid materi- serve as synthetic precursors for molecular UO3 ana- als. logs. • Synthesized and analyzed a series of transition metal dithiophosphinate complexes using X-ray absorption • Submit manuscripts relevant to the study of uncon- spectroscopy to lay the foundation for analyzing f- ventional superconductivity and exotic magnetism in element dithiophosphinate complexes. a series of Pu-based heavy-fermion superconductors. Complete the study of at least four compounds, spe- • Synthesized a novel dithiophosphinate extractant that cifically PuCl6, PuCoGa5, PuSb2, and δ-Pu. we anticipate will provide better separation factors for separation trivalent actinides from lanthanides. • Develop procedure for manufacture of thin foils of This could be useful for the recycle of actinides in an Pu-239 and other actinides, commission a detection advanced nuclear energy system. system for materials sputtered from the actinide foils, • We have investigated multiple routes to generat- & perform first characterizations using the ultracold 41

Page 42

neutron source. materials. 2012. MRS Proceedings. 1444 (2012): 123. • Perform a series of measurements in which the slow Beaux, M. F., T. Durakiewicz, L. Moreschini, M. Grioni, dynamics of defects in Pu metal is studied via evolution F. Offi, G. Monaco, G. Panaccione, J. J. Joyce, E. D. of elastic moduli of a crystal. Bauer, J. L. Sarrao, M. T. Butterfield, and E. Guziewicz. Electronic structure of single crystal UPd3, UGe2, • Continue photoluminescent measurements on Cs2N- and USb2 from hard X-ray and angle-resolved pO2Cl4 doped into a Cs2UO2Cl4 matrix. Synthesize photoelectron spectroscopy. 2011. Journal of Electron crystals of Cs2PuO2Cl4 doped into a Cs2UO2Cl4 ma- Spectroscopy and Related Phenomena. 184 (8-10): trix. Perform theoretical calculations on Cs2PuO2Cl4. 517. Bejger, C., Y. H. Tian, B. J. Barker, K. S. Boland, B. L. Scott, • Evaluate impact of the spectroscopic overlap of stron- E. R. Batista, S. A. Kozimor, and J. L. Sessler. Synthesis tium Kα line on plutonium quantification. Investigate and characterization of a tetrathiafulvalene-salphen potential approaches to compensate for the strontium actinide complex. 2013. Dalton Transactions. 42: 6716. interference. Conduct laboratory validation, including assessment of precision, accuracy, & limit of detection Boland, K., S. Conradson, A. Costello, A. Gaunt, S. for use of monochromatic wavelength dispersive XRF Kozimor, I. May, S. Reilly, and D. Schnaars. Stabilising (MWDXRF) & micro XRF (MXRF). pentavalent actinides-visible-near infrared and X-ray absorption spectroscopic studies of the utility of the • Expand efforts to enhance the detection of plutonium (Np3W4O15)(H2O)(3)(MW9O33)(3)(M = Sb, Bi) structural type. 2012. DALTON TRANSACTIONS. 41 (7): & americium in thermal ionization mass spectrometry 2003. measurements, focusing on construction & testing of 2nd generation sources to characterize increased Chen, X. Y., G. S. Goff, B. L. Scott, and W. Runde. sample detection abilities/needs for various materials. Comparison of structural Variations of Ln(III) Exploit improved sample utilization to address current compounds with (pyrazol-1-yl)acetic acid. To appear in challenges in plutonium chronometry. Polyhedron. • Synthesize f-element complexes with dithiophosphi- Chen, X., G. S. Goff, M. Quiroz-Guzman, D. P. Fagnant, Jr., J. F. Brennecke, B. L. Scott, and W. Runde. Directed nate ligands. Analyze samples from f-metal-ion reac- nucleation of monomeric and dimeric uranium(VI) tions with dithiophosphinates by EXAFS. Complete the complexes with a room termperature carboxyl- synthesis of a new family of extractants that we antici- functionaled phosphonium ionic liquid. 2012. Chemical pate will provide better actinide separation factors. Communications. 49 (19): 1903. Conclusion Chen, X., G. S. Goff, W. C. Ewing, B. L. Scott, and W. Runde. Solid-state and solution-state coordination It is expected that between 12 and 16 individual research chemistry of Lanthanide(III) complexes with alpha- tasks will be active on average during the period of the hydroxyisobutyric acid. 2013. Inorganic Chemistry. 52 project. Each is expected to produce peer-reviewed publi- (6): 3217. cations and an internal LANL seminar each year for which the sub-task is active. It is intended that these sub-tasks Chen, X., G. S. Goff, W. C. Ewing, B. L. Scott, and W. cover all areas of actinide science that are relevant to LANL Runde. Solid-state and solution-state coordination missions. It is important to note that not all areas will be chemistry of Lanthanide(III) complexes with alpha- covered at any one time, and that precedence for under- hydroxyisobutyric acid. 2012. Inorganic Chemistry. 51: represented areas will be given. 13254. Coleman, M. E., E. M. Bond, W. A. Moody, and L. Publications Tandon. The analysis of uranium-232: comparison Ashley, K. R., S. M. Bowen, J. W. Oldham, Jr., A. C. Olson, S. of radiochemical techniques and an improved A. Kozimor, and A. R. Schake. Quantifying abundances method by alpha spectrometry. 2013. JOURNAL OF of long-lived 155Eu and 151Sm fission products RADIOANALYTICAL AND NUCLEAR CHEMISTRY. 296 (1): using gamma- and liquid scintillation counting 483. spectroscopies. Analytical Chemistry. Dakovski, G., Y. Li, S. Gilbertson, G. Rodriguez, A. Balatsky, Beaux, M. F., J. J. Joyce, T. Durakiewicz, K. S. Graham, E. D. J. Zhu, K. Gofryk, E. Bauer, P. Tobash, A. Taylor, J. Bauer, J. N. Mitchell, P. H. Tobash, and S. Richmond. Sarrao, P. Oppeneer, P. Riseborough, J. Mydosh, and Electronic structure, localization, and 5f occupany in Pu T. Durakiewicz. Anomalous femtosecond quasiparticle 42

Page 43

dynamics of hidden order state in URu2Si2. 2011. 2013. INTERNATIONAL JOURNAL OF PLASTICITY. 49: PHYSICAL REVIEW B. 84 (16): 161103. 185. Daly, S., J. Keith, E. Batista, K. Boland, D. Clark, S. Kozimor, Knezevic, M., I. J. Beyerlein, T. Nizolek, N. A. Mara, and T. and R. Martin. Sulfur K-edge X-ray absorption M. Pollock. Anomalous basal slip activity in zirconium spectroscopy and time-dependent density under high-strain deformation. 2013. Materials functional theory of dithiophosphinate extractants: Research Letters. 1 (3): 133. Minor actinide selectivity and electronic structure correlations. 2012. Journal of the American Chemical Knezevic, M., J. S. Carpenter, and R. J. McCabe. Society. 134 (35): 14408. Deformation behavior of the cobalt-based superalloy Haynes 25: Experimental characterization and crystal Daly, S., J. Keith, E. Batista, K. Boland, S. Kozimor, R. Martin, plasticity modeling. To appear in Acta Materialia. and B. Scott. Probing NiS2PR2 Electronic Structure to Generate Insight Relevant to Minor Actinide Knezevic, M., R. A. Lebensohn, O. Cazacu, B. Revil-Baudard, Extraction Chemistry. 2012. INORGANIC CHEMISTRY. G. Proust, S. Vogel, and M. E. Nixon. Modeling bending 51 (14): 7551. of α-titanium with embedded crystal plasticity and analytical yield surface formulations in implicit finite Daly, S., J. Klaehn, K. Boland, S. Kozimor, M. MacInnes, elements. 2013. Materials Science and Engineering A. D. Peterman, and B. Scott. NMR spectroscopy and 564 (1 March): 116. structural characterization of dithiophosphinate ligands relevant to minor actinide extraction processes. Knezevic, M., R. J. McCabe, C. N. Tome, R. A. Lebensohn, 2012. Dalton Transactions. 41 (7): 2163. S. R. Chen, C. M. Cady, G. T. Gray, and B. Mihaila. Modeling mechanical response and texture evolution Gianetti, T. L., G. Nocton, S. G. Minasian, N. C. Tomson, A. L. of alpha-uranium as a function of strain rate and D. Kilcoyne, S. A. Kozimor, D. K. Shuh, T. Tyliszczak, R. G. temperature using polycrystal plasticity. 2013. Bergman, and J. Arnold. Diniobium inverted sandwich INTERNATIONAL JOURNAL OF PLASTICITY. 43: 70. complexes with μ-η6:η6-arene ligands: synthesis, kinetics of formation, and electronic structure. 2012. Knezevic, M., R. J. McCabe, R. A. Lebensohn, C. N. Tome, Journal of the American Chemical Society. 135 (8): C. Liu, M. L. Lovato, and B. Mihaila. Integration of self- 3224. consistent polycrystal plasticity with dislocation density based hardening laws within an implicit finite element He, H. M., D. A. Andersson, D. D. Allred, and K. D. Rector. framework: Application to low-symmetry metals. Determination of the Insulation Gap of Uranium 2013. JOURNAL OF THE MECHANICS AND PHYSICS OF Oxides by Spectroscopic Ellipsometry and Density SOLIDS. 61 (10): 2034. Functional Theory. 2013. JOURNAL OF PHYSICAL CHEMISTRY C. 117 (32): 16540. Knezevic, M., R. J. McCabe, R. A. Lebensohn, C. N. Tome, and B. Mihailia. Modeling mechanical response and He, H., P. Wang, D. D. Allred, M. P. Wilkerson, J. Majewski, texture evolution of α-uranium as a function of strain and K. D. Rector. Characterization of chemical rate and temperature using polycrystal plasticity. 2013. speciation in ultra thin uranium oxide films. 2012. International Journal of Plasticity. 43 (April): 70. Analytical Chemistry. 84 (23): 10380. Koutroulakis, G., H. Chudo, H. Yasuoka, P. H. Tobash, H. N. Jiang, Yu, C. H. Booth, P. H. Tobash, K. Gofryk, M. A. Torrez, Mitchell, E. D. Bauer, and J. D. Thompson. Microscopy F. Ronning, E. D. Bauer, and J. D. Thompson. Magnetic properties of the heavy-fermion superconductor Frustration Effects in Uranium Intermetallics. 2011. PuCoIn5: An NQR study. New Journal of Physics. In International Conference on Strongly Correlated Electron Systems (SCES 2010) ; 27 June-2 July 2010 ; Koutroulakis, G., H. Chudo, H. Yasuoka, P. H. Tobash, J. N. Santa Fe, NM, USA. Vol. 273, p. 012036 (4 pp.). Mitchell, E. D. Bauer, and J. D. Thompson. Itinerant weak antiferromagnetism in PuIn3 explored by 115In Jilek, R., L. Spencer, R. Lewis, B. Scott, T. Hayton, and J. NQR. Journal of Physics - Condensed Matter. Boncella. A Direct Route to Bis(imido)uranium(V) Halides via Metathesis of Uranium Tetrachloride. 2012. McCleskey, T. M., E. Bauer, Q. X. Jia, A. K. Burrell, B. L. Scott, Journal of the American Chemical Society. 134 (24): S. D. Conradson, A. Mueller, L. Roy, X. D. Wen, G. E. 9876. Scuseria, and R. L. Martin. Optical band gap of NpO2 and PuO2 from optical absorbance of epitaxial films. Knezevic, M., I. J. Beyerlein, D. W. Brown, T. A. Sisneros, 2013. JOURNAL OF APPLIED PHYSICS. 113 (1): -. and C. N. Tome. A polycrystal plasticity model for predicting mechanical response and texture evolution Minasian, S. G., J. M. Keith, E. R. Batista, D. L. Clark, S. A. during strain-path changes: Application to beryllium. Kozimor, R. L. Martin, and D. K. Shuh. The importance 43

Page 44

of ϕ-orbital mixing in thorocene and uranocene Sarrao, J. D. Thompson, and L. H. Greene. Observation determined from carbon K-edge X-Ray absorption of the Hybridization Gap and Fano Resonance in the spectroscopy and time-dependent density functional Kondo Lattice URu₂Si₂. 2012. theory. Agnewandte Chemie. Physical Review Letters. 108 (24): 246403 (5 pp.). Minasian, S. G., J. M. Keith, E. R. Batista, K. S. Boland, J. Rim, K. R., D. Eom, S. W. Chan, M. Flytzani- A. Bradley, S. R. Daly, W. W. Lukens, J. Nordlund, G. T. Stephanopoulow, G. W. Flynn, and E. R. Batista. Seidler, D. Sokaras, G. L. Wagner, T. C. Weng, P. Yang, S. Scanning tunneling microscopy and theoretical study A. Kozimor, R. L. Martin, D. K. Shuh, and T. Tyliszczak. of water adsorption on Fe3O4: Implications for Orbital mixing in metal-oxygen multiple bonds catalysis. 2012. Journal of the American Chemical evaluatied using oxygen K-edge spectroscopy and DFT. Society. 134 (46): 18979. Nature Chemistry. Romanchuk, A. Y., S. N. Kalmykov, A. V. Egorov, Y. Minasian, S., J. Keith, E. Batista, K. Boland, D. Clark, S. V. Zubavichus, A. A. Shiryaev, O. N. Batuk, S. D. Conradson, S. Kozimor, R. Martin, D. Schwarz, D. Shuh, Conradson, D. A. Pankratov, and I. A. Presnyakov. G. Wagner, M. Wilkerson, L. Wolfsberg, and P. Yang. Formation of crystalline PuO2+xnH2O nanoparticles Determining relative f and d orbital contributions to upon sorption of Pu(V,VI) onto hematite. 2013. M-Cl covalency in MCl₆^2- Geochimica et Cosmochimica Acta. 121: 29. (M = Ti, Zr, Hf, U) and UOCl₅ <sup>-</ sup> using Cl K-edge X-ray absorption spectroscopy Shekhter, A., B. J. Ramshaw, R. Liang, W. N. Hardy, D. A. and time-dependent density functional theory. 2012. Bonn, F. F. Balakirev, R. D. McDonald, J. B. Betts, S. C. Journal of the American Chemical Society. 134 (12): RIggs, and A. Migliori. Bounding the pseudogap with 5586. a line of phase transitions in YBa2Cu3O6+δ . 2013. Nature. 498: 75. Minasian, S., K. Boland, R. Feller, A. Gaunt, S. Kozimor, I. May, S. Reilly, B. Scott, and D. Shuh. Synthesis and Spencer, L. P., S. G. Minasian, R. E. Jilek, E. R. Batista, K. Structure of (Ph4P)(2)MCl6 (M = Ti, Zr, Hf, Th, U, Np, S. Boland, J. M. Boncella, S. D. Conradson, D. L. Clark, Pu). 2012. INORGANIC CHEMISTRY. 51 (10): 5728. R. L. Gdula, T. W. Hayton, S. A. Kozimor, R. L. Martin, M. A. McInnes, A. C. Olson, B. L. Scott, D. K. Shuh, Monreal, M. J., R. K. Thomson, T. Cantat, N. E. Travia, B. L. and P. Yang. Tetrahalide complexes of the [U(NR)2]2+ Scott, and J. L. Kiplinger. UI4(1,4-dioxane)(2), UCl4(1,4- ion: synthesis, theory, and Cl K-edge X-ray absorption dioxane), and UI3(1,4-dioxane)(1.5): stable and spectroscopy. 2013. Journal of the American Chemical versatile starting materials for low- and high-valent Society. 135 (6): 2279. uranium chemistry . 2012. Organometallics. 30 (7): 2031. Stanley, F. E., K. J. Spencer, D. S. Schwartz, M. G. Watrous, and J. E. Delmore. Investigating enhanced thorium Monreal, M. J., R. K. Thomson, T. Cantat, N. E. Travia, B. L. ionization in TIMS using Re/Pt porous ion emitters. Scott, and J. L. Kiplinger. UI4(1,4-dioxane)(2), UCl4(1,4- Journal of Radioanalytical and Nuclear Chemstry. dioxane), and UI3(1,4-dioxane)(1.5): Stable and Versatile Starting Materials for Low- and High-Valent Teng, B. T., F. M. Wu, W. X. Huang, X. D. Wen, L. H. Zhao, Uranium Chemistry. 2011. ORGANOMETALLICS. 30 (7): and J. F. Luo. A DFT study of the structures of aux 2031. clusters on a CeO2(III) surface. 2012. A European Journal of Chemical Physics and Physical Chemistry. 13 Nerikar, P. V., L. A. Casillas Trujillo, D. A. Andersson, C. Unal, (5): 1261. B. P. Uberuaga, C. R. Stanek, D. C. Parfitt, R. W. Grimes, and S. B. Sinnott. Segregation of xenon to dislocations Thomson, R., C. Graves, B. Scott, and J. Kiplinger. and grain boundaries in uranium dioxide. 2011. Journal Straightforward and efficient oxidation of Name: Physical Review. B, Condensed Matter and tris(aryloxide) and tris(amide) uranium(III) complexes Materials Physics. 84 (17): Medium: X; Size: page(s) using copper(I) halide reagents. 2011. INORGANIC 17410. CHEMISTRY COMMUNICATIONS. 14 (11): 1742. Olson, A. C., J. M. Keith, E. M. Batista, K. S. Boland, S. R. Thomson, R., M. Monreal, J. Masuda, B. Scott, and J. Daly, S. A. Kozimor, M. M. MacInnes, R. L. Martin, and Kiplinger. Lutetium gets a crown: Synthesis, structure B. L. Scott. Using solution- and solid-state S K-edge and reaction chemistry of the separated ion pair X-ray absorption spectroscopy with Density Functional complex, [Li(12-crown-4)₂][(C<sub>5</ Theory to evaluate M-S bonding for MS42. (M = Cr, sub>Me ₅)₂LuMe<sub>2</ Mo, W) dianions. Inorganic Chemistry. sub>]. 2011. Journal of Organometallic Chemistry. 696 (25): 3966. Park, W. K., P. H. Tobash, F. Ronning, E. D. Bauer, J. L. 44

Page 45

Tobash, P., F. Ronning, J. D. Thompson, B. L. Scott, P. J. CONDENSED MATTER. 25 (2): 025501. Moll, B. Batlogg, and E. D. Bauer. Single crystal study of the heavy-fermion antiferromagnet CePt <sub>2</ Wen, X. D., R. L. Martin, L. E. Roy, G. E. Scuseria, S. P. Rudin, sub>In₇. 2012. Journal of Physics E. R. Batista, T. M. McCleskey, B. L. Scott, E. Bauer, J. J. Condensed Matter. 24 (1): 015601. Joyce, and T. Durakiewicz. Effect of spin-orbit coupling on actinide dioxides, AnO2 (An=Th, Pa, U, Np, Pu and Travia, N. E., M. J. Monreal, B. L. Scott, and J. L. Kiplinger. Am): A screened hybrid density functional study. 2012. Thorium-mediated ring-opening of tetrahydrofuran Journal of Chemical Physics. 137 (15): 154707. and the development of a new starting material for thorium chemistry: preparation, crystal structure, and Wen, X. D., R. L. Martin, T. M. Henderson, and G. E. chemistry of ThI4(DME)4. 2012. Dalton Transactions. Scuseria. Density functional theory studies of the 2012 (41): 14514. electronic structure of solid state actinide oxides. 2013. Chemical Review. 113 (2): 1063. Ueland, B. G., C. F. Miclea, Y. Kato, O. Ayala-Valenzuela, R. D. McDonald, R. Okazaki, P. H. Tobash, M. A. Torrez, F. Wen, X. D., S. P. Rudin, E. R. Batista, D. L. Clark, G. E. Ronning, R. Movshovich, Z. Fisk, E. D. Bauer, I. Martin, Scuseria, and R. L. Martin. Rotational rehybridization and J. D. Thompson. Controllable chirality-induced and the high temperature phase of UC2. 2012. geometrical Hall effect in a frustrated highly-correlated Inorganic Chemistry. 51 (23): 12650. metal. 2012. Nature Communications. 3: 1067. Wen, X., T. Yang, R. Hoffmann, N. W. Ashcroft, R. Martin, Vasudevan, K. V., B. L. Scott, R. Marti-Arbona, and J. C. S. Rudin, and J. Zhu. Graphane nanotubes. 2012. ACS Gordon. Decamethylsamarocene-mediated reductive Nano. 6 (8): 7142. coupling of the (1R,2R)-N,N’-Bis(2- pyridylmethylene) Yang, Y., J. Saiers, N. Xu, S. Minasian, T. Tyliszczak, S. cyclohexane-1,2-diamine ligand . European Journal of Kozimor, D. Shuh, and M. Barnett. Impact of natural Inorganic Chemistry. organic matter on uranium transport through Vasudevan, K. V., N. C. Smythe, B. L. Scott, and J. C. Gordon. saturated geologic materials: From molecular to Metallopolymer formation using the (1R,2R)-N,N’- column scale. 2012. Environmental Science and bis(2-pyridylmethylene)- cyclohexane-1,2-diamine Technology. 46 (11): 5931. (BPID) ligand class. 2012. Dalton Transactions. 42 (14): Yasuoka, H., G. Koutroulakis, H. Chudo, S. Richmond, D. K. 4768. Veirs, A. I. Smith, E. D. Bauer, E. D. Thompson, G. D. Vasudevan, K., B. Scott, and J. Gordon. Main-group Jarvinen, and D. L. Clarks. Observation of <sup>239</ element compounds derived from the (1R,2R)-N,N sup>Pu Nuclear Magnetic Resonance. 2012. Science. ‚Äô-bis(2-pyridylmethylene)cyclohexane-1,2-diamine 336 (6083): 901. (BPID) ligand. 2012. MAIN GROUP CHEMISTRY. 11 (1, Zhao, Y., B. T. Teng, X. D. Wen, L. H. Zhao, and M. F. Luo. SI): 45. A theoretical evaluation and comparison of MxCe1− Vasudevan, K., N. Smith, B. Scott, B. Bennett, R. xO2−δ(M = Au, Pd, Pt, and Rh) catalysts. 2012. Muenchausen, and J. Gordon. Ionic liquid mediated Catalysis Communications. 27: 63. routes to polydentate oxygen-donor adducts of Zhao, Y., B. T. Teng, X. D. Wen, L. H. Zhao, and M. F. Luo. cerium(iii) bromide. 2012. Dalton Transactions. 41 (7): Superoxide and peroxide species on CeO2(111), 1924. and their oxidation roles. 2012. Journal of Physical Wang, S. W., M. Tang, L. L. Zhang, G. L. Xiao, K. S. Brinkman, Chemistry. 116: 15986. and F. L. Chen. Irradiation effect on the structure change for Sr2Fe1.5Mo0.5O6-delta perovskite ceramic. 2013. JOURNAL OF ALLOYS AND COMPOUNDS. 578: 170. Wen, X. D., R. Hoffmann, and N. W. Ashcroft. Two- dimensional CdSe nanosheets and their interaction with stabilizing ligands. 2012. Advanced Materials. 25 (2): 261. Wen, X. D., R. L. Martin, G. E. Scuseria, S. P. Rudin, E. R. Batista, and A. K. Burrell. Screened hybrid and DFT plus U studies of the structural, electronic, and optical properties of U3O8. 2013. JOURNAL OF PHYSICS- 45

Page 46

Chemistry and Material Sciences Directed Research Continuing Project Accelerating Materials Certification Through Co-design Jack S. Shlachter 20120707DR Introduction velop a paradigm for accelerated materials certification The goal of this project is to make progress towards the for fusion energy. grand challenge goal of predictive process-aware materi- als performance through a focus on “co-design.” Co-de- Progress sign traditionally refers to the simultaneous optimization On the major experimental/modeling effort under this of hardware and software in computer science; here, project, the Dynamic Materials Performance effort, we we broaden the definition of co-design to include the have significant findings from polycrystal experiments integrated coupling of multi-scale theory, computation/ and single crystal simulations on how phase transforma- information science & technology techniques (includ- tion is preceded by deformation modes and the depen- ing data visualization), and experiment. We propose a dence on shock direction. Through this work we have series of four coupled thrusts to achieve our goal. We identified a need for single crystal experiment measure- will i) build on past successes to pursue the co-design of ments. For the balance of this year, we will focus on micro-mechanical models to improve predictive capabil- single crystal measurements + estimation of barrier ity in regimes of high-pressures and strain rates with a energies + preliminary results from our microstructural focus on Zr and Ti, ii) develop, in parallel with thrust i, a model. co-design framework for multi-scale models focused on radiation damage evolution, in particular creep-related Our Linac Coherent Light Source (LCLS) jumpstart activi- phenomena, iii) utilize ion beam irradiation as a means ties, crucial to demonstrating use of an x-ray free elec- of accelerating materials damage, including providing tron laser (XFEL), the experiment has been completed, key data for thrust ii, and iv) explore the foundations and the data analysis will be continued until it is ready of coherent x-ray diffractive imaging in order to trans- for publication. In late breaking news – the proposal to late physical observables into model inputs in support use the LCLS for Run 8 was ranked in the top group and of thrusts i and ii. In this thrust structure, the project is expected to receive beamtime. leadership’s principal responsibility is fostering integra- tion and collaboration among these thrusts. This man- Measurement of temperature in extreme environments agement plan will yield synergies not possible through a is a major R&D activity for developing advanced certifi- series of discrete projects. cation of materials. In the first of our three thermom- etry activities, the spontaneous Raman approach, we are Benefit to National Security Missions currently performing continuous wave measurements to The need to predict and control materials in extremes 900 K accompanied by writing automated data acquisi- spans the full range of DOE missions from weapons per- tion and analysis software. Pulsed measurements on formance to energy security. Further, successful nuclear statically heated samples will be performed over the weapons program approaches to certification of materi- next few months. We have constructed a cell capable of als performance (and the associated quantification of testing static samples from liquid nitrogen to >900 K. We performance margins and uncertainty) have not e.g., have assembled a Stokes/anti-Stokes Raman Detection been translated to nuclear energy materials certifica- system, written LabView acquisition software, and tested tion challenges. In this project, we will address specific several algorithms for fluorescent background subtrac- process-aware materials performance challenges in our tion and real time (few Hz) conversion to temperature. enduring nuclear weapons stockpile and work to de- This is being tested with CW lasers currently, but is 46

Page 47

mobile for transition to the microsecond and nanosecond ii, and iv) explore the foundations of coherent x-ray diffrac- laser laboratories. We have measured Raman spectra/tem- tive imaging in order to translate physical observables into peratures of quartz to 900 K. We have performed mainte- model inputs in support of thrusts i and ii. nance to restore operation to the microsecond laser, which is now performing normally. Conclusion The need to predict and control materials in extremes Calculations relating experimental observations to theo- spans the full range of DOE missions from weapons per- retical predictions of parametric effects have begun. The formance to energy security. Further, successful nuclear static temperature test cell, spectrometers, detectors, and weapons program approaches to certification of materials lasers were all assembled from available equipment for performance (and the associated quantification of perfor- cost savings. mance margins and uncertainty) have not e.g., been trans- lated to nuclear energy materials certification challenges. A second approach to measuring temperatures involves Through this project, we will develop theoretical models analyzing subtleties in the diffraction signals. This work and identify physical observables that can be measured by is titled Separation of Deformation, Particle Shape and experiment to test these models that when taken together Temperature Factors from Nuclear Motion Data. Initial will accelerate the rate at which we can design and discov- analysis on Cu and Ta samples has shifted to Zr simula- er new materials to meet these certification challenges. tions in the Dynamic Materials Performance effort though the phase transformation is complicating our analysis (a very large number of defects structures are generated that excessively blur the peaks). The transfer of initial amounts of this data has recently occurred and we have begun our SOMERIF structural analysis on this. One staff member had previously written a more sophisticated diffraction calcula- tion program that integrates over many time steps and can include particle shape factors and realistic detector scenar- ios. This is very well suited for our application. The initial temperature analysis will likely revert back to Cu samples for the proof of principle. We are now agreed on a path forward and will be ramping up the efforts. On the radiation front, our LANL/TAMU collaboration continues its activities in the Ion Beam Materials Labora- tory (IBML). The corrosion chamber design and off-site test is being done at the TAMU in collaboration with INL and LANL. The chamber will be installed in the IBML in July/August and the irradiation experiment is scheduled in the Aug./Sept. time frame. Finally, our UCSD collaboration is developing. We estimate that it will take some time to design, install, and conduct first wall materials research on this unique and novel capability. Future Work Work in FY14 will build on and extend scope initiated in FY12 and continued in FY13. In particular, we will i) pursue the co-design of micro-mechanical models to improve pre- dictive capability in regimes of high-pressures and strain rates with a focus on Zr and Ti, ii) develop a co-design framework for multi-scale models focused on radiation damage evolution, in particular creep-related phenomena, iii) utilize ion beam irradiation as a means of accelerating materials damage, including providing key data for thrust 47

Page 48

Chemistry and Material Sciences Directed Research Continuing Project Fighting Carbon with Carbon: All-Carbon Nanomaterial Photovoltaics Stephen K. Doorn 20130026DR Introduction photovoltaic devices. iii) Pursue these objectives within Our pioneering development of graphene oxide and the context of the interfacial interactions dictated by two compositionally-defined bundles of carbon nanotubes primary device architectures we view as the most prom- as new optical materials has revealed new photophysical ising for realizing all-CNM photovoltaics. These include a behaviors that indicate these carbon nanomaterials pos- layered thin-film platform and a dye-sensitized solar cell sess many properties of ideal light harvesters for thin- configuration, where our CNM become the active dye. film photovoltaic applications. These benefits include more efficient light absorption than molecular chromo- Benefit to National Security Missions phores, tunability of optical and electrochemical prop- Our effort introduces a promising new approach that will erties to match the solar emission spectrum and allow enhance LANL’s portfolio in renewable energy research, optimization of interfacial charge flow between other which must continue to grow and strengthen to meet device materials, and extremely efficient photogenera- the critical energy challenges facing the nation. Success tion of long-lived free charges available for collection will contribute to reducing climate disruption threats as electrical power. Our findings create a compelling while ensuring our nation’s future energy security and opportunity to create new photovoltaic device types. thus national security. Our approach does not rely on These will be designed to specifically exploit the poten- rare, isolated, or politically inaccessible resources. We tial our newly-introduced materials hold for overcoming thus will be pioneering new earth-abundant materials current bottlenecks in solar-energy harvesting perfor- to solve energy problems, while developing/controlling mance imposed by the limitations of currently used new multifunctional semiconductor and photonics mate- light-harvesting chromophores. By project end we will rials. Beyond energy-harvesting applications, advances harness these properties within functioning phtovoltaic from our effort will also impact threat reduction needs device prototypes based on two complementary plat- through detector and sensors development and photon- forms suitable for follow-on development. In establish- ics applications, all of which rely on similar photophysical ing these prototype platforms, our primary goals will be processes as will be studied here. to: Our discovery position in the emerging areas of gra-

1. Demonstrate for the first time the benefits of carbon phene oxide and defined-composition nanotube bundles nanomaterials (CNM) as active light harvesting com- as optical nanomaterials also places us at the forefront ponents. of such areas as probing their photophysical properties,
2. Demonstrate that CNM can be used to overcome key understanding and controlling energy and charge flow performance limitations of solar cells. in these systems, and determining how these dynamical processes are defined by the interfacial interactions of
3. Demonstrate the potential of integrated CNM-based multiple materials. With relevant behaviors and inter- photovoltaic devices. actions arising and playing out across nm to um length scales and dynamical processes occurring from fs to ms To meet these goals we will: i) Establish the fundamen- timescales, these questions go to the heart of under- tal photophysical and electrochemical behaviors of CNM standing functional behaviors that are defined at the chromophores and the means to control and tune the mesoscale. Thus, our studies also address DOE Office associated energy and charge flows. ii) Establish the of Science needs for improved basic understanding of principles for CNM materials integration into functioning 48

Page 49

materials and fundamental chemistry for materials control. to correlate PL spectra directly to chemical species. An especially exciting result is finding GO photoconductivity Progress (PC) in devices: a prerequisite to PV function. Correlated spatial mapping of photoconductivity surprisingly reveals Carbon Nanotube (CNT) PV Architecture response is located away from collection electrodes, Several devices have been prepared from CNT enriched in showing an internal potential is developed. Paired with (6,5) structures generated via density gradient purification. electrostatic force microscopy (EFM) measurements, we While the sample is nearly pure (6,5), sufficient metal con- correlate PC response with topological features and see tamination in the material has been found to effectively that topological defects create built-in potentials that drive short-out our PV devices through production of metallic photogenerated charge carriers to collection electrodes networked percolation pathways. The devices, however, (manuscript in preparation). The results provide a strategy have proved valuable in indicating film thickness require- for device architectures based on interleaved GO and re- ments and demonstrating the need for enhanced purifica- duced-GO regions we can generate via laser heating-based tion. New separations approaches have been hit hard and lithographic patterning. GO thermodynamics relevant to include gel chromatography, which shows good metals activity of functional groups and potential O-migration are removal, but yields short tubes. Polymer-based separa- being explored theoretically. The results also tie to oxygen tions also show good metal removal and selectivity, but doping of CNTs as model systems for charge trapping in give low yields and tricky processing. One highlight is the devices. polymer MehPPV, which shows interesting potential for light collection sensitization at wavelengths not absorbed Dye Sensitized Solar Cell (DSSC) Architectures by our CNTs. Our newest approach is based on two-phase A primary challenge in CNM-based DSSC devices arises extraction and is already giving pure (6,5) and significant from the optimized TiO2 films having pore sizes of ~20 metal reduction in one separation step (manuscript in nm, constraining how we integrate CNMs for optimizing preparation). Samples of enriched material from each ap- interactions for charge transfer reactions. We focus on proach are now being incorporated into second generation three routes. Co-sintering of long CNTs is being enabled devices. by optimizing CNT surface chemistry. Acid-based cutting is being used to reduce CNT lengths to the same scale as the Fundamental Exciton Behavior TiO2 pores and is looking to be a very promising approach. Understanding transport of optical excitations and free Cutting introduces carboxylate groups, ideal for binding to charges is essential for optimizing PV devices. We have TiO2. Furthermore, low-level surface passification is found revealed for the first time via our ability to directly image to turn on CNT PL, demonstrating their optical properties exciton diffusion that CNT exciton transport proceeds via are retained. We also are pursuing nanoscale graphene an incoherent hopping mechanism (published in Nano flakes as light-harvesting chromophores. Their synthesis Letters). We have also revealed the detailed nature of is a complex multi-step process. We have just finished syn- environmental dipole interactions that lead to exciton thesis of all precursors. Finally, we are pursuing a chemical dissociation (published in J. Am. Chem. Soc.). Related vapor deposition route for synthesis as well. theory is showing an enhanced free carrier generation in the presence of low-level dopants (published in J. Phys. New Capability Development Chem. Lett.) and points towards an approach to enhance All optical components are assembled for our super- free-carrier generation from photo-excitations. Direct resolution photocurrent imaging apparatus, which can imaging techniques are also revealing novel biexciton now be used for standard correlated photocurrent imag- behaviors. We are also probing optical states in single ing. Testing of the anti-phase beam shaping to give 50 nm chirality bundles. New Raman results show evidence for resolution is proceeding. Our transient absorbance system long-lived excited states oriented perpendicular to typi- is now assembled. We have installed a new state-of-the- cal CNT excitations that can enhance charge transport to art Raman triple monochromator to further explore CNT collection interfaces. LANL-unique nonadiabatic excited bundle interactions. state dynamics modeling is revealing unique localization of optical excitations in CNT analogues. Future Work Graphene Oxide (GO) Our continuing goals are organized by three interconnect- We are probing the nature of GO optical states and how ed areas: Materials Development and Characterization, they respond to changes in their chemical environment Development of Fabrication/Integration Procedures, and using photoluminescence (PL) imaging of single flakes. Studies of Exciton and Charge Transfer/Transport. Follow-on studies are targeting site-selective spectroscopy In the area of Materials Development and Characteriza- 49

Page 50

tion, we will target generation of nanotube and graphene Publications oxide materials and films with specific characteristics. For Arnold, M. S., J. L. Blackburn, J. J. Crochet, S. K. Doorn, J. G. graphene oxide the effort will include development of Duque, A. Mohite, and H. Telg. Recent developments methods for isolation of flakes with defined dimensions us- in the photophysics of single-walled carbon nanotubes ing density gradient approaches and defining film morphol- for their use as active and passive material elements in thin film photovoltaics. 2013. Physical Chemistry ogies from such isolated materials. In parallel, theoretical Chemical Physics. 15: 14896. and experimental efforts will be aimed at understanding and tuning graphene oxide optical and electrochemical be- Crochet, J. J., J. G. Duque, J. H. Werner, B. Lounis, L. Cognet, havior. For nanotubes, effort will be directed at improved and S. K. Doorn. Disorder limited exciton transport in separations for limiting metallic content in semiconducting colloidal single-wall carbon nanotubes. 2012. Nano films, and defining morphology and makeup of nanotube- Letters. 12: 5091. based films. In support of development of dye sensitized Duque, J. G., L. Oudjedi, J. J. Crochet, S. Tretiak, B. Lounis, cell architectures, we will continue development of solid S. K. Doorn, and L. Cognet. Mechanism of electrolyte- state electrolytes, mapping of relevant nanotube band induced brightening in single-wall carbon nanotubes. structures and probing charge transfer reactions at TiO2 2013. Journal of the American Chemical Society. 135: interfaces. 3379. For Development of Fabrication/Integration Procedures Sau, J. D., J. J. Crochet, S. K. Doorn, and M. L. Cohen. our separations methods will provide metal-free nanotube Multiparticle exciton ionization in shallow doped material for devices. For our film-based devices, we will carbon nanotubes. 2013. Physical Chemistry Letters. 4: then focus on defining film morphologies, optimizing initial 982. band offsets, controlling materials content, and devising Solenov, D., C. Junkermeier, R. L. Reinecke, and K. A. bulk heterojunction geometries for optimized materials Velizhanin. Tunable adsorbate-adsorbate interactions interactions. For dye sensitized cells, we will be optimizing on graphene. 2013. Physical Review Letters. 111: strategies for chemical coupling of nanotubes, graphene 115502. oxide, and nanographene flakes to nanostructured TiO2. For Exciton/Charge Transfer and Transport Studies, while test devices are progressing, we will in parallel focus on putting in place the proposed capabilities for capacitive photo-current imaging spectroscopies paired with super- resolution techniques. Once in place we will then proceed with characterization of charge transport behavior in the first-generation test devices. Paired with theory, we will use the results to help define directions for optimization towards second-generation devices. Conclusion By project’s end we expect to i) establish the fundamental photophysical behaviors of our new optical nanomateri- als for use as active light-harvesting materials and how to tune these behaviors, ii) establish the principles for carbon nanomaterials integration into functioning photovoltaic devices, and iii) establish the most promising device archi- tectures for follow-on development. The result will be a promising new approach to photovoltaic energy harvesting that overcomes many current limitations. The effort will also create new understanding of the optical behaviors of our new class of materials to advance other applications while also developing powerful new techniques for nano- materials characterization. 50

Page 51

Chemistry and Material Sciences Directed Research Continuing Project Design Principles for Materials with Magnetic Functionality Joe D. Thompson 20130052DR Introduction ous energy-efficient technologies, but their useful func- Advanced materials have been and will continue to be tion presently relies on rare and expensive elements. essential for the country to remain globally competi- We will discover scientific principles that will allow tive and to ensure our energy and defense security. designed control of materials that are strongly magnetic Magnetism underpins many energy-efficient technolo- and free of critical rare elements. The same techniques gies, notably wind turbines and ‘green’ automobiles, for rational materials design may eventually benefit the but magnetism is an intrinsically complex and delicate weapons program. collective-electron function that is reflected in the multiple electronic properties of materials that must Progress be satisfied simultaneously to produce useful function. In a closed loop of computation, theoretical model- Magnets presently used in these technologies rely on ing, materials synthesis, and experimental validation, nationally critical rare and expensive elements, and we have studied in detail a family of rare-earth-free despite decades of making-and-measuring to find new materials with compositions T23-xT’xZ6, where T is a magnetic materials that are free of those critical ele- 3d-element (Cr, Mn, Fe, Co and Ni), T’ is a heavier 4d- or ments, this approach has not been successful. Progress 5d-element and Z=B, C, or Al. Density functional elec- requires a strategy based on scientific principles that tronic structure and ordered magnetic moment have ultimately will allow the design of suitable replacement been calculated for 35 new variants of this family, and magnetic materials. With remarkable recent advances in based on these calculations, the most promising materi- quantum chemistry and in modeling the consequences als have been made and their magnetic properties have of complex electronic properties as well as our ability to been measured. Trends in magnetic properties predicted test predictions with new experimental techniques, the from the calculations are supported qualitatively, and discovery, understanding and application of principles often quantitatively, by experiments. Though many of for control of strong magnetic properties in metals rich these materials have large saturated magnetic moments in earth-abundant elements is within grasp. In a closed and high ferromagnetic ordering temperatures, which loop of state-of-the-art computation, theoretical model- are desirable characteristics, none of these materials has ing, materials synthesis, and in-depth measurements, a large coercive field. As we have shown from theoretical we will determine and test scientific principles that will modeling, the relatively strong hybridization between allow the designed control of magnetic properties in T and T’ electrons and the absence of strong crystalline advanced materials suitable for application in energy- electric fields in the cubic structure of these materials efficient technologies. cause the spin-orbit coupling, and hence coercive mag- netic field, to be small, which renders the magnetic fig- Benefit to National Security Missions ure of merit of this family of materials to be insufficient The computationally motivated, experimentally validat- for technological applications. This family will not be ed design of new materials with specified functionality pursued further, but other materials hold more promise has emerged as a major scientific challenge for materials as we have found from the over 100 density functional physics and chemistry, a challenge that must be met to calculations on approximately 50 different materials. satisfy ever increasing demands for advanced materials Motivated by these calculations, we have synthesized 65 that underpin the Nation’s energy security and competi- samples and measured the magnetic properties of the tiveness. Strongly magnetic materials are key to numer- 34 most promising materials. This study is on-going. 51

Page 52

In a parallel effort, we have focused on understanding conditions that favor strong ferromagnetic properties in the origin of a modestly large coercive magnetic field in new and known rare-earth-free materials that are identi- YCo5, which has been known for many years; however, the fied through calculations and simulations. Specifically, mechanism responsible for its coercive field is not known. we will explore theoretically and experimentally families From theoretical modeling and density functional calcula- of materials that form in hexagonal or tetragonal crystal tions on YCo5, we have three important predictions: (1) structures and that are rich in magnetic Co and Fe. We will hexagonal or tetragonal crystal symmetry favors a large predict their magnetic properties and validate those pre- orbital magnetic moment and hence coercive field; (2) the dictions by combinations of magnetization, magnetic force electronic configuration of Co also lends itself to a substan- microscopy, magnetic circular dicrosim and photoemission tial orbital moment; and, (3) a Coulomb interaction of 2eV experiments. In parallel, we will develop new computa- or greater is important for inducing a large orbital moment tional algorithms that more accurately predict the intrinsic on the Co electrons. Preliminary photoemission experi- spin-orbit coupling, and hence magnetic anisotropy energy, ments support the third prediction and planned x-ray circu- that must be optimized in materials to have a sufficiently lar dicroism experiments will test the second one. On the large magnetic figure of merit to be technologically useful. basis of these guiding principles, we have begun a study Results of these calculations will be tested experimentally. of two new families of materials that form in the Mo5SiB2 By iteratively closing compute, make, model and measure and Ti3Co5B2 structure types as well as used total energy loops, we expect to refine the initial set of design prin- and magnetic anisotropy energy calculations to predict ciples developed in the first year of this project for predict- that a new material ScCo5 might be very promising. Experi- ing strongly ferromagnetic properties of material devoid of ments are underway to test this prediction. In addition, rare-earth elements. an entirely new approach has begun that uses the local density approximation plus dynamical mean field theory Conclusion (LDA+DMFT) to calculate realistic Coulomb, crystal-field, State-of-the-art calculations of electronic structure and and spin-orbit interactions on multiple d-electron orbitals. magnetic anisotropy, coupled with the solution of theo- Such a calculation on multiple orbitals has never been at- retical models of complex electronic interactions, will be tempted but will be a powerful predictive tool of magnetic tested experimentally to establish scientific principles that functionality. allow the designed control of the crystal structure and chemical make-up of new materials with magnetic proper- Besides scientific progress, this project also has pursued ties favorable for strong magnetism in metallic compounds programmatic funding opportunities. In response to the that do not contain nationally critical elements. DOE call for a Critical Materials Hub proposal, we partici- pated actively in the Laboratory’s response to that call. Publications Though the proposal was among the few finalists, it was Ronning, F., and J. L. Sarrao. Material’s prediction scores a not selected for funding. In coordination with Laboratory hit. Physics Viewpoint. ARPA-E program managers, the concept of this project was discussed with two APRA-E program offices, the Rare- Earth Alternatives in Critical Technologies and the Vehicle Technology office. Both considered our concept attractive, but both also judged that it was premature to invest in it at this time. Finally, one team member has been requested to serve as a guest editor of the special issue of the Jour- nal of Physics: Condensed Matter that will be devoted to rare-earth replacement magnets. Another team member will contribute an invited review to this special issue that summarizes our scientific approach to developing design principles for the predictive control of magnetic functional- ity and the success that this approach has realized. Future Work We will apply our state-of-the-art capabilities of electronic structure calculations, many-electron modeling, materi- als synthesis and detailed experimental measurements to develop a predictive understanding of crystal-chemical 52

Page 53

Chemistry and Material Sciences Directed Research Continuing Project Non-Precious Metal Electrocatalysts for Clean Energy Piotr Zelenay 20130065DR Introduction LDRD Energy & Earth Systems Challenge, including the In order to fully realize the clean energy conversion and need to utilize earth-abundant elements in place of storage technologies (fuel cells, metal-air batteries, wa- precious metals. Ultimately, this proposed work will also ter electrolyzers for hydrogen generation), highly active, position LANL to target major new initiatives in clean- durable, and inexpensive non-precious metal electro- energy catalysis in the future. catalysts are needed for oxygen reduction reaction (ORR) and oxygen evolution reaction (OER). In this project, we Benefit to National Security Missions respond to that grand challenge in electrocatalysis by This research directly supports the Los Alamos LDRD combining experimental and multi-scale modeling ap- Energy & Earth Systems Grand Challenge “concepts and proaches for the purpose of (1) elucidating the nature of materials for clean energy”. In particular, it tackles the the ORR and OER active catalytic sites on the surface of top research priorities of “new materials, for energy non-precious metal catalysts, (2) probing the interplay applications, containing earth-abundant elements that between meso- and nanoscale controls on electrocata- mimic properties of rare and expensive materials” and lyst performance, and (3) designing and synthesizing “energy generation and efficiency”. This research will catalysts with the structure optimized for maximum enable sustainable and low-cost energy devices with oxygen reduction/evolution activity and required perfor- enhanced energy efficiency/capacities, which will posi- mance stability. tion LANL to target new initiatives in meso-scale cataly- sis, scheduled to be announced by DOE-BES in the near Active site characterization from density functional the- future. ory and model system experiments will provide insights into ORR & OER reaction mechanisms and the molecular In particular, this research is A-relevant to the DOE-EERE nature of the active site. Mesoscale/multiphysics studies mission of developing alternative energy sources, espe- using lattice Boltzmann method (LBM) will guide the syn- cially for transportation and stationary applications (fuel thesis towards materials with appropriate physicochemi- cells, hydrogen generation). By being directly relevant to cal properties. Ultimately, information generated from renewable energy this research earns the highest mark the theory and characterization of model systems will be in the energy security category. Finally, this effort fo- used to tune precursors for the design and synthesis of cuses on the fundamental understanding of materials for ultra-high performance catalysts. Advanced characteriza- two possibly most challenging reactions in electrochem- tion using nitric oxide (NO) probes, iron-specific nuclear istry, oxygen reduction and evolution, using a combina- resonance vibrational spectroscopy (NRVS), and other tion of modeling, experimentation with model systems, methods will be combined with theory to further tune and catalyst development. This makes this project very catalyst performance. Stabilization and promotion of highly relevant to both basic understanding of materials optimal active sites will be investigated by the employ- and fundamental chemistry. ment of carbon and non-carbon (i.e. oxide) template structures. Progress Model Systems Research in this project will advance the fundamentals To achieve the goal of rational synthesis of non-precious of oxygen electrocatalysis and develop novel materi- metal catalyst (NPMC) systems for oxygen reduction re- als and concepts for energy applications. The proposed action (ORR), we have focused our efforts on the synthe- effort directly tackles two top research priorities in the 53

Page 54

sis of model systems based on graphene (G) and graphene (i) graphene and (ii) carbon meso- and nanotubes. We oxide (GO). We know that N-doped GO shows higher ORR have established the synthesis-structure-activity correla- activity compared to pure GO. Therefore, we have devel- tions by varying the synthesis conditions: support, nitro- oped processes to make nitrogen-doped graphene oxide gen precursor, heating temperature, and transition metal (nGO), characterized by a variety of techniques including and its content. We have experimentally verified that by XPS, thermo gravimetric analysis, electron microscopy and controlling the level and type of nitrogen functionalities we electrochemical techniques. The data obtained has been can substantially improve the ORR activity and durability used to guide subsequent syntheses of the next-generation of nGO-rich structures. This generates much needed input nGOs. Our results indicate that the total amount of nitro- for the model system design and modeling efforts (above) gen, typically 5-12 at%, has minimal effect on ORR activ- in terms of the impact doped nitrogen functionalities have ity. However, the location of the nitrogen in the graphene on the catalyst ORR activity. We have also studied elec- lattice seems to have more influence on catalytic activity, trode structures of the graphene and tube-based catalysts which is consistent with the theory. We also know that to understand the relationship between the 3D mesoscale reduction of nGO starts above 200°C, but higher synthesis catalyst structure, mass-transport effects, and active-site temperatures (850°C) are required to improve its catalytic kinetics. This data is used as an input for theoretical effort properties towards ORR. Electron microscopy collected based on DFT calculations and LBM. shows that as the samples are heated the graphene sheets become more isolated, i.e. exfoliated, indicating that sur- The high-temperature route provides relatively little con- face area increases with temperature, which could explain trol over the exact structure and population of the ORR some the catalytic activity improvements. active sites. Therefore, in our third approach, we have focused on designing molecules as active-site mimics, char- Theory acterizing these sites spectroscopically and ultimately gain- Static and dynamic density functional theory calculations ing more control over the active site preparation. We have have been used to investigate the properties of the active been building molecular mimics of the active sites by coor- site(s) in model systems. ORR pathways have been deter- dinating extended aromatic ligands with N-donor atoms to mined for a series of transition metals and transition metal Fe. Initially, we attempted to use relatively simple ligands, clusters incorporated into the carbon-nitrogen framework. such as phenanthroline, with weak-field co-ligands incor- From this analysis we have been developing generalized porated to encourage the complex to be high-spin and to scaling laws so that a descriptor-based framework can be confer square-planar geometry. This has been extended rapidly used to optimize catalyst active site design. Ab ini- further to slightly larger ligands which similar properties, tio molecular dynamics simulations have been performed such as isoeilatin and dibenzoeilatin. While it is clear that to investigate the role of solvent in modifying the reaction these ligands complex Fe, full characterization has been pathway and/or the identity of the active site, in particu- hampered by low solubility and mixed stoichiometry of the lar probing the stability of these sites in aqueous media. reaction products. We are currently trying to incorporate Kinetic models based on the detailed molecular-level reac- additional functional groups to increase the solubility and tion pathways are being developed to support the lattice- thereby improve the characterization. Boltzman mesoscale model (LBM). Future Work We have developed a new lattice Boltzmann framework Research in this project will be performed in three main for pore-scale investigation of multicomponent multiphase tasks, with the following goals in the first year: reactive transport that can handle large density ratios be- tween liquid and gas phases. It can predict vapor to liquid Molecular Level Insights into the Nature of Non-Precious condensation, as well as oxygen and water vapor transfer Metal Active ORR and OER Sites and heterogeneous surface chemical reactions. The activa- We will establish a capability in the synthesis of model tion energy obtained from the DFT simulations will be used electrocatalyst systems based on nitrogen-doped graphene to calculate the reaction rate constant that will in turn be and graphene oxide (GO) functionalized with non-precious used in the LBM simulation, leading to a true multiscale metals and synthesize first samples of both types. Those model. samples will be characterized using spectroscopic and im- aging tools to identify the most active reaction sites. Catalyst Synthesis In the part of effort devoted to the modeling of graphene We pursue three approaches to the development of ORR systems we will start using density functional theory (DFT) catalysts with enhanced performance. The first two ap- calculations to assign experimental spectra and interro- proaches are based on high-temperature routes, involving gate active-site geometry and electronic structure. Based 54

Page 55

on the understanding obtained from DFT, modifications Transactions. will be made to the graphene/GO systems to test par- Chen, L., Q. Kang, Y. L. He, and W. Tao. A critical review ticular aspects of the theory. Together, this will provide of the pseudopotential multiphase lattice Boltzmann a preliminary scientific model critical for optimization of model: methods and applications. International non-precious metal ORR, OER, and bifunctional ORR/OER Journal of Heat and Mass Transfer. electrocatalysts. Chen, L., Y. L. He, Q. Kang, and W. Tao. Coupled Theory and Optimization of the Mesoscale Catalyst numerical approach combining finite volume and Structure lattice Boltzmann methods for multi-scale multi- In this task, fundamental kinetic quantities for active sites, physicochemical processes. 2013. Journal of Computational Physics. 255: 83. based on DFT, will be reconciled with mass-transport phenomena via the mesoscale lattice Boltzmann method Chung, H. T., J. H. Won, and P. Zelenay. Active and (LBM). The initial model will be developed and validated stable carbon nanotube/nanoparticle composite using experimental techniques developed at LANL. One electrocatalyst for oxygen reduction. 2013. NATURE class of measurements will provide model inputs, and COMMUNICATIONS. 4: -. includes porosimetry (hydrophobic/hydrophilic pores), TEM (particle size distribution), X-ray tomography (gas Elbaz, L., G. Wu, and P. Zelenay. Heat-treated non-precious- diffusion layer mesostructure). Another class will be used metal-based catalysts for oxygen reduction. 2013. In Electrocatalysis in fuel cells: A non- and low-platinum to validate outputs, including electrochemical impedance approach. Edited by Shao, M.. , p. 231. London: spectroscopy, helium/oxygen diffusivity, and neutron Springer-Verlag. radiography. Holby, E. F., G. Wu, P. Zelenay, and C. D. Taylor. Modeling Rational Design and Synthesis of Next-Generation NPMCs non-precious metal matalyst mtructures and mheir relationship to ORR activity. To appear in ECS Based on work derived from the initial studies of the mod- Transactions. el graphene electrocatalysts, as well as past results, the Hussey, D. S., D. L. Jacobson, D. Liu, B. Khaykovich, M. design, synthesis and optimization of new non-precious V. Gubarev, D. Spernjak, G. Wu, J. Fairweather, R. metal electrocatalysts will begin. Carbon and non-carbon Mukundan, R. Lujan, P. Zelenay, and R. L. Borup. templates (i.e. oxides) will be investigated for the stabiliza- Neutron imaging of water transport in polymer- tion and promotion of active sites. electrolyte membranes and membrane-electrode assemblies. To appear in ECS Transactions. Conclusion Wu, G., and P. Zelenay. Nanostructured Nonprecious The primary outcome of this project will be a fundamen- Metal Catalysts for Oxygen Reduction Reaction. 2013. tal understanding of non-precious metal ORR and OER ACCOUNTS OF CHEMICAL RESEARCH. 46 (8): 1878. catalysis in electrochemical energy systems. It will include reaction mechanisms, active-site structures, and synthesis- Yamaguchi, H., J. Granstrom, W. Nie, H. Sojoudi, T. Fujita, structure-reactivity correlations. In particular, integration D. Voiry, M. Chen, G. Gupta, A. D. Mohite, S. Graham, of modeling of the active site with mesoscale theory will and M. Chhowalla. Reduced graphene oxide thin films allow coupling of intrinsic site reactivity with mesostruc- as ultrabarriers for organic electronics. 2013. Advanced ture-sensitive kinetics, providing “real-world” activity Energy Materials. : DOI: 10.1002/aenm.201300986. directly comparable to the experimental measurements. First-principles calculations will assist in interpreting spec- troscopic and electrochemical signatures emerging from characterization studies. The fundamental understanding of electrocatalysis will be used to design, synthesize, and demonstrate next-generation Non Precious Metal Catalysts with high performance and durability. Publications Chen, L., Q. Kang, Y. L. He, W. Tao, and P. Zelenay. Pore-scale simulation of sater vapor generation, condensation, and transport in the cathode of a proton exchange membrane fuel cell. To appear in ECS 55

Page 56

Chemistry and Material Sciences Directed Research Continuing Project Phase Stability of Multi-Component Nanocomposites under Irradiation Blas P. Uberuaga 20130118DR Introduction Benefit to National Security Missions We propose to design multi-component nano-composite Our aim is to develop a basic understanding of a primary materials that resist radiation-induced solute redistri- damage failure mode in materials under irradiation: bution (RISR) and thus maintain stable nano-phases at radiation-induced solute redistribution. This leads to extremes of high dose irradiation at elevated tempera- phase separation, dissolution of important secondary tures. The next generation of nuclear energy systems phases, and changes in chemical composition that ac- will not be possible without the development of such celerate materials failure in nuclear energy systems. By new materials. Further, the accident at the nuclear plant developing a basic understanding of these materials, at Fukushima demonstrates the need for advanced clad we will formulate design principles that will allow us to materials that not only can sustain high burn-up but are propose future nuclear energy materials that resist such also resistant to other phenomena such as corrosion. damage and failure. The design of materials that are essentially “immune to That this is a fundamental materials approach, this is radiation”, as called for by DOE, is one of the grand chal- clearly relevant to the DOE Office of Science, particularly lenges in the materials science discipline. This is because Basic Energy Sciences (BES). If successful, the types of energetic radiation damages a material by a variety materials we would propose would also be of interest of very different mechanisms, such as void swelling, to Fusion Energy Sciences (FES) with in the Office of embrittlement, irradiation creep, and phase instabili- Science, as well as the Office of Nuclear Energy, which is ties from radiation induced solute redistribution (RISR). more focused on fission energy systems. Typically, the solution for one problem either has no effect on or exacerbates the other damage mechanisms. This work would enhance energy security by enabling For example, using interfaces as sinks for vacancies and future generation nuclear energy systems that are more interstitials can reduce void swelling but the same inter- robust and thus more politically and socially acceptable faces that mitigate void swelling can increase RISR. as replacements for fossil energy sources. Further, by developing more radiation tolerant claddings, nuclear RISR causes phase instabilities that lead to catastrophic fuel could be burned to greater amounts, reducing the failure through embrittlement, fracture and corro- quantity of spent fuel in storage and the consequent sion. We propose an innovative solution to the chronic load on any future nuclear waste repository. problem of radiation-induced phase instability: nano- Fuel cycle research of this sort builds capability for composite materials design via interfaces that minimize nuclear nonproliferation in general. defect fluxes, which is the origin of RISR. We hypothesize several factors we can use to control defect fluxes and Progress suppress RISR. We propose a materials co-design R&D We have begun work on two main efforts. approach where experiments and multi-scale modeling The first involves model systems of Ni-Al alloys, to mimic are closely integrated to test our hypotheses, develop Inconel materials. This involves two primary tasks. The a fundamental understanding of RISR to interfaces, and first is atomistic scale modeling, which consists of ac- advance design principles for next-generation materials celerated molecular dynamics (AMD) simulations of for nuclear energy applications that maintain chemical diffusion in disordered alloys, atomistic simulations of composition profiles that are critical for many materials Al segregation to Ni grain boundaries, an assessment properties in extreme environments. 56

Page 57

of potentials for the Ni-Al phase diagram, and atomistic prove useful for studying complex alloys under irradiation. modeling of the role of alloying elements on the stability of We are also in the initial stages of planning atom probe Ni3Al precipitates within Ni. This is complemented by ex- tomography experiments, with the University of Michigan, perimental studies which involve the growth, characteriza- to complement our TEM characterization efforts at LANL. tion, and implantation studies of Ni/Ni3Al multilayers. The Emmanuelle Marquis (UMich) will visit in July to coordi- experimental effort just began as the experimental postdoc nate these activities. Finally, we are also exploring the use has just joined LANL. In the AMD simulations, we find of mechanical alloying coupled with levitation melting to that defects interact only weakly with Al atoms, and that synthesize three-dimensional nanocomposites of the same diffusion must occur via random encounters rather than composition as our multilayer morphologies. the formation of defect complexes. The atomistic simula- tions of segregation show that Al only weakly segregates to In addition, to perform some of the irradiations, we had model boundaries, but once it does, begins repelling other to complete modifications of the Tandem accelerator in Al, forming somewhat complex ordered structures at the the IBML in order to perform higher energy implantations/ boundary that can be validated with transmission electron irradiations at higher temperatures that are more relevant microscopy (TEM). for nuclear energy applications. We also installed a pyrom- eter to measure the temperature of the sample during the The second effort involves model metal/oxide systems to implantations/irradiations. mimic nanostructured ferritic alloys (NFAs). Again, there are two primary tasks. The first involves multiscale model- We have put in place two subcontracts (MIT and UMich) ing, using density functional theory (DFT) to examine the and hired multiple new postdocs to perform the work on structure and chemistry of the interface between Fe and this project. We are also organizing a course on the use of TiO2 and Fe and Y2O3, the two model systems we have Idaho National Laboratory’s MARMOT phase field simula- focused on initially. The output from these calculations tion code which will provide the necessary framework to will be used to parameterize a phase field model of alloy perform the phase field simulations. redistribution under irradiation. In conjunction, we are performing experimental studies on multilayer materials Future Work such as Fe/Y2O3/Fe/TiO2/Fe/MgO, in which the Fe layers During the next fiscal year, we will focus on two primary are doped with Cr, to understand how Cr segregates to tasks. these interfaces. We will soon irradiated these materials in the Ion Beam Materials Laboratory (IBML) at LANL and The first is developing a multiscale model that incorporates characterize how the Cr redistributes under irradiation. atomistic information into a mesoscale description of the In conjunction with these efforts, we have performed material to understand how solute redistribution changes neutron reflectometry of Fe/TiO2 and Fe/Y2O3 interfaces as a function of microstructure during irradiation. This will and will be analyzing that data this summer. Regarding the be done for two metallic model systems, Fe and Ni alloys, multilayer materials, we have characterized the atomic which have real engineering counterparts in nuclear en- structure of the Fe(Cr)/MgO, Fe(Cr)/Y2O3, and Fe(Cr)/TiO2 ergy systems. A critical component of this modeling effort interfaces. We are currently determining the local oxygen will be to determine to what extent a simplified descrip- content at those interfaces using analytical microscopy tion of the chemical interactions within the alloy system techniques. The DFT calculations have pointed to several can be used to develop an accurate kinetic model of the high-level governing rules that determine the propensity solutes within the defect fields produced during irradia- of alloying element segregation to metal/oxide interfaces, tion, as this will be key for the next step: the development based on Hume-Rothery Rules and Ellingham Diagrams. of such a model for metal/oxide composites. The model- This provides a quick predictive estimate of which ele- ing development will be focused on four areas: 1) defect ments will segregate to a given interface. thermodynamics in and around boundaries in the differ- ent model materials, 2) the coupling of solutes with those In parallel, we are characterizing real engineering materi- defects, 3) the redistribution of solutes near a given grain als, particularly 14YWT, to compare with the results on our boundary, and 4) incorporation the atomistic information model materials described above. into a mesoscale model that determines how the solutes redistribute in a complex microstructure. There are three side projects related to these main tasks. First, we are developed spatially parallel acceler- The second task will be to perform experiments on the ated molecular dynamics methods to extend the system exact same model systems as are being examined theo- size that these methods can be applied to, which should retically. We will synthesize polycrystalline materials, both 57

Page 58

Fe and Ni, with various amounts of solute elements. We will characterize the materials pre and post irradiation to understand how the elements have redistributed. This will include SIMS, XRD, and TEM, the later particularly focused on determining how the segregation depends on boundary type. These results will then be compared to the modeling to determine to what extent the modeling has captured the important phenomena and develop design principles on what features of the microstructure, as detailed by the hypotheses in our proposal, control radiation-induced solute redistribution. Conclusion The goal of this project is to understand and control the atomic scale processes that lead to redistribution of ele- ments in complex alloys during irradiation. The key results of this project will be (i) development of a multi-length and multi-time scale model to predict solute/elemental redis- tribution in a given alloy as a function of interface proper- ties and spacing, (ii) tests of the hypotheses correlating the interface properties to RISR in nanocomposites, and (iii) development of figures-of-merit from the integrated results of all the hypotheses. 58

Page 59

Chemistry and Material Sciences Directed Research Final Report Radioparagenesis: Robust Nuclear Waste Form Design and Novel Materials Discovery Christopher R. Stanek 20110009DR Abstract the impact of daughter product formation on the stabil- Of the outstanding issues currently preventing expan- ity of solids comprised of radioactive isotopes? In order sion of nuclear energy, nuclear waste disposal possesses to answer this question, we require a multidisciplinary the highest potential to benefit from technological approach integrating first principles modeling with advances. By verifying the concept of radioparagenesis, the synthesis and characterization of small and highly we will introduce a new way of thinking about nuclear radioactive samples. Once we understand the effects of waste disposal. Specifically, not only can robust waste radioparagenesis we can use the insight to evaluate and forms be designed to withstand radiation damage and design waste forms. We have pursued three tasks that chemical evolution, but also computational tools pre- correspond to each of the following goals: (1) develop dicting the performance of waste in a repository set- a new branch of physics and chemistry associated with ting can be made significantly more reliable. The goal our recent discovery of radioparagenesis, (2) utilize the of this project is to validate and verify the new concept insights gained to design robust nuclear waste forms, of “radioparagenesis.” Radioparagenesis was recently and (3) discover and characterize heretofore unobserved discovered by our team, and is the formation of uncon- materials with potentially interesting properties. We ventional crystal structures (e.g. rocksalt BaCl) for com- have designed these goals to guide and focus the explo- pounds formed during the chemical transmutation that ration that occurs after a significant discovery – in this occurs during radioactive decay. The technical approach case the discovery of radioparagenesis, which we define of this project consists of experimental and theoretical as: the formation of compounds, often unconventional components. The experimental component relies on the (e.g. BaCl in the rocksalt structure), due to the chemical Isotope Production Facility (IPF) at LANL to synthesize transmutation that occurs during radioactive decay. accelerated fission product analogues that decay quickly For over thirty years, crystalline ceramic nuclear waste enough (as opposed to actual fission products of interest forms have been proposed for the encapsulation of spe- to the nuclear energy community, whose decay dictates cific fission products. Faced with the necessity to predict hundred year experiments) for observation. We employ the long-term behavior of these waste forms, research- complementary first principles calculations to predict ers have focused on candidate compositions that exhibit: the result of the accelerated experiment. Verification (1) geological analogues, thus suggesting long-term sta- of this phenomenon will allow us to drastically improve bility, e.g. Synroc, (2) resistance to radiation damage (ei- the predictability of nuclear waste form performance ther using accelerated isotopes or by ion irradiation) and as well as the backward design robust waste forms. In (3) leaching resistance. However, these studies have not addition to providing a potential innovative solution to yet adequately accounted for the chemical and physical the nuclear waste problem, radioparagenetic phases changes that will occur in the waste form during fission may also exhibit interesting materials properties for non- product transmutation, i.e. the conversion of one ele- nuclear applications. ment or isotope into another, in this case during radio- active decay. Indeed, for many fission products, waste Background and Research Objectives form performance may be dominated by transmutation In order to develop a predictive capability to design effects. The selection criteria for a candidate waste form radiation tolerant and chemically robust nuclear waste discussed above may address the performance of the forms (and in the process, remove a significant obstacle waste form very early in its life, but lack the predictive limiting the expansion of nuclear energy), we must first power required to evaluate the performance of the solve a fundament materials science question: What is 59

Page 60

waste form far in to the future. project, we made several slight course corrections. For example, Task 3 was refocused to more generally study the For as long as ceramic waste forms have been considered, implications of radioparagenesis to include novel materi- researchers have recognized what Vance, nearly 30 years als and new directions subtasks. Furthermore, our original ago, called the “transmutation problem,” i.e. the deleteri- emphasis on X-ray diffraction in Task 2 was augmented to ous effects on waste form performance due to the forma- include electron microscopy in order to take advantage of tion of chemically disparate daughter products. In the reduced sample volume (in order to minimize radioactiv- same year as the Vance work, which outlined potential ity). deleterious effects due to transmutation (for example due to the different valence and ionic radius of parent and Scientific Approach and Accomplishments daughter isotopes), Gray published in Nature the results of The major accomplishment of this project has been the an experimental attempt to determine the role of daughter establishment of a new capability to study accelerated product formation on candidate waste form stability. Since chemical aging samples. This has required the integra- consideration of actual fission products would require ex- tion of radiochemistry (for sample synthesis) and materi- periments lasting one hundred years or more, Gray identi- als science (for sample characterization). The ability to fied two approaches to accelerate the phenomenon using synthesize very small samples for characterization via e.g. either (1) separation of short-lived isotopes, or (2) neutron electron microscopy in relatively open areas (such as the irradiation of non-radioactive material to activate short- Materials Science Laboratory) has allowed for unique stud- lived isotopes. Gray pursued the second option - claiming ies to be performed. For example, we have very recently that the first was “prohibitively expensive” - and found no studied isotopically pure Lu-177 samples, the details of significant change in waste form properties. However, only which are below. a small fraction of transmutation (~10% of that expected in actual waste) had occurred and, the experiment did Lu-177 decays primarily through 0.497 MeV beta radiation, not reproduce typical waste form conditions (e.g. neu- and has a half-life of 6.65 days, and is currently used in nu- tron damage was annealed). Since the work of Vance and clear medicine f. It decays to Hf-177, which is stable. From Gray, the transmutation problem has remained essentially a coordination chemistry perspective, Lu-177 meets the dormant, not because the problem has been solved, but requirements we have defined for accelerated aging stud- rather, because there has been no clear path forward for ies, namely: (1) half-life must be conducive to experimental further investigation. In fact, an important point about investigation, (2) the daughter product must be chemically these previous studies is that although the investigators distinct, (3) the radioactivity must be sufficiently low such were certainly aware of the transmutation issue, they were that it can be handled, and (4) the isotope must be avail- only able to ask rhetorical questions. With new experi- able. In this case, Lu is a rigidly 3+ cation while Hf is rigidly mental (e.g. Isotope Production Facility (IPF)) and theoreti- 4+, thus satisfying criterion 2. The beta radiation associat- cal tools (e.g. density functional theory (DFT)) that exist ed with Lu-177 is relatively low, and can be shielded espe- today, we have an important opportunity to not only ask cially for small transmission electron microscopy samples, these questions, but also provide answers via methods which satisfies criterion #3. Importantly, we negotiated that were recently thought to be impossible (e.g. we have with colleagues at the Missouri University Research Reac- started Gray’s “prohibitively expensive” experiment in an tor (MURR) for us to receive Lu-177 for only the shipping LDRD feasibility study). Furthermore, IPF experiments are cost, which very elegantly satisfies requirement #4. The “cleaner” in that the samples are isotopically purer, free very short half-life of Lu-177 however requires extreme from neutron radiation induced damage (not present in choreography of experimental activities. As an example, actual waste) and more representative of the separations for the first experiment we performed on Lu-177, the and nuclear waste form processes. sample arrived via FedEx from MURR on 6/18/13. The sample was purified in the TA-48 hot cells before dawn To maximize the potential for success of this LDRD DR on 6/19/13 (to remove daughter product grown in during project, we originally organized it to consist of three inte- shipping). The sample was oxidized (to Lu2O3) and the grated components, relying upon unique LANL strengths TEM sample produced in the evening of 6/19/13 (Figure in radiochemistry and characterization of radioactive 1). The sample was shipped from TA-48 to the Materials material; earth and energy systems; and computational Science Laboratory in TA-3 early on 6/20/13, at which time materials science, respectively. These tasks are: (1) Experi- we began to accumulate TEM date. This experiment has mental Verification of Radioparagenesis, (2) Robust Waste required significant RCT and line management support. Form Design and Evaluation, and (3) Exploration of Novel Materials and Their Properties. During the course of the 60

Page 61

Figure 1. Laura Wolfsberg (C-IIAC) and Jeff Aguiar (MST-8), rehearsing the loading procedure of the Lu-177 TEM sample in to a retractable holder through a shielded plexiglass box. Note the camera providing real time imagery of the sample in the laptop screen. We are still in the process of analyzing the Lu-177 data, Figure 3. The black circles in the top image correspond to lattice and have performed a second experiment in order to constant of 177Lu2-xHfxO3 sample, plotted as a function of time confirm initial results. However, the preliminary results (Lu-177 half-life = 6.65 days). The concentration of Lu and Hf are are very exciting. We have obtained diffraction patterns, shown as open red squares and triangles respectively. It is clear chemistry and even atomic resolution images (Figure 2) from this figure that there are at least 4 separate phases that for the sample over the course of 6 half lives. From the occur. These phases are compared to the equilibrium phase dia- diffraction data, it seems as though the phases present do gram represented in the lower image. It is also clear that there is not correspond exactly to equilibrium phase diagrams, see not agreement between the measured data and the equilibrium Figure 3. In fact, there is a particular phase that we believe phase diagram, suggesting the transmutation is resulting in non- we have observed that has not been observed in nature equilibrium behavior. – although further analysis (including comparison to DFT In addition to accelerated chemical aging, we have also calculations) is required to confirm. studied the impact of daughter product formation on the stability of waste forms in a repository setting. This has been done by linking DFT results of surface energetics and reactivity, with Lattice-Boltzmann pore scale models. For example, we have calculated the segregation of Ba to CsCl surfaces (Ba is the daughter product of Cs-137, an impor- tant short-lived fission product) see Figure 4, the influence of Ba on the formation of surface defects, and the interac- tion of water on both clean CsCl and Ba-doped surfaces. Importantly, we have found that the incorporation of Ba in the subsurface dramatically reduces the energy to form surface defects or, alternatively, to rip them off the surface. This effect results in enhanced leaching when the daugh- ter is present. This effect has been demonstrated using Lattice-Boltzmann approaches, see Figure 5. This result demonstrates the dramatic effect that daughter products can have on waste form performance, even when not resulting in a phase change. Several other related studies have been conducted related to the chemical effects associated with transmutation or Figure 2. STEM image of 177Lu2-xHfxO3 sample. daughter product formation. For example, we have syn- 61

Page 62

thesized Fe-55 rich samples for experimental characteriza- Impact on National Missions tion. In addition to studying the effects of the formation of A critical problem for DOE-NE is providing a solution for Mn (daughter product), we have also loaded these samples nuclear waste disposal. This is evidenced by the recent in to a range of diamond anvil cells to study the combined Blue Ribbon Commission report on nuclear waste disposal effect of chemical strain due to daughter product and ap- as well as guidance from DOE to focus on “ long-term, plied pressure. We have also use density functional theory science-based R&D of technologies with the potential to to examine the effect of C-14 transmutation on DNA stabil- produce beneficial changes to the manner in which the ity. The human body contains 20 ng of radioactive 14C, nuclear fuel cycle and nuclear waste is managed.” This generating 4000 decays per second of which 20–30 occur project directly fits this description. This project also ad- in nuclear DNA. We have found differences in behavior dresses issues of fundamental materials science, using between DNA and RNA that might suggest justification computer simulation to design materials with engineered for genetic code evolution. Finally, we are working with functionality. colleagues at Idaho National Laboratory to examine a ~50 year old sample 137CsCl in order to compare our acceler- In addition to civilian nuclear waste, In the 1970s and ated chemical aging results to actual fission product waste 1980s large quantities of 137Cs (t 1/2 = 30 yr) and 90Sr forms of a certain age. This study will continue in FY14 via (t1/2 = 29 yr) were extracted from spent nuclear fuel at INL LDRD support. Hanford in Washington State, and synthesized into cap- sules of CsCl and SrF2. Even though many of the capsules have aged for more than one half-life, they still account for around one-third of the radioactivity at the Hanford site and are the most significant source of radiation in the United States outside of a reactor. Stored underwater due to the radiation field, monitoring primarily consists of engineering-type “clunk-tests” which assess swelling associated with radiation damage, gas evolution, and/or phase decomposition. Aside from health-and-safety con- siderations, the Hanford capsules are scientifically interest- ing from a solid-state chemistry viewpoint, since the stable daughter isotopes differ markedly in oxidation state and size to the parent. These large changes in local chemistry have the potential to compromise a nuclear waste form over time by inducing large volume changes and disrupt- Figure 4. Schematic of atomistic simulations performed on the ing the crystalline lattice. Again, studies performed in this effect of Ba (daughter product) on CsCl surface stability in the project directly address the transmutation issue present in presence of water. this defense waste. Publications Costa, G. C. C., H. W. Xu, and A. Navrotsky. Thermochem- istry of Barium Hollandites. 2013. JOURNAL OF THE AMERICAN CERAMIC SOCIETY. 96 (5): 1554. Jiang, C., B. P. Uberuaga, K. E. Sickafus, F. M. Nortier, J. J. Kitten, N. A. Marks, and C. R. Stanek. Using “radiopara- genesis’’ to design robust nuclear waste forms. 2010. ENERGY & ENVIRONMENTAL SCIENCE. 3 (1): 130. Figure 5. Dissolution of CsCl into water at 20○C. Reaction rate Jiang, C., C. R. Stanek, N. A. Marks, K. E. Sickafus, and B. P. constant is calculated without considering Ba effect (E=1.4eV, Uberuaga. Radioparagenesis: The formation of novel top); and is calculated considering Ba effect (bottom). From left compounds and crystalline structures via radioactive to right: time=0, 10, 20, and 30 years. The frequency factor (A) decay. 2010. PHILOSOPHICAL MAGAZINE LETTERS. 90 in the Arrhenius equation is 1.e11; initial domain is 10cmX10cm, (6): 435. and initial diameter of the CsCl particle is 4 cm. Initial concentra- tion of CsCl is 0.1mol/L; concentration at the left boundary is set Jiang, C., C. R. Stanek, N. A. Marks, K. E. Sickafus, and B. P. at 0.1 mol/L; the right boundary is the exit; and top and bottom Uberuaga. Predicting from first principles the chemical boundaries are nonreactive walls. evolution of crystalline compounds due to radioactive 62

Page 63

decay: The case of the transformation of CsCl to BaCl. 2009. PHYSICAL REVIEW B. 79 (13): -. Marks, N. A., D. J. Carter, M. Sassi, A. L. Rohl, K. E. Sickafus, B. P. Uberuaga, and C. R. Stanek. Chemical evolution via beta decay: a case study in strontium-90. 2013. JOURNAL OF PHYSICS-CONDENSED MATTER. 25 (6): -. Sassi, M., B. P. Uberuaga, C. R. Stanek, and N. A. Marks. Transmutation in (SrF2)-Sr-90: A density functional theory study of phase stability in ZrF2. 2012. PHYSICAL REVIEW B. 85 (9): 094104. Stanek, C. R., B. P. Uberuaga, B. L. Scott, R. K. Feller, and N. A. Marks. Accelerated chemical aging of crystalline nu- clear waste forms. 2012. CURRENT OPINION IN SOLID STATE & MATERIALS SCIENCE. 16 (3, SI): 126. Uberuaga, B. P., C. Jiang, C. R. Stanek, K. E. Sickafus, N. A. Marks, D. J. Carter, and A. L. Rohl. Implications of trans- mutation on the defect chemistry in crystalline waste forms. 2010. NUCLEAR INSTRUMENTS & METHODS IN PHYSICS RESEARCH SECTION B-BEAM INTERACTIONS WITH MATERIALS AND ATOMS. 268 (19): 3261. Winkler, B., C. R. Stanek, W. A. Taylor, L. Wolfsberg, B. L. Scott, R. M. Dickerson, W. Morgenroth, and L. Bayar- jargal. Synthesis and characterization of (Fe2O3)-Fe-55 for the investigation of radioparagenesis. 2013. SOLID STATE SCIENCES. 18: 58. Zhang, F., J. D. Gale, B. P. Uberuaga, C. R. Stanek, and N. A. Marks. Importance of dispersion in density functional calculations of cesium chloride and its related halides. 2013. PHYSICAL REVIEW B. 88 (5): -. 63

Page 64

Chemistry and Material Sciences Directed Research Final Report Hydrogen Effects in Delta-Stabilized Pu Alloys: Fundamental Thermodynamics and Interactions at Reduced Dimensionality Daniel S. Schwartz 20110011DR Abstract Vacancy (SAV) phenomenon. Our goal in this project was Hydrogen is known to strongly interact with plutonium to investigate the effects of hydrogen in Pu-Ga alloys, as and its alloys. Our data from a variety of diffraction tech- a function of both hydrogen and Ga content. We were niques coupled with immersion density measurements particularly interested to search for SAV behavior in shows that hydrogen induces a high vacancy concentra- these alloys. tion, 10-3- 10-4, in Pu-2 atomic % Ga alloys. We feel that Scientific Approach and Accomplishments the high vacancy concentration is an incipient form of the Superabundant Vacancy phenomenon. Modeling of With the cooperation of the plutonium manufacturing hydrogen-vacancy structures using density functional team in the plutonium facility at LANL, we prepared techniques suggest that the complex formed by intro- three different alloys for this project in kilogram quanti- ducing multiple hydrogen atoms into a Pu vacancy is a ties. The alloys included pure Pu, which was electro- stable, low-energy structure. Increasing the Ga concen- refined twice to maximize purity, a Pu-2 atomic % Ga tration to 7 atomic % effectively shuts down the vacancy- (Pu-2Ga) alloy, and a Pu-7 atomic % Ga (Pu-7Ga) alloy. stabilizing ability of hydrogen. Ga is commonly added to Pu alloys to stabilize the face centered cubic δ-Pu phase, which is a much more duc- Background and Research Objectives tile, machinable, and generally easier to handle form of Plutonium-based alloys exhibit complex behaviors that Pu. At room temperature, pure Pu has the monoclinic have puzzled materials scientists since they were first crystal structure (the α-Pu phase) which is quite brittle. produced in the 1940’s. LANL Pu scientists are tasked The Pu-7Ga alloy (Figure 1) is approximately at the Ga with cutting through this complexity and developing concentration where δ-Pu is most stable at room tem- an understanding of Pu alloys that will allow effective perature [2]. The ability to make significant quantities management of the Nation’s aging nuclear stockpile. of Pu alloys is a unique LANL capability, and we consider This project focused specifically on expanding our under- the preparation of our project alloys a great success. standing of the influence of dissolved hydrogen on the The alloys were exposed to hydrogen (or deuterium) structure and behavior of Pu alloys. The urgency for the using a Sieverts’ apparatus, which is a furnace capable project is the fact that much of the Pu in the stockpile of heating samples while they are exposed to precisely contains significant amounts of hydrogen, with 0.5 – 1.0 measured quantities of gas. The temperature and pres- atomic % being typical. Despite the high amounts of sure are accurately measured, yielding the pressure-tem- hydrogen in our Pu alloy stream, our understanding of perature diagram for the specimen, from which it can be the effects of hydrogen dissolved in the Pu alloys lattice determined exactly how much hydrogen was absorbed is rudimentary. by the specimen. As the hydrogen pressure is increased Hydrogen is known to have a profound effect on defect over a Pu specimen, three stages are typically observed: properties in many metal systems, where it both cre- 1) hydrogen being absorbed in the Pu lattice, 2) the ates and stabilizes vacancy structures at high concen- nucleation and growth of PuH2 in the Pu specimen, and trations [1]. Vacancies, i.e. sites in the crystal lattice 3) the completion of the hydriding process, where the where atoms are missing, can be a trap for hydrogen in entire specimen is PuH2. In this program, we exclusively the lattice, where the hydrogen atoms and the vacancy focused on the first region, where hydrogen is in solution defect mutually stabilize each other. In some metals, this in the Pu lattice, but no hydride has formed. effect can be extreme and is called the Superabundant 64

Page 65

bulk density was measured (~0.45% decrease), while the density based on XRD measurements increased (~0.4%). For further confirmation, we performed similar lattice parameter measurements on H-charged and pure Pu-2Ga using both neutron diffraction at LANL’s neutron accelera- tor and synchrotron diffraction at the Stanford Synchrotron Radiation Lightsource. In both cases, 1 atomic % hydrogen in Pu-2Ga resulted in a ~0.4% density increase, based on the measured lattice parameters. This is a classic indication that the lattice contains a significant quantity of vacancies. We estimate from these numbers that the vacancy concen- tration is in the 10-3 to 10-4 range. This is a strikingly high vacancy concentration. For comparison, well-annealed gold at room temperature will have a vacancy concentra- tion on the order of 10-17! (A) (B) a0 a0´ Figure 2. (A) Schematic of a perfect Pu lattice. The lattice param- Figure 1. A portion of the Pu-7 atomic % Ga casting produced for eter, a0, can be measured by XRD. (B) When a vacancy forms this project. (marked by □) a Pu atom moves from its site to a different loca- tion in the Pu lattice, leaving a vacant lattice site. The vacancy Our first goal was to quantify the concentration of vacan- causes the lattice to collapse slightly, so XRD measures a smaller cies in our alloys, as function of Ga and H content. We lattice spacing (a0´ < a0). However, the overall volume of the used a method first developed to quantify the vacancy ensemble of atoms becomes larger, because the atom formerly concentration in metals as they approach their melting at the vacancy finds a new site to occupy. Therefore, the XRD point [3]. This technique involves comparing the bulk measurements show a densification of the lattice, while bulk den- density of a specimen to its density determined by X-ray sity measurements show a decrease in density. This differential indicates the presence of vacancies. Diffraction (XRD). The comparison is sensitive to vacan- cies because the atomic lattice in a metal contracts in the As a further check, we took the hydrogen-charged speci- presence of vacancies, i.e. there is an apparent increase in men back into the Sieverts’ apparatus and removed the density, while the bulk density actually decreases because hydrogen by vacuum annealing at 450°C. We subsequently many lattice sites are vacant (Figure 2). Our bulk density remeasured the XRD lattice parameter for the dehydroge- measurements are made using the Archimedes immer- nated specimen, and it had returned to a value close to its sion method, which is highly accurate for specimens in the initial, hydrogen-free value. The results indicate that the 1 - 10 gram weight range. XRD measurements yield a very vacancy concentration had returned to its normal level accurate value for the lattice parameter, i.e. the atomic after hydrogen removal. This series of measurements is the spacing in the sample’s lattice, typically on the order of first distinct evidence for hydrogen-induced vacancies in Pu 200 ppm error. We first held several Pu-2Ga specimens at alloys. 450°C in vacuum for several days to uniformly distribute the Ga throughout the specimens. After slow cooling, we In parallel with our experimental efforts, we performed measured the density and XRD lattice parameter of the a variety of calculations to examine the energetics and specimens. Then we used the Sieverts’ apparatus to charge atomic structure of hydrogen-vacancy complexes in the Pu the specimens with 1 atomic % hydrogen, and remeasured system. Density functional theory (DFT) was used to cal- the density and XRD lattice parameter. A clear decrease in culate the energy of different plausible hydrogen-vacancy 65

Page 66

structures. DFT calculations are based on fundamental SAV systems described by Fukai [1], which can approach 30 quantum mechanics and yield electronic configurations at. %. We are operating at far lower hydrogen pressures, and energies. The calculations are complex for atoms with but the basic processes responsible for the SAV phenom- large numbers of electrons like Pu, where f-orbitals must enon are active, i.e. hydrogen-vacancy complexes having be included in the process, and the computation cells are lower energy than isolated vacancies and interstitial hydro- limited to ~30 atoms. The effect of introducing hydrogen gen. We consider the hydrogen-vacancy structures in our atoms into a vacancy in Pu was systematically examined by Pu alloys an incipient form of the SAV phenomenon. putting from 1 to 8 hydrogen atoms into a cell of Pu atoms We revived the Positron Annihilation Lifetime Spectros- with a vacancy site at its center (Figure 3). The energies of copy (PALS) instrument at LANL for this project. PALS is one the resulting hydrogen-vacancy complexes were calculated, of the few analytical techniques capable of giving direct and the atoms were allowed to move to minimize the cell information about atomic level defects with volumes of energy and reveal where the hydrogen atoms preferential- ~1 nm3 and is therefore ideal for probing our hydrogen- ly bind to the Pu vacancy. The hydrogen atoms were found vacancy complexes. Positrons directed into a material to avoid the center and bind to the edges of the vacant will penetrate hundreds of micrometers into the bulk and site. Several hydrogen atoms may coexist in the vacant site annihilate preferentially at regions of excess volume in the with only weak hydrogen-hydrogen atom interactions. It crystal lattice (e.g. vacancies). The lifetime of the positron was not until 4 hydrogen atoms were introduced into the can be measured accurately, and is directly related to the vacancy that significant crowding (i.e. strong electronic volume of the annihilating defect. Lifetime measurements interaction) of the hydrogen atoms took place (Figure 3e). are coupled with DFT calculations of the electron and posi- Hydrogen atoms were observed to monotonically reduce tron density for expected defects, such as single vacancies the energy of formation for vacancy-hydrogen complexes. and hydrogen-vacancy complexes. Accurate calculations of The calculations thus support the idea that hydrogen stabi- the positron lifetime can be made from the charge den- lizes vacancies in Pu through the formation of a low-energy sity values, and a good match of measured lifetime to a hydrogen-vacancy complex, most likely with multiple calculated value is convincing evidence that the defects are hydrogen atoms associated with each complex. the calculated type. We successfully developed the ability to perform such calculations at LANL, and calculated very accurate lifetimes for aluminum as a test. While we had good success resuscitating the PALS system and performing the necessary calculations, our results on Pu specimens were not satisfactory. We performed PALS on Pu-2Ga and Pu-7Ga alloys, both hydrogen-free and with 1 atomic % hydrogen. Our results were not consistent with calculated lifetimes, which were significantly lower than the measured values. In addition, we did not see a mean- ingful effect on lifetime due to the addition of hydrogen. We believe that a mixture of instrument errors and the electronic complexity of Pu are confusing our results. Nevertheless, we support the future use of PALS to probe hydrogen-vacancy complexes in Pu, as it should be able to give us information about the volume of the complexes which is otherwise unattainable. Hydrogen that is dissolved in Pu alloys must first pass through the sample surface. Therefore, a key part of understanding hydrogen in Pu alloys is understanding how hydrogen interacts with the Pu surface. X-ray Photoemis- Figure 3. Calculational cells for increasing numbers of hydrogen sion Spectroscopy (XPS) is a powerful tool for examining atoms (0 to 8 H atoms) in a Pu vacancy used for density function- atomic species and charge states in the top few nm of the al calculations. The Pu atoms are gray, and the hydrogen atoms are circled. specimen surface. XPS was used to study the surface of hydrogen-free Pu-7Ga. This alloy was chosen because we The concentration of hydrogen-stabilized vacancies that believed its well-known oxidation resistance and phase sta- we measure in our Pu alloys is much lower than that in the bility would make it a simple starting point, from which we 66

Page 67

could better understand the behavior of the more complex scientists have been tasked with assessing the state of the hydrogen-charged materials. However, the surface of Pu- Pu stockpile as it ages. While these efforts have been quite 7Ga was not simple. As received, the surface was a com- rigorous over the past decades, they have not dealt with plex mixture of PuO2 and Ga oxide, along with fluorine, or recognized the prevalence of hydrogen in the Pu metal. carbon, and other elements picked up from the glovebox Relatively straightforward and conservative calculations environment where the Pu-7Ga specimens are prepared. suggest that the damage field produced by alpha-decay in Surfaces are prepared for XPS analysis by sputter cleaning 239Pu will interact with ~2% of the hydrogen-vacancy de- with argon ions, which removes incidental contaminants, fects per year in a specimen with the typical concentration and sputtering was successful in removing fluorine, carbon, of 10-3 that we measured. Even without any understand- and copper from the surface. However, even extended ing of what results from this interaction, it seems clear that sputtering did not remove the oxide layer. We conclude the intrinsic defect structure of the stockpile Pu needs to that much of the surface oxide comes from oxygen in the be considered to understand Pu aging at the fundamental Pu bulk moving to the surface, even at room temperature. level. We used a related technique, Photoemission Spectroscopy Our project also had an impact on the development and (PES) to get information about the electronic structure of maintenance of Pu facilities at LANL. The positron annihila- δ-phase Pu, and δ-Pu in the presence of hydrogen. A key tion system in particular was resurrected and used for Pu finding was that δ-Pu has a multivalent electron structure, measurements for the first time at LANL. Additionally, this i.e. the f-orbital contains a mixture of 5f 5 and 5f 6 elec- project involved three early career scientists, supporting tronic states. With this more detailed understanding of the goal of maintaining a stream of plutonium scientists for pure δ-Pu electronic structure, we addressed the question the future. of how hydrogen interacts with Pu to modify the electronic structure. PES spectra were taken from δ-Pu in the pres- References ence of hydrogen gas at a range of temperatures from 20K 1. Fukai, Y.. Formation of superabundant vacancies in to 300K. At the lowest temperatures, H and Pu weakly M-H alloys and some of its consequences: a review. interact indicative of simple physiosorption of hydrogen on 2003. JOURNAL OF ALLOYS AND COMPOUNDS. 356: the Pu surface. At 300K, there are clear signs of chemical 263. bonding between H and Pu, specifically in the disruption of 2. Timofeeva, L. F.. Phase transformations in Pu-Ga and the 5f 5 component of the multivalent f-orbital. The chemi- Pu-Al alloys Effects of pressure and temperature on cal bonding of H and Pu at 300K was detectable at very low kinetics of delta-phase decomposition. 2007. JOURNAL concentrations of hydrogen, so it is clear that the ingress of OF ALLOYS AND COMPOUNDS. 444: 124. hydrogen into the Pu lattice is easy and involves an elec- tronic interaction between the two species. 3. SIEGEL, R. W.. VACANCY CONCENTRATIONS IN METALS. 1978. JOURNAL OF NUCLEAR MATERIALS. 69-7 (1-2): The system we are studying is a ternary with three com- 117. ponents: Pu, Ga, and H. We found that the effects of 4. Hecker, S. S., D. R. Harbur, and T. G. Zocco. Phase stabil- hydrogen in Pu cannot be separated from Ga. Increasing ity and phase transformations in Pu-Ga alloys. 2004. Ga in the system to 7 atomic % effectively switched off PROGRESS IN MATERIALS SCIENCE. 49 (3-4): 429. the vacancy-stabilizing ability of hydrogen. We measured no increase in vacancy concentration when Pu-7Ga was Publications charged with hydrogen. In general, Pu with high amounts Beaux, M. F., J. J. Joyce, T. Durakiewicz, K. S. Graham, E. D. of Ga behaves in a more normal way. We determined that Bauer, J. N. Mitchell, P. H. Tobash, and S. Richmond. many of the well-known [4] anomalous thermodynamic Electronic Structure, Localization and 5f Occupancy in properties of Pu-2Ga disappear at high Ga concentrations. Pu Materials. 2012. ACTINIDES AND NUCLEAR ENERGY We developed computational frameworks in this project MATERIALS. 1444: 123. with promise to help understand the anomalous effects of Joyce, J. J., T. Durakiewicz, K. S. Graham, E. D. Bauer, D. P. Ga, which will continue to influence the field well beyond Moore, J. N. Mitchell, J. A. Kennison, R. L. Martin, L. the term and scope of this project. E. Roy, and G. E. Scuseria. Pu Electronic Structure and Photoelectron Spectroscopy. 2011. In International Impact on National Missions Conference on Strongly Correlated Electron Systems The findings of this project have important ramifications (SCES 2010) ; 27 June-2 July 2010 ; Santa Fe, NM, USA. for the understanding of how plutonium in the Nation’s nu- Vol. 273, p. 012023 (5 pp.). clear weapons stockpile is aging. LANL and LLNL plutonium Joyce, J. J., T. Durakiewicz, K. S. Graham, M. F. Beaux II, E. 67

Page 68

D. Bauer, D. P. Moore, J. N. Mitchell, P. H. Tobash, and Venhaus, T.. X-ray photoelectron spectroscopy of 7-at.% S. Richmond. Delta-Pu electronic structure and va- gallium δ-stabilized plutonium. Presented at Actinides lence. Phys. Rev. Lett.. 2013, Karlsruhe, Germany. (Karlsruhe, 21-26 July, 2013). Lee, T., M. I. Baskes, A. C. Lawson, S. D. Conradson, S. P. Chen, A. Caro, S. M. Valone, and C. D. Taylor. Atomistic modeling of thermodynamic properties of Pu-Ga alloys based on Invar mechanism. Phys. Rev. B. Mitchell, J. N., F. J. Freibert, D. S. Schwartz, and T. E. Mitch- ell. Diverse phase transformation behavior in pure plu- tonium. Presented at Plutonium Futures - The Science 2012. (Cambridge UK, 15-20 July). Mitchell, J. N., F. J. Freibert, and D. S. Schwartz. Phase Transitions in Pure Plutonium. 2012. ACTINIDES AND NUCLEAR ENERGY MATERIALS. 1444: 159. Richmond, S., D. S. Schwartz, A. I. Smith, and A. L. Costello. New observations supporting the existence of the su- perabundant vacancy phenomenon in delta-Pu metal. Presented at Plutonium Futures - The Science 2012. (Cambridge UK, 15-20 July 2012). Schwartz, D. S., S. Richmond, A. I. Smith, A. Costello, and C. D. Taylor. Hydrogen-Vacancy Effects in Pu-2 at. % Ga Alloys. 2012. ACTINIDES AND NUCLEAR ENERGY MATE- RIALS. 1444: 183. Schwartz, D. S., S. Richmond, A. I. Smith, and W. D. Peach. Hydrogen effects in Pu-Ga Alloys: preliminary thermo- dynamic and microstructural analysis of Pu-7 at. % Ga. Presented at Plutonium Futures 2012. (Cambridge, UK, 15-20 July). Smith, A. I., S. Richmond, K. L. Page, J. Siewenie, and T. Saleh. First look at local structural evolution of hydro- gen loaded delta 239Pu-2at.%-Ga alloys. Presented at Plutonium Futures - The Science 2012. (Cambridge UK, 15-20 July). Taylor, C.. Surface segregation and adsorption effects of iron-technetium alloys from first-principles. 2011. Jour- nal of Nuclear Materials. 408 (2): 183. Taylor, C. D., S. C. Hernandez, M. F. Francis, D. S. Schwartz, and A. K. Ray. Hydrogen trapping in delta-Pu: insights from electronic structure calculations. 2013. JOURNAL OF PHYSICS-CONDENSED MATTER. 25 (26): -. Valone, S.. Quantum Mechanical Origins of the Iczkowski- Margrave Model of Chemical Potential. 2011. JOUR- NAL OF CHEMICAL THEORY AND COMPUTATION. 7 (7): 2253. Valone, S. M., S. R. Atlas, and M. I. Baskes. Fragment Ham- iltonian model potential for nickel: metallic character and defects in crystalline lattices. Modeling and Simu- lation in Materials Science and Engineering. 68

Page 69

Chemistry and Material Sciences Directed Research Final Report First Reactions: Simple Molecule Chemistry Behind the Shock Front Dana M. Dattelbaum 20110012DR Abstract were better understood, our predictive capabilities for Understanding the behaviors of materials in extreme the dynamic response of materials would be dramati- environments is foundational for predicting the response cally improved, and further, shock-driven reactivity could of weapons materials, creating environmentally tolerant be designed into new molecules for use under these properties, and potentially exploiting extreme conditions conditions. However, there remain significant challenges to tailor materials functionality. Furthermore, the in- to understanding and predicting chemical reactivity creasing diversity in long-term NNSA mission drivers, and under these conditions; most notably the harsh in situ needs for increasing fidelity of weapons physics models, conditions, short temporal existence of reactants and will continue to drive research into dynamic materials intermediates, and single-event nature of shock com- performance under the extreme conditions encountered pression experimentation. under shock loading. Significant gaps in our knowledge Foundational work[10-21] established that many organic remain regarding the general principles of shock-induced materials react readily under shock compression, even chemical reactions, and the specifics of bond-changing those that do not readily react under thermal or catalytic steps under these conditions. Through in situ measure- conditions such as styrene, tetrahydrofuran, benzene, ments of shock wave structures and spectroscopic sig- and tetrachloroethylene. The goals of this project are natures, this synergistic LDRD project has shed light on to predict, measure and quantify the details associated the details of chemical reactions behind the shock front with the first reactions behind a shock front for a selec- in simple chemical structures. These include reaction tion of relevant functional groups (such as O-O, O-H, thresholds/rates and their state sensitivities, as well as C=C, C≡C, C≡N, and ring structures). Theory and experi- electronic changes leading to and associated with chemi- ment are matched in length and time scales, with differ- cal bond changes, and predicted reaction mechanisms. ent shock compression platforms (gas gun-driven plate The outcome of this work is a better understanding of impact, and direct laser shock drive) providing sustained how chemical reactions proceed under shock compres- shocks on duration from 100s ps to microseconds in sion. time. The objectives of the project are to quantify the time scales of processes behind the shock front leading Background and Research Objectives to and associated with chemical bond breaking, predict Molecular bonding is defined by strong intramolecu- and measure electronic states, volume changes and tem- lar covalent bonds, and weaker intermolecular forces peratures for direct comparison to experiment, and to (hydrogen bonding, van der Waals, dipole-dipole). Ap- understand reaction activation barriers and mechanisms plication of pressure results in dramatic volume reduc- under shock compression. tions, overcoming repulsive intermolecular interactions, and potentially new chemical reaction pathways and Scientific Approach and Accomplishments products[1-4]. Today, the interrogation of new chemical A synergistic theoretical-experimental effort focusing structures at static high pressures is commonplace[1, on shock-driven reactivity of simple chemical struc- 3-4]. Shock compression creates extreme compression tures has been established to quantify the mechanisms, conditions on a rapid (ps) timescale, invoking additional length- and timescales of the processes associated with effects on local chemical structures, including unique chemical reactions behind a shock front. This approach steric and electronic structures, shock heating, and combines new computational methods in many-atom transient non-equilibrium partitioning of kinetic energy reactive molecular dynamics, with application of pico- behind the front[1,2,5,8]. If the first chemical reactions 69

Page 70

second-to-microsecond time-resolved experimental diag- 3. We have developed a novel quantum reactive mo- nostics during shock compression. In this project, we have lecular dynamics approach (LATTE) with significant focused on predicting and interrogating shock-induced improvements in energy conservation and scaling, chemical reactions in three classes of molecular com- parameterized the potentials for C, H, N, and O, and pounds: oxygen-rich molecules (H2O2, HCOOH), unsaturat- simulated shock-driven chemistry in simple molecules, ed structures (e.g. C=C, C≡C, C≡N), and ring (saturated and validated against and guiding experiments on the same unsaturated) structures. We have chosen simple chemical timescale; structures because they are computationally tractable, and 4. We have demonstrated a bridging of the temporal and in many cases, their thermodynamic properties and static spatial scales of chemical reactions behind the shock high-pressure responses are reasonably well-characterized, front through shock temperature-sensitive global reac- offering benchmarks for the dynamic experiments. They tion rates. represent functional groups present in most molecular solids, and offer an opportunity to establish a foundational Reaction thresholds and dynamics via wave evolution knowledge of comparative group reactivity. The mechanical variables (Us, up, P, V) associated with shock-induced chemical reactions, and their kinetic evolu- Shock-compression experiments were performed using tion have been quantified under sustained shock condi- both gas gun-driven and laser-driven shock compression tions using LANL large bore single- and two-stage light platforms, launching shock waves into the materials with gas guns. Reaction threshold conditions on the principal sustained durations ranging from ~350 picoseconds to sev- Hugoniot have been quantified by in situ measurements eral microseconds. Experiments and theory were matched of the reactive flow, including the first wave state on the in length and time scale to the processes occurring follow- unreacted Hugoniot.[5, 25, 26] In doing so, we have ing shock compression. At the earliest time/length scales, established an order of reactivity for a variety of chemical laser-driven shock compression with ultrafast (ps) probes functional groups and simple molecules as a function of of electronic and bonding changes[7, 28-30] were matched shock pressure on the principal Hugoniot, including –C=C-, to quantum molecular dynamics (LATTE) simulations[8, -C≡C-, -C-H, -C=O, and O-O motifs.[5, 6, 7] For example, 22-23] at the same conditions. From nanoseconds-to- unsaturated structures (such as C=C and C≡C) were found microseconds behind the front, measurement of in situ to react at modest shock input conditions (4-7 GPa). Aro- wave dynamics, and global reaction rates were interpreted matic structures were more stable, reacting at > 10 GPa, with calculated temperatures from inert and product and saturated molecules appear to be the most stable equations of state.[5,6,25,27] The dynamic experiments (including C-C, C-H, O-O). Multiple wave structures result- were benchmarked to static high pressure-temperature ing from volume changes associated with shock-driven experiments of pressure- and temperature-driven chemi- chemical reactions, provide information about reaction cal reactions using in situ synchrotron-based diagnostics thresholds, densities of intermediate and product states, of structure and bonding (x-ray diffraction, infrared spec- and global reaction rates. Figure 1a shows example shock troscopy).[24, 26] As a result of this focused LDRD effort, wave profiles (in particle velocity vs. time) of one of the the major accomplishments toward understanding shock- molecules studied, acrylonitrile (CH3C≡N) shocked to driven reactivity are: 7.6 GPa. As seen in the Figure 1a, over measurable time

1. We have quantified many of the details of shock-driven and distance, the input shock wave evolves into a 3-wave reactions for a number of simple molecular struc- structure, indicating at least 2 higher density products are tures. These include defining the timescales of induc- formed behind the shock front with volume reductions of 5 tion times, electronic structure, and bonding changes and 6% for the first and second reaction step, respectively. behind the front. We have also defined the reaction [5] The threshold or “cusp” condition for acrylonitrile was thresholds on the principal shock adiabat (Hugoniot) found to be at ~ 5.6 GPa and 850 K. for a large variety of chemical functional groups, vol- From the evolution of shock wave profiles as a function of ume changes leading to and associated with chemical varying shock input strength, reaction induction times and reactions, and global reaction rates as a function of global reaction rates for transformation of initial molecules shock input condition; to states of higher density were quantified. Figure 1(b)
2. We have developed and applied in situ spectroscopic shows example wave profiles from 5 different shock experi- diagnostics to measure electronic and structural ments on acrylonitrile above 6 GPa; shock stress is given changes in shocked organic liquids from ps-to-micro- (in GPa) next to each profile. It is observed that the wave second timescales behind the shock front.; dynamics, and thus the reaction rates, vary substantially over a small range of shock input conditions; e.g. from 6 70

Page 71

to 8.9 GPa. The risetimes of the reactive waves, Figure 1b, Reactive molecular dynamics simulations: LATTE are related to the global reaction rates of the reactions. A new reactive molecular dynamics capability based on For acrylonitrile, the global reaction rates are ~1.0-1.7 × semi-empirical methods (self-consistent field-tight binding 107 s-1 near the reaction threshold, with reactions occur- level) has been developed to allow for long-duration (ns) ring over 10s of nanoseconds at 6-8 GPa.[5] However, the molecular dynamics simulations of molecular chemistry reactions are readily accelerated by shock input condition, under simulated shock conditions.[8, 22, 23] Optimal reaching 109 s-1 by 18 GPa[5, 7]. This is due to the state scaling of computation cost with system size was achieved sensitivities of the reactions to shock pressure, and more through sparse matrix algebra methods,[8] and by porting importantly, shock temperature. Figure 2 shows the mea- the computationally-expensive algorithms to high perfor- sured and calculated reaction rates for the two reactions in mance, many-core graphics cards. These novel approaches acrylonitrile as a function of calculated shock temperature, allow for realization of simulations with the thousands of illustrating the shock state (P,T) sensitivities of the reaction atoms that are required for explicit resolution of a shock rates[5]. front, and the transient non-equilibrium phenomena it in- Probes of chemistry in the first ½ nanosecond vokes.[8] Shock simulations were performed on the same Ultrafast spectroscopic probes coupled to laser-driven molecular compounds as focused on by the experimental shock compression allows for evolution in electronic and effort, with specific focus on acrylonitrile, phenylacetylene, bonding structures behind the shock front at early times. formic acid, and benzene. Figure 4 shows the results from Shocked states were diagnosed using ultrafast dynamic LATTE simulations on acrylonitrile. The simulations in Fig- ellipsometry, and evidence of “cusps” or non-linearities in ure 4a show the computational cell before, and at simulat- the shock adiabat due to chemical reaction were observed ed shock compression at >10 GPa. Shock compression at at higher input conditions than in the gas gun-driven this P,T condition results in bond cleavage and polymeriza- experiments, due to the shorter timescales sampled by tion of acrylonitrile to form polyacrylonitrile. The simula- the laser shock experiments[7, 28-30]. Ultrafast tran- tions results, Figure 4b and 4c, show HOMO-LUMO gap sient absorption measurements were performed for the reduction following shock compression, and the number same series of simple molecular structures as described of atoms incorporated into the growing polyacrylonitrile above[7]. The transient absorption spectra reveal new product. Following shock compression, an induction time insights into electronic structure changes following shock is observed, and reduction of the HOMO-LUMO gap (by compression. For example, a continuous red shift of the over an eV) associated with closure of the chemical poten- absorption band, consistent with shock-induced band gap tial, and reaction to form the product(s). The simulations closure, was not observed for any compound studied [7]. are directly consistent with the observed densification Rather, transient absorption features in the range 400-800 reactions observed in the gas gun and laser experiments, nm were observed for molecules undergoing chemical re- and visible-centered absorption feature that appears as action, such as acrylonitrile shocked above 12 GPa, Figure reaction proceeds. 3, and phenylacetylene shocked above 10 GPa. Coincident LATTE was further extended to extract Raman spectra “on- with the growth of the visible absorption feature in phen- the-fly” from simulations for the first time in order to track ylacetylene, the transmitted shock wave measurements reaction pathways and their kinetics via coupled theory exhibited a volume collapse that occurred exclusively in and experiment. Spectra were computed by the exten- the high stress region of the shock. This is consistent with sion of the tight-binding formalism to incorporate into the a densification reaction observed in gas gun embedded Hamiltonian a dependence on an external electric field. gauge measurements of shocked phenylacetylene, but the The polarizability tensor was computed at intervals as the stress state was much higher and the reaction time much gradient of the total energy with respect to the applied faster in the laser shock experiments.[7] In addition to field and Raman intensities by the Fourier transform of its direct observation of electronic absorption changes as- time correlation function. This new capability is of tremen- sociated with compression and chemical reaction to form dous significance to objectives in understanding chemistry products, the transient absorption experiments indicate in dynamic extreme conditions, because for the first time that at modest pressures, induction times of 50-100 pico- theory and experiment will coincide in terms of temporal seconds are observed behind the shock front. An ultrafast (ps to ns) and spatial (multi-nm) scales as well as in the CARS (Coherent Anti-Stokes Raman Spectroscopy) diag- diagnostic information that can be extracted. nostic was also coupled to the laser-driven shock platform for probing bonding changes behind the front. The CARS High pressure synthesis and “dynamic” diamond anvil cell diagnostic was demonstrated with single shot sensitivity experiments on micron-thick shocked samples. Definition of the static P/T phase diagrams, and high pres- 71

Page 72

sure reactivity of a large number of simple organic mol- ecules was performed using x-ray microdiffraction tech- niques at the HP-CAT beamline at the Advanced Photon Source, and the U2A beamline at the National Synchrotron Light Source to aid in interpretation of the dynamic ex- periments. The materials studied to date include: t-butyl acetylene, ethynyl trimethylsilylacetylene, hydrazine, acrylonitrile, benzonitrile, formic acid, nitrobenzene, and aniline.[24, 26, 31] Overall, reaction thresholds of pres- sure-driven chemistry were found to be greater than under shock conditions by at least 4 GPa, due to the temperature Figure 1. (A) Particle velocity wave profiles measured in acrylo- nitrile shocked to 7.6 GPa. A three-wave structure indicative of and strain rate differences. Raising the temperatures at the formation of two higher density products is observed. (B) high pressures was found to decrease the reaction thresh- Particle velocity wave profiles from five shock compression ex- olds in pressure to be close to shock conditions. New periments on acrylonitrile at varying shock input strengths (listed polymer materials were also synthesized by high pressure next to the profile in GPa). The shock-driven reactions are quite and/or high P/T conditions in diamond anvil cells, and sensitive to the shock input condition and span a range of wave recovered to ambient conditions. Examples include the dynamics from 6.0- 8.9 GPa. glassy poly(t-butyl acetylene), and opaque poly(ethynyl tri- methylsilane) polymers.[26] Furthermore, using the new 3.5 generation synchrotron PETRA-III (Hamburg, Germany), organic materials were rapidly ramp-compressed through the liquid-solid transition and reaction (polymerization) thresholds while simultaneously measuring time-resolved full angle-dispersive x-ray diffraction patterns at strain rates up to 10-2 s-1. Impact on National Missions Predicting the evolution of chemical reactions in extreme environments underpins our ability to perform weapons assessments through predicting how materials behave under dynamic compression, and the development of realistic EOS and reactive burn models that capture shock- driven chemistry. An improved fundamental knowledge is also the only means of enabling the design of new materi- Figure 2. Measured and predicted reaction time (in nanoseconds) als. As such, the themes of this LDRD project are directly for acrylonitrile as a function of predicted initial shock tempera- aligned with Stockpile Stewardship and related National ture. The reactions are found to be highly state sensitive. Security missions. Nationally, this work can also be tied to Office of Science priority research directions in thermo- mechanical extremes, as outlined by the recent Materials under Extreme Environments report,[1] and National user facilities such as the Linear Coherent Light Source (LCLS), HP-CAT and DCS beamlines[2,3] at the Advanced Photon Source, and National Synchrotron Light Source. New diag- nostics of matter in extreme conditions, including ultrafast transient absorption and coherent anti-Stokes Raman spectroscopy, and transient Raman spectroscopies coupled to single- and two-stage light gas guns at Los Alamos, have been developed, as well as new triggering methods for timing high latency diagnostics (flash lamp-pumped lasers, accelerator-generated x-rays and protons) to shock condi- tions.[9] Time-resolved (few-100s ms) x-ray diffraction at strain rates up to 102 s-1 within high pressure diamond anvil cells was also demonstrated. 72

Page 73

2. New Research Opportunities in Dynamic Compression Science. 2012. Dynamic Compression Sector at the Advanced Photon Source, Washington State University and DOE/NNSA.
3. Advances in Matter under Extreme Conditions - HPCAT Workshop Report. 2013. High Pressure Collaborative Access Team, Advanced Photon Source, Argonne Na- tional Laboratory.
4. Hemley, R. J.. Effects of high pressure on molecules.
5. Annual Reviews of Physical Chemistry. 51: 763.
6. Dattelbaum, D. M., and S. A. Sheffield. Shock-induced chemical reactions in simple organic molecules. 2012. (Chicago, IL, 26 June - 01 July 2011). Vol. 1426p. 627. Chicago, IL: American Institute of Physics.
7. Manner, V. W., S. A. Sheffield, D. M. Dattelbaum, and D. B. Stahl. Shock compression of formic acid. 2012. (Chicago, IL, 26 June - 01 July, 2011). Vol. 1426p. 201. Chicago, IL: American Institute of Physics. Figure 3. Time-resolved transient absorption spectra (in wave-
8. Dang, N. C., C. A. Bolme, D. S. Moore, and S. D. Mc- length) as a function of time following shock compression of Grane. Shock Induced Chemistry In Liquids Studied acrylonitrile. A visible absorption is observed at shock input With Ultrafast Dynamic Ellipsometry And Visible conditions above 13 GPa associated with shock-driven reactions. Transient Absorption Spectroscopy. 2012. Journal of Physical Chemistry A. 116 (42): 10301.
9. Cawkwell, M. J., and A. M. N. Niklasson. Energy Con- serving, Linear Scaling Born-Oppenheimer Molecular Dynamics. 2012. Journal of Chemical Physics. 137:
11. Goodwin, P. M., B. Marshall, G. D. Stevens, and D. M. Dattelbaum. A Non-Invasive Method for Tracking the Position and Velocity of Gas Gun-Launched Projectiles Using External-Surface Transient Strain Measurements with Optical Fiber-Bragg Gratings. 2013. Review of Scientific Instruments. 84 (3): 035002.
12. Dremin, A. N., and L. V. Babare. The shock wave chemistry of organic substances. 1982. AIP Conference Proceedings. 78: 27. Figure 4. (A) LATTE simulation of acrylonitrile before (left) and after (right) simulated shock compression. The shock compres- 11. Dremin, A. N., and L. V. Babare. On the shock polym- sion event results in a decrease in the HOMO-LUMO gap (B) erization process. 1984. Le Journal de Physique Collo- with time, and polymerization of acrylonitrile with time (C). (C) ques. 45: C8. indicates the number of atoms in the polymer chain as a function to time to > 50 picoseconds. 12. Vakushev, V. V., S. S. Nabatov, and O. B. Yakusheva. Physical properties and conversion of acrylonitrile at References high dynamic pressures. 1974. Fizika Goreniya i Vzryva.
13. Wadsworth, J., G. Crabtree, and R. J. Hemley. Materi- 10: 583. als under Extreme Environments. 2007. Department of Energy Office of Science, Basic Energy Sciences report. 13. Sheffield, S. A., and G. E. Duvall. Response of liquid car- bon disulfide to shock compression: equation of state 73

Page 74

at normal and high densities. 1983. Journal of Chemi- 26. Chellappa, R. S., D. M. Dattelbaum, S. A. Sheffield, cal Physics. 79: 1981. and D. L. Robbins. Pressure-induced polymerization in substituted acetylenes. 2012. (Chicago, IL, 26 June-01 14. Sheffield, S. A.. Response of liquid carbon disulfide to July, 2011). Vol. 1426p. 1421. Chicago, IL: American shock compression. II. Experimental design and mea- Institute of Physics. sured Hugoniot information. 1984. Journal of Chemical Physics. 81: 3048. 27. Dattelbaum, D. M., S. A. Sheffield, J. D. Coe, M. A. Margevicious, and M. J. Cawkwell. Shock-driven chem- 15. Dick, R. D.. Shock wave compression of benzene, car- istry in phenylacetylene. to be submitted to Journal of bon bisulfide, carbon tetrachloride, and liquid nitro- Physical Chemistry. gen. 1970. Journal of Chemical Physics. 52: 6021. 28. Schulze, P. A., N. C. Dang, C. A. Bolme, K. E. Brown, and 16. Dick, R. D.. Shock compression data for liquids. I. Six S. D. McGrane. Shock hugoniot equations of state for hydrocarbon compounds. 1979. Journal of Chemical binary ideal (toluene/fluorobenzene) and non-ideal Physics. 71: 3203. (ethanol/water) liquid mixtures. 2013. Journal of Physi- cal Chemistry A. 117 (29): 6158. 17. Graham, R. A., and B. W. Dodson. Shock-induced or- ganic chemistry. 1981. APS conference on shock waves 29. McGrane, S. D., D. S. Moore, D. J. Funk, and R. L. Ra- in condensed matter. : 0. bie. Spectrally modified chirped pulse generation of sustained shock waves. 2002. Applied Physics Letters. 18. Graham, R. A., and B. W. Dodson. Bibliography on : 3919. shock-induced chemistry. 1980. Sandia National Labo- ratory Report SAND-80-1642. 30. McGrane, S. D., D. S. Moore, and D. J. Funk. Sub-pico- second shock interferometry of transparent thin films. 19. Nellis, W. J., F. H. Ree, R. J. Trainor, A. C. Mitchell, and 2003. Journal of Applied Physics. : 5063. M. B. Boslough. Equation of state and optical luminos- ity of benzene, polybutene, and polyethylene shocked 31. Manner, V. W., R. S. Chellappa, S. A. Sheffield, Z. Liu, to 210 GPa (2.1 Mbar). 1984. Journal of Chemical Phys- and D. M. Dattelbaum. High pressure far-infrared spec- ics. 80: 2789. troscopic studies of hydrogen bonding in formic acid. 2013. Applied Spectroscopy. 67 (9): 1080. 20. Sheffield, S. A., D. M. Dattelbaum, R. R. Alcon, D. L. Robbins, D. B. Stahl, and R. L. Gustavsen. Shock- Publications induced chemical reaction in organic and silicon-based Bishop, M. M., R. S. Chellappa, M. Pravica, J. D. Coe, Z. Liu, liquids. 2006. AIP Conference Proceedings. : 921. D. M. Dattelbaum, and Y. Vohra. 1,1-diamino-2,2-dini- troethylene under high pressure-temperature. 2012. 21. Sheffield, S. A., and R. R. Alcon. Shock-induced reaction Journal of Chemical Physics. 137 (17): 174304. in several liquids. 1989. American Physical Society topi- cal conference on shock compression of condensed Cawkwell, M. J., E. J. Sanville, S. M. Mniszewski, and A. M. matter. : 683. N. Niklasson. Self-Consistent Tight-binding Molecular Dynamics Simulations of Shock-Induced Reactions 22. Niklasson, A. M. N.. Extended Born-Oppenheimer mo- in Hydrocarbons. 2012. In American Physical Society lecular dynamics. 2008. Physical Review Letters. 100: Shock Compression of Condensed Matter Topical 123004. Group Meeting. (Chicago, IL, 26 June-01 July 2011). Vol. 1426, p. 1295. Chicago, IL: American Institute of 23. Odell, A., A. Delin, B. Johansson, M. J. Cawkwell, and A. Physics. M. N. Niklasson. Geometric integration in Born-Oppen- Cawkwell, M. J., E. J. Sanville, S. M. Mniszewski, and A. M. heimer Molecular Dynamics. 2011. Journal of Chemical N. Niklasson. Computing the density matrix in elec- Physics. 135: 224105. tronic structure theory on graphics processing units. 2012. Journal of Chemical Theory and Computation. 24. Chellappa, R., and D. M. Dattelbaum. Hydrogen bond- 2012 (8): 4094. ing in hydrazine to 20 GPa. Journal of Physical Chemis- try. Cawkwell, M. J., and A. M. N. Niklasson. Energy Conserv- ing, Linear Scaling Born-Oppenheimer Molecular 25. Dattelbaum, D. M., S. A. Sheffield, J. D. Coe, and M. A. Dynamics. 2012. Journal of Chemical Physics. 137: Margevicius. Shock-induced chemistry of phenylacety- 134105. lene. J. Phys.: Conf. Series. 74

Page 75

Chellappa, R. S., D. M. Dattelbaum, N. Velisavljevic, and S. Position and Velocity of Gas Gun-Launched Projectiles A. Sheffield. The phase diagram of ammonium nitrate. Using External-Surface Transient Strain Measurements 2012. Journal of Chemical Physics. 137: 064504. with Optical Fiber-Bragg Gratings. 2013. Review of Sci- entific Instruments. 84 (3): 035002. Chellappa, R. S., D. M. Dattelbaum, S. A. Sheffield, and D. L. Robbins. Pressure-induced polymerization in sub- Manner, V. W., R. S. Chellappa, S. A. Sheffield, Z. Liu, and stituted acetylenes. 2012. In American Physical Soci- D. M. Dattelbaum. High pressure far-infrared spec- ety Shock Compression of Condensed Matter Topical troscopic studies of hydrogen bonding in formic acid. Group meeting. (Chicago, IL, 26 June-01 July, 2011). 2013. Applied Spectroscopy. 67 (9): 1080. Vol. 1426, p. 1421. Chicago, IL: American Institute of Physics. Manner, V. W., S. A. Sheffield, D. M. Dattelbaum, and D. B. Stahl. Shock compression of formic acid. 2012. In Chellappa, R., and D. M. Dattelbaum. Hydrogen bonding in American Physical Society Shock Compression of Con- hydrazine to 20 GPa. Journal of Physical Chemistry. densed Matter Topical Group meeting. (Chicago, IL, 26 June - 01 July, 2011). Vol. 1426, p. 201. Chicago, IL: Dang, N. C., C. A. Bolme, D. S. Moore, and S. D. McGrane. American Institute of Physics. Shock Induced Chemistry In Liquids Studied With Ul- trafast Dynamic Ellipsometry And Visible Transient Niklasson, A. M. N., and M. J. Cawkwell. Fast Quantum Absorption Spectroscopy. 2012. Journal of Physical Molecular Dynamics. 2012. Physics Review B. 86 (17): Chemistry A. 116 (42): 10301. 174308. Dang, N. C., C. A. Bolme, D. S. Moore, and S. D. McGrane. Odell, A., A. Delin, B. Johansson, M. J. Cawkwell, and A. M. Temperature measurements in condensed phases us- N. Niklasson. Geometric integration in Born-Oppen- ing non-resonant femtosecond stimulated Raman scat- heimer Molecular Dynamics. 2011. Journal of Chemical tering. 2013. Journal of Raman Spectroscopy. 44 (3): Physics. 135: 224105. 433. Schulze, P. A., N. C. Dang, C. A. Bolme, K. E. Brown, and Dattelbaum, D. M., R. S. Chellappa, P. R. Bowden, J. D. Coe, S. D. McGrane. Shock hugoniot equations of state for and M. A. Margevicious. Chemical stability of molten binary ideal (toluene/fluorobenzene) and non-ideal 2,4,6-trinitrotoluene at high pressure. Applied Physics (ethanol/water) liquid mixtures. 2013. Journal of Physi- Letters. cal Chemistry A. 117 (29): 6158. Dattelbaum, D. M., S. A. Sheffield, J. D. Coe, M. A. Margevi- cious, and M. J. Cawkwell. Shock-driven chemistry in phenylacetylene. to be submitted to Journal of Physical Chemistry. Dattelbaum, D. M., S. A. Sheffield, J. D. Coe, and M. A. Mar- gevicius. Shock-induced chemistry of phenylacetylene. J. Phys.: Conf. Series. Dattelbaum, D. M., and S. A. Sheffield. Shock-induced chemical reactions in simple organic molecules. 2012. In American Physical Society Shock Compression of Condensed Matter Topical Group meeting. (Chicago, IL, 26 June - 01 July 2011). Vol. 1426, p. 627. Chicago, IL: American Institute of Physics. Davidson, A., R. Chellappa, D. M. Dattelbaum, and C. S. Yoo. Pressure-induced isostructural metastable phase transition of ammonium nitrate. 2011. Journal of Phys- ical Chemistry. 115: 11889. Davidson, A., R. Dias, C. S. Yoo, and D. M. Dattelbaum. “Stubborn” Triaminotrinitrobenzene (TATB): Unusually High Chemical Stability of a Molecular Solid to 150 GPa . 2011. J. Chem. Phys.. 135: 174507. Goodwin, P. M., B. Marshall, G. D. Stevens, and D. M. Dat- telbaum. A Non-Invasive Method for Tracking the 75

Page 76

Chemistry and Material Sciences Directed Research Final Report Innovative and Validated Sub-micron to Meso-scale Modeling of the Evolution of Interface Structure and Properties under Extreme Strains Irene J. Beyerlein 20110029DR Abstract the fraction of grain boundaries can lead to several-fold Nanocomposite materials are an exciting class of ma- increases in strength [2,3]. Impressive though this result terials that have proven to be ten times stronger, more may be, the grain boundaries created by mechanical corrosion and radiation resistant than traditional metals processing are disordered (high energy, tangled net- [1,2]. These are bulk-sized composites comprised inter- works of defects) [3] and unstable in high temperatures. nally of nano-sized metallic crystals and a high density Consequently, they do not retain their superior strength of bi-metal interfaces. As materials for next generation in high temperature. applications are expected to push the envelope and Our objective in this project was to develop a multi-scale exhibit predictable lifetime performance in extreme model for processing of two-phase composites. The environments for longer service-lives than are attainable model aimed to predict the processing route that pro- with today’s materials, nanostructured composite metals vides for the ‘right’ interfaces—those that are stable in have become an attractive candidate. strain and at high temperature. Unlike prior work using Designing nanostructural materials for operation in SPD on one metal, we applied a SPD process to a com- extreme environments for future energy and defense posite consisting of two dissimilar metals. The idea is applications requires the development of a predictive that compared to grain boundaries, the bimetal interfac- capability for both material synthesis and mechanical es would be much more stable under elevated tempera- performance. In this program, we developed innova- tures and when spaced several nanometers apart, would tive synthesis and multi-scale modeling techniques for be effective barriers to dislocation motion. Ultimately, two-phase metal composites. Both techniques will for the model designs the interfaces needed for creating a the first time permit development of scientifically guided strong and thermally stable nanostructured metal. synthesis routes needed for a targeted set of desired Scientific Approach and Accomplishments interfacial properties. Our scientific approach consisted of developing a novel Background and Research Objectives processing technique for making bulk nanomaterials and Metals need to be far stronger and more thermally a multi-scale processing model for designing the routes stable than those in use today in order to have a pro- taken by this technique. found impact on energy efficiency in future automotive, Regarding this first, we fabricated Cu-Nb nanocomposite aircraft, power generation, and nuclear and non-nuclear materials in bulk form (>cm3) with a processing tech- military and defense applications. A promising way for- nique called accumulative roll bonding (ARB) [2,4]. To ward has been to shrink the grain size of metals down to prevent the introduction of oxides to the interfaces, a nanoscale dimensions. specially designed ARB process was used [2]. Our ARB Nanostructuring via severe plastic deformation (SPD) has process delivers two-phased (Cu-Nb) samples with become a popular way of fabricating nanomaterials in controllable layer thicknesses from submicron to the structural size scales [3,4]. This grain refining technique nanoscale (down to 10 nm) [5,6]. Figure 1 shows many is a top-down process that can transform a traditional typical nanolayered microstructures manufactured by coarse-grained metal to a nanocrystalline metal with- this technique. The samples shown differ in the spac- out changing the original dimensions of the sample [3]. ing between the Cu-Nb interfaces [5-7]. The size of the Experiments have shown that the resulting increases in samples from which these micrographs were taken was 76

Page 77

several orders of magnitude larger (cm3). Therefore, the metals in sizes suitable for car bodies and aircraft and ARB process combines an unprecedented level of control structures in nuclear reactors. of nm-scale structure with the ability to fabricate large Accomplishment #2: The evolution to a predominant volumes of material. interface seen above is not anticipated. It is not expected For the second technique, we developed the first two- that an SPD process can actually lead to ordered interfac- phase composite crystal plasticity finite element model es. We hypothesized that the interface characters that are [8,9]. Displacement and traction continuity are enforced mechanically stable (that is preferred in SPD) are associat- at the interface throughout deformation. In this way, ed with plastically stable orientations on both sides of the the interface provides a ‘kinematic constraint’ on the interface and low formation energy interfaces. These two deformation of the two adjoining crystals, just as in the things need to be satisfied concurrently, plastic stability in experiment. While the total deformation state is two-di- both orientations and creation of low energy interfaces, mensional (i.e., plane strain condition of the ARB process), which imposes a severe constraint on which interfaces will the deformation of the crystals is fully three-dimensional be stable. (out-of-plane plastic deformation is allowed). The consti- To study the plastic stability of interfaces during rolling tutive laws used for both Cu and Nb follow an anisotropic of Cu-Nb, our multi-scale model was employed to model elastic and rate-dependent crystal plasticity formulation, texture evolution during rolling [9]. An elastically aniso- which is given in [9]. Plastic deformation is accommodated tropic and rate-dependent plasticity law was employed by dislocation glide and normal slip is assumed, wherein for both Cu and Nb [9]. Cu was assumed to deform on the slip system activation is correlated with the resolved shear \{111\}<110> slip systems and Nb on both the \{110\}<111> stress. Our model employs slip system flow kinetics based and \{112\}<111> slip systems. on the notion of thermally activated dislocation glide. In this classical scenario, dislocations glide from one pinning For validation, we first tested the model on traditional point to another and use a combination of thermal activa- composites, containing relatively thick micron scale Cu and tion and mechanical stress to overcome each pinning site. Nb layers, where several grains span an individual layer, In the Cu-Nb bicrystal simulations that follow, we make measured textures corresponded to the theoretical tex- available the 12 \{111\}<110> slip systems in Cu and the 12 tures of rolling Cu or Nb alone [4]. The CPFE model [9] in \{110\}<111> slip systems and 12 \{112\}<111> slip systems in Figure 3 was able to effectively predict this texture devel- Nb [28]. The thermal activation slip rate model introduces opment in both the Cu and Nb phases within a two-phase material parameters for each allowable slip family in Cu Cu-Nb multilayer, containing only a few grains through- and in Nb. The calibrated material parameters are given in thickness. More details on the model development and [9] and were determined by fitting to a large suite of tests experiments used for input and validation can be found in with widely varying temperature and strain rate. [9]. Together with these two techniques we made the follow- Once validated, we applied the model to the nanolayered ing three accomplishments: Cu-Nb composite comprised of alternating single crystal- line layers [8]. Consistent with experiment, the interfaces Accomplishment #1: We discovered that nanocomposites were enforced to maintain compatibility at all times dur- with stable interfaces can evolve from extreme mechani- ing deformation. Using this model geometry, the rolling cal deformation. The textures of these nanocomposites stability of many interface characters was tested [8] and were measured by neutron diffraction at LANL’s LANSCE of these a handful were found to be mechanically stable facility [4]. The remarkable findings were that the textures and maintain their character to within 10 degrees. Spe- were unusually sharp and significantly different from the cifically, for the \{112\}fcc||\{112\}bcc KS interface the Cu theoretical rolling textures expected of Cu or Nb alone. We crystal tilted ~6º and the Nb ~3º about the TD. As con- further analyzed via TEM the nanocomposite interfaces at firmation, deviations of up to 10º from \{112\}Cu||\{112\} different stages in the SPD process [10]. Remarkably, in Nb KS interfaces have been reported in the actual sam- spite of exposure to extreme straining, the same interface ples [11,12]. The most stable interface tested was the prevailed over the entire sample and exhibited a regular \{110\}<001>Cu||\{112\}<110>Nb interface, which begins structure at the atomic level (Figure 2). We also proved to emerge at the finest scales, 10 nm [10] (Figure 4). For that the interfaces were stable up to 600 C, which is half these same interfaces, we calculated the corresponding the melting temperature of Cu [10]. Our demonstration formation energy using molecular dynamics. In plotting that nanocomposites with stable interfaces can evolve on a map the formation energy and plasticity stability for from extreme mechanical deformation promotes devel- all interfaces, see Figure 5, we found that the most stable opment of innovative manufacturing methods for novel 77

Page 78

interfaces seen experimentally lay in a ‘regime’ of stability and Technology very recently issued the “Report to the with low formation energy and high plastic stability. This President on Capturing Domestic Competitive Advantage newly discovered “regime of mechanical interface stabil- in Advanced Manufacturing”. Our research is aligned with ity” can be explored to design synthesis techniques for their recommendation to increase R&D funding in “Ad- nanocomposites vanced Materials Design, Synthesis and Processing” and “Nanomanufacturing” for ultra-strong, lightweight mate- Accomplishment #3: Based on the two foregoing accom- rials that could have future implications for private busi- plishments, we were able to further demonstrate that ness and address manufacturing challenges in the private it is possible to achieve atomic-level tuning of material sector. Additionally, this research more than ever answers interfaces via macro-level control of the manufacturing the recent 2012 DOE-BES report entitled “From Quanta process. Combining our new modeling tools and synthesis to the Continuum: Opportunities for the Mesoscale” and techniques, we predicted that either atomically faceted their priority research directions is “Mastering Defect or atomically flat interfaces can be successfully created Mesostructure and its Evolution” with science that controls in bulk nano-laminar Cu-Nb composites via severe plas- meso-scale behavior by understanding defect structures tic deformation (SPD). As mentioned earlier, this would and their underlying mechanisms. not have been thought possible before since usually SPD techniques leave internal interfaces thermally unstable and structurally disordered. We confirmed our hypoth- esis by demonstrating in the laboratory that interfaces can be tailored to be atomically regular in structure. We then carried out mechanical and thermal stability tests to prove that the nanostructured materials we designed possessed a combination of superior strength and thermal stability. This third accomplishment implies that improve- ment in thermal stability of the material can be achieved by optimizing the manufacturing processes. The ability to alter the manufacturing route to create two significantly different interfaces that produce outstanding performance is unprecedented. Impact on National Missions Figure 1. Multilayered Cu-Nb composite fabricated by our spe- The results of our project will have the following impacts cially designed accumulative roll bonding technique. on national missions. First, our demonstration that interfaces can be tuned via extreme mechanical deformation promotes development of innovative manufacturing methods for novel metals eventually in sizes suitable for car bodies and aircraft. Second, the knowledge and modeling capability provided by this program will enable design and synthesis of a new class of advanced materials by methods that are no longer prohibitively expensive or incapable of producing sufficient material quantities. Third, the predictive tools developed will accelerate the Figure 2. (a) high resolution transmission electron micrograph of development of interface-dominant materials by reduc- the \{112\}fcc||\{112\}bcc <110>fcc||<111>bcc interface prevailing ing the number of experimental iterations and laboratory in the ARB Cu-Nb nanolayered composites and a (b) schematic of tests required to bring them into practice, impacting a this interface. wide range of industries: automotive, aircraft and power generation turbines, and both nuclear and non-nuclear military and defense applications. Fourth, the President’s Council of Advisors for Science 78

Page 79

References

1. Misra, A., and R. G. Hoagland. Effects of elevated temperature annealing on the structure and hardness of copper/niobium nanolayered films. 2005. Journal of Materials Research. 20: 2046.
2. Hansen, B. L., J. S. Carpenter, S. D. Sintay, C. A. Bronk- horst, R. J. McCabe, J. R. Mayeur, H. M. Mourad, I. J. Beyerlein, N. A. Mara, S. R. Chen, and G. T. Gray. Mod- eling the texture evolution of Cu/Nb layered compos- ites during rolling. 2013. INTERNATIONAL JOURNAL OF Figure 3. Crystal plasticity finite element calculation of Cu-Nb PLASTICITY. 49: 71. multilayer composite in rolling (a) Inhomogeneous deformation fields during deformation. Calculation shown assumes that the 3. Valiev, R. Z., and T. G. Langdon. Principles of equal- Cu-Nb interfaces are kinematically constrained to remain bonded channel angular pressing as a processing tool for grain during deformation and (b) agreement with measurements of refinement. 2006. PROGRESS IN MATERIALS SCIENCE. texture development. 51 (7): 881.
3. Carpenter, J. S., S. C. Vogel, J. E. LeDonne, D. L. Ham- mon, I. J. Beyerlein, and N. A. Mara. Bulk texture evolution of Cu-Nb nanolamellar composites during accumulative roll bonding. 2012. ACTA MATERIALIA. 60 (4): 1576.
4. Zheng, S. J., I. J. Beyerlein, J. Wang, J. S. Carpenter, W. Z. Han, and N. A. Mara. Deformation twinning mecha- nisms from bimetal interfaces as revealed by in situ straining in the TEM. 2012. ACTA MATERIALIA. 60 (16):

Figure 4. Most stable interface predicted and made by our pro- 6. Carpenter, J. S., R. J. McCabe, I. J. Beyerlein, T. A. gram Wynn, and N. A. Mara. A wedge-mounting technique for nanoscale electron backscatter diffraction. 2013. JOURNAL OF APPLIED PHYSICS. 113 (9): -. 7. Beyerlein, I. J., N. A. Mara, J. S. Carpenter, T. Nizolek, W. M. Mook, T. A. Wynn, R. J. McCabe, J. R. Mayeur, K. Kang, S. J. Zheng, J. Wang, and T. M. Pollock. Interface- driven microstructure development and ultra high strength of bulk nanostructured Cu-Nb multilayers fab- ricated by severe plastic deformation. 2013. JOURNAL OF MATERIALS RESEARCH. 28 (13): 1799. 8. Mayeur, J. R., I. J. Beyerlein, C. A. Bronkhorst, H. M. Mourad, and B. L. Hansen. A crystal plasticity study of heterophase interface character stability of Cu/Nb bicrystals. 2013. INTERNATIONAL JOURNAL OF PLAS- TICITY. 48: 72. 9. Hansen, B. L., J. S. Carpenter, S. D. Sintay, C. A. Bronk- Figure 5. Map plotting a measure of plastic stability against horst, R. J. McCabe, J. R. Mayeur, H. M. Mourad, I. J. interface formation energy. The most stable interfaces are found Beyerlein, N. A. Mara, S. R. Chen, and G. T. Gray. Mod- in the top left hand corner of the map. In agreement, interfaces eling the texture evolution of Cu/Nb layered compos- produced by our synthesis technqiue fall in this regime of the ites during rolling. 2013. INTERNATIONAL JOURNAL OF map. PLASTICITY. 49: 71. 79

Page 80

10. Zheng, S. J., I. J. Beyerlein, J. S. Carpenter, K. W. Kang, Chu, H. J., E. Pan, J. Wang, and I. J. Beyerlein. Three-dimen- J. Wang, W. Z. Han, and N. A. Mara. High-strength and sional elastic displacements induced by a dislocation of thermally stable bulk nanolayered composites due to polygonal shape in anisotropic elastic crystals . 2011. International Journal of Soilds and Structures. 48: twin-induced interfaces. 2013. NATURE COMMUNICA-

TIONS. 4: -. Chu, H. J., E. Pan, X. Han, J. Wang, and I. J. Beyerlein. Elastic 11. Kang, K., J. Wang, and I. J. Beyerlein. Atomic structure fields of dislocation loops in three-dimensional aniso- variations of mechanically stable fcc-bcc interfaces. tropic bi-materials. 2012. Journal of the Mechanics and 2012. JOURNAL OF APPLIED PHYSICS. 111 (5): -. Physics of Solids. 60: 418. 12. Carpenter, J. S., X. Liu, A. Darbal, N. T. Nuhfer, R. J. Chu, H. J., J. Wang, C. Z. Zhou, and I. J. Beyerlein. Misfit McCabe, S. C. Vogel, J. E. LeDonne, A. D. Rollett, K. Strain Relaxation Mechanisms in Core/Shell Nanow- Barmak, I. J. Beyerlein, and N. A. Mara. A comparison ires. 2013. Journal of Materials. 64: 1227. of texture results obtained using precession electron Chu, H. J., J. Wang, C. Zhou, and I. J. Beyerlein. Self-energy diffraction and neutron diffraction methods at dimin- of elliptical dislocation loops in anisotropic crystals and ishing length scales in ordered bimetallic nanolamellar its application for defect-free core/shell nanowires. composites. 2012. SCRIPTA MATERIALIA. 67 (4): 336. 2011. Acta Materialia. 59: 7114. Publications Chu, H. J., J. Wang, I. J. Beyerlein, and E. Pan. Anisotropic Al-Maharbi, M., I. Karaman, I. J. Beyerlein, D. Foley, K. T. Bimaterial Models Incorporating Atomistic-scale Phys- Hartwig, L. J. Kecskes, and S. N. Mathaudhu. Micro- ics for the Interfaces-Dislocation Interactions using structure, crystallographic texture, and plastic anisot- Green’s Function Method. 2013. International Journal ropy evolution in an Mg alloy during equal channel of Plasticity. 41: 1. angular extrusion processing. 2011. Materials Science and Engineering A. 528 (25-26): 7616. Chu, H. J., J. Wang, and I. J. Beyerlein. Anomalous reactions of a supersonic coplanar dislocation dipole: bypass or Beyerlein, I. J., N. A. Mara, J. Wang, J. S. Carpenter, S. J. twinning?. 2012. Scripta Materialia. 67: 69. Zheng, W. Z. Han, R. F. Zhang, K. Kang, T. Nizolek, and T. M. Pollock. Structure-Property-Functionality of Bi- Hansen, B. L., J. S. Carpenter, S. D. Sintay, C. A. Bronkhorst, metal Interfaces. 2012. Journal of Materials. 64: 1192. R. J. McCabe, J. R. Mayeur, H. M. Mourad, I. J. Beyer- lein, N. A. Mara, S. R. Chen, and G. T. Gray III. Modeling Carpenter, J. S., S. C. Vogel, J. LeDonne, D. L. Hammon, I. the Texture Evolution of Cu/Nb Layered Composites J. Beyerlein, and N. A. Mara. Bulk texture evolution of during Rolling . 2013. International Journal of Plasticity. Cu-Nb nanolamellar composites during accumulative 49: 71. roll bonding. 2012. Acta Materialia. 60: 1576. Hunter, A., I. J. Beyerlein, T. C. Germann, and M. Koslowski. Carpenter, J. S., X. Liu, A. Darbal, N. T. Nuhfer, R. J. McCabe, Influence of the stacking fault energy surface on ex- S. C. Vogel, J. E. LeDonne, A. D. Rollett, K. Barmak, I. tended partials with a 3D phase field dislocations mod- J. Beyerlein, and N. A. Mara. A comparison of texture el. 2011. Physical Review B. 84 (14): 144108. results obtained using precession electron diffraction and neutron diffraction methods at diminishing length Kang, K., J. Wang, S. J. Zheng, and I. J. Beyerlein. Minimum scales in ordered bimetallic nanolamellar composites. energy structures of faceted, incoherent interfaces. 2012. Scripta Materialia. 67: 336. 2012. Journal of Applied Physics. 112: 073501. Chen, X., I. J. Beyerlein, and L. C. Brinson. Bridged crack Kang, K., J. Wang, and I. J. Beyerlein. Atomic structure vari- models for the toughness of composites reinforced ations of mechanically stable interfaces. 2012. Journal with curved nanotubes. 2011. Journal of the Mechan- of Applied Physics. 111: 053531. ics and Physics of Solids. 59 (9): 1938. Lee, S. B., J. E. LeDonne, S. C. V. Lim, I. J. Beyerlein, and A. Chu, H. J., C. Zhou, J. Wang, and I. J. Beyerlein. An analyti- D. Rollett. The five-parameter Heterophase Interface cal model for the critical shell thickness in core/shell Character Distribution (HICD) of physical vapor-depos- nanowires based on crystallographic slip. 2012. Journal ited and accumulative roll-bonded Cu-Nb multilayer of the Mechanics and Physics of Solids. 60: 418–431. composites. 2012. Acta Materialia. 60: 1747. Chu, H. J., E. Pan, J. Wang, and I. J. Beyerlein. Three-di- Mara, N. A., I. J. Beyerlein, J. S. Carpenter, and J. Wang. mensional elastic fields due to a dislocation loop in an Interfacially driven deformation twinning in bulk Cu-Ag anisotropic bi-material . 2012. Journal of Applied Me- composites. . 2012. Journal of Materials. 64: 1218. chanics. 79: 021011. 80

Page 81

Mayeur, J. R., I. J. Beyerlein, C. A. Bronkhorst, H. M. Mou- rad, and B. L. Hansen. A crystal plasticity study of inter- facial stability of CuNb bicrystals. 2013. International Journal of Plasticity. 48: 72. Wang, J., I. J. Beyerlein, N. A. Mara, and D. Bhattacharyya. Interface-facilitated deformation twinning in copper within submicron Ag-Cu multilayered composites. 2011. Scripta Materialia. 64: 1083. Wang, J., K. Kang, R. F. Zhang, S. J. Zheng, I. J. Beyerlein, and N. A. Mara. Structure and property of interfaces in ARB Cu/Nb laminated composites. 2012. Journal of Materials. 64: 1208. Xu, X. F., K. Hu, I. J. Beyerlein, and G. Deodatis. Statistical strength of hierarchical carbon nanotube composites. 2011. International Journal for Uncertainty Quantifica- tion. 1 (4): 279. Zhou, C., I. J. Beyerlein, and R. LeSar. Plastic deformation mechanisms of FCC single crystals at small scales. 2011. Acta Materialia. 59 (20): 7673. 81

Page 82

Chemistry and Material Sciences Directed Research Final Report Next Generation Ionic Liquids for Plutonium Science, Separation, and Production George S. Goff 20110083DR Abstract tible to radiation damage. Pyrochemical processing was Most experts agree that nuclear energy will continue to developed during the 1940-50s to produce Pu metal for play a significant role throughout the 20th century and weapons applications. Later, this technology was inves- beyond, yet the key to sustainable nuclear power lies tigated for use in partitioning the actinides from metal in the proper management of nuclear waste. Whether reactor fuels, and is one of the three production-scale looking at the direct disposal of used nuclear fuel (UNF) techniques currently used at the Plutonium Facility at or reprocessing and recycling of U (and/or Pu and the TA-55 for the production of plutonium metal [1]. Pyro- minor actinides), disposition and stability of Pu waste processing of nuclear materials generally involves dis- forms, formation of advanced reactor fuels, or the puri- solution into molten salts (generally a mixture of alkali fication of Pu for weapons applications, understanding or alkaline earth chlorides or fluorides) at temperatures the underlying chemistry of Pu is vital for an accurate above 500°C, followed by selective recovery of actinides evaluation of nuclear fuel cycle options and responsible by electroreduction at the cathode [2]. Several disad- stewardship of our existing stockpile of nuclear weap- vantages have prevented pyroprocessing from replacing ons. PUREX including high temperature operations, batch op- eration, highly corrosive solvents, and lower decontami- Ionic liquids (ILs) are low-melting salts, often contain- nation factors [2]. More recently, ILs have been proposed ing an asymmetrical organic cation and an organic or as alternative solvents for liquid-liquid extractions [3-6] inorganic anion. There is an emerging interest in ILs for and as electrolyte solutions for the electrodeposition of potential applications in the separation of actinides from transition metals and actinides including reprocessing UNF, electrochemical separation of actinide ions, and of used nuclear fuel [5, 7, 8]. Identifying ILs with suf- solid-state coordination studies of the actinides. Due to ficient radiation stability, solubility, and electrochemical the wide range of tunable physical properties, ILs have windows for performing U and Pu electroplating would the potential to stabilize unusual oxidation states and revolutionize conventional pyroprocessing technologies. unique structural motifs not readily accessible in con- ventional aqueous media. The chemistry of Pu in ionic Nearly all early studies of actinides in ILs used the first liquids remains a highly undeveloped field, with the generation haloaluminate ILs [5, 9, 10]. These ILs gen- potential to discover revolutionary separation schemes erally have a low solubility for U, and electrochemical to displace currently used separation technologies. Our windows that are insufficient to electroplate either U or research program combined experimental measure- Np metals, and thus are not suitable for process appli- ments and theoretical modeling to develop the requi- cations.[5] Since the discovery of water-stable second site structural, thermodynamic, and kinetic parameters generation ILs in the mid 1990s, ILs have been studied needed to evaluate the feasibility of using ionic liquids to be used as solvents for separations, catalysis, and in for separating plutonium from a variety of matrices. electrolytic processes, due to their negligible vapor pres- sure, non-flammability, low corrosivity, and wide range Background and Research Objectives of thermal stability and electrochemical windows (up to Conventional processing of used nuclear fuel is based 6 V compared to 2-3 V for water) [11]. on the 60-year-old, aqueous/organic PUREX process to In contrast to aqueous and conventional organic solu- partition the actinides and fission products. This process tions, solvation of metal cations is a critical process in ILs is highly complex, has a large facility footprint, and em- that determines the effective charge and thermodynam- ploys organic diluents and extractants which are suscep- 82

Page 83

ic properties of the cationic metal complex. Model predic- Synthesis and Characterization of Novel ILs and Actinide tions indicate solvation of the metal cations occurs via the Compounds: This project has synthesized and character- formation of subsequent alternating shells of IL-anions and ized a number of novel ILs containing functional groups to cations, but there is currently no experimental validation complex actinides. One approach was to incorporate into of these results. Solvation mechanisms have been studied the structure of the ionic liquids ligands that have tradi- by ab initio molecular dynamics simulations for U and Eu, tionally been used as actinide or lanthanide separating but no studies have been performed for the transuranium agents. Oxalate has long been used as a precipitating agent elements. for Pu, and a number of ILs containing the oxalate anion were synthesized (manuscript in preparation). These ILs Studies on the radiation stability of ILs remain rare and had very high viscosities, making them impractical for most radiolysis products for most ILs are not well known. An applications, which was largely overcome by decreasing early study with 1,3-alkylmethylimidazolium chloride (C2, the size of the cation or protonating the oxalate anion to C4, and C6mmim) and nitrate showed a 15 times higher the monovalent bioxalate anion. Alpha-hydroxyisobutyric stability than tributyl phosphate/kerosene solvents (used acid (HIBA) is a characteristic human metabolite which in PUREX) when subjected to alpha and gamma radia- has long been used to separate individual actinides and tion [12]. More recently, imidazolium, pyridininium, and lanthanides from complex mixtures. Despite being used quaternary ammonium ILs showed less than 1% decompo- in lanthanide separations for over 55 years, prior to our sition when exposed to gamma-radiation dose equivalent work there was only one reported crystal structure for to 1 year of processing used nuclear fuel [13]. While the lanthanide-HIBA complex. We were able to isolate lantha- bulk IL showed strong radiation stability, all ILs investigated nide-HIBA complexes across the lanthanide series which showed an increase in the UV-vis absorption spectra and were characterized by both single crystal and powder X-ray a narrowing of the electrochemical window. Data in the diffraction studies [14]. ILs were synthesized using HIBA literature are insufficient to understand radiolysis mecha- as the anion, and also tethered to a quaternary amine in nisms or predict stability for a given structure. the cation. Preliminary results indicate coordination of the actinides to the HIBA group, which would allow us to tune Scientific Approach and Accomplishments whether the metal center is incorporated into the cation or This LDRD project focused on understanding the funda- the anion of the IL. mental chemical and physical phenomena such as solva- tion, solubility, electronic structure of f-element complex- A number of other novel ILs were synthesized which con- es, and electrochemical reduction and plating processes. tain a carboxyl group in the cation. Most previous work Specifically, this project focused on identifying ionic liquids was focused on ammonium-based cations, which gave suitable for electrochemically separating Pu from the other fairly narrow electrochemical windows and possess melt- light actinides (Th, U, Np, Am, Cm) as well as the light ing points above room temperature. We improved upon actinides from the lanthanides. We employed a multi- these ILs by changing the nitrogen center in the cation to faceted approach to study the structural, thermodynamic, a phosphorous center [15]. The carboxyl-functionalized and kinetic behavior of Pu in ILs in order to demonstrate phosphonium ionic liquid (IL), [HCTMP][Tf2N], enabled the feasibility of using ILs for advanced Pu separations. Our the directed nucleation of either monomeric or dimeric approach combined single-crystal X-ray diffraction stud- uranyl(VI) compounds, shown in Figure 1. This new IL is ies with a variety of spectroscopic tools (UV-vis-NIR, FT-IR, the first carboxyl-functionalized IL which is liquid at room Raman, NMR) to probe the behavior of actinides in ILs. Our temperature and exhibits a wider electrochemical window efforts utilized both conventional ILs as well as functional- and lower melting point than its ammonium analogue. ized ILs that incorporated metal centers into the structure Melting points for the monomeric and dimeric uranium of the IL to improve solubility and tune the electrochemi- compounds were both below 100°C, making these the first cal window for separations. Molecular modeling was confirmed uranium-based ILs reported in the literature. integrated with experiments by using measured results to Molecular Dynamics of Actinides in ILs: Molecular dy- validate calculations, and calculations were used to identify namics (MD) simulations of actinide-containing materials appropriate IL candidates for experimental studies. Quan- can provide valuable information about the interactions tum mechanical calculations enabled us to understand the between actinides and ionic liquids which can be difficult stability, solvation, redox behavior, spectroscopic signa- to obtain experimentally. MD simulations can directly com- tures, and electronic structures of actinide complexes. The pute properties such as viscosities and electrical conductiv- following sections briefly describe some of the scientific ities, phase equilibria (liquid-liquid or solid-liquid), and the highlights of our work. partitioning of ions between different phases. The accuracy 83

Page 84

of these simulations depends solely on the ability of the ILs for a wide range of applications beyond Pu separations, force field (model interaction potential) to mimic the true including other areas of interest such as lithium batteries interaction potential of the system. The number of force and fuel cells. fields available to model molecular interactions in common Electrochemistry of Actinides and Lanthanides in ILs: actinide systems is minuscule to non–existent and most Electrochemical studies performed under this LDRD project were developed and optimized to study solid uranium ox- had several specific goals including identifying ILs with ides. In collaboration with the University of Notre Dame, a sufficient electrochemical windows to allow for deposi- set of classical potentials was developed to model interac- tion of actinides, understanding how to manipulate ILs to tions with water with uranium, neptunium, plutonium, and extend the ECWs to enable electroplating, and identifying americium in the +V and +VI oxidation states [16]. These minimum potentials and temperatures for An plating from are the first reported potentials for Np, Pu, and Am. These these ILs. We have identified a number of ILs with suffi- potentials were utilized in MD simulations to predict the ciently broad electrochemical windows to enable electro- interactions of uranyl(V/VI) and plutonyl(V/VI) with wa- recovery of actinides. Most studies were performed using ter in ionic liquids,[17] and the results were validated by the bistriflimide anion, which has exceptional (electro) spectroscopic measurements. Figure 2 shows the compe- chemical stability and only weakly coordinates the ac- tition between water and the anion of the IL to complex tinide centers. Cyclic voltammetry studies for U and Pu in the actinide over a range of water concentrations. This piperidinium- and imidazolium-based ILs reveal a series of project established the ability to perform MD simulations ill-defined reduction peaks and quasi-reversible stripping on actinides in ILs, as well as to successfully predict the peaks. Bulk deposition experiments indicate that both U speciation of an actinide in a complex multi-component and Pu can be recovered from ILs. Under most conditions environment for the first time. the bulk deposits, such as those shown in Figure 3, are Prediction of Electrochemical Windows of ILs: At the amorphous. Annealing the samples improves the crystal- beginning of this project, computational methods existed linity, and X-ray diffraction and SEM/EDX characterization to predict certain physical properties of ILs, but there reveal this deposit to be UO2. Studies with lanthanides, were no methods to predict the electrochemical stability. which are important fission products in used nuclear fuel, The electrochemical window corresponds to the range of show that the lanthanides are significantly more difficult voltages that can be applied to an IL before it irreversibly to electroplate, and studies with mixed samples of U and decomposes (e.g. the cation will reduce and the anion lanthanides indicate the feasibility of selectively recover- will oxidize). Having a predictive capability to accurately ing the U while the lanthanides remain in solution. This describe the redox behavior of an IL is key to being able to provides the scientific basis for a possible IL-based electro- identify and screen and develop novel ILs with an electro- recovery process for processing UNF. chemical window wide enough to allow the electroplat- Radiation Stability of ILs: While a significant amount of ing of Pu. Modeling IL systems is more challenging than work has been reported on the stability of ILs subjected to standard polarizable solvents (such as water), which can be gamma-radiation, only one previous report examined the adequately described with a polarizable continuum model. effect of alpha-radiolysis on the stability of ILs, which is the Due to the low dielectric constants and the ionized nature primary decay mode for many actinides [12]. Radiolysis of of ILs, the response due to a change in oxidation state is organic materials is known to produce hydrogen gas, which the rearrangement of the solvent environment rather than can present significant safety hazards in industrial appli- a change in polarization. Our methodology for predict- cations. Very few studies have quantified H2 production, ing the electrochemical window uses a combination of and those only focused on gamma-radiation. Experiments quantum-mechanics and molecular mechanics (QM/MM) performed under this LDRD with collaborators at the approaches in order to be able to explicitly simulate a large University of Notre Dame Radiation Laboratory measured region of the solvent. A single anion-cation pair is simu- the yield of H2 production using 5 MeV He ions (for alpha- lated using QM and the surrounding environment is mod- radiation) and a Shepherd Co-60 gamma source for a range eled with MM, for which force-fields have been developed of IL and IL related compounds [19]. The results indicated previously by other authors. With this approach we have that the ILs behaved like conventional organic materials, successfully calculated electrochemical windows in good with H2 yield decreasing for aromatic compounds. While agreement with experimental data for a range of conven- the H2 production was higher for alpha radiation, the rates tional ILs [18], and a manuscript is in preparation expand- were relatively independent for aliphatic amines. In fact ing this methodology to ILs containing functional groups quaternary ammonium ILs showed significantly lower H2 such as carboxylate. This work successfully established production than their analogous aliphatic hydrocarbon the methodology to predict electrochemical windows for 84

Page 85

compounds, indicating that this family of ILs could posses superior radiation stability. Impact on National Missions This project addresses LANL’s weapons and energy secu- rity missions. Understanding actinide behavior in non- conventional solvents can lead to revolutionary nuclear material processing technologies that can reduce costs, proliferation risks, and environmental impacts, and eventu- ally contribute to a sustainable nuclear energy portfolio. Expanding the scientific understanding of Pu and actinide chemistry is central to LANL’s leadership in Pu science and its function as the NNSA Plutonium Center of Excel- Figure 2. Simulation snapshots showing coordination of lence. Developing advanced separation technologies for water and bistriflimide to plutonium(VI) in the first solva- used fuel, other nuclear materials, or high value isotopes tion shell. Plutonium (orange), oxygen (red), hydrogen such as Pu-242 supports LANL’s Plutonium Science and (white), nitrogen (blue), sulfur (yellow), and CF3 (cyan). Figure 2: Simulation snapshots showing coordination of water and bistriflimide to plutonium(VI) Research Strategy, and is of interest to NNSA and several in the first solvation shell. Plutonium (orange), oxygen (red), hydrogen (white), nitrogen (blue), DOE Offices. The scientific challenge of understanding the Figure 3. Electrodeposited U from the IL [PMP][Tf2N] (left) radiation-induced degradation of ionic liquids in extreme ansdu l[fEuMr (IyMe]ll[oTwf2)N, a] n(rdig ChFt3) a(ncyda Pnu). f rom the IL [EMIM][Tf2N] radiation environments as found in used nuclear fuel (right). reprocessing is of great important to the MaRIE mission. References Training a new generation of actinide chemists as postdocs is a critical aspect identified across the DOE complex. This 1. Clark, D., G. Jarvinen, C. Kowalczyk, J. Rubin, and M. project provided partial or full funding for 9 postdocs, and Stroud. Plutonium Processing at Los Alamos. 2008. six students. Actinide Research Quarterly. 3: 1. 2. Nuclear Technology Review 2008. 2008. 3. Huddleston, J. G., and R. D. Rogers. Room temperature ionic liquids as novel media for ‘clean’ liquid-liquid ex- traction. 1998. Chemical Communications. (16): 1765. 4. Visser, A. E., R. P. Swatloski, S. T. Griffin, D. H. Hartman, [UO2(CTMP)3(H2O)2][Tf2N]2 and R. D. Rogers. Liquid/liquid extraction of metal ions in room temperature ionic liquids. 2001. Separation Science and Technology. 36 (5 & 6): 785. 5. Cocalia, V. A., K. E. Gutowski, and R. D. Rogers. The coordination chemistry of actinides in ionic liquids: A [UO2(CTMP)3]2[Tf2N]4 review of experiment and simulation. 2006. Coordina- Figure 1: Photographs and thermal ellipsoids showing the structure of the cationic units in the Fi m g o u no r m e e 1 ric . P c h o o m t p o ou g nd r ap [U h O s2 (C a T n M d P) 3t ( h H2e O r )2m ]2+ al ( u e p l p l e ip r) so a i n d d s s th h e o w di i m n e g ric t h c e o mpound tion Chemistry Reviews. 250 (7-8): 755. [UO2(CTMP)3]2 4+ (lower). The hydrogen atoms have been removed for clarity. structure of the cationic units in the monomeric compound 6. Dietz, M. L., and J. A. Dzielawa. Ion-exchange as a [UO2(CTMP)3(H2O)2]2+ (upper) and the dimeric compound mode of cation transfer into room-temperature ionic [UO2(CTMP)3]24+ (lower). The hydrogen atoms have been re- liquids containing crown ethers: implications for the moved for clarity. ‘greenness’ of ionic liquids as diluents in liquid-liquid extraction. 2001. Chemical Communications. (20): 2124. 7. Schubert, T., S. Z. El Abedin, A. P. Abbott, K. J. McK- enzie, K. S. Ryder, and F. Endres. Electrodeposition of metals. 2008. 8. Venkatesan, K. A., T. G. Srinivasan, and P. R. Rao. A review on the electrochemical applications of room 85

Page 86

temperature ionic liquids in nuclear fuel cycle. 2009. J. Liquids. 2013. J. Phys. Chem. B. 117 (22): 6782. Nucl. Radiochem. Sci.. 10 (1): R1. Publications 9. Trulove, P. C., and R. A. Mantz. Electrochemical proper- Carney, B. D., X. Chen, G. S. Goff, H. He, W. Runde, B. L. ties of ionic liquids. 2008. Scott, and L. A. Seaman. Ionic Liquids for Actinide Pro- cessing. Invited presentation at 2013 Chemistry Capa- 10. Binnemans, K.. Lanthanides and Actinides in Ionic Liq- bility Review. (Los Alamos, NM, 9-12 June 2013). uids. 2007. Chemical Reviews. 107 (6): 2592. Chen, X., G. S. Goff, B. L. Scott, M. T. Janicke, and W. Runde. 11. Forsyth, S. A., J. M. Pringle, and D. R. MacFarlane. Ionic Solid-state and solution-state coordination chemistry Liquids-An Overview. 2004. Aust. J. Chem.. 57 (2): 113. of lanthanide(III) complexes with (pyrazol-1-yl)acetic acid. 2013. Inorganic Chemistry. 52: 3217. 12. Allen, D., G. Baston, A. E. Bradley, T. Gorman, A. Haile, Chen, X., G. S. Goff, B. L. Scott, and W. Runde. Comparison I. Hamblett, J. E. Hatter, M. J. Healey, B. Hodgson, of structural variations of Lanthanide(III) compounds R. Lewin, K. V. Lovell, B. Newton, W. R. Pitner, D. W. with (pyrazol-1-yl)acetic acid. Polyhedron. Rooney, D. Sanders, K. R. Seddon, H. E. Sims, and R. C. Thied. An investigation of the radiochemical stability of Chen, X., G. S. Goff, B. L. Scott, and W. Runde. Comparison ionic liquids. 2002. Green Chem.. 4 (2): 152. of structural variations of Ln(III) compounds with (pyr- azol-1-yl)acetic acid. To appear in Polyhedron. 13. Bridges, N., A. Visser, M. Williamson, J. Mickalonis, and T. Adams. Effects of gamma radiation on electrochemi- Chen, X., G. S. Goff, M. Quiroz-Guzman, D. P. Fagnant, Jr, B. cal properties of ionic liquids. 2010. Radiochimica Acta. L. Scott, and W. Runde. Directed nucleation of mono- 98: 243. meric and dimeric uranium(VI) complexes with a room temperature carboxyl-functionalized phosphonium 14. Chen, X., G. S. Goff, W. C. Ewing, B. L. Scott, and W. ionic liquid. 2013. Chemical Communications. 49 (19): Runde. Solid-State and Solution-State Coordina- 1873. tion Chemistry of Lanthanide(III) Complexes with Chen, X., G. S. Goff, W. C. Ewing, B. L. Scott, and W. Runde. α-Hydroxyisobutyric Acid. 2012. Inorg. Chem.. 51 (24): Solid-state and solution-state coordination chemistry 13254. of lanthanide(III) complexes with alpha-hydroxyisobu- tyric acid. 2012. Inorganic Chemistry. 51 (24): 13254. 15. Chen, X., G. S. Goff, M. Quiroz-Guzman, D. P. Fagnant, J. F. Brennecke, B. L. Scott, and W. Runde. Directed nucle- Dhiman, S. B., G. S. Goff, W. Runde, and J. A. LaVerne. Hy- ation of monomeric and dimeric uranium(VI) complex- drogen Production in Aromatic and Aliphatic Ionic Liq- es with a room temperature carboxyl-functionalized uids. 2013. Journal of Physical Chemistry B. 117: 6782. phosphonium ionic liquid. 2013. Chemical Communica- Dhiman, S., and J. A. LaVerne. Hydrogen production in the tions. 49 (19): 1903. gamma-ray and helium-ion radiolysis of ionic liquids 16. Maerzke, K. A., G. S. Goff, W. H. Runde, W. F. Schneider, . Presented at Radiation Chemistry Gordon Research Conference. (Andover, NH, July 29 - August 3, 2012). and E. J. Maginn. Structure and Dynamics of Uranyl(VI) and Plutonyl(VI) Cations in Ionic Liquid/Water Mixtures Fagnant, D. P., G. S. Goff, B. L. Scott, W. Runde, and J. F. via Molecular Dynamics Simulations. 2013. J. Phys. Brennecke. Switchable phase behavior of [HBet][Tf2N]- Chem. B. 117 (37): 10852. H2O upon neodymium loading: implications for lan- thanide separations. 2013. Inorganic Chemistry. 52 (2): 17. Pomogaev, V., S. P. Tiwari, N. Rai, G. S. Goff, W. Runde, 549. W. F. Schneider, and E. J. Maginn. Development and application of effective pairwise potentials for UO2n+, Goff, G. S., X. Chen, K. M. Long, B. L. Scott, G. D. Jarvinen, NpO2n+, PuO2n+, and AmO2n+ (n = 1, 2) ions with W. Runde, D. F. Fagnant, Jr., and J. F. Brennecke. Ionic water. 2013. Phys. Chem. Chem. Phys.. 15 (38): 15954. Liquids for Advanced Nuclear Fuel Cycles. Invited pre- sentation at 2012 LDRD Day. (Pojoaque, NM, 23 Oct. 18. Tian, Y., G. S. Goff, W. H. Runde, and E. R. Batista. 2012). Exploring Electrochemical Windows of Room-Temper- Goff, G. S., X. Chen, K. M. Long, W. C. Ewing, G. D. Jarvinen, ature Ionic Liquids: A Computational Study. 2012. J. and W. Runde. Development of ionic liquids for the Phys. Chem. B. 116 (39): 11943. electrochemical separations of actinides. Presented at 36th Actinide Separations Conference. (Chattanooga, 19. Dhiman, S. B., G. S. Goff, W. Runde, and J. A. LaVerne. TN, 22-24 May, 2012). Hydrogen Production in Aromatic and Aliphatic Ionic 86

Page 87

Jarvinen, G. D., and D. L. Clark. Fundamental Studies of Metal Complexation and Higher-Order Structures in Separation Systems. Invited presentation at 37th Ac- tinide Separations Conference. (Spokane, WA, 25-27 June 2013). Joyce, E. L., and L. D. Schulte. Plutonium reprocessing re- quirements and considerations for ionic liquid develop- ment . 2011. Los Alamos National Laboratory Report, LA-CP 11-01479. Long, K. M., G. S. Goff, and W. H. Runde. Exploring the chemistry of transuranic elements in ionic liquids. Pre- sented at 242nd ACS National Meeting & Exposition. (Denver, CO, United States, August 28-September 1, 2). Maerzke, K. A., G. S. Goff, W. Runde, W. F. Schneider, and E. J. Maginn. Structure and Dynamics of Uranyl and Plutonyl in Ionic Liquid/Water Mixtures via Molecular Dynamics Simulations. 2013. Journal of Physical Chem- istry B. 117 (37): 10852–10868. Maerzke, K. A., W. F. Schneider, and E. J. Maginn. Molecular dynamics simulations of uranyl and plutonyl coordina- tion in water/ionic liquid mixtures. Presented at Ameri- can Chemical Society Annual Meeting. (New Orleans, LA, April 2013). Pomogaev, V., N. Rai, S. P. Tiwari, W. Schneider, and E. Mag- inn. Formation energies and PES of uranyl and plutonyl hydrated complexes to develop effective potentials for condensed phase simulations. Presented at 14th In- ternational Congress of Quantum Chemistry. (Boulder, CO, 25-30 June, 2012). Pomogaev, V., S. P. Tiwari, N. Rai, G. S. Goff, W. Runde, W. F. Schneider, and E. J. Maginn. Development and Ap- plication of Effective Potentials for UO2n+, NpO2n+, PuO2n+, and AmO2n+ (n = 1,2) Ions with Water. 2013. Physical Chemistry Chemical Physics. 15 (38): 15954. Tian, Y., G. S. Goff, W. Runde, and E. R. Batista. Exploring electrochemical windows of room-temperature ionic liquids: a computational study. 2012. Journal of Physi- cal Chemistry B. 116 (39): 11943. Tian, Y., and E. R. Batista. Exploring electrochemical win- dows of room-temperature ionic liquids by hybrid QM- MM method and thermodynamics calculations. Pre- sented at 243rd ACS National Meeting & Exposition. (San Diego, CA, March 25-29, 2012). 87

Page 88

Chemistry and Material Sciences Directed Research Final Report Functional Materials by Design: Explosives with Tailored Optical Response Properties Robert J. Scharff 20130006DR Abstract flash of light [11, 12]. For laser initiation of pentaerythri- We offer an innovative approach to establish design tol tetranitrate (PETN) [3-10], several established routes principles and a scientific methodology for generating to couple a laser field to the explosive are utilized [13]: photoactive energetic materials with controllable optical (i) the beam directly impinges on an explosive target functionality. The project aims to increase the controlla- resulting in a thermal response [9,10]. (ii) The beam bility of chemical dynamics in novel photoactive materi- directly impinges on an explosive target that contains ab- als through a concerted experimental and theoretical sorbing particles used to create local “hot spot” regions effort that characterizes the dynamics of energy localiza- where chemical reactions develop [4-6]. (iii) The beam is tion, chemical bond activation, and chemical reactions. fiber-optically coupled to the explosive through a metal This approach will provide critical insight on how to ma- layer sputter coated to the end of the fiber [3, 7, 8]. The nipulate electronic structure through synthesis in order laser pulse vaporizes the metal layer to create a plasma to generate the desired material response to overcome that inputs a quasi isentropic compression wave into the mechanisms that limit controllability. We have proposed explosive. In the all instances, the mechanism of initia- a three-part solution for the design of photoactive en- tion is not a direct photochemical process, but rather an ergetic materials for quantum optical initiation. The first indirect, shock-less deflagration (burning)-to-detonation part involves the development of new explosives de- transition (DDT). The problems with current optical rivatized with an optical chromophore. The optimization initiation schemes are: (i) high energies are required, (ii) of optical response properties of the photon-absorbing initiation is “indirect”, and (iii) initiation of the chemical chromophore enables efficient coupling of the laser reaction is not “controllable”. field to the explosive and facilitates control of excited While a programmatic need exists for the development state dynamics. The second part addresses the need for of optical isolation techniques to enhance the safety of the elucidation of electron-vibrational dynamics follow- the enduring U.S. nuclear stockpile, even greater pro- ing photoexcitation. Our feasibility studies support our grammatic needs exist for quantum controlled explosive main idea that exothermic chemical reactions could be initiation (QCI) in future surety applications and other achieved photochemically. The third part of the project technologies. In order to achieve QCI, we have proposed uses femtosecond laser pulse-shaping quantum control a key innovation: engineer new explosive materials spe- to optimize photodissociation quantum yields and pho- cifically optimized for quantum control of photochemical toinitiation of exothermic explosive chemistry. decomposition. Optical control can only be effective in newly designed materials that are synthesized to take Background and Research Objectives advantage of such control. To date, our efforts have fo- Laser ignition of explosives materials is of considerable cused on making new explosive materials with energetic interest for both commercial and defense purposes [1- optical chromophores, calculating the non-linear optical 3]. Current efforts to initiate explosives with lasers are response properties, and validating those calculations driven by the increase in electrical safety requirements with ultrafast optical techniques. These results have es- of weapons. Ideally, an insensitive (e.g., to spark, fric- tablished the platform from which we can now envision tion, fire, lightning flash) explosive could only be initiat- creating new explosives with controllable optical func- ed with a laser. While it is possible to initiate explosives tionality for next generation explosive applications. with a laser, the process is indirect and requires high la- ser energies [3-10]. There are also applications that use Photochemistry is the key to control. The term photo- very sensitive explosives that can be initiated with only a 88

Page 89

chemistry applies to chemical reactions that are induced optical susceptibilities are ideal for photoactive energetic by light-matter interactions. Upon absorption of light, the materials due to their potential for rapid energy redistribu- ground state population in a molecular system can trans- tion during relaxation from excited states above S1. Also, fer to an electronically excited state where a subsequent we have formulated a novel theoretical algorithm that chemical reaction ensues. Classic examples of 1-photon describes coherent photoexcited state dynamics, a critical induced photochemistry are the Norrish reactions with innovation necessary to model quantum control experi- ketones and aldehydes [14, 15]; 3D microfabrication with ments [28]. Characterization and control: In FY13, we have nonlinear photoinitiator polymers is an example of 2-pho- developed an excited state femtosecond stimulated Ra- ton induced photochemistry [16-23]. Broadly defined, man spectroscopy diagnostic (ES-FSRS) [29] that we have the term “nonlinear” includes simultaneous multi-photon coupled to our existing UV/near-IR shaped pulse quantum absorption, cascaded 1-photon absorptions, and “pump- control experiment. This represents a unique, advanced dump” processes; all of which are physical mechanisms spectroscopic capability for characterization and control of that can be optimized to control chemical dynamics. excited state molecular dynamics. Preliminary results on nitromethane demonstrate our new capability to measure Our project develops an alternate initiation scheme that real-time molecular changes due to excited state photo- uses ultrashort (<100 fs) optical pulses to trigger and chemistry, and also illustrate how “off the shelf” explosives control a net exothermic photochemical-to-DDT reaction. lack large optical cross-sections that lead to decomposition Our scheme provides a significant improvement to current (Figure 1). laser ignition schemes because of the direct absorption of the laser field by the explosive and increased optical sensitivity that allows the energy of the excitation source to be reduced. Furthermore, the wide spectral bandwidth 520 of ultrashort optical pulses makes it possible to control chemical reactions [24-26] responsible for explosive de- 480 composition. 440 Scientific Approach and Accomplishments Understanding and controlling excited state dynamics (spa- tial energy transfer, excitation localization/delocalization, 0 40 80 120 and/or charge separation) lies at the heart of all our efforts Probe delay (ps) to design photoactive materials with desired functionality. While this design approach has become the standard for many technological applications (e.g., solar energy harvest- ing), a similar approach will be required to achieve quan- tum control of photoactive energetic materials. Through FY13, we have amassed theoretical and experimental results that demonstrate key elements of our hypothesis, mitigate the level of risk, and provide feasibility for our research plan. We have utilized LDRD FY13 Reserve to (i) synthesize new explosive materials with energetic optical chromophores, (ii) validate their optical response proper- ties through experiment and simulation, and (iii) develop critical experimental capabilities to measure time depen- dent photoexcited molecular dynamics. Synthesis: We have expanded our chemical basis set for added flexibility in the preparation of photoactive energetics for quantum control. Preliminary results on chromophore-derivatized explosive prototypes indicate favorable trends in the com- puted and measured two photon absorption spectra, and current excited state molecular dynamics calculations dem- onstrate the plausibility of our hypothesized mechanism for photoinitiation [27]. Theory: Our theoretical modeling suggests that materials possessing large linear or nonlinear 89 )mn( htgnelevaW 0.04 0.03 0.02 0.01 0.00 1.0 0.8 0.6 0.4 0.2 0.0 niaG namaR 10 ∆ R a 5 m a n 0 G a -5 ni x -10 1 0 -15 3- 1000 2000 3000 4000 -1 Wavenumbers (cm ) Figure 1. Nitromethane photochemistry. Time resolved transient absorption (top) quantifies excited state dynamics, but the broad features generally do not definitively identify molecular species. The FSRS spectrum (bottom, black) and the ES-FSRS spectrum (red curve is FSRS of products – reactants) have sharp peaks cor- responding to specific bonds, and small changes in bonding are observed.

Page 90

Impact on National Missions ticles at elevated temperatures. 2010. Russian Journal Nuclear stockpile aging and enhanced surety are two of of Physical Chemistry B. 4 (3): 452. the primary concerns for the Weapons Program. The de- 7. Stewart, D. S., K. A. Thomas, S. Clarke, H. Mallett, E. velopment of photoactive energetic materials that initiate Martin, M. Martinez, A. Munger, and J. Saenz. On the through the quantum control of photochemistry addresses initiation mechanism in exploding bridgewire and laser future surety themes outlined in the FY12 SSMP and pro- detonators. 2006. AIP Conference Proceedings. 845 vides an enabling technology for future Directed Stockpile (1): 471. Work [30, 31]. The theory-guided synthesis of energetic materials tailored for optical response would be unprec- 8. Nagayama, K., K. Inou, and M. Nakahara. Initiation of edented and unique to LANL, paving the way for the future PETN high explosive by pulse laser-induced high-tem- creation of novel optical explosives that perform like RDX perature plasma. 2001. In Impact Engineering and Ap- (1,3,5-trinitro-1,3,5-triazacyclohexane) yet are insensitive plication, Vols I and II. Edited by Chiba, A., S. Tanimura, to mechanical and thermal stimuli. Since there are no in- and K. Hokamoto. , p. 515. Oxford: Elsevier. sensitive photoactive explosives currently, producing new insensitive photoactive explosives will generate licensable 9. Tarzhanov, V. I., A. D. Zinchenko, V. I. Sdobnov, B. B. technology opportunities for direct optical initiation deto- Tokarev, A. I. Pogrebov, and A. A. Volkova. Laser initia- nators, quantum control detonators, and conventional light tion of PETN. 1996. Combustion Explosion and Shock sensitive explosive applications. The design of photoactive Waves. 32 (4): 454. materials for quantum controlled initiation offers modern technology that will address future requirements [30] 10. Bykhalo, A. I., E. V. Zhuzhukalo, N. G. Kovalskii, A. for “direct optical initiation or other advanced initiation N. Kolomiiskii, V. V. Korobov, A. D. Rozhkov, and A. I. approaches” above and beyond those of current optical Yudin. INITIATION OF PETN BY HIGH-POWER LASER- isolation techniques [1, 3, 7]. More safe and secure control RADIATION. 1985. Combustion Explosion and Shock of explosives also has Department of Defense applications. Waves. 21 (4): 481. 11. Benham, R.. Preliminary experiments using light-initiat- References ed high explosive for driving thin flyer plates. 1980.

1. Clarke, S. A., C. D. Landon, M. J. Murphy, M. E. Marti- nez, T. A. Mason, and K. A. Thomas. Detonator Per-
2. Benham, R.. A new LIHE (Light Initiated High Explosive) formance Characterization Using Multi-frame Laser test capability for spherical targets. 1987. Schlieren Imaging. 2009. APS Topical Conference on Shock Compression of Condensed Matter. 1195: 663. 13. Bourne, N. K.. On the laser ignition and initiation of explosives. 2001. Proceedings of the Royal Society A -
3. Akinci, A. A., S. A. Clarke, K. A. Thomas, and A. C. Mathematical, Physical, and Engineering Sciences. 457 Munger. Optical initiation spot size effects in low-den- (2010): 1401. sity PETN. 2006. SPIE Optical Technologires for Arming, Safing, Fuzing, and Firing II. 6287: 628709. 14. Norrish, R. G., and C. H. Bamford. Photodecomposition of aldehydes and ketones. 1936. Nature. 138: 1016.
4. Akinci, A., K. Thomas, A. Munger, L. Nunn, S. Clarke, M. Johnson, J. Kennedy, and D. Montoya. On the develop- 15. Norrish, R. G., and C. H. Bamford. Photo-decomposi- ment of a laser detonator. 2005. SPIE Optical Technolo- tion of aldehydes and ketones. 1937. Nature. 140: 195. gies for Arming, Safing, Fuzing, and Firing. 5871: 1.
5. Baldacchini, T., H. Chen, R. A. Farrer, M. J. Previte, J.
6. Aluker, E. D., A. G. Krechetov, A. Y. Mitrofanov, and D. Moser, M. J. Naughton, and J. T. Fourkas. Multiphoton R. Nurmukhametov. Photochemical and Photothermal photopolymerization with a Ti : sapphire oscillator. Dissociation of PETN during Laser Initiation. 2011. Rus- 2002. Edited by G. S. Edwards, J. Neev, A. Ostendorf, sian Journal of Physical Chemistry B. 5 (4): 658. and J. C. Sutherland. Vol. 4633, p. 136.
7. Aluker, E. D., G. M. Belokurov, A. G. Krechetov, A. Y. 17. Baldacchini, T., R. A. Farrer, J. Moser, J. T. Fourkas, and Mitrofanov, and D. R. Nurmukhametov. Laser initiation M. J. Naughton. Efficient multiphoton polymerization of PETN containing light-scattering additives. 2010. for the fabrication of 3-dimensional microstructures. Technical Physics Letters. 36 (3): 285. 2003. Synthetic Metals. 135 (1-3): 11.
8. Aduev, B. P., D. R. Nurmukhametov, and A. V. Puzynin. 18. Belfield, K. D., and K. J. Schafer. Two-photon photoiniti- Laser initiation of a mixture of PETN with NiC nanopar- ated polymerization. 2003. Edited by K. D. Belfield, and 90

Page 91

J. V. Crivello. Vol. 847, p. 464. second stimulated Raman spectroscopy. 2013. Annual Review of Physical Chemistry. 19. Engelhardt, S., Y. Hu, N. Seiler, D. Riester, W. Meyer, H. Krueger, M. Wehner, E. Bremus-Koebberling, and A. 30. FY12 Stockpile Stewardship and Management Plan, U. Gillner. 3D-Microfabrication of Polymer-Protein Hybrid S. Department of Energy. 2011. Structures with a Q-Switched Microlaser. 2011. Journal 31. Guidice, S. J.. Weapons Engineering Capability Review. of Laser Micro Nanoengineering. 6 (1): 54. 2012. 20. Gu, J., W. Yulan, W. Chen, X. Dong, X. Duan, and S. Kawata. Carbazole-based 1D and 2D hemicyanines: synthesis, two-photon absorption properties and appli- cation for two-photon photopolymerization 3D lithog- raphy. 2007. New Journal of Chemistry. 31 (1): 63. 21. Ovsianikov, A., A. Deiwick, S. Van Vlierberghe, P. Du- bruel, L. Moeller, G. Draeger, and B. Chichkov. Laser Fabrication of Three-Dimensional CAD Scaffolds from Photosensitive Gelatin for Applications in Tissue Engi- neering. 2011. Biomacromolecules. 12 (4): 851. 22. Ovsianikov, A., M. Malinauskas, S. Schlie, B. Chichkov, S. Gittard, R. Narayan, M. Loebler, K. Sternberg, K. P. Schmitz, and A. Haverich. Three-dimensional laser micro- and nano-structuring of acrylated poly(ethylene glycol) materials and evaluation of their cytoxicity for tissue engineering applications. 2011. Acta Biomateria- lia. 7 (3): 967. 23. Steenhusen, S., T. Stichel, R. Houbertz, and G. Sextl. Multi-photon polymerization of inorganic-organic hy- brid polymers using visible or IR ultra-fast laser pulses for optical or (opto-)electronic devices. 2010. Edited by W. V. Schoenfeld, J. J. Wang, M. Loncar, and T. J. Suleski. Vol. 7591. 24. Brumer, P., and M. Shapiro. Control of unimolecular reactions using coherent light. 1986. Chemical Physics Letters. 126 (6): 541. 25. Tannor, D. J., R. Kosloff, and S. A. Rice. Coherent pulse sequence induced control of selectivity of reactions: Exact quantum mechanical calculations. 1986. The Journal of Chemical Physics. 85 (10): 5805. 26. Tannor, D. J., and S. A. Rice. Control of selectivity of chemical reaction via control of wave packet evolution. 1985. Journal of Chemical Physics. 83 (10): 5013. 27. Nelson, T., and S. Tretiak. NA-ESMD calculation of explosive prototype. 2012. Theoretical Division, LANL, unpublished work. : 1. 28. Mozyrsky, D., V. Gorshkov, and S. Tretiak. manuscript under review. Nature Physics. 29. Kukura, P., D. W. McCamant, and R. A. Mathies. Femto- 91

Page 92

Chemistry and Material Sciences Directed Research Final Report Alternative Explosives Signatures for Detection David S. Moore 20130086DR Abstract that a suite of technologies must be deployed in order The events at the Boston Marathon finish line this past to cover the multitude of possible threat scenarios. A April graphically illustrate the difficulties we face as a na- serious gap is evident in this suite of technologies – that tion to find and neutralize explosive devices before they of detecting the bulk explosive using non-ionizing radia- are used. Our toolset is massive and diverse, from intel- tion – the gap that the work described in this proposal is ligence, counter-intelligence, and persistent surveillance intended to fill. The characteristic of explosives that we to regulation of explosives and precursors, X-ray imaging, exploit is the intrinsic property that defines an explo- and metal detection, yet it fails in many situations, in- sive – its metastable chemical energy that can be quickly cluding venues without controlled access like the streets released on command (shock, friction, spark). The work of any city. Our adversaries are agile and adapt by coun- addressed the high-risk aspect of this novel explosive tering every new technology that achieves some level detection in the GHz to THz electromagnetic (EM) of success (metal detectors, RF jammers, hand swipes/ frequency range by performing proof-of-principle stud- IMS technology, X-ray imaging). Our project aims to fill a ies. First, we measured reflected power from pressed major gap in our arsenal, by detecting the bulk explosive powder samples of HMX and sugar (each with as close to itself. This proposed capability is potentially disruptive to the same particle size distribution as feasible, which was current evasion-of- detection tactics, because it exploits measured using micro-computed X-ray tomography) as the nonlinear coupling of penetrating GHz-to-THz elec- a function incident EM amplitude. Second, the reflected tromagnetic radiation to explosives and detection of the power data were related to theoretical results. These alternative signatures that are generated. utilized macroscopic models for the EM response of crystalline materials using effective medium theory, with We performed proof-of-principle studies that show a the difference between inerts and explosives being an connection between local chemical energy release and activation energy barrier that can be overcome in small local nonlinearities that can be exploited to differentiate local regions by sufficiently strong excitation, thereby inerts from explosive materials. The experiments and releasing energy and locally increasing the temperature. theory were designed to correlate the changes in effec- This response in turn modifies the effective electric per- tive permittivity caused by the presence or absence of mittivity and thereby the reflected power, providing the a local source term with activation energy at hot spots. connection to the experiment. The ultimate goal is to develop fundamental principles that define the processes and signatures, which will Scientific Approach and Accomplishments guide the design of a prototype detection system with We performed experiments and developed theory in- area-scanning capability from safe standoff distances. tended to show a connection between local chemical en- ergy release and local nonlinearities that can be exploit- Background and Research Objectives ed to differentiate inerts from explosive materials using The detection of explosives is a pressing current prob- GHz frequency stimulation. Experimentally, we mea- lem, both at home and abroad. The Washington Post sured the reflected power from pressed powder samples series “Left of Boom” succinctly outlines the latest evolu- of inert and explosive materials prepared with the same tion of the use of IEDs (improvised explosive devices) particle size distribution. We observed distinctly differ- especially in Iraq and Afghanistan, and the progress of ent time dependent GHz reflectivity from pressed HMX counter-IED technologies [1]. Each new technology has and pressed sugar – chemically similar except one is an been countered by our adversaries with great agility, so explosive and the other an inert (Figure 1). Note the very 92

Page 93

similar behavior of four different samples of HMX and the Figure 2. Reflected power versus time for 11 GHz irradiation at very similar behavior of the three different samples of sug- three different incident powers. The HMX shows a nonlinear ar, along side the significantly different behavior of HMX increase in reflected power with increasing incident power, while the sugar does not. compared to sugar. Pre- and post-mortem micro-CT images of the two materials show voids in the HMX sample caused The power levels involved are less than or similar to ev- by local reactivity due to the locally high temperatures un- eryday EM exposures. For a detection system with output der GHz irradiation. Note that the GHz source was only on power of 10 kW, pulse duration of 1 µs and a spot size for a short time, so that even though the peak power was on target of 50 cm diameter, the peak intensity is greater high, the duty cycle and therefore also the deposited aver- than 2 W/cm2. However, thousands of these 1 µs pulses age power was low. A simple theory was developed based a minute at this intensity do not exceed the FCC average on macroscopic Maxwell equations and effective medium irradiance allowance of 5 mW/cm2. In addition, a typical theory for anisotropic materials, resulting in effective per- cell phone emits about 20 mW/cm2 at a distance of 2.2 cm mittivity tensors that reveal the frequency and tempera- from the antenna, larger than the irradiance from these ture dependent response of the materials to external EM stimulation sources. Finally, finite element modeling of waves (dotted lines in Figure 1). these experiments shows only small temperature increases (even a much longer and stronger 100 ms pulse at 43 W/ cm2 increases an HMX sample temperature by less than 2 ˚C), so the probability of accidentally setting off the explo- sive during detection is very low. Impact on National Missions The results described above are the first step towards explosive detection in frequency regimes capable of pen- etrating non-metallic packaging, clothing, or camouflage to interrogate bulk explosives, filling a capability gap in the Figure 1. Micro-CT image of a pristine (left top) and an exposed current suite of counter-IED technologies, which includes (left bottom) HMX sample showing void growth; Reflected power X-ray imaging, metal detection, trace analysis, intelligence, versus time for 11 GHz irradiation at 500 W (simple theory = and persistent surveillance. These GHz range EM frequen- dots) showing distinctly different behavior of HMX and sugar. cies have not been investigated for standoff detection until recently. The ultimate aim of the work is to obtain We also obtained reflected power data on these same the characteristics and parameters needed for forward samples as a function of excitation power, showing dif- deployment of an actual detector system. Subsequent ferent behavior for the explosive compared to the inert LDRD work (new start FY2014) will jump-start new model- (Figure 2). ing capabilities at LANL that will enable predictive theory of elasto-electric coupling with chemical energy release at hot spots in explosives. We at LANL are breaking ground for the detection of alternate signatures using GHz-to-THz frequencies integrated with modeling of elasto-electric coupling including chemical energy release at hot spots and thermal diffusion. This project represents a transformational approach to un- covering the direct explosive signal type applicable to the Discover Signatures and Revolutionize Measurement com- ponents of the Science of Signatures pillar. The overarch- ing goal of the project directly supports LANL’s long-term objective of discovery of the next generation of materials signatures of explosives within its global security mission. The deep fundamental understanding of energy absorp- tion and subsequent responses are applicable to MaRIE’s “Decadal Challenges for Predicting and Controlling Materi- als Performance in Extremes” for the design and control of 93

Page 94

energy release in explosives. This work along with that to be performed in the FY2014 new start Directed Research project will develop a new framework for a unified descrip- tion of explosives’ hot spots coupled to the local mechani- cal, thermal and electromagnetic signatures. Understand- ing and controlling the material functionality of defects and crystalline interfaces is important for hot spots and their signatures and underpins the Materials for the Future pillar with the focus area Defects and Interfaces, as well as the priority area 2 for advanced (THz) spectroscopies. There are a number of agencies interested in the results of this research, including the Office of Naval Research (ONR – Code 30), Department of Homeland Security (DHS – S&T Explosives), and Joint IED Defeat Organization (JIEDDO). Jon Schoonover, a LANL Global Security program manager, is mentoring the project and helping with transition. The results and data obtained in FY2013 resulted in a full LDRD FY2014 new start DR project, whose results will eventu- ally enable transition to a product for controlled access portals or vehicle mounted systems including stimulation source(s), antenna arrays and thermal detectors, or as an add-on to ground penetrating radar systems. References

1. Atkinson, R.. About Left of Boom: The Fight Against Roadside Bombs. 2007. The Washington Post. 94

Page 95

Chemistry and Material Sciences Early Career Research Continuing Project Room Temperature Oxidation and Corrosion of Plutonium Alison L. Pugmire 20130738ECR Introduction to ensure the reliability and performance of the nation’s Currently, a systematic investigation of plutonium cor- weapons systems currently in the stockpile. Interest in rosion under room temperature, pressure and humid- the health of the nuclear stockpile has been heightened ity has yet to be performed. This work proposes to use in recent months due to the Department of Energy a carefully controlled environment to form the oxide (DOE) and LANL’s recent commitment to pit reuse. As layer on well-characterized Pu metal samples. At vari- stated by LANL Director Charlie McMillan in a June 27, ous times during the corrosion process, a combination 2012 All-Hands Meeting, “Pit reuse is now a certification of spectroscopic tools will probe both the surface layer ‘grand challenge, one of the most complicated we have (spectroscopic ellipsometry, SE; Auger electron spectros- ever faced.” Plutonium corrosion has been identified as a copy, AES; x-ray photoelectron spectroscopy, XPS) and key issue in pit aging and reuse, and the current state of the bulk material (x-ray diffraction, XRD; x-ray absorption corrosion knowledge has been recognized as limited at fine structure, XAFS). This work will also aim to address a best. Thus, this research directly addresses the laborato- long-standing issue with plutonium metal experiments: ry’s mission to en-sure the effectiveness of the stockpile variability in the sample preparation and resultant repro- and the challenges facing certification of pits for reuse. ducibility. These issues will be addressed by utilizing a consistent and controlled preparation method and con- Progress ducting the experiments on multiple samples to ensure Year 1 focused on developing a sample preparation repeatable and, more importantly, reliable results. method, utilizing in-house techniques to structurally and Many questions remain regarding the room tempera- chemically characterize the samples, and submit propos- ture/pressure mechanism of plutonium corrosion. What als for off-site XAFS experiments for Year 2. is the chemical composition of the corrosion products (PuO2, Pu2O3, etc.)? Is the surface layer oxide crystalline Accomplishments from the first half of Year 1 (FY2013 PuO2 or an amorphous oxide mixture as recent inves- 2nd, 3rd quarters) are listed and organized in accordance tigations indicate? Is Pu2O3 present at the PuO2-metal with the Year 1 objectives. interface? Why are seemingly different results obtained for ultra-high vacuum conditions versus “real world”, Task 1: Determine the optimal samples and sample ambient conditions? The proposed work aims to answer preparation method these fundamental questions with a comprehensive sur- • The first set of stock material (Sample Set 1) was face and bulk layer study at conditions most relevant to identified and collected the storage and use of plutonium: near room tempera- • A preparation method was chosen and tested on ture and pressure. Sample Set 1 (mechanical polishing followed by electropolishing) Benefit to National Security Missions • A duplicate sample of one exposure condition was Los Alamos National Laboratory (LANL) is the premier also selected to test reproducibility institution for plutonium science, with a rich history of excellence in actinide research and development, pit • Based on the results of Tasks 2-4, the second set of manufacturing, and surveillance. The laboratory’s mis- stock materials (Sample Set 2) was identified to be sion is to “ensure the safety, security, and effectiveness tested in the latter part of Year 1/Year 2 of the U.S. nuclear deterrent.” The core of this mission is 95

Page 96

Task 2: Determine the optimal exposure conditions Task 2: Exposure Experiments • Three optimal sample exposure conditions were cho- • Expose additional stock materials to same exposure sen (wet air, dry air, wet argon environment) time initially tested (~30 days) • An optimal sample exposure time was selected (~30 • Based on initial experiments, determine additional days) based on the samples ident-fied in Task 1 and relevant exposure times literature review • Characterize samples using in-house techniques • Set up of exposure containers was completed and • Incorporate more characterization techniques not em- tested (Task 3) ployed in first set of exposures (XRD, Auger) Task 3: Exposure Experiments Task 3: Analyze Spectroscopic Data and Interpret results • The first set of exposure experiments were completed • Analyze data collected throughout FY13 and FY14 (30 day cycle) (XRD, XAFS, XPS, Auger) • Samples were characterized using various x-ray spec- troscopies (x-ray photoelectron, XPS, and x-ray absorp- Task 4: Determine Future Work and Develop a Proposed tion, XAFS, in collaboration with the project’s collabo- Corrosion Mechanism rator) • Based on the results from Tasks 2 and 3, determine op- timal structural and characteriza-tion techniques most • Surface quality of Sample Set 1 was determined using appropriate and most useful for potential future work XPS and XAFS • Based on the results from Tasks 2 and 3, propose a • Technique limitations (probe depth, etc) were identi- room temperature corrosion mechanism fied and will guide and aid in the co-ordination of mul- tiple techniques in end of Year 1/Year 2 experiments Conclusion • XPS and XAFS data analysis from these experiments is The goals of this research are to study the ambient oxida- on-going, but initial results are guiding the prepara- tion/corrosion mechanism of plutonium using a systematic tion/setup/exposure of the second set of experiments approach to prepare the specimens; develop the oxide with the second set of stock material (Sample Set 2) layer as a function of humidity, oxygen content, and expo- sure; and study the oxide layer using multiple techniques. Expected results include determination of the chemical Task 4: Submit Beam Time Proposals (e.g. PuO2, Pu2O3, and/or PuOxCy) and structural com- • An off-site XAFS beam time proposal was prepared and position (e.g. crystalline or amorphous) of the surface and submitted at the first available deadline (June 1, 2013) bulk oxide layer. These results will help develop a compre- hensive picture of the corrosion mechanism, impacting our Future Work understanding of Pu aging in the stockpile. Year 2 will continue the structural and chemical character- ization of the corrosion process using in-house capabilities Publications and off-site synchrotron experiments. The tasks and goals Booth, C. H., Y. Jiang, S. A. Medling, D. L. Wang, A. L. to be accomplished in the next fiscal year (FY14 1st, 2nd, Costello, D. S. Schwartz, J. N. Mitchell, P. H. Tobash, E. 3rd, and 4th quarters) are listed. D. Bauer, S. K. McCall, M. A. Wall, and P. G. Allen. Self- irradiation damage to the local structure of plutonium Task 1: Determine additional samples and other potential and plutonium intermetallics. 2013. JOURNAL OF sample preparation methods APPLIED PHYSICS. 113 (9): -. • Identify and collect additional stock materials (Sample Set 2 (see progress section), Sample Set 3, etc.) Pugmire, A. L.. Spectroscopic Analysis of Actinyl Systems. To appear in Actinide Research Quarterly. • If relevant or needed, test other preparation methods (e.g. mechanical polishing with no electropolishing) Pugmire, D. L., H. G. Garcia Flores, and A. L. Costello. • Identify the advantages/disadvantages of preparation The room temperature oxidation/corrosion of delta methods based on “real world” surface relevance, plutonium. Presented at XIII International Workshop: reproducibility, and experimental results Fundamental Properties of Plutonium. (Sarov, Russia, 8-13 Sept. 2013). • Determine 1) surface quality 2) sample pedigree of dif- ferent stock materials and sample preparation 96

Page 97

Chemistry and Material Sciences Early Career Research Continuing Project Novel Mesoscale Modeling Approach for Investigating Energetically Driven Nanoscale Defect/Interface Interactions Abigail Hunter 20130745ECR Introduction sible with current mesoscale methods. In nanomaterials (grain sizes less than 100nm), experi- mental studies and atomistic simulations show that Benefit to National Security Missions when the characteristic length scale (ex: layer thickness, Understanding the effects of defects and interfaces on grain size) decreases many deformation mechanisms material behavior as materials age becomes increasingly arise that are not active in bulk and coarse-grained ma- important for determining the performance, reliability, terial counterparts, such as dislocation emission and ab- and safety of weapon systems. Los Alamos National sorption at grain boundaries and interfaces, dislocation Laboratory (LANL) has advanced atomistic and con- pile-ups, confined layer slip, and deformation twinning. tinuum simulation tools, however, it is not clear how to These alternative deformation mechanisms, along with connect information on these two very different length grain boundaries and bimetallic interfaces, can result in scales. Hence, there remains a computational gap at the unique material behavior including high strength and mesoscale that translates into a gap in our understand- improved ductility. As applications for nano-devices ing of material deformation. The phase field dislocation increases, there is not only a growing need to ensure dynamics (PFDD) model aims to fill this gap, and produce reliability, but also the opportunity to engineer materials predictive multiscale simulations crucial for stockpile for specific applications, however a predictive model to stewardship. address these issues still does not exist. The goals of this proposal include innovative extensions This project is focused on investigating nanoscale defect/ of the PFDD model that will not only be new in the sci- interface interactions, such as nucleation and growth entific community, but will also answer many questions of deformation twins, and dislocation dynamics as about dislocation behavior on the nanoscale, particu- controlled by interfaces through development of an in- larly at interfaces. The proposed model advancements novative three-dimensional (3D) phase field dislocation will be extensions that have not yet been achieved by dynamics (PFDD) mesoscale model that links atomic- current phase field dislocation models. These advance- scale and nanoscale physics. This model is unique at Los ments will enable the PFDD model to investigate many Alamos National Laboratory (LANL), and has great poten- defect/interface deformation mechanisms not possible tial to answer many questions about the effects of de- with current mesoscale methods. formation twinning, and the impact of grain boundaries and heterophase interfaces on dislocation and material Progress strength behavior. The goals of this project include in- The primary goal of this project is to investigate na- novative extensions of the 3D PFDD model that will not noscale defect/interface interactions, such as nucle- only be new in the scientific community, but will also an- ation and growth of deformation twins, and dislocation swer many questions about dislocation behavior on the dynamics as controlled by interfaces through develop- nanoscale, particularly at interfaces. Model advance- ment of an innovative three-dimensional (3D) phase ments as part of this project include: incorporation of field dislocation dynamics (PFDD) mesoscale model that temperature dependence, implementation of a complex links atomic-scale and nanoscale physics. This model description of nanolayer interfaces, and reformulation to is unique to Los Alamos National Laboratory (LANL) account for body-centered cubic (bcc) crystal structures. and the novel extensions of the PFDD model achieved Development of this computational tool will enable the as part of this project will be both new to the scientific study of deformation mechanisms at interfaces not pos- 97

Page 98

community and answer many questions about dislocation complex description for extended dislocation cores. behavior on the nanoscale, particularly at interfaces. Three This project has resulted in several collaborations including specific goals are highlighted in this project: (1) investigate those spanning divisions/groups within LANL, and external the effects of grain size variation and temperature depen- academic relationships. Several publications are currently dence on deformation twinning in face-centered cubic (fcc) underway, in addition to those already published or sub- materials, such as copper, nickel, gold, and silver, (2) study mitted: dislocation behaviors in fcc/fcc multilayers, and (3) account for dislocation behavior in base-centered cubic (bcc) mate- Three other papers are currently being written that ad- rials, such as tantalum, steel, chromium, and niobium. dress two-layer twin formation, an analytical model that describes dislocation formation from grain boundaries, and Over the past year, much progress has been made toward combined DFT-PFDD simulations that investigate partial achieving these three goals. While work in all three areas dislocation dependence on material variations. has begun, focus of the project in the first year has been surrounding fcc/fcc interfaces, which is the primary goal Future Work of the project (collaborator: Irene Beyerlein- T-3). This The goals of this projectl include innovative extensions of topic initiated a collaboration with Dr. Marisol Koslowski the three-dimensional (3D) phase field dislocation dynam- at Purdue University. A two-layer PFDD model has been ics (PFDD) model that will not only be new in the scientific developed that accounts for differences in material prop- community, but will also answer many questions about erties between two metallic layers. Currently, this new dislocation behavior on the nanoscale, particularly at inter- algorithm is being extended to a parallel PFDD code and faces. Three specific goals are highlighted in this project: numerical considerations are being addressed. In parallel with modification to the PFDD code itself, MD simulations

1. Investigate the effects of grain size variation and of copper/nickel interfaces have been completed (col- temperature dependence on deformation twinning laborator: Ruifeng Zhang – Iowa State). This also included in face-centered cubic (fcc) materials, such as copper, development of a cross-potential for nickel (collaborator: nickel, gold, and silver. This goal requires that tem- Xiang-Yang (Ben) Liu-MST-8). These simulations provide perature dependence be included in the current PFDD key information for the development of a complex de- model framework through the material parameters scription of the interface region between two nanolayers. and energy functional. With this model advancement, Furthermore, information from these simulations will be it will be possible to study the transition between slip directly incorporated into the PFDD model in order to ac- and deformation twinning deformation mechanisms. curately describe dislocation behavior across the interface.
2. Study dislocation behaviors in face-centered cubic This summer (July 2013) a Purdue student, Yifei Zeng, (fcc)/fcc multilayers. Modeling dislocation behavior will come to LANL and focus on the implementation of a across nanolayer interfaces has never been attempted complex energy landscape into the PFDD model in order to with a phase field model. A complex description of a investigate defect behavior in this region. bi-material interface must be implemented into the current PFDD framework in order to study defect/ Studying deformation twinning is another key goal of this interface interactions. This model advancement will project (collaborator: Irene Beyerlein- T-3). Current efforts present opportunities to study other interfaces in fcc include simulations studying the formation of deforma- materials such as grain boundaries. tion twins in copper, nickel, and silver. More specifically, we are investigating the emergence of twins from grain 3. Account for dislocation behavior in base-centered boundaries in several different initial defect configurations cubic (bcc) materials, such as tantalum, steel, chromi- proposed in literature. This work also investigates the ef- um, and niobium. This final goal requires a complete fects of grain size variation, and material variation on the reformulation of the current 3D PFDD model, making emergence of deformation twins in fcc metals. it the first model of its kind applied to bcc materials. Describing the dislocation core in bcc materials will The final goal of this project requires reformulation of the be both the most difficult challenge to overcome in PFDD model to account for a different material crystal achieving this goal, and the most important compo- class, bcc. Initial reformulation of the model has begun, in nent for accurately modeling dislocation behavior in particular, modification of crystallographic considerations bcc materials. The PFDD model has a direct connec- such as number of slip planes and general geometry. Next tion with atomistic simulation tools, which is particu- steps include simulations of perfect dislocations in bcc larly advantageous in addressing this problem. metals, followed by development and implementation of a 98

Page 99

Conclusion The primary goal of this proposal is to investigate na- noscale defect/interface interactions, such as nucleation and growth of deformation twins, and dislocation dynam- ics as controlled by interfaces through development of an innovative three-dimensional (3D) phase field dislocation dynamics (PFDD) mesoscale model that links atomic-scale and nanoscale physics. Model advancements as part of this project include: incorporation of temperature de- pendence, implementation of a complex description of nanolayer interfaces, and reformulation to account for body-centered cubic (bcc) crystal structures. Development of this computational tool will enable the study of defor- mation mechanisms at interfaces not possible with current mesoscale methods. Publications Hunter, A., R. F. Zhang, I. J. Beyerlein, T. C. Germann, and M. Koslowski. Dependence of equilibrium stacking fault width in fcc metals on the gamma-surface. 2013. MODELLING AND SIMULATION IN MATERIALS SCIENCE AND ENGINEERING. 21 (2): -. Hunter, A., R. F. Zhang, and I. J. Beyerlein. Universal scaling law for the dislocation core size in fcc metals. Physical Review Letters. Hunter, A., and I. J. Beyerlein. Unprecedented grain size effect on stacking fault width. 2013. Applied Physics Letters Materials. 1: 032109. 99

Page 100

Chemistry and Material Sciences Early Career Research Continuing Project Magnetic Field Effects on Convection-Modified Solid-Liquid Interfaces Amy J. Clarke 20130755ECR Introduction and microstructure evolution. Application of a magnetic In-situ characterization techniques are now affording field during solidification provides an opportunity to con- direct interrogation of opaque materials during syn- trol alloy melt fluid flow characteristics and the interface thesis and processing, in part because of pioneering shape and dynamic equilibrium at that interface. Ma- first experiments performed by this team to monitor terials for the future, including the creation of materials metallic alloy melting and solidification using x-rays and with application-tailored properties to achieve controlled protons. Feedback from in-situ characterization will en- functionality by directed synthesis and processing, is one able in-process adjustments to control solidification and of three science pillars at Los Alamos National Labora- microstructure evolution and permit the advancement of tory. Controlling materials synthesis and processing solidification theory and the development of predictive and the design of materials functionality through the solidification and microstructure evolution models. This exploitation of defects and interfaces, understanding project extends our methodology to explore the influ- the dynamic evolution of materials, defect structures, ence of magnetic field on the solid-liquid interface of and interfaces, and the use of novel characterization metallic alloys during solidification, demonstrating a new techniques, especially at extremes, are priorities listed in approach to further control microstructure evolution Los Alamos National Laboratory’s Directed Research and during processing. In the proposed project, we wish to Development Defects and Interfaces in Materials (DIM) directly observe the impact of controlled alloy melt fluid Exploratory Research Category. These priorities are also flow on the solid-liquid interface during solidification. reflected in the laboratory’s Matter-Radiation Interac- Melt flows interact with solid interfaces at a variety of tions in Extremes (MaRIE) Signature Facility vision, and length scales, from the macro-scale down to the micro- are deeply rooted in nationally recognized needs identi- scale dendritic tips to influence solute segregation and fied by the U.S. DOE Basic Energy Sciences, The National microstructure evolution. The dynamic equilibrium pres- Academies, and the White House Office of Science and ent at the solid-liquid interface is critical to morphology Technology Policy’s Materials Genome Initiative (MGI). selection, and we intend to probe the selection criteria for important types of flow geometries. Specific flows in Progress metallic systems will be achieved by controlling geom- The broad goal of this program is to better understand etry, heat flux, and the application of magnetic field. The the impact of fluid flow on solid-liquid interfaces during solid-liquid interface shape, local composition, and the solidification and melting by using magnetic field as a influence of magnetic field on the microstructure evolu- way to change the flow characteristics. To satisfy these tion during processing will be determined from in-situ goals, furnaces and crucibles must be built; upon initia- imaging using synchrotron x-ray radiography at Argonne tion of this program in January of 2013, design of these National Laboratory’s Advanced Photon Source. parts began. We first re-examined the flow conditions from the dynamic imaging data which prompted the idea Benefit to National Security Missions for the proposed work and reviewed relevant informa- Convection of an alloy melt during casting mold filling tion available in the literature. Three types of flow condi- plays a critical role in the solute segregation that devel- tions have been selected for further study. Two involve ops at the macro-scale, whereas localized convection the development of natural convection patterns in the cells that develop during solidification drastically influ- melt, and a third flow condition will be based upon ence micro-scale solute segregation and morphological forced fluid flow, which will have direct relevance to ap- 100

Page 101

plications ranging from casting to microfluidics. The first will be studied. We will use synchrotron x-ray imaging to iteration of one of our experimental designs will promote create movies of the solid-liquid interface during solidifica- 2D convection cells and is currently being constructed. tion, and will demonstrate the potential of magnetic field Potential designs for a forced-flow crucible are currently to modify the interface characteristics. The use of mag- being modeled to determine desired heat flow conditions. netic field during solidification will provide a set of broadly A portable 0.3T permanent magnet has been acquired and applicable microstructure and property control schemes will be used for the planned experiments. An adjustable- for the creation of structural alloys. power electromagnet is currently being installed for access to high fields up to 1.5T. Considering the progress in the design of crucible geometries and the acquisition of suf- ficient magnetic field capabilities, we are currently on pace with our original schedule. Future Work The application of magnetic field has significant potential to influence alloy melt fluid flow, heat transfer, and the shape of the solid-liquid interface to affect morphological (and microstructure) evolution during solidification. The role of liquid motion on solid-liquid interface character- istics will be examined and real-time x-ray imaging at a national user facility will capture the effects. Fiscal Year 2013 will be dedicated to developing the experimental capabilities needed for these experiments. Furnace and casting geometry design will commence immediately. Geometry and heat flow conditions will be designed at Los Alamos National Laboratory utilizing multiple characteriza- tion methods, including experimental validation, imaging, and finite element simulation. The goal for the Fiscal Year 2013 design phase and experimentation is to develop a response surface of characteristics, such as liquid veloci- ties, as a function of our control parameters, so that a high degree of reproducibility is achieved. Ex-situ microstruc- tural analysis will provide the preliminary validation. Along with experimental platform design, candidate materials will also be selected based upon the experimental results. In this way, the sampling ranges may be optimized to fully explore the theoretical response space. We anticipate identification of at least two alloy or composite systems to use as the primary focus of experimentation during the second year of this project. The specific Fiscal Year 2013 tasks include: the design and fabrication of furnace geome- tries; validation of designs and measurement of achievable flow velocity ranges; and the selection of candidate alloys. The real-time imaging experiments will be performed in Fiscal Year 2014; analysis of the image sequences and data visualization will start in the same year. Conclusion In this project, we will explore the role of magnetic field on solid-liquid interfaces during solidification. The influ- ence of different fluid flow regimes on solid-liquid inter- face shape, solute segregation, and morphology evolution 101

Page 102

Chemistry and Material Sciences Early Career Research Continuing Project Understanding The Catalytic Conversion of Oligosaccharides to Fuels and Chemical Feedstocks Andrew Sutton 20130757ECR Introduction 1. Identification and isolation of possible intermediates A procedure to produce branched alkanes containing ten 2. Use of isotopically enriched substrates for mechanis- and eleven carbon atoms from starch in two steps using tic investigation mild acidic conditions with lanthanide and palladium 3. Theoretical calculations of likely reaction pathways catalysts is known. The reaction proceeds well at 200 oC with input from experimental work, and validation under moderate pressure of H2 (200 psi); however, the of these pathways through Tasks 1 and 2 mechanism and the identity of any reaction intermedi- ates are unknown. If the steps involved in this general 4. Use of knowledge obtained from Tasks 1, 2 and 3 reaction can be better understood, routes that obviate to start developing NPMCs for the hydrogenation the need for expensive catalysts as well as the use of and HDO reaction chemistries, designed to achieve ambient reaction conditions might be achievable. transformations to the intermediates identified in Tasks 1 and 2 Benefit to National Security Missions 5. Task 4 will also address the isolation of intermedi- The work to be carried out develops the understanding ates with applicability to feedstocks of fundamental chemical reactions (i.e. Fundamental Chemistry Mission) with direct relevance to biomass Conclusion derived substrates to produce fuels and chemical feed- We expect to gain a more detailed understanding of stocks. Development of competitive alternative energy hydrodeoxygenation reaction mechanisms involved sources and feedstocks directly impact our national in alkane production from biomass and to develop an security mission. insight into potential new catalysts for performing these transformations. Progress We have made progress identifying a selection of inter- Publications mediates in the HDO reaction of the triketone. We have King, A. E., A. D. Sutton, and E. Batista. Mechanistic In- used labeled products to follow the reaction pathway sights into the Palladium-Catalyzed Reduction of Penta- and we now have a clearer understanding of the initial none to Pentane: Insights into a Biomass-to-Fuel Conver- deoxygenation steps. This has been applied to the Gar- sion Process. TBD. cia-Gonzales reaction and we have further modified the HDO of that reaction to progress in a step-wise reaction exhibiting much more control. We have investigated a range of different reaction conditions and have screened different Bronsted acids to perform the transformations. We have also moved to industrially relevant solvents which have further improved the process. Future Work Due to the two year time frame for the project, all com- ponents will be performed in tandem. These tasks are: 102

Page 103

Chemistry and Material Sciences Early Career Research Continuing Project Effects of Joining Processes on Bimetal Interface Content and Radiation Damage Resistance John S. Carpenter 20130764ECR Introduction Benefit to National Security Missions A fundamental study of the effect of pre-existing inter- Fossil Energy, DOD, Commerce and Transportation facial structure on the joinability and resulting mate- If joining of bulk nanostructured metals (BNMs) is rial properties of bulk nanostructured metals (BNM) is successful, this class of materials could reduce vehicle proposed. BNMs, due to the nanometer length scale, weight significantly without compromising and, per- have material properties and behaviors that arise as a haps, enhancing safety. These materials are lightweight, result of two factors: 1) interfacial type and 2) interfacial strong, and resistant to shock damage making BNMs density. Success has been achieved in fabricating strong useful in reducing weight of both civilian passenger cars and lightweight BNMs with industrial methods that and military vehicles. This reduction in weight would exhibit properties that far exceed those of bulk met- reduce the amount of fossil fuels needed providing more als through purposeful manipulation of both interfacial fossil energy security. density and interface type. In order for BNMs to be used in engineering applications, however, the reaction of the Nuclear Energy interfacial density and interface type to industrial joining Materials in nuclear fuel reactors experience extreme processes must be evaluated. radiation and temperatures over their lifetimes. BNMs have exhibited ‘self-healing’ interfaces that both remove The Cu-Nb bimetallic multilayer nanocomposite system radiation damage and provide thermal stability. These has been selected as the model system for the proposed interfaces would enable significantly improved safety, work. Controlled interfacial structures and controlled performance, and lifetime length for future advanced nanoscale interfacial densities in a bulk sheet form ap- reactor and fuel cycle design. propriate for joining studies have been demonstrated at LANL. Understanding the physics that govern the Basic Understanding of Materials dynamic and thermal stability of interfaces in Cu-Nb Although it is clear that interfacial type and density con- nanolamellar material will be valuable for the joining trol properties in BNMs, the stability of these interfaces of any material, BNM or otherwise, where interfaces under joining conditions has not been well-studied. comprise the primary control on material behavior and Only recently has a BNM with tailored interfacial types properties. and densities been fabricated. With this material in hand, concentrated studies on the relative stability and Cu-Nb multilayers also exhibit outstanding material dynamics of various interfacial structures can com- behavior such as high strength, thermal stability, and mence. ultrahigh resistance to both shock and radiation damage. Given the properties exhibited by Cu-Nb composites, Matter-Radiation Interactions in Extremes (MaRIE) a range of potential applications exist such as nuclear Signature Facility fuel cladding, armor for defense applications, and as a This study will investigate the stability of a well-con- strong, lightweight structural material in transportation. trolled and well-characterized microstructure under A secondary goal of this proposal, therefore, is to ex- the extreme conditions of radiation and joining. The plore methods for joining Cu-Nb nanolamellar material current study will make use of ex situ characterization in order to unlock its potential as a lightweight manufac- techniques that provide static properties. These results turable nanolayered composite. could form a seed study for future work at the MaRIE 103

Page 104

facility where in situ techniques could be used to manufac- a higher density (material B) was then attempted. The ture specific microstructures under dynamic control. tooling needed to be redesigned and significant changes in weld parameters were also required to successfully weld Progress this material. Refinement was again noted in the weld nugget but the length scale of the refined microstructure The overall goal of the project is to study the effect of pre- was now less than 80 nm. When similar weld parameters existing interfacial structure on the joinability and result- were used on material C, successful welding was not ing material properties of bulk nanostructured metals achieved as the material tore, instead. Further work is (BNM). Success has been achieved in fabricating strong needed in tool redesign and weld parameter selection that and lightweight BNMs with industrial methods that exhibit will allow joining of the 30 nm material. properties that far exceed those of bulk metals through purposeful manipulation of both interfacial density and Investigate stability of interfacial types and densities interface type. In order for BNMs to be used in engineer- under the effects of joining ing applications, however, the reaction of the interfacial Due to the length scale involved, the interfacial types and density and interface type to industrial joining processes densities in the welded zone are currently being studied must be evaluated. using a combination of electron microscopy based tech- niques. This work is ongoing and will form the bulk of the In order to study this, two welding techniques – fusion work over the final part of the first year. welding and friction stir welding (FSW) - were used on Cu- Nb bimetallic multilayer composites with varying nominal Investigate stability of mechanical properties after joining layer thicknesses. Cu and Nb are immiscible and the con- Mechanical properties were tested in the parent material, trolled geometry of the composites provides an embedded in the weld affected zone, and in the welded zone, or nug- measurement tool allowing microstructural changes due to get, with nanoindentation. Measured hardnesses within welding to be evaluated. Over the course of the first five the weld nugget for the material B were 200% greater than months of funding, work has commenced towards achiev- the parent material and 20% greater than the strongest ing the four goals seen in the technical description. Prog- Cu-Nb multilayer composite fabricated thus far (a Cu-Nb ress to this point will be described in the context of these multilayer with nominal layer thicknesses of 7 nm). This four goals. In addition, three Cu-Nb multilayer materials increase in strength is consistent with a grain size < 5nm have been investigated to this point – material A) multi- in the welded zone and might indicate that performing layer with nominal layer thicknesses (h) of 20 um, material FSW on a whole sheet of Cu-Nb multilayer nanocomposite B) multilayer with h = 300 nm, and material C) multilayer could lead to improved mechanical properties. with h = 30 nm. Publicize results Identify differences in welding parameters that can be Two talks (one invited) will be given at two materials associated with interfacial type and density science conferences over the next six months. A peer- Eight different sets of weld parameters were utilized during reviewed journal publication detailing the results obtained fusion welding in order to investigate how material A react- to this point is in preparation. ed to standard welding techniques. Delamination between layers as well as substantial porosity was noted for all sets of weld parameters. In addition, complete melting of the Future Work Cu phase occurred leading to a loss of layered geometry The goals of this project will be accomplished through in the weld and heat affected zone. These results showed completing the following tasks: that standard welding techniques do not present a usable technique for joining nanostructured metals while main- • Utilize two joining techniques, friction stir welding and taining the integrity of the overall material and the unique fusion welding, to fabricate three joined sheets of Cu- properties that stem from the nanostructured grains. Nb multilayer nanocomposite with each sheet exhibit- ing differences in interfacial type or density prior to FSW was performed on the same material utilized in the welding. fusion welding studies. Successful welding was achieved • Characterize interfacial type and density after joining with results indicating that Cu did not melt in order to through the use of diffraction, electron microscopy and accomplish joining. Instead, the layered structure was optical microscopy techniques. refined and mixed in the welded area from layers on the • Utilize nanoindentation in order to characterize the rel- order of 20 um thick to layers ~200 nm thick. Welding ative hardness between joined and unjoined regions. of Cu-Nb material with similar interfacial structure, but 104

Page 105

• Submit results to a peer-reviewed journal. • Present results in a conference talk. Conclusion The technical goal is to join two sheets of Cu-Nb bimetal- lic multilayer nanocomposites such that properties like strength, radiation damage resistance, and thermal stabil- ity are maintained or enhanced. Joining will be attempted on Cu-Nb materials that exhibit three different interfacial types and two different interfacial densities. Through this study the relative effect of interfacial type and density on joinability can be studied as well as the stability of various interfaces under extreme strain. In addition, if joining is successful while maintaining the material properties, the last hurdle inhibiting bulk metallic nanocomposites from use as structural materials will have been overcome. 105

Page 106

Chemistry and Material Sciences Early Career Research Continuing Project Probing Interface Reactions of Calcite Nanocrystals at Elevated Temperatures and Pressures Katharine L. Page 20130772ECR Introduction and temperature sample environment development for Calcite (the most stable form of calcium carbonate, local structure measurements, and differential neutron CaCO3) is an important mineral phase in both environ- total scattering measurements with fluid-solid mixtures. mental systems and energy applications. Calcium car- Along the way we will advance techniques based on neu- bonate is considered the long-term sink for atmospheric tron diffraction for characterizing the atomic structure at carbon dioxide (CO2), and thus the process by which the interfaces of fluids and nanomaterials. CO2 is trapped and transformed to stable carbonate minerals is of great fundamental and practical interest. Benefit to National Security Missions The main goal of this project will be to isolate and char- Understanding the behavior of mineral interfaces in geo- acterize the atomic interface structures that facilitate thermal fluids is important for expanding many of the precipitation and dissolution of carbonate species under earth’s energy options, including geothermal production, conditions found in underground reservoirs for CO2 hydrocarbon extraction, and geologic carbon sequestra- sequestration. tion. Successful study of the interface reactions in the CaCO3/D2O/CO2 system directly supports the missions There are many challenges to probing the relevant of reducing climate and energy impacts, and enabling interface reactions experimentally: in bulk carbon- environmental remediation and restoration. ate systems, the solid/fluid interface represents a very small proportion of the total system, resulting in low Sample environment capabilities and modeling meth- signal strength for many experimental measurements. odology developed with this work will also provide new Additionally, standard diffraction (the usual means of or improved research tools for probing complex materi- determining atomic structure) is intrinsically insufficient als, relevant to the MARIE and NNSA missions of LANL due to the amorphous nature of the fluid phase and the and the DOE Office of Science, as well as DOE Fossil non-periodic/highly disordered interface with the solid Energy, and Nuclear Energy Programs. For instance, the phase. As a further complication, the sequestering of temperature and pressure regimes and length scales of CO2 in underground reservoirs occurs under hot, pres- interest in this study are directly applicable to optimizing surized conditions. Appropriate sample environments hydrocarbon extraction from oil and gas reservoirs and further reduce signal from the interface. All these fac- understanding corrosion and scaling in nuclear reactors. tors have hindered progress in experimentally probing the atomic interface structures controlling precipitation Progress and dissolution of carbonates at crustal conditions. The project scope has four tasks associated with it: (1) We will use advanced scattering techniques based on careful synthesis and characterization of size dependent neutron total scattering measurements to target the nanocrystalline carbonate materials; (2) development atomic structure at the interface between the min- of high pressure and temperature cells compatible with eral calcite (and other carbonate species) and water/ local structure measurements; (3) neutron total scatter- carbon dioxide fluids at elevated temperatures and ing measurements of calcite-fluid solutions at elevated pressures. The approach will involve a size dependent pressure and temperature; and (4) a data simulation and (and therefore surface area dependent) set of carefully data modeling effort to interpret experimental results. characterized nanocrystalline samples, high pressure 106

Page 107

Task 1: In the last several months we have identified that Task 4: Neutron and X-ray data simulation at ambient con- very small and monodisperse calcite nanocrystal samples ditions (to establish and benchmark signal sensitivity for are quite difficult to synthesize. Nonetheless, a large CaCO3-D2O/CO2 fluid/solid interfaces) is in early stages. (100 gram) calcite nanoparticulate sample with an aver- Two data modeling approaches and software suites are age particle diameter of 40 nm has been prepared at the currently being compared and evaluated for their utility. Lujan Center by a senior adviser for this project, Dr. Luke Additional Accomplishments: In a recent unrelated visit Daemon. This sample has been characterized by labora- to Oak Ridge National Laboratory (ORNL), the PI became tory x-ray diffraction, thermogravimetric analysis, transmis- aware of relevant research led by Dr. David Wesolowski, Dr. sion electron microscopy (for morphology), and surface Andrew Stack, and Dr. Gernot Rother in the Geochemistry area measurements. Plans for both mechanical milling and Interfacial Science Group within the Chemical Science and digestive ripening of this sample have been made in Division. The idea of following calcite and carbonate min- hopes to produce smaller and larger average particle size eral growth within controlled pore glass (CPG) substrates samples, respectively. In addition, an external collabora- with capabilities developed in this project was nucleated. tion has been formed with Dr. Andrew Villegas-Jimenez The lack of experimental data for tracking the role of at the Carnegie Institute of Washington to leverage his water on the structural stability and transformation of the recent successes in producing high purity calcium carbon- metastable hydrous amorphous calcium carbonate (ACC) ate samples of homogenous nanometer size without using to calcite was also discussed as an area the present capa- chemical additives. bilities could address. Plans for collaborative and follow-on research are being discussed. Task 2: The PI, in consultation with the Lujan Center’s sample environment, high pressure, and mechanical design Finally, a postdoctoral associate offer has been accepted teams, has completed the engineering design of a series of by Dr. Hsiu-Wen Wang, a skilled mineralogist, geochemist, high pressure and temperature cells compatible with neu- and neutron scatterer, to join the PI in completing this proj- tron total scattering measurements (which require a very ect. She will begin work at LANL in August 2013, pending low and stable background scattering signal). The suite of completion of all LANL hiring protocols. gas pressure cells includes both mechanically weaker but proven vanadium material designs and mechanically stron- Future Work ger but unproven titanium-zirconium marerial designs, as Tasks and goals for the next fiscal year include: well as several different cell thicknesses. Finite element analysis was applied to the final designs to validate safety Task 1: completion of the synthesis and characterization of factors for experimental measurements. Specifically, we size dependent nanocrystalline calcite materials. We will have designed a 200 bar pressure cell for operation at complete preparation of a size-dependent series of five or room temperature and below, a 400 bar pressure cell for more homogenous and narrow size-distribution carbon- operation up to 500 K, and an 800 bar pressure cell for ate nanocrystal powders. The size series will provide an operation up to 750 K. The three generations of designs increasing availability of surface area for liquid-surface will be assembled and tested in July. Successful prototypes interaction per sample. The materials will be characterized will be tested in situ on the NPDF beamline in August and by laboratory x-ray diffraction, thermogravimetric analysis, September of this year, in hopes to identify the best com- transmission electron microscopy (for morphology), and promise of high temperature/pressure conditions and low surface area measurements. and stable background scattering. Task 2: We will continue development of high pressure Task 3: Four days of neutron scattering beamtime have and temperature cells compatible with local structure been awarded through the Lujan Center’s competitive measurements. Specifically, in the comping year three proposal review process to investigate the first series of design prototypes will be commissioned and evaluated for nanoparticulate calcite samples with the commissioned compatibility with the project goals. Any needed modifica- high temperature and high pressure cells on the NPDF tions will be addressed in the upcoming 2013 run cycle of instrument. Measurements will be scheduled for Sep- the Lujan Neutron Scattering Center. tember and October of this year. Complementary X-ray total scattering experiments have also been scheduled at Task 3: Neutron total scattering measurements will be beamline 11-ID-B at the Advanced Photon Source, Argonne completed for the initial series of calcite nanocrystal National Laboratory for the middle of August. The X-ray samples, and for calcite-fluid solutions at elevated pressure measurements will only be completed at ambient condi- and temperature by differential scattering techniques (by tions. subtracting the signals separately collected for the solid 107

Page 108

particles and D2O/CO2 fluid from the scattering signal of each solution). Complementary x-ray total scattering mea- surements will also be completed for the series of nano- crystal samples at ambient conditions. Task 4: A data simulation and data modeling effort will be initiated to interpret the experimental results. Interpre- tation of the results will involve both data simulation at ambient conditions (to establish and benchmark signal sensitivity for CaCO3-D2O/CO2 fluid/solid interfaces) and data modeling at all experimental conditions accessed. Two data modeling approaches and software suites will be evaluated for their utility. The feasibility of studying carbonate-mineral growth within controlled pore glass (CPG) substrates with developed capabilities will also be investigated. If feasible, the trans- formation of the metastable amorphous calcium carbonate (ACC) species to stable calcite species may be tracked. Conclusion We will isolate and characterize the atomic interface structures that determine precipitation and dissolution of carbonate species under conditions found in underground reservoirs. Conversion of CO2 into solid carbonate phases will provide the largest storage capacity within porous sandstone saline aquifers, and prevent leakage through cracks and fissures in shale and salt capping layers. Under- standing the species (composition and structure), condi- tions (temperature and pressure), and processes involved in the formation of stable carbonate minerals will aid mod- elers in building more realistic simulations, help to predict the success of geologic repositories, and assist in identify- ing candidate sites for sequestration. 108

Page 109

Chemistry and Material Sciences Early Career Research Final Report Exploiting Non-Innocent Ligands in Catalysis: New Base Metal Catalysts for the Reduction of Carbon Dioxide Susan K. Hanson 20110537ECR Abstract be possible to enhance the reactivity of earth-abun- Transition metal catalysts could have a transforma- dant metals by using “non-innocent” ligands capable tive impact on renewable energy technologies, but the of metal-ligand cooperativity. In addition, the results majority of effective catalysts are based on scarce and underscore the potential of cobalt complexes to replace expensive precious metals. The replacement of precious precious metals as catalysts for energy-related trans- metal catalysts with earth-abundant metals is thus a formations. By enabling more effective and sustainable key goal of sustainable and green chemistry, and was a renewable energy technologies, earth abundant metal primary focus of this project. This project evaluated the catalysts could have an important impact on domestic hypothesis that by designing catalysts that incorporated energy security. “non-innocent” ligands (ligands which could participate Background and Research Objectives in the reaction as proton donors or acceptors via metal- ligand cooperativity), it would be possible to enhance Amidst concerns over increasing carbon dioxide emis- the reactivity of earth-abundant metals. sions, worldwide energy consumption is projected to nearly double over the next 30 years. To meet growing Several different iron, nickel, and cobalt complexes were worldwide demands for energy without further increas- synthesized and the reactivity of these complexes was ing atmospheric carbon dioxide levels, it will be neces- evaluated. The most promising results were obtained sary to more effectively utilize non-petroleum based with nickel and cobalt complexes of a PNP pincer ligand. feedstocks. However, the conversions of renewable Two different nickel hydride complexes were prepared feedstocks to fuels, including the reduction of carbon and shown to catalyze the hydrogenation of olefins un- dioxide, are challenging reactions, requiring a catalyst der mild conditions. Both cationic and neutral cobalt(II) to perform a complex series of chemical steps. New and alkyl complexes of the cooperative pincer ligand PNHPCy more effective catalysts could improve our ability to har- (PNHPCy = bis(dicyclo-hexyl¬ethyl)amine) were synthe- ness renewable resources. sized. The cationic cobalt(II) alkyl compound was shown to be a catalyst for the hydrogenation of both olefins Transition metal catalysts promise to enable key trans- and substrates with polar multiple bonds (ketones, formations in renewable energy technologies, such aldehydes, and imines). In addition, the cobalt catalyst as biofuel production, carbon dioxide reduction, and showed a surprising and remarkable tolerance for differ- hydrogen storage and generation. However, current ent functional groups and water, suggesting its promise catalytic technologies are based on rare and expensive for energy related transformations. This discovery of a precious metals (Pt, Pd, Ir, Rh, and Ru). For widespread versatile, earth abundant metal cobalt hydrogenation applications in energy technologies, there may not be catalyst was an important breakthrough. Subsequent enough of these precious metals available in the world work demonstrated that the cobalt complex could also to meet the anticipated demand. Consequently, a major catalyze the acceptorless dehydrogenation of alcohols, goal of sustainable chemistry is to develop alternatives a reaction that has applications in hydrogen storage to precious metal catalysts based on inexpensive, read- and generation. Detailed studies of the reactivity of the ily available, and earth-abundant metals such as iron, cobalt catalyst provided important insights into the reac- cobalt, or nickel. tion mechanism. The goal of this proposal was to synthesize a new family Overall, the results of these studies confirm that it may of earth-abundant metal complexes of Ni, Co, and Fe, 109

Page 110

and develop these complexes as catalysts for hydrogena- complexes and their catalytic activity. tion reactions and the reduction of carbon dioxide. The Mechanistic experiments were performed to gain insight initial hypothesis was that incorporating a tethered base into the cobalt-catalyzed reactions. The role of metal- (proton acceptor) in the metal coordination sphere would ligand cooperativity in the reactions was assessed by com- increase catalytic activity or enable new reaction modes. paring the reactivity of cobalt catalyst 4 with the analogue The objective of this work was to prepare complexes of 5, where the central nitrogen of the ligand is substituted such “non-innocent” ligands and investigate if altering with a methyl group (Figure 4). If the N-H group on the the ligand framework around a non-precious metal center ligand plays an important role in the catalytic cycle, a sup- could result in better catalysts. pression in catalytic activity would be expected using 5. In Scientific Approach and Accomplishments fact, catalyst 5 showed diminished reactivity relative to 4 in the low temperature hydrogenation of ketones, suggest- In initial work, a number of different ligand candidates ing an important role for metal-ligand cooperativity and were evaluated in reactions with iron, cobalt, and nickel the “non-innocence” of the ligand. The full set of experi- metals. The most promising approach involved the synthe- ments and mechanistic insights gained from this work are sis of nickel and cobalt complexes of the ligand PNHPCy described in a full paper published in the Journal of the (PNHPCy = bis(dicyclohexylethyl)amine). Use of the PNHP- American Chemical Society. Cy ligand posed several advantages; including a meridonal tridentate coordination mode and large steric bulk, both Additional work explored the potential of cobalt complexes of which could help stabilize highly reactive species or to catalyze the acceptorless dehydrogenation of alcohols, intermediates. The PNHPCy ligand also had the potential to and the transfer hydrogenation of C=O and C=N bonds. participate in the reaction in a “non-innocent” fashion via These reactions have application in energy related technol- the basic central nitrogen of the pincer ligand (Figure 1). ogies, including hydrogen storage and generation and as a selective and low-temperature route to generate hydrogen Using the ligand PNHPCy, two new nickel hydride com- from biomass-derived alcohols and carbohydrates. The full plexes were synthesized (complexes 1 and 2, Figure 2). results were published in the journals Organic Letters and These complexes were tested as catalysts for the hydroge- Chemical Communications. nation of substrates with C=C and C=O bonds. The cationic nickel(II) hydride complex was an active catalyst for the In conclusion, using a “non-innocent” ligand was evaluated hydrogenation of olefins, although only limited reactivity as a strategy for enhancing the reactivity of earth-abun- was observed with substrates with C=O bonds. Mechanis- dant metals. Using such ligands, new and active examples tic experiments indicated that the increased reactivity of of nickel and cobalt hydrogenation catalysts were devel- the nickel complex with olefins may be due to an increased oped. The development of the cobalt catalyst was particu- propensity of olefins to undergo insertion into the Ni-H larly significant, given its high activity under mild condi- bond (C=O insertion into the nickel-hydride bond was not tions and robust functional group tolerance. These features observed). A full paper describing this work was published of the cobalt catalyst suggest that it could have important in the European Journal of Inorganic Chemistry. applications in energy-related technologies. Overall, the design of effective earth-abundant metal catalysts for the Neutral and cationic cobalt(II) alkyl complexes 3 and 4 hydrogenation of C=C and C=O bonds is a key step forward of the ligand PNHPCy were synthesized and evaluated as in our ability to more efficiently and cost-effectively utilize hydrogenation catalysts (Figure 2). The cationic cobalt(II) renewable resources like carbon dioxide. alkyl complex 4 was a highly active catalyst for the hy- drogenation of olefins. Carbon-heteroatom (C=O, C=N) Impact on National Missions bonds were also reduced under mild conditions using 4 as In a recent 2013 address to the nation, President Obama a catalyst. Additional experiments probed the functional stated that “the only way for America’s energy supply to be group tolerance of the cobalt catalyst. Surprisingly, cata- truly secure is by permanently reducing our dependence lyst 4 was not completely deactivated by water or oxygen on oil…We have to discover and produce cleaner, renew- functional groups. These results suggest the promise of 4 able sources of energy with less of the carbon pollution as an earth-abundant metal catalyst, since the ability to that threatens our climate. And we have to do it quickly.” tolerate oxygenated substrates will be critical for applica- In order to effectively use renewable sources of carbon tion in energy-related technologies. Details of this work (non-food biomass or carbon dioxide) as fuels, new cata- were published and featured on the cover of the journal lytic technologies are needed. In demonstrating that it is Angewantde Chemie International Edition (Figure 3). A pat- possible to use inexpensive, earth-abundant metals like ent application was filed describing the nickel and cobalt nickel and cobalt as catalysts for these applications, this 110

Page 111

work has provided an important conceptual and techno- logical advance. The work is relevant to Office of Basic Energy Sciences mission priorities. Figure 4. Comparison of the reactivity of cobalt complexes 4 and 5 provided insights into the role of metal-ligand cooperativity and the “non-innocent” ligand in the catalytic reactions. The two complexes differ in the substitution at the central nitrogen of the pincer ligand (N-H vs. N-methyl). Publications Lucas, M., S. Hanson, G. Wagner, D. Kimball, and K. Rector. Evidence for room temperature delignification of wood Figure 1. (a) The “non-innocent” PNP pincer ligand and (b) its using hydrogen peroxide and manganese acetate as a catalyst. 2012. Bioresource Technology. 119: 174. proposed ability to participate in a reaction by metal-ligand cooperativity. Vasudevan, K. V., B. L. Scott, and S. K. Hanson. Alkene Hy- drogenation Catalyzed by Nickel Hydride Complexes of an Aliphatic PNP Pincer Ligand. 2012. European Journal of Inorganic Chemistry. : 4898. Zhang, G., B. L. Scott, and S. K. Hanson. Mild and homoge- neous cobalt-catalyzed hydrogenation of C=C, C=O, and C=N bonds. 2013. Angewandte Chemie International Edition. 51: 12102. Zhang, G., K. V. Vasudevan, B. L. Scott, and S. K. Hanson. Understanding the mechanisms of cobalt-catalyzed hydrogenation and dehydrogenation reactions. 2013. Journal of the American Chemical Society. 135: 8668. Zhang, G., and S. K. Hanson. Cobalt-catalyzed acceptorless alcohol dehydrogenation: synthesis of imines from al- Figure 2. New nickel(II) hydride and cobalt(II) alkyl complexes cohols and amines . 2013. Organic Letters. 15: 650. synthesized and found to be active catalysts for hydrogenation reactions. Zhang, G., and S. K. Hanson. Cobalt-catalyzed transfer hydrogenation of C=O and C=N bonds. Chemical Com- munications. Thesubstitutionofcobaltforpreciousmetalsincatalysisisavariationonthe ancientalchemicalthemeoftransmutingbasemetalsintopreciousones.Intheir Communication(DOI:10.1002/anie.201206051),S.K.Hansonetal.describethe cobalt-catalyzedhydrogenationofalkenes,aldehydes,ketones,andiminesunder mildreactionconditions.Theprecatalystisacationiccobalt(II)alkylcomplex. (TheartworkisbyJoshSmith,LosAlamosNationalLaboratory.) Figure 3. Cobalt-catalyzed hydrogenation reactions featured on the cover of Angewandte Chemie International Edition. 111

Page 112

Chemistry and Material Sciences Early Career Research Final Report Advanced Materials for Exploring New Electronic and Optical Behaviors in Single-walled Carbon Nanotube Systems Juan Duque 20120502ECR Abstract tube surface, and reactivity with other organic molecules A general challenge in generating functional materials can be controlled. Recent reports have shown that from nanoscale components is integrating them into multi-functional materials may be obtained by tailoring useful composites that retain or enhance the properties interactions between tubes and other species. Of par- of interest. Because the unique electronic and optical ticular interest is to study the charge transfer (CT) and properties of single-walled carbon nanotubes (SWCNT) energy transfer (ET) behavior in these multi-component are highly sensitive to environmental effects, incorporat- materials. When studying ET behaviors, important mate- ing SWCNTs into useful composites which retain and/or rial functionality can be extracted including interstitial enhance the photophysical properties for optoelectron- spacing, inter-particle coupling, and percolation. For in- ics and sensing applications is still a major challenge. stance, Förster resonant ET (FRET) in bundles of different bandgap semiconducting tubes has resulted in diameter Background and Research Objectives dependent exciton ET. Similar FRET observations have The photophysical properties of SWNTs are highly stable been reported with quantum dots, polymers, and metal and excitonic in nature, with discrete levels below the porphyrins. However, a major drawback is the lack of single-particle excitation spectra. While long radiative control of the bundle size, tube type, and interactions. lifetimes (~ns) suggest higher photoluminescence (PL) Thus, the general challenge of generating functional quantum yields (QY), only modest QY ~7-20% have been materials from nanoscale components and integrating reported, as consequence of the lack of homogene- them into useful composites that retain or enhance the ity in the optical properties at the ensemble level and properties of interest still remains. Nevertheless, ET is the strong environmental sensitivity. Emerging data the basis for understanding how to build up more com- indicates high mobility of the exciton in 1-D along the plex interactions within systems. SWNT surface efficiently quenches PL by rapidly carrying If properly prepared, aerogels can offer the alternative excitons to defect sites where they dissociate into free of having a solution- and surfactant-free environment charge carriers. Moreover, environmental perturbations where the inter-particle interactions may be controlled. can also couple emissive states (dark) with optically al- Aerogel preparation methods can be modified to op- lowed states (bright) affecting exciton behavior. Although timize interstitial distances and optical density, as well these photo-generated carriers are of interest for energy as tune the pore size distribution, thus allowing signifi- harvesting, PL QY is compromised since exciton binding cant degrees of freedom for engineering of the desired and exciton diffusion behaviors are directly affected. A properties. Although SWNT-aerogel composites have key capability in meeting these challenges is the ability been investigated for their mechanical and electronic to tune interfacial behaviors and obtain photolumines- properties, no reports to date have demonstrated SWNT- cent solution- and surfactant-free systems with minimal aerogels with an observable emission signal. Recently, and/or controllable perturbations from the local envi- here at LANL, we developed SWNT-aerogel nanocom- ronment; the presence of aqueous media and suspend- posites with tunable local environmental interactions. ing agents such as surfactants and polymers masks the These nanocomposites exhibit a behavior characteristic intrinsic photophysical behavior of the tubes. of non-interacting samples in the absence of solution The ability to tune interfacial behaviors presents a rich and surfactants. These silica matrices are of particular opportunity for engineering and optimizing SWNT opti- importance because they provide the first bulk-phase cal response since surfactant structure, access to the platform to develop a fundamental understanding of the 112

Page 113

link between SWNT interfacial chemistry and underlying derstood. The effect is due to binding of REI to surfactant photophysical behaviors, while also providing a suitable molecules followed by formation of a closed shell micelle, matrix for development of relevant applications. This work completely screening REIs from the water molecules. The pioneers the currently unexplored area of SWNTs in solu- process is shown to be slowed down in the crowded envi- tion- and surfactant-free systems for basic research as well ronment of a hydrogel. as for energy and optoelectronics applications. We studied behavior of Tb ions in different environments. Engineering material functionality through interfacial inter- We found that Tb ions quickly form long-lived PL com- actions is a major component in developing the next gen- plexes in solution containing surfactant (DOC) molecules. eration of composites for energy harvesting, sensing, and This process is faster than the diffusion through the DOC/ photonic applications. A general challenge in generating SWNT silica hydrogel. The PL lifetime in these complexes functional materials from nanoscale components is inte- increases to 2.3 ms for DOC/SWNT hydrogel with time. grating them into useful composites that retain or enhance Similar changes of the shape of Tb lines of steady-state the properties of interest. Here, we aim to design and PL spectra in DOC bulk solution and DOC/SWNT silica gel, engineer the functionality of nano-structured materials, compared to TbCl3 water solution, suggest the formation specifically SWNT nanocomposites, by tuning interfacial of DOC micelles in both cases (Figure 2). behaviors and interactions (such as surfactant structure, dopants, and matrix composition), thus controlling inter- FRET tube and matrix interactions. • Control over energy transfer • Surfactant dependent FRET Scientific Approach and Accomplishments We used an innovative method developed here at LANL by our team to obtain highly fluorescent and low density SWCNT-aerogel mesocomposites. The approach provides PL Lifetime a rationale for tailoring SWCNT photophysical response via interfacial chemistry. In gel • Large surface area of tubes act as templates for Tb ions • Crowded environment of gel slows Using this mesomaterials we monitored the interaction of down ion mobility molecules inside and outside carbon nanotubes and the ef- In water fect of confinement. Moreover, we examined the changes in dielectric environment due to local environment (Figure 1). Figure 2. Lifetime of rare-earth ions inside hydrogels. (12,1) As result of this work we have establish several collabora- (11,3) tions inside LANL but also with universities, which have Vacuum Open resulted in data for upcoming publications. Closed Nitrogen Open Impact on National Missions Closed (10,5) This work focused on the development of new mesocom- (9,7) posites with tunable and desire properties for sensing and photonics applications. Through my early career project I was able to establish and maintain two capabilities very 210 220 230 240 important to the mission of the laboratory: Raman Shift (cm-1) Figure 1. Raman spectra of carbon nanotubes empty and filled The unique and extensive resonant Raman excitation capa- with water. bility (with wavelengths extending from 350 to 1000 nm) at LANL’s Integrated Spectroscopy Laboratory (residing in We also studied the energy transfer of nanotubes and Chemistry Division) is recognized worldwide. This capabil- rare-earth ions. Formation and evolution of highly photolu- ity has fostered both internal and external (national and minescent complexes of rare-earth ions (REI) and bile salt international) collaborations, which have resulted in signifi- molecules is observed in a crowded environment of silica cant discoveries appearing in well-over 100 peer-reviewed hydrogels. By comparing the behavior of REI complexes in publications in high-profile journals including Nature jour- bulk solution and small compartments inside the hydro- nals, Phys. Rev. Lett., J. Am. Chem. Soc., etc. The impacted gel, the nature of the 5x-longer lifetime of REIs was un- subject areas have spanned many disciplines and include 113

Page 114

nanoscience (carbon nanotubes, graphene, quantum-dots, M. Tu, M. Zheng, and S. K. Doorn. Quantum Interfer- nanowires), thin film development, actinide chemistry, ence between the Third and Fourth Exciton States in plasmonic materials, biophysics, and inorganic photo- and Semiconducting Carbon Nanotubes Using Resonance Raman Spectroscopy. 2012. Physical Review Letters- coordination chemistry among others, putting the labora- Physical Review Letters. 108 (11). tory at the vanguard in nanophotonics, optoelectronics, materials, chemistry, and biosciences research. Duque, J. G., L. Oudjedi, J. J. Crochet, S. Tretiak, B. Lounis, S. K. Doorn, and L. Cognet. Mechanism of Electrolyte- Nevertheless, a major drawback with the system was Induced Brightening in Single-Wall Carbon Nanotubes. the lack of a microscope that couples all of the excita- 2013. Journal of the American Chemical SocietyJournal tion wavelengths, namely a continuous excitation, high- of the American Chemical Society. 135 (9): 3379. resolution Resonance Raman microscope. The ability to spatially resolve Raman features with multiple excitation Duque, J., and e. et al. Fluorescent Single-Walled Carbon wavelengths in one experiment will allow correlation of a Nanotube Aerogels in Surfactant-free Environments. 2011. ACS Nano. 5 (8): 6686. wide range of optical features with morphological dif- ferences in materials. This is vital to understanding the Haroz, E. H., J. G. Duque, B. Y. Lu, P. Nikolaev, S. Arepalli, R. optical, electronic, and catalytic properties, among others, H. Hauge, S. K. Doorn, and J. Kono. Unique Origin of of materials, particles, and molecules. The addition of a Colors of Armchair Carbon Nanotubes. 2012. Journal confocal microscope in conjunction with this project is at of the American Chemical SocietyJournal of the Ameri- the forefront of Raman research, opening the door to new can Chemical Society. 134 (10): 4461. groundbreaking studies, and makes the new spatially cor- Haroz, E. H., J. G. Duque, W. D. Rice, C. G. Densmore, J. related Raman, Fluorescence and Absorption Microscope a Kono, and S. K. Doorn. Resonant Raman spectroscopy vital tool to materials researchers in support of our nation- of armchair carbon nanotubes: Absence of broad G(-) al security mission. feature. 2011. Physical Review BPhysical Review B. 84 (12). Publications Arnold, M. S., J. L. Blackburn, J. J. Crochet, S. K. Doorn, J. G. Haroz, E. H., J. G. Duque, X. M. Tu, M. Zheng, A. R. Walker, Duque, A. Mohite, and H. Telg. Recent developments R. H. Hauge, S. K. Doorn, and J. Kono. Fundamental in the photophysics of single-walled carbon nanotubes optical processes in armchair carbon nanotubes. 2013. for their use as active and passive material elements NanoscaleNanoscale. 5 (4): 1411. in thin film photovoltaics. 2013. Physical Chemistry Chemical PhysicsPhysical Chemistry Chemical Physics. Telg, H., J. G. Duque, M. Staiger, X. M. Tu, F. Hennrich, M. 15 (36): 14896. M. Kappes, M. Zheng, J. Maultzsch, C. Thomsen, and S. K. Doorn. Chiral Index Dependence of the G(+) and G(-) Crochet, J. J., J. G. Duque, J. H. Werner, B. Lounis, L. Cog- Raman Modes in Semiconducting Carbon Nanotubes. net, and S. K. Doorn. Disorder Limited Exciton Trans- 2012. Acs NanoAcs Nano. 6 (1): 904. port in Colloidal Single-Wall Carbon Nanotubes. 2012. Nano LettersNano Letters. 12 (10): 5091. Crochet, J. J., J. G. Duque, J. H. Werner, and S. K. Doorn. Photoluminescence imaging of electronic-impurity- induced exciton quenching in single-walled carbon nanotubes. 2012. Nature NanotechnologyNature Nan- otechnology. 7 (2): 126. Crochet, J., J. Duque, J. Werner, and S. Doorn. Photolu- minescence imaging of electronic-impurity-induced exciton quenching in single-walled carbon nanotubes. 2012. NATURE NANOTECHNOLOGY. 7 (7): 126. Duque, J. G., G. Gupta, L. Cognet, B. Lounis, S. K. Doorn, and A. M. Dattelbaum. New Route to Fluorescent Sin- gle-Walled Carbon Nanotube/Silica Nanocomposites: Balancing Fluorescence Intensity and Environmental Sensitivity. 2011. Journal of Physical Chemistry CJour- nal of Physical Chemistry C. 115 (31): 15147. Duque, J. G., H. Telg, H. Chen, A. K. Swan, A. P. Shreve, X. 114

Page 115

Chemistry and Material Sciences Early Career Research Final Report Valence, Coordination and Reactivity at Interfaces Christopher D. Taylor 20120510ECR Abstract particles’ size.[1] Redox reactions and interdiffusion at Metal/oxide interfaces are critical to technological appli- heterophase junctions can change the nature of charge cations in heterophase junctions, metal/support interac- transport, modify surface dipoles and create new com- tions in catalysis, stability of electrocatalysts, the corro- pounds at the interphase.[2] Furthermore, it has been sion of materials, and thin-film technologies. The physics known for some time that metal stability is conferred by controlling such interfaces is opaque: band-bending and formation of passive oxide films.[3] interface dipoles, redox chemistry, interdiffusion, and High-Resolution Characterization: In situ electrochemi- interfacial energy considerations obfuscate the construc- cal scanning tunneling microscopy (EC-STM) and X-ray tion of clear predictive models. Furthermore, the com- asdorption spectroscopy (XAS) provide information plexity of the interface structure and redox chemistry about the structure and composition of materials/envi- places it at the very limits of what we can tackle using ronment interfaces. EC-STM has been applied to charac- first-principles, electronic-structure based models. Ac- terize monolayer growth, epitaxial relations and elec- cordingly, we propose that theory should begin by simu- trochemical potentials associated with passivation.[4, 5] lating known configurations of the metal/oxide interface, The atomic resolution provides insights into preferred obtained from well-defined, in situ, high-resolution coordination at these interfaces, but does not answer surface science measurements. Several such case stud- questions regarding valence or transport. Pugmire and ies have appeared in the recent literature, for the cases co-workers demonstrated with XAS that the Pu metal/ of oxide films grown on Mg, Ni, and Pu metal. Conse- oxide interface consists of a sub-stoichiometric Pu2O3- quently, we have constructed first-principles models for y phase that lies somewhere between metallic Pu and these experimentally determined interfacial structures the sesquioxide Pu2O3.[6] Valence and coordination and quantify the interfaces in terms of valence (charge information of thin oxide layers have been gleaned for states at the interface), coordination (structure and local platinum metal nanoparticles in fuel cells via X-ray probe bonding at the interface) and reactivity (interfacial ener- techniques.[7] Experimental characterization provides a gies). In doing so, we provide fundamental information starting point for theory to probe the nature of interme- concerning the charge states of ions that exist at this diate states and the kinetics of transformations. interface (somewhere between a metallic and an ionic state), demonstrate new capabilities for understanding Modeling Oxide Growth: The process of oxidation has interfacial (two-dimensional) systems, and open up new been described by several models, the most definitive materials design concepts for exercising control over of which is the point defect model (PDM) of Macdonald interfaces in metal/oxide and metal/ceramic systems. et al.[8, 9] Fundamental is the rate-dependence of oxide This first-principles data is foundational for developing film growth via transport of vacancies across the film, new capabilities in the molecular dynamics simulation from the metal to the environment. Although PDM is of charge-transfer systems relevant to electrocatalysis, “deterministic”,[10] there is, at present, no easy way to metal/ceramic interfaces, and energy generation. determine the rates of many of these processes from first-principles, other than by electronic structure com- Background and Research Objectives putations. Surface states and interfaces critically define materials behavior and response. Electrocatalysis by nanoparticles, Such computations applied to metal/oxide interfaces for example, can be limited by oxidation states that are ambitious as system representations must be large emerge under electrochemical conditions unique to the 115

Page 116

enough to take into account relative orientations of the These oxygen atoms attract and form hexagonal clusters two interfaces and the presence of defects. Coarse-grained that co-exist with clean magnesium surface[25]. This be- approaches, such as combining interatomic potentials for havior is distinct from the general outline above, suggest- metals with Coulomb potentials,[11] can provide some ing an alternative model for oxidation.[37, 38]. Therefore, means for simulating defect transport,[12] but the physics Mg oxidation was considered an attractive model system that defines the intermediate states – lying somewhere in to begin application of our method. between that of a metal and an ion – lies beyond the scope When comparing our findings to the four step process of these methods. While innovative frameworks for captur- described above distinctives for Mg emerge immediately: ing these “superposition” (ionic+metallic) states are being The subsurface tetrahedral (tet-1) is, in fact, most stable: developed by Valone in MST-8/LANL[13] the implementa- adsorption at the surface sites (atop, bridge, and hcp) is tion of such models requires verifiable electronic structure unstable. In addition to this distinctive, O*-O* interactions data to construct the model Hamiltonian. By studying are unique in that they are attractive, not repulsive. In the selected systems for which high-resolution measurements progression from single, to three- and seven-atom clusters, of the interface are available, we have created a first-prin- there is an accumulative effect of progressively stronger ciples framework. binding, i.e. a thermodynamic driving force accelerates Scientific Approach and Accomplishments MgO cluster nucleation. When the work function of these structures was calculated, the tetrahedral-1 clus- First-principles studies of electrochemical interfaces: Our ters uniquely demonstrate a reduced work function. This research team had begun systematic development of first- explains the previous determination of a diminished work principles approaches for modeling the materials/environ- function upon initial surface oxidation obtained using both ment interface.[14-18] Using first-principles calculation, photoelectric and Kelvin probe techniques[37, 38, 41]. atomistic simulation, MD and constrained optimization we investigated factors controlling metal dissolution.[19, 20] The differential binding provides further insight into the The strength of metal-metal interactions was related to mechanisms of oxide nucleation (Figure 1). The differential local coordination and alloy composition. In this work, we binding energies indicate that the smallest clusters grow generalized this approach to look not only at dissolution of laterally, and the 7-atom clusters grow subsurface. The 10- metals (Fe alloys) but also the formation of oxides on met- atom nucleus grows in a non-trivial pattern, intercalating als (Mg, Ni, Pu). further oxygen atoms within the Mg plane. The multilayer oxide has been previously studied using experiment and Incipient Magnesium Oxidation theory[42, 43]. Schröder et al. used x-ray photoelectron Oxide initiation on metal surfaces has been characterized diffraction, single scattering cluster[44], and DFT to study in four-steps[21-33]. O2 dissociates over a metal surface oxide structure on Mg(0001).[39, 40] They found that the to form chemisorbed oxygen[34-36]. These oxygen atoms multilayered oxide was a mixed 2 and 3 ML structure on sit over high-coordination hollow sites on the atomisti- top of an almost undistorted Mg(0001) surface. The Mg-O cally corrugated surface planes[21]. Repulsive interac- layers were reconstructed into a “graphite-like” structure. tions between surface O atoms results in a 2-dimensional Our work shows that this structure emerges from oxidation phase[25-29]. Further oxygen exposure enhances oxygen of the O10 nuclei. coverage, leading to new 2-dimensional phases that ac- commodate pair-wise interactions. Once a critical surface Applying first principles thermodynamics[45-47] we find coverage occurs, oxygen atoms ingress[33] sub-surface[30, that attractive Mg-O interactions result in phase separa- 31]. In the third step subsurface oxygen atoms experi- tion. This prediction is supported by experiment - transmis- ence mutual attraction and combine to form oxide nu- sion electron microscope images of O2/Mg(0001) show clei[30-32]. Further uptake of oxygen at the surface leads formation of MgO platelets separated by regions of free to bulk oxide formation[21]. This process is self-limiting, metal[48, 49]. Phase separation results in a renewal of re- with a critical concentration required before the system active metal surface and persistence of oxidation until the can pass to the next stage of oxide formation. surface is completely populated with oxide nuclei: there is no immediate monolayer or sub-monolayer pre-passiv- While magnesium may be considered one of the simplest ation, hence the strong reactivity of Mg to oxidation. metals (i.e. 2s2 electron configuration) its oxide formation is not consistent with this scheme [25, 37-40]. In the first Extensions beyond LDRD include study of Mg-alloy sys- stages O2 dissociates to form chemisorbed oxygen, O*. tems, critical to development of lightweight alloys for Rather than adsorbing on the surface, the oxygen atoms biomedical, transportation and aerospace. Opportunities appear to immediately chemisorb subsurface [17, 18]. are being pursued with NIH and Boeing. 116

Page 117

Multi-layer Oxides on Nickel to become oxide-like. Generally corrosion-resistant, nickel alloys are susceptible From the energetic analysis, nickel oxidation adopts the to localized corrosion. Proper elucidation of the problem conventional scheme described above. Oxidation initi- requires detailed understanding of the passive film, which ates at three-fold hollow sites, higher exposure leads to can be begun by establishing a first-principles program to step-site migration and geometries that maximize oxygen- simulate oxide/metal interfaces on nickel alloys. Oxide for- oxygen distances and/or provide metallic screening. As mation on nickel surfaces was studied using first principles bilayers form, the system adopts a non-oxygen terminated electronic structure calculations. As for Mg, our desire was surface that prevents reaction with the oxide layer and to understand the transition from nickel to nickel oxide, further surface oxidation. and quantify oxide formation in terms of valence, coordi- nation and reactivity. Initial Oxidation of Plutonium Passivation of plutonium is relevant to laboratory missions. In the case of nickel, DFT methods need to be well-tested The sequence of chemisorption steps leading up to mono- to ensure electronic structure is determined correctly. For layer oxide formation on δ-Pu(111) was simulated via DFT. instance, it is known that band gaps for nickel oxide are As in nickel, chemisorption begins by occupation of oxygen incorrectly determined from conventional DFT, but self- at a three fold hollow site, a configuration maintained up energy corrections remedy this problem.[50] Conversely, to 0.75 ML. At 1.0 ML, significant reconstruction takes metallic phases of nickel are well-described by conven- place, pushing Pu into an α-like configuration. The oxygen- tional DFT, but self-energy corrections introduce significant plutonium binding relationships move towards a prefer- errors. ence for four-fold coordination (Figure 2). We first studied single oxygen atom adsorption on a (2x2) The sequential chemisorption steps are exothermic in na- unit cell of Ni(111). Single oxygen adsorption at an fcc site ture, suggestive of the high reactivity towards oxygen and on the Ni(111) was the most stable when performing both the difficulty encountered in surface science studies when PBE (conventional DFT) and PBE+U (DFT with self-energy deoxygenated surfaces are desired.[56, 57] Thermody- corrections). Higher charge transfer from nickel to oxy- namic analysis implies a preference for uniform oxidation gen was observed in all PBE+U calculations. Three oxygen as opposed to islanding, consistent with the experimental atoms were placed along the (111) plane on the Ni(111) finding that there is a continuum in oxide stoichiometries surface. There was no reorganization in the PBE calculated across the metal/oxide interface.[58] geometry, maintaining (111) site coordination, akin to ex- perimental geometries with hydrated surfaces. The PBE+U Differences in local coordination and structure are cor- calculations led to a reconstruction of the Ni(111) surface. related with charge transfer. The Pu-O interaction involves A nickel atom terminated octopolar reconstruction was ob- charge transfer, mostly from s and d states of Pu to oxygen, served, similar to the p(2 x 2) reconstruction predicted by but also involving back-bonding from oxygen back to the Wolf and has subsequently been observed experimentally. Pu p and f states. Classical electrostatic models overes- [51-55] This suggests that the nickel atoms at the surface timate the extent to which oxygen polarizes the surface. do adopt more of a nickel-oxide like structure, since PBE+U The quantum mechanical (Bader [59-61]) analysis provides leads to a structure that resembles experiment. a consistent charge state for oxygen throughout all the interactions – from low to high coverage – and charge We considered a Ni(111) surface covered by an entire transfer is principally confined to the top monolayer of Pu monolayer of NiO. We observed a reconstruction of the atoms and the chemisorbed O. The maximum charge state surface using PBE that resulted in an oxygen terminated induced on Pu for the reconstructed 1.0 ML system is +1.8 octopolar structure. Two of the three PBE+U variations re- electrons, suggestive of an emerging Pu2O3 phase. Experi- sulted in oxygen terminated octopolar-like structures, but ment suggests that there exist a range of oxidation states with nickel clusters visible. Calculations with all interface in the passive oxide film: PuO (an oxidation state of (II), Ni atoms treated with the self-energy correction resulted Pu2O3 (an oxidation state of (III)) and, finally, PuO2.[57, in the preservation of the (111) plane ordering with higher 58, 62] binding energy and charge transfer to the oxygen atoms. The preference for no reconstruction tracks more with the The variety of charge states emerging from this quantum OH termination of nickel oxidation seen in ex-situ environ- mechanically based study, and the flexibility to reconstruc- ments as being stable and more favorable than oxygen tion provide some insight as to how the diversity in stoi- terminated surfaces.[4, 5] The agreement between these chiometries comes about and provides the foundation for latter calculations and the STM results[4, 5] implies that all further studies of metal/multilayer systems.[63, 64] interface atoms undergo a transition in electronic structure 117

Page 118

Coordination, Valence and Reactivity on Iron before-break” mechanism observed for copper dissolution The process of dissolution entails a change in the charge [65]. state of the atom at the interface between a material and its environment. The dissolution process on iron-based Impact on National Missions alloys was simulated by vertically displacing a surface The interplay of valence, coordination and reactivity that adatom above the lowest energy Fe(110) surface. corresponds to the transformation of metals into oxides or dissolved ions touches aspects of LANL Mission and The energy curve that controls the reactivity of metal Capabilities ranging from aging of actinide materials, atoms during dissolution can be divided into three distinct through to thin film fabrication, catalysis, energy genera- regions or stages.(Figure 3a) In the early stage of displace- tion, and nuclear fuel materials. The techniques and theory ment (stage I – elevation of adatoms is less than 1.0 Å), that we have demonstrated and advanced in this work are the energy is controlled by strong metal surface bonding anticipated to open up new directions to explore the role and the work is paid to the potential energy required for of processing techniques such as alloying, grain-boundary stretching the interatomic bonds. If the height is not con- engineering, surface texturing and thin-film fabrication to strained, there is a tendency for adatoms to be drawn back create materials with “smart interfaces.” Already, there to the surface, therefore the bond stretching in this stage have been impacts from this work in the study of oxide- could be regarded as elastic, i.e. reversible. If the adatom is strengthened metal alloys, corrosion simulations devel- constrained it tends to draw surface atoms towards itself. oped in partnership with US industries, and in a recently The middle part of the energy curve (stage II – adatom’s el- awarded LDRD-DR effort aimed at exploring surface chem- evation between 1.0 and 2.5 Å) corresponds to the region istry mechanisms on plutonium surfaces. of bond breaking. During this stage, a certain amount of energy is increasingly required for breaking the bonds and separating the adatom from the surface. In the last part of the curve (stage III - elevation is greater than 2.5 Å), the energy curve is stable, indicating a “free zone” in which the adspecies energy is determined by the formation of bonds with surrounding solvents (if any). The work of separation increases in the order of Cr < Fe < Mo adatom. In other words, the dissolution of adsorbed Mo is relatively more difficult than that of Fe and Cr in the same circumstance. With no solvents, the native adsorbed atom (Fe in this study) is neutrally charged (Figure 3c). When the material is an alloy, electronegativity differences result in charge transfer. Cr and Mo, respectively, have positive and nega- tive charges in adsorbed states on the Fe surface. The valence orbitals of Mo are more fully occupied by electrons than those of Cr are. As a result, Mo adsorption is more Figure 1. Differential binding energy of oxygen clusters on stable. Mg(0001) shows the tendency for oxygen to cluster, forming ar- eas on the surface that are oxygen rich (passivated) and oxygen We simulated the role of water as a source of OH ions. deficient (reactive). Chemically, one would expect this to lower the energy penalty as some of the Fe atom’s dangling bonds will be saturated through bonding to OH. Fe is considered to bind to two hydroxyl ions and the simulation procedure conducted as above. In this case, the shortest Fe-Fe bond of the adatom is 2.40 Å, which is longer than that of the stand-alone adatom but a little shorter than that of bulk atoms. When the elevation is less than 1.0 Å, the energy changes of Fe and Fe(OH)2 are almost identical (Figure 3a). If the altitude of Fe(OH)2 is raised further, the formation of bonds with hydroxide groups alleviates the bond-breaking process and results in a noticeable reduction of dissolution energy. This observation is consistent with the “make- 118

Page 119

References

1. 1, J. K. A.U. Goonewardene. An STM study of the oxidation of Mg(0001). 2002. Surface ScienceSurface Science. 501: 9.
2. Barbier, A., C. Mocuta, H. Kuhlenbeck, K. F. Peters, B. Richter, and G. Renaud. Atomic structure of the polar NiO(111)-p(2 x 2) surface. 2000. Physical Review Let- tersPhysical Review Letters. 84 (13): 2897.
3. Barbier, A., G. Renaud, C. Mocuta, and A. Stierle. Struc- tural investigation of the dynamics of the NiO(111) surface by GIXS. 1999. Surface ScienceSurface Science. 433: 761.
4. Behler, J., B. Delley, S. Lorenz, K. Reuter, and M. Schef- fler. Dissociation of O_\{2\} at Al(111): The Role of Spin Selection Rules. 2005. Physical Review LettersPhysical Review Letters. 94 (3): 036104. Figure 2. Chemisorption of oxygen on delta-Pu(111) at 0.25 (top), 5. Bungaro, C., . , P. Ballone, and W. Kress. Early Oxidation 0.50 (middle) and 0.75 (bottom) monolayers, shown in side (left) Stages of Mg(0001): A Density Functional Study. 1997. and top (right) views. Physical Review LettersPhysical review letters. 79 (22):
6. Campbell, T. J., G. Aral, S. Ogata, R. K. Kalia, A. Nakano, and P. Vashishta. Oxidation of aluminum nanoclusters.
7. Phys. Rev. BPhys. Rev. B. 71 (205413-2).
8. Carley, A. F., P. R. Davies, K. R. Harikumar, R. V. Jones, and M. W. Roberts. Oxygen States at Magnesium and Copper Surfaces Revealed by Scanning Tunnel- ing Microscopy and Surface Reactivity. 2003. Topics in CatalysisTopics in Catalysis. 24 (1): 51.
9. Carley, A. F., P. R. Davies, R. V. Jones, K. R. Harikumar, and M. W. Roberts. Atom resolved evidence for a defective chemisorbed oxygen state at a Mg(0001) surface. 2002. Chemical CommunicationsChemical Communications. (18): 2020.
10. Chao, C. Y., L. F. Lin, and D. D. Macdonald. A point defect model for anodic passive films: I. Film growth kinetics. 1981. J. Electrochem. Soc.J. Electrochem. Soc.. 128: 1187. Figure 3. (a) Energies of dissolution (bond-breaking) of iron,
11. Cox, L.. X-ray photoemission study of δ-Pu stabilized chromium and molybdenum adatoms in proximity to a (1 1 0) by 1 wt.% Ga: Electronic structure and sputter-induced iron surface, (b) Force acting on the adatoms calculated from desorption energy, (c) charges determined by Bader method for surface phase transformation. 1988. Physical Review iron, chromium and molybdenum adatoms while being ‘dragged’ BPhysical Review B. 37 (14): 8480. away from the (1 1 0) iron surface. Fig. (c) has the same legend as in (a). 11. Dinh, L. N., J. M. Haschke, C. K. Saw, P. G. Allen, and W. McLean II. Pu. 2011. J. Nuc. Mater.J. Nuc. Mater.. 408:

119

Page 120

12. Dudarev, S. L., G. A. Botton, S. Y. Savrasov, C. J. Hum- 24. ., , . , and . . Abstractive dissociation of oxygen over phreys, and A. P. Sutton. Electron-energy-loss spectra Al(111): A nonadiabatic quantum model. 2012. Journal and the structural stability of nickel oxide: An LSDA+U of Chemical PhysicsJournal of Chemical Physics. 120 study. 1998. Physical Review BPhysical Review B. 57 (8): 19. (3): 1505.
13. Gileadi, E.. Can an electrode reaction occur without
14. Ebensperger, C., and B. Meyer. First-principles study electron transfer across the metal/solution interface?. of the reconstruction and hydroxylation of the polar 2004. Chemical Physics LettersChemical Physics Let- NiO(111) surface. 2011. Physica Status Solidi B-Basic ters. 393 (4-6): 421. Solid State PhysicsPhysica Status Solidi B-Basic Solid
15. Haschke, J. M., and T. H. Allen. Plutonium hydride, State Physics. 248 (10): 2229. sesquioxide and monoxide monohydride: pyrophoric-
16. Fadley, C.S.. 1990. ity and catalysis of plutonium corrosion. 2001. J. Alloys. Comp.J. Alloys. Comp.. 320: 58.
17. Fehlner, F. P., and M. J. Graham. Thin oxide formation on metals. 2002. 27. Hawkins, J. M., J. F. Weaver, and A. Asthagiri. Density functional theory study of the initial oxidation of the
18. Flores, H. G., and D. L. Pugmire. The growth and evolu- Pt(111) surface. 2009. Physical Review BPhysical Re- tion of thin oxide films on δ-plutonium surfaces. 2010. view B. 79 (12): 125434. IOP Conference Series: Materials Science and Engi- neeringIOP Conference Series: Materials Science and 28. Hayden, B. E., E. Schweizer, R. Kötz, and A. M. Brad- Engineering. 9: 012038. shaw. The early stages of oxidation of magnesium single crystal surfaces. 1981. Surface ScienceSurface
19. Flores, H. G., P. Roussel, D. P. Moore, and D. L. Pugmire. Science. 111 (1): 26. Characterization and stability of thin oxide films on plu- tonium surfaces. 2011. Surface ScienceSurface Science. 29. Henkelman, G., A. Arnaldsson, and H. Jonsson. A fast 605 (3-4): 314. and robust algorithm for Bader decomposition of charge density. 2006. Computational Materials Sci-
20. Friebel, D., D. J. Miller, C. P. O’Grady, T. Anniyev, J. enceComputational Materials Science. 36 (3): 354. Bargar, U. Bergmann, H. Ogasawara, K. T. Wikfeldt, L. G. Pettersson, and A. Nilsson. In situ X-ray probing reveals 30. Holby, Edward F., . , and . . Thermodynamics and Hys- fingerprints of surface platinum oxide. 2011. Phys. teresis of Oxide Formation and Removal on Platinum Chem. Chem. Phys.Phys. Chem. Chem. Phys.. 13: 262. (111) Surfaces. 2012. The Journal of Physical Chemistry CThe Journal of Physical Chemistry C.
21. Fu, Q., and T. Wagner. Interaction of nanostructured metal overlayers with oxide surfaces. 2007. Surf. Sci. 31. Lee, D.D., J.H. Choy, and J.K. Lee. Computer generation Rep.Surf. Sci. Rep.. 62: 431. of binary and ternary phase diagrams via a convex hull method . 1992. Journal of Phase EquilibriaJournal of
22. Ganduglia-Pirovano, M. V., K. Reuter, and M. Scheffler. Phase Equilibria. 13 (4): 7. Stability of subsurface oxygen at Rh(111). 2002. Physi- cal Review BPhysical Review B. 65 (24): 245426. 32. Lin, L. F., C. Y. Chao, and D. D. Macdonald. A point defect model for anodic passive films: II. Chemical
23. Flores, H. G. Garcia, P. Roussel, D. P. Moore, and D. L. breakdown and pit initiation. 1981. J. Electrochem. Pugmire. Characterization and stability of thin oxide Soc.J. Electrochem. Soc.. 128: 1194. films on plutonium surfaces. 2011. Surf. Sci.Surf. Sci.. 605: 314. 33. Lindroos, M., H. Pfnur, G. Held, and D. Menzel. Ad- sorbate induced reconstruciton by stron chemisorp-
24. Gesell, T. F., and E. T. Arakawa. Work function changes tion - Ru(001) p(2x2)-O. 1989. Surface ScienceSurface during oxygen chemisorption on fresh magnesium Science. 222 (2-3): 451. surfaces. 1972. Surface ScienceSurface Science. 33 (2):
25. 34. Macdonald, D. D.. The history of the point defect model for the passive state: a brief review of film
26. Getman, Rachel B., . , and William F. Schneider. Ther- growth aspects. 2011. Electrochim. ActaElectrochim. modynamics of Environment-Dependent Oxygen Acta. 56: 1761. Chemisorption on Pt(111). 2008. Journal of Physical Chemistry CJournal of Physical Chemistry C. 112: 14. 35. Namba, H., J. Darville, and J. M. Gilles. A model for the 120

Page 121

oxidation of Mg(0001) based upon LEED, AES, ELS and 45. Schwoebel, R. L.. Oxide Formation on Magnesium work function measurements. 1981. Surface Science- Single Crystals. I. Kinetics of Growth. 1963. Journal of Surface Science. 108 (3): 446. Applied PhysicsJournal of Applied Physics. 34 (9): 2776. 36. Pfnur, H., G. Held, M. Lindroos, and D. Menzel. Oxygen 46. Schwoebel, R. L.. Oxide Formation on Magnesium induced reconstruction of close-packed surface - a Single Crystals. II. Structure and Orientation. 1963. LEED0IV study on Ru(001)-p(2x1)O. 1989. Surface Sci- Journal of Applied PhysicsJournal of Applied Physics. enceSurface Science. 220 (1): 43. 34 (9): 2784. 37. Reuter, K.. Nanometer and sub-nanometer thin oxide 47. Seyeux, A., V. Maurice, L. H. Klein, and P. Marcus. In films at surfaces of late transition metals. 2004. Los situ scanning tunneling microscopy study of the initial Alamos National Laboratory, Preprint Archive, Con- stages of growth and of the structure of the passive densed MatterLos Alamos National Laboratory, Pre- film on Ni(111) in 1 mM NaOH(aq). 2005. J. Solid State print Archive, Condensed Matter. : 1. Electrochem.J. Solid State Electrochem.. 9: 337. 38. Reuter, K., M. V. Ganduglia-Pirovano, C. Stampfl, and 48. Streitz, F. H., and J. W. Mintmire. Electrostatic potential M. Scheffler. Metastable precursors during the oxida- for metal-oxide surfaces and interfaces. 1994. Phys. tion of the Ru(0001) surface. 2002. Physical Review Rev. BPhys. Rev. B. 50: 11996. B (Condensed Matter and Materials Physics)Physical 49. Surnev, L., G. Rangelov, and G. Bliznakov. Interaction of Review B (Condensed Matter and Materials Physics). oxygen with a Ru(OO1) surface. 1985. Surface Science- 65 (16): 165403/1. Surface Science. 159 (2–3): 299. 39. Reuter, K., and M. Scheffler. First-Principles Atomis- 50. Tang, L., B. Han, K. Persson, C. Friesen, T. He, K. Si- tic Thermodynamics for Oxidation Catalysis: Surface eradzki, and G. Ceder. Electrochemical stability of Phase Diagrams and Catalytically Interesting Regions. nanometer-scale Pt particles in acidic environments. 2003. Physical Review LettersPhysical Review Letters. 2010. J. Am. Chem. Soc.J. Am. Chem. Soc.. 132: 596. 90 (4): 046103/1. 51. Tang, W., E. Sanville, and G. Henkelman. A grid-based 40. Reuter, K., and M. Scheffler. Oxide formation at the Bader analysis algorithm without lattice bias. 2009. surface of late 4d transition metals: insights from first- Journal of Physics-Condensed MatterJournal of Phys- principles atomistic thermodynamics. 2004. Applied ics-Condensed Matter. 21 (8). Physics A: Materials Science & ProcessingApplied Phys- ics A: Materials Science & Processing. 78 (6): 793. 52. Taylor, C. D.. The transition from metal-metal bond- ing to metal-solvent interactions during a dissolution 41. Richmond, S., J. S. Bridgewater, J. W. Ward, and T. H. event as assessed from electronic structure. 2009. Allen. The solubility of hydrogen and deuterium in al- Chem. Phys. Lett.Chem. Phys. Lett.. 469: 99. loyed, unalloyed and impure plutonium metal. 2010. IOP Conference Series: Materials Science and Engi- 53. Taylor, C. D.. Cohesive Relations for Surface Atoms in neeringIOP Conference Series: Materials Science and the Iron-Technetium Binary System. J. Alloys Comp.J. Engineering. 9: 012036. Alloys Comp.. Submitted. 42. Sanville, E., S. D. Kenny, R. Smith, and G. Henkelman. 54. Taylor, C. D., M. Janik, M. Neurock, and R. G. Kelly. Ab Improved grid-based algorithm for Bader charge allo- Initio Simulations of the Electrochemical Activation of cation. 2007. Journal of Computational ChemistryJour- Water . 2006. Mol. Sim.Mol. Sim.. 33: 429. nal of Computational Chemistry. 28 (5): 899. 55. Taylor, C. D., R. G. Kelly, and M. Neurock. First-Princi- 43. ., , . , and . . O adsorption and incipient oxidation of the ples Calculations of the Electrochemical Reactions of Mg(0001) surface. 2004. Physical Review B (Condensed Water at an Immersed Ni(111)/H2O Interface. 2006. J. Matter and Materials Physics)Physical Review B (Con- Electrochem. Soc.J. Electrochem. Soc.. 153: E207. densed Matter and Materials Physics). 69: 8. 56. Taylor, C. D., R. G. Kelly, and M. Neurock. First-prin- 44. ., , . , and . . Mg(0001) surface oxidation: a two-dimen- ciples prediction of equilibrium potentials for water sional oxide phase. 2004. Physical Review B (Con- activation by a series of metals. 2007. J. Electrochem. densed Matter and Materials Physics)Physical Review B Soc.J. Electrochem. Soc.. 154 (12): F217. (Condensed Matter and Materials Physics). 69: 4. 121

Page 122

57. Taylor, C. D., M. Neurock, and J. R. Scully. First-prin- Rossi, M., and C. Taylor. Atomistic modeling of ISCC in fuel- ciples investigation of the fundamental corrosion pellet interactions. 2011. In 242nd National Meeting properties of a model Cu38 nanoparticle and the (111), of the American-Chemical-Society (ACS) ; 20110828 - 20110901 ; Denver, CO. Vol. 242. (113) surfaces. 2008. J. Electrochem. Soc.J. Electro- chem. Soc.. 155: C407. Taylor, C.. Connections between the energy functional and interaction potentials for materials simulations. 2011.
58. Taylor, C. D., S. Wasileski, J. W. Fanjoy, J. Filhol, and M. ABSTRACTS OF PAPERS OF THE AMERICAN CHEMICAL Neurock. First principles reaction modeling of the elec- SOCIETY. 242. trochemical interface: Consideration and calculation of a tunable surface potential from atomic and electronic structure. 2006. Phys. Rev. B.Phys. Rev. B.. 73: 165402.
59. Todorova, M., W. X. Li, M. V. Ganduglia-Pirovano, C. Stampfl, K. Reuter, and M. Scheffler. Role of subsurface oxygen in oxide formation at transition metal surfaces.
60. Physical Review LettersPhysical Review Letters. 89 (9): 961031.
61. Todorova, M., K. Reuter, and M. Scheffler. Density-func- tional theory study of the initial oxygen incorporation in Pd(111). 2005. Physical Review B: Condensed Matter and Materials PhysicsPhysical Review B: Condensed Matter and Materials Physics. 71 (19): 195403/1.
62. Valone, S. M., and S. R. Atlas. Energy dependence on fractional charge for strongly interacting subsystems.
63. Phys. Rev. Lett.Phys. Rev. Lett.. 97: 256402.
64. Wei-Xue, L., C. Stampfl, and M. Scheffler. Subsurface oxygen and surface oxide formation at Ag(111): a density-functional theory investigation. 2003. Physical Review B (Condensed Matter and Materials Physics) Physical Review B (Condensed Matter and Materials Physics). 67 (4): 45408.
65. Wiame, F., V. Maurice, and P. Marcus. Initial stages of oxidation of Cu(111). 2007. Surf. Sci.Surf. Sci.. 601:
67. Wolf, D.. Reconstruction of NaCl surfaces from a dipo- lar solution to the Madelung problem. 1992. Physical Review LettersPhysical Review Letters. 68 (22): 3315.
68. Zhang, W. B., and B. Y. Tang. Stability of the polar NiO(111) surface. 2008. Journal of Chemical Physics- Journal of Chemical Physics. 128 (12). Publications Francis, M. F., and C. D. Taylor. Materials Science and Engi- neering, Los Alamos National Labs, Los Alamos, New Mexico, USA. 2013. First-principles insights into the structure of the incipient magnesium oxide and its in- stability to decomposition: Oxgen chemisorption to Mg (0001) and thermodynamic stability. Physical Review B . 87 (87): 13. 122

Page 123

Chemistry and Material Sciences Early Career Research Final Report Plasmas of Magnetic Monopoles in Artificial Spin Ice: Designing Frustration, Engineering Magnetricity Cristiano Nisoli 20120516ECR Abstract nano-islands whose shape anisotropy defines Ising-like At the end of this project we have realized our goal: spins arranged along the sides of a regular lattice. Unlike proving that artificial spin ice contains bona fide mono- naturally occurring magnetically frustrated materials poles in artificial spin ice, emerging as relevant degrees such as the spin ice pyrochlores, ASI remarkably allows of freedom of its low-energy dynamics, endowed of imaging of its microstate, therefore providing a direct magnetic charge, and interact through Coulomb force. experimental benchmark to theoretical tre¬atments, We have shown formation of crystallites of magnetic and can be engineered into different geometries, to charges in kagome spin ice much similar to an ionic explore new phenomena and effects. Since its introduc- crystal. Moving beyond our initial aim, we have also tion, ASI has been used successfully to study frustration proposed new topologies meant to predict and control [1-10], and extension of thermodynamics to granular monopole dynamics via the concept of “emergent” systems governed by non-trivial interactions [2, 5, 8] to- frustration. We have designed these new geometries, pological defects [13, 17], information encoding [6], and studied them theoretically, analytically and numerically, anomalous topological hall effect connected to emerging and realized them experimentally. The experiments chirality [18]. It also became preferred ground for direct completely validated out predictions: this lead to the imaging of a new striking fractionalization phenomenon: proof of emergent frustration, of screening of mono- “magnetic monopoles” [18-26]. This novel excitation, poles by embedded magnetic charges and of formation endowed with a quantized magnetic charge, introduced of magnetic polarons. We have published our theoreti- by C. Castelnovo and collaborators [19] in 2007, and cal and experimental results on Nature, Phys. Rev. Lett., then detected in rare heart pyrochlores [20] (including New Journal of Physics (Invited). Further work is under currents of magnetic monopoles [19-25]) was soon after submission to Nature Physics. The PI has also cemented directly visualized in ASI [13-15]. his leadership in this field, which he cofounded, by In this project we had posed the following series of sci- authoring a Review of Modern Physics. This LDRD grant entific questions: funded an undergraduate student from UNV Reno, Muir Morrison, who has worked with the PI for a full year and Are observed ASI monopoles “real” magnetic mono- was later admitted in prestigious graduate programs, poles? While there is little doubt that magnetic mono- including Cornell, Caltech, Berkley (he is at Caltech as of poles of spin ice rightfully deserve their name [13-20], July 2013). The PI has also collaborated with a graduate the situation for ASI is entirely different. ASI observed student (Tammie Nelson, T-1, now Director’s Fellow) and topological excitations do correspond structurally to the with Gia-Wei Chern (T-4, Oppenheimer Fellow). Some magnetic monopoles of spin ice (Figure 3), insofar as one of the results of this work have been reported in a LANL can attribute somehow formally a “net magnetic charge” press release. to the excited vertices. Yet questions remains: do they interact as coulomb charges? Are there signatures of this Background and Research Objectives interaction in the experimental data? (..) are ASI mag- In 2006, a joint theoretical and experimental collabora- netic monopoles “physical”? tion between the PI and P. Schiffer’s group (now at UI Urbana) at Penn State University, presented a new class Of course monopoles as elementary particles have never of frustrated magnetic artificial materials, called “Artifi- been found in Nature, despite a long search since Dirac’s cial Spin Ice” [1-17] (Figure 1). Artificial spin ice is a two- early predictions. We asked here whether one could dimensional array of elongated single-domain permalloy prove the existence of magnetic monopoles, as el- 123

Page 124

ementary excitations and low-energy degrees of freedom, in parallel to exploit this new technique. We concentrated emerging in artificial spin ice arrays as a particular case on a material-by-design approach that would allow us to of fractionalization, or deconfinement of quasi-particles control the properties of these magnetic charges. All the that together are seen as comprising the fundamental unit studies of artificial spin ice so far had used simple geom- of the system, in this case the north and south poles of a etries (square, hexagonal, ladder). These were not chosen nanomagnet. for their austere beauty, but because of limits in nanofab- rication. As limits were progressively lifted, the PI saw that We therefore proposed: the time was ripe to introduce novel topologies. These take advantage of emergent rather than direct frustration. Our primary goal is to show theoretically and experimen- tally that ASI monopoles do indeed interact as coulomb Frustration in artificial spin ice comes from the impos- charges and respond to external magnetic fields just as sibility of minimizing all the interactions between near- electrical charges respond to electrical fields (…). We will est neighboring nanoislands. However because of the prove the existence of magnetic monopoles in Artificial anisotropy of the magnetic interaction, the degeneracy of Spin Ice (ASI). this frustration is lifted in most geometric arrangement. Degeneracy of the manifold is essential for magnetic Scientific Approach and Accomplishments charges: one can prove that, in a non-degenerate phase, Our main goal was to prove the bona-fide (see above) magnetic monopoles are energetically confined by their nature of ASI monopoles. We believed small signatures of Dirac strings. monopoles were buried into old data and we proposed to refine our out of equilibrium theory [2, 5, 8] to reveal The PI devised a new way to achieve degeneracy, through those signals. While we pursued this route, new develop- an emergent frustration. There each vertex of the lattice ments in the field suggested to us a better way to achieve has a defined minimal energy, however it is impossible our results in a more solid way. Since the beginning (2006) to allocate all vertices in their ground state: excitations artificial spin ice was disordered through a protocol of AC are topologically protected (Figure 3). Without indulging demagnetization, which we had developed especially to too much on the details, one can show that this differ- lower the interisland magnetic interaction [5]. The PI had ent, emergent frustration leads to a different behavior of previously demonstrated [3, 6] that this method produces monopole excitations. They can effectively be screened by a well-defined statistical ensemble, which can be predicted an embedded magnetic charge and the monopole dressed in terms of an effective thermodynamics [2, 8]. Meanwhile by the charge forms a bound magnetic polaron. researchers at Leeds attempted true thermalization during We offered general mathematical criteria to assess emer- growth [14]. The PI explained those results (New Journal of gent vertex-frustration, and we studied in depth (analyti- Physics, dedicated issue, 2012, invited). An idea emerged: cally and numerically) some of the lattices. For one of our reaching true thermalization by heating the material in design we assisted experimentalists in fabrication and anal- the narrow temperature window (about 10 Celsius) below ysis. The “Shakti” lattice was then realized experimentally the Curie temperature of the nanoisland, but above their (to be submitted to Nature Physics) , thermally annealed blocking temperature. By engineering the material such with our new technique, and all of our predictions (submit- that the blocking temperature would sit low enough in ted to Phys. Rev. Lett.) were validated. The emergent frus- the phase diagram of artificial spin ice, it was possible to tration revealed the predicted manifold, and monopoles create a thermal spin ensemble able to access low entropy indeed formed magnetic polarons (Figures 4, 5). regions. For the kagome geometry, the odd coordination of the lattice means that each vertex in the ice phase sup- In conclusion ports a magnetic charge, either positive or negative. One thus expects (and can prove) that such charges would rear- • We developed a new experimental design that directly range in a sort of ionic crystal as the entropy is lowered. revealed monopoles! (New Journal of Physics 2012) We were able to witness such charge crystallization (Figure 2), which provided a much clearer proof of the existence • We actually SAW monopoles interacting and reorganiz- and interaction of these emergent magnetic charges than ing (Nature, 2013) the route we initially detailed in the proposal. This result • As a consequence, we realized the need of “Mono- was published in Nature on August 2013 and featured in a pole-Oriented Geometries” (New Journal of Physics LANL press release. 2013, Phys. Rev. Lett, submitted) Confident that we would reach our initial goal (which re- • One of these geometries have been realized and viewers had considered perhaps too ambitious) we worked 124

Page 125

showed monopole screening (to be submitted to Na- ture Physics) Impact on National Missions Magnetic technology generally concerns itself with ma- nipulation of localized dipolar degrees of freedom. The ability to build materials containing delocalized monopolar charges is very exciting with possible technological impli- cations in data storage and computation, and can lead to materials with needed and emergent functionality includ- ing unprecedented sensitivity for sensing and control in national security missions. This work demonstrates a direc- tion in condensed matter physics that is quite opposite to what has been done in the last six decades or so. Instead of imagining an emergent theoretical description to model the behavior of a naturegiven material and validating it indirectly, we engineer materials of desired emergent properties that can be visualized directly. DOE and LDRD alike have largely invoked this material-by-design approach Figure’ 2:’ Emergent’ domains’ of’ ordered’ magnetic’ charges’ in’ that replaces serendipity with design, assisted by a meso- aFirgtiufricei a2l.’ Ekmageorgmeen’t sdpoimn’aiicnes. ‘o(f ao,‘rdce)r’eMd FmMa’ginmeatigce csh’aorfg’eths einr maratill-y’ scopic analysis. annealed’artificial’kagome’spin’ice’with’lattice’parameter’of’260’ ficial kagome spin ice. (a, c) MFM images of thermally annealed nm’and’490’nm.’(b,‘d)‘Maps’of’the’distribution’of’magnetic’charge’ artificial kagome spin ice with lattice parameter of 260 nm and in’(a,‘c).’Red’and’blue’dots’correspond’to’vertices’belonging’to’ e4a9c0h n’mof.’ (tbh, ed’) tMwao’psd eogf etnheer daitset’rimbuatigonne toicf’ mchaagrngeetiFoc rcdhearregde’ isnt a(ate, s’ dc)e. sRceridb aedn’din b’tluhee ‘dteoxtts. ‘cGorrereensp’aonndd’y teol lvoewrti’dcoetss ‘breelpornegsienngt ‘ttoh ee’a+c3h’a onfd ’ tF3h’e tewxcoi tdaetgioennse.‘raFteig muraeg’neSti11c ‘chianr’geS-uoprpdleermede nsttaartey’s dIensfocrrimbeadti oinn ’ reproduces’these’maps’but’also’includes’vectors’representing’the’ the text. Green and yellow dots represent the +3 and -3 excita- magnetic’moments’of’the’islands’comprising’the’lattices’(Nature’ tions. Figure S11 in Supplementary Information reproduces these 2013).’ maps but also includes vectors representing the magnetic mo- ments of the islands comprising the lattices (Nature 2013). ! Figure’1.‘Artificial’Spin’Ice’(shown’on’top’left’as’an’AFM’image)‘allows’ Ffoigru’drier e1c.t ‘Avirstiufiacliizaalt Siopnin’o fI’cites ‘(mshagonwenti co’dne tgorpee lse’ftof ‘afrse aedno AmF’tMhr iomugahg’e) aMllFoMw’(st ofopr’c denirteecrt) ‘vainsdu’aotlihzearti’toench onfi qituse ms’(asegen’beetilco wde’ing’rtehees’t eoxft )f.r’eedom Energetically’favorable’and’unfavorable’dipole’interactions’are’ through MFM (top center) and other techniques (see below in described’in’the’top’right.‘Bottom:”four’legged’vertices’have ’ tphoes stiebxlte)‘.m Eonmeergnet’ticocnaflilgyu fraavtioornasb,‘lweh aicnhd’s uepnafaravtoer’ianbtole’f oduipr’oslyem i !mnteetrraycG- tidoisntisn actr’ety dpeessc,‘srhiboewdn i’nhe trhee’w tiothp’ trhige’hrte.l aBtoivtteo’frmeq: u feonucry l,‘ewghgiechd!’! ve=rti!c”es hcaorvree 2sp^o4n=d1s6’t op’roasnsdibolme ‘masosimgneantti ocno’onffi’tghue’rmatiomonesn,t sw.‘Nhoicthe ‘tsheapta’ervaetne’ in’ the’low’energy’configurations,‘some’interactions’are’frustrated,‘but’ into four symmetry-distinct types, shown here with the relative the’lack’of’degeneracy’beside’spin’inversion’in’Type’I’vertices’ fpreroqvuiednesc’ya,n w’ohrdicehre cdo’arrnetisfpeorrnodmsa tgon reatincd’goromu nads’ssitgantea’[ti1o].n’ of the mo- Figure 3. Vertex-frustrated and non-frustrated lattices based on ments. Note that even in the low energy configurations, some in- perpendicular vertices. Lattices on the left are frustrated while teractions are frustrated, but the lack of degeneracy beside spin ! those at the centeer, despite the apparent similarity of the lat- inversion in Type I vertices provides an ordered antiferromagnetic tices, are not. The unhappy vertices are circled (blue), while red ground state [1]. spins may be chosen in either direction without changing the energy of the configuration: that is not the only source of degen- eracy, as other allocations of the unhappy vertices are degener- ate. The ‘staggered brickwork’ lattice (top left) has a ‘trivial’ degeneracy, which can be mapped into a system of independent spins. In contrast, the ‘pinwheel’ (middle left) and ‘shakti’ lattices (lower left) possess more complex degeneracy. The ice manifold of the shakti lattice can be mapped into an emergent six vertex 125

Page 126

model. On the top right the Santa Fe lattice can be mapped into References an eight vertex model. Both show monopole crystallization. On

1. Wang, R. F., C. Nisoli, R. S. Freitas, J. Li, W. McConville, the bottom right, the Tetris lattice show a sliding phase in its ice B. J. Cooley, M. S. Lund, N. Samarth, C. Leighton, V. manifold (New Journal of Physics 2013). H. Crespi, and P. Schiffer. Artificial ‘spin ice’ in a geo- metrically frustrated lattice of nanoscale ferromagnetic islands. 2006. NATURE. 439 (7074): 303.
2. Nisoli, C., R. F. Wang, J. Li, W. F. McConville, P. E. Lam- mert, P. Schiffer, and V. H. Crespi. Ground state lost but degeneracy found: The effective thermodynamics of artificial spin ice. 2007. PHYSICAL REVIEW LETTERS. 98 (21): -.
3. Wang, R. F., J. Li, W. McConville, C. Nisoli, X. Ke, J. W. Freeland, V. Rose, M. Grimsditch, P. Lammert, V. H. Crespi, and P. Schiffer. Demagnetization protocols for frustrated interacting nanomagnet arrays. 2007. JOUR- NAL OF APPLIED PHYSICS. 101 (9): -.
4. Wang, R. F., C. Nisoli, R. S. Freitas, J. Li, W. McConville, B. J. Cooley, M. S. Lund, N. Samarth, C. Leighton, V. Figure 4. Top: MFM images of thermally annealed shakti 2 lat- H. Crespi, and P. Schiffer. Artificial ‘spin ice’ in a geo- tice s with lattice constant a = 360 nm (left) and 800 nm (right). metrically frustrated lattice of nanoscale ferromagnetic Most of the four-island vertices in the 360 nm lattice are in the islands (vol 439, pg 303, 2006). 2007. NATURE. 446 Type I configuration; there are a few higher-energy Type II (blue circle) and Type III (green circle) vertices. Note how the (light (7131): 102. contrast) Type III vertex circled in green is surrounded by four
5. Ke, X., J. Li, C. Nisoli, P. E. Lammert, W. McConville, R. F. three-island vertices with dark contrast. This is an example of the monopole screening described in the text. Bottom: a portion of Wang, V. H. Crespi, and P. Schiffer. Energy minimization completely annealed Shakti 1 (right) employed to reveal emer- and ac demagnetization in a nanomagnet array. 2008. gent frustration in the ice manifold; the reticulate of turquois PHYSICAL REVIEW LETTERS. 101 (3): -. lines define the plaquettes, each supporting two and only two unhappy vertices, Type II’; the assignation of topologically pro- 6. Lammert, P. E., X. L. Ke, J. Li, C. Nisoli, D. M. Garand, tected defect on each plaquette corresponds to an emergent six- V. H. Crespi, and P. Schiffer. Direct entropy determina- vertex, ice-rule model (right) in which red arrows point toward a tion and application to artificial spin ice. 2010. NATURE defect, whereas yellow arrows point away from it (unpublished). PHYSICS. 6 (10): 786.
6. Li, J., S. Zhang, J. Bartell, C. Nisoli, X. Ke, P. E. Lammert, V. H. Crespi, and P. Schiffer. Comparing frustrated and unfrustrated clusters of single-domain ferromagnetic islands. 2010. PHYSICAL REVIEW B. 82 (13): -.
7. Nisoli, C., J. Li, X. L. Ke, D. Garand, P. Schiffer, and V. H. Crespi. Effective Temperature in an Interacting Vertex System: Theory and Experiment on Artificial Spin Ice.
8. PHYSICAL REVIEW LETTERS. 105 (4): -.
9. Li, J., X. Ke, S. Zhang, D. Garand, C. Nisoli, P. Lammert, V. H. Crespi, and P. Schiffer. Comparing artificial frus- trated magnets by tuning the symmetry of nanoscale permalloy arrays. 2010. PHYSICAL REVIEW B. 81 (9): -.
10. Ke, X. L., J. Li, S. Zhang, C. Nisoli, V. H. Crespi, and P. Figure 5. Top: Our theoretical predictions and experimental data Schiffer. Tuning magnetic frustration of nanomagnets on the vertex population of the Shakti lattice. Bottom: Charge in triangular-lattice geometry. 2008. APPLIED PHYSICS screening of magnetic monopoles: theory vs. experiment. LETTERS. 93 (25): -. 126

Page 127

11. Ke, X., J. Li, C. Nisoli, P. E. Lammert, W. McConville, R. F. 703. Wang, V. H. Crespi, and P. Schiffer. Energy minimization
12. Giblin, S. R., S. T. Bramwell, P. C. W. Holdsworth, D. and ac demagnetization in a nanomagnet array. 2008. Prabhakaran, and I. Terry. Creation and measurement PHYSICAL REVIEW LETTERS. 101 (3): -. of long-lived magnetic monopole currents in spin ice.
13. Wang, R. F., J. Li, W. McConville, C. Nisoli, X. Ke, J. W. 2011. NATURE PHYSICS. 7 (3): 252. Freeland, V. Rose, M. Grimsditch, P. Lammert, V. H.
14. Fennell, T., S. T. Bramwell, D. F. McMorrow, P. Manuel, Crespi, and P. Schiffer. Demagnetization protocols for and A. R. Wildes. Pinch points and Kasteleyn transi- frustrated interacting nanomagnet arrays. 2007. JOUR- tions in kagome ice. 2007. NATURE PHYSICS. 3 (8): 566. NAL OF APPLIED PHYSICS. 101 (9): -.
15. Bramwell, S. T.. A spin on cool complexity. 2006. NA-
16. Mengotti, E., L. J. Heyderman, A. F. Rodriguez, F. Nolt- TURE PHYSICS. 2 (4): 219. ing, R. V. Hugli, and H. B. Braun. Real-space observa- tion of emergent magnetic monopoles and associated
17. Bramwell, S. T., S. R. Giblin, S. Calder, R. Aldus, D. Prab- Dirac strings in artificial kagome spin ice. 2011. NA- hakaran, and T. Fennell. Measurement of the charge TURE PHYSICS. 7 (1): 68. and current of magnetic monopoles in spin ice. 2009. NATURE. 461 (7266): 956.
18. Morgan, J. P., A. Stein, S. Langridge, and C. H. Marrows. Thermal ground-state ordering and elementary excita-
19. Fennell, T., P. P. Deen, A. R. Wildes, K. Schmalzl, D. tions in artificial magnetic square ice. 2011. NATURE Prabhakaran, A. T. Boothroyd, R. J. Aldus, D. F. McMor- PHYSICS. 7 (1): 75. row, and S. T. Bramwell. Magnetic Coulomb Phase in the Spin Ice Ho2Ti2O7. 2009. SCIENCE. 326 (5951):
20. Ladak, S., D. E. Read, G. K. Perkins, L. F. Cohen, and W.

R. Branford. Direct observation of magnetic monopole defects in an artificial spin-ice system. 2010. NATURE Publications PHYSICS. 6 (5): 359. Chern, G. W., M. Morrison, and C. Nisoli. Engineering De- 16. Farhan, A., P. M. Derlet, A. Kleibert, A. Balan, R. V. generacy: a Critical Ground State for Artificial Spin I. Chopdekar, M. Wyss, L. Anghinolfi, F. Nolting, and L. J. Physical Review Letters. Heyderman. Exploring hyper-cubic energy landscapes Gilbert, I., G. W. Chern, C. Nisoli, and P. Schiffer. Emergent in thermally active finite artificial spin-ice systems. Frustration in Artificial Spin Ice: Magnetic Polarons and 2013. NATURE PHYSICS. 9 (6): 375. Charge Ordering. . 2013. To be submitted to Nature Physics. 17. Mengotti, E., L. J. Heyderman, A. F. Rodriguez, F. Nolt- ing, R. V. Hugli, and H. B. Braun. Real-space observa- Lammert, P., V. Crespi, and C. Nisoli. Gibbsianizing nonequi- tion of emergent magnetic monopoles and associated librium dynamics of artificial spin ice and other spin Dirac strings in artificial kagome spin ice. 2011. NA- systems. 2012. New Journal of Physics. 14 (4): 045009 TURE PHYSICS. 7 (1): 68. (19 pp.). 18. Branford, W. R., S. Ladak, D. E. Read, K. Zeissler, and L. Morrison, M., T. R. Nelson, and C. Nisoli. Unhappy vertices in artificial spin ice: new degeneracies from vertex- F. Cohen. Emerging Chirality in Artificial Spin Ice. 2012. frustration. (Invited). 2013. New Journal of Physics. 15 SCIENCE. 335 (6076): 1597. (4): 045009 . 19. Castelnovo, C., R. Moessner, and S. L. Sondhi. Magnetic Nisoli, C.. On thermalization of magnetic nano-arrays at monopoles in spin ice. 2008. NATURE. 451 (7174): 42. fabrication. 2012. New Journal of Physics. 14: 35017. 20. Morris, D. J. P., D. A. Tennant, S. A. Grigera, B. Klemke, Nisoli, C., R. Moessner, and P. Schiffer. Artificial spin ice: C. Castelnovo, R. Moessner, C. Czternasty, M. Meissner, Designing and imaging magnetic frustration . To appear K. C. Rule, J. U. Hoffmann, K. Kiefer, S. Gerischer, D. in Review of Modern Physics. Slobinsky, and R. S. Perry. Dirac Strings and Magnetic Monopoles in the Spin Ice Dy2Ti2O7. 2009. SCIENCE. Zhang, S., I. Gilbert, C. Nisoli, G. W. Chern, M. J. Erickson, L. O’Brien, C. Leighton, P. E. Lammert, V. H. Crespi, and 326 (5951): 411. P. Schiffer. Crystallites of magnetic charges in artificial 21. Bramwell, S. T.. MAGNETIC MONOPOLES Magnetricity spin ice. 2013. NATURE. 500 (7464): 553. near the speed of light. 2012. NATURE PHYSICS. 8 (10): Zhang, S., J. Li, I. Gilbert, J. Bartell, M. Erickson, Y. Pan, P. 127

Page 128

Lammert, C. Nisoli, K. K. Kohli, R. Misra, V. Crespi, N. Samarth, C. Leighton, and P. Schiffer. Perpendicular magnetization and generic realization of the ising mod- el in artificial spin ice. 2012. Physical Review Letters. 109 (8): 087201. Zhang, S., J. Li, J. Bartell, X. Ke, C. Nisoli, P. Lammert, V. Crespi, and P. Schiffer. Ignoring your neighbors: Mo- ment correlations dominated by indirect or distant interactions in an ordered nanomagnet array. 2011. Physical Review Letters. 107 (11): 117204. 128

Page 129

Chemistry and Material Sciences Early Career Research Final Report Valence-Fluctuation-Mediated Superconductivity in Heavy-Fermion Materials Marc Janoschek 20120525ECR Abstract in mediating unconventional superconductivity. The belief that room temperature superconductiv- Background and Research Objectives ity – the holy grail of condensed matter physics – is within reach, originated in the surprising discovery of A common theme found in many materials exhibiting unconventional superconductivity in the high-Tc cuprate unconventional superconductivity is the existence of a materials at temperatures up to 160 K, and has been zero-temperature magnetic instability. At the instability, recently refuel by the high-Tc ¬iron-pnictides. Here our a nonmagnetic state can be smoothly transformed into a current understanding of superconductivity is based on magnetically ordered state as function of a non-thermal a theory by Bardeen, Cooper and Schrieffer (BCS) who control parameter such as chemical substitution, or were able to show that superconductivity develops pressure. On the nonmagnetic side of the instability, the in materials that exhibit an arbitrarily weak attractive magnetic moments are randomly oriented and there is force between electrons (that usually strongly repel no correlation between neighboring moments. When each other), that consequently leads to the formation the magnetic state is approached, magnetic fluctuations of electron pairs, the so-called Cooper pairs. These pairs arise that correlate moments over increasing length can travel through a solid without resistance due to the scales. The fluctuations are strongest right at the mag- remarkable properties of quantum physics. While we netic instability and the associated correlation length understand that in conventional superconductors collec- becomes infinite resulting in the onset of magnetic tive atomic vibrations (phonons) provide the required order (Figure 1). It is precisely these fluctuations that are pairing interaction, this also implies that their supercon- thought to mediate the unconventional superconductiv- ducting transitions temperatures are very low. The high ity. In this picture, a symmetric dome of superconductiv- superconducting transition temperatures frequently ity emerges around the instability. The highest transition found in unconventional superconductors showed that a temperature is expected at the instability, where the new “super-glue” must mediate unconventional super- fluctuations become strongest (Figure 1(a)) [1-4]. conductivity. Because unconventional superconductiv- While it is clear that proximity to a magnetic instability ity is usually found in vicinity of magnetism, the major and the associated magnetic fluctuations play an impor- contender for the new pairing glue are magnetic fluctua- tant role in mediating unconventional superconductiv- tions (magnetic analog of phonons). However, a “smok- ity, there are also growing experimental and theoretical ing gun”-experiment for that scenario is still missing, and indications that magnetic fluctuations by themselves notably, the abundance of emergent electronic phenom- may not be a sufficient condition for unconventional su- ena that is generally found in conjunction with uncon- perconductivity: (i) In many cases, the highest supercon- ventional superconductivity suggests that other pairing ducting transition temperature (TS) occurs far away from glues should exist. In this early-career project we have where the magnetic fluctuations are the strongest. (ii) investigated the recently proposed alternative mecha- Only for few materials such as CeCu2Si2 [5] and YBa2C- nism of valence-fluctuation-driven superconductivity. u3O6.6 [6], it has been reported that magnetic fluctua- We have used a new neutron scattering technique called tions are strong enough to mediate unconventional Larmor diffraction to shed some light on this issue. Our superconductivity. On the other hand there are many results show that Larmor diffraction is a useful technique calculations that demonstrate that magnetic fluctuations to study the physics of strongly correlated electron com- alone may be too weak [7, 8]. (iii) Various ordered states pounds more broadly, and more specifically add to the and exotic behavior emerge in the vicinity of supercon- evidence that valence fluctuations may indeed play role 129

Page 130

ductivity that has been proposed to drive unconventional rated by various transport and thermal measurements that superconductivity. Examples for other superconducting agree with the theory of valence fluctuations [13,17]. glues include: charge loop-and stripe order [9,10], elec- A similar case is the material CeRhIn5 that shows magnetic tronic nematic ordering [11,12], and valence fluctuations order below a temperature TN = 3.8 K [21]. As illustrated in [13,14]. The main obstacle in identifying such alternative Figure 3b the magnetism is abruptly suppressed for pres- pairing mechanisms is that it is generally not possible to sures larger than P1 = 1.77 GPa, while superconductivity measure the associated fluctuations that are thought to appears in a broad dome that roughly extends between mediate the SC. However, the proposal of valence fluctua- 1.5 and 4.5 GPa and reaches the maximal superconducting tion mediated superconductivity may represent a tangible transition temperature Tc = 2.3 K at P2 = 2.3 GPa [22]. The way to overcome this issue. difference between P1 and P2 is peculiar; if the supercon- Valence fluctuations occur frequently in materials that ductivity were mediated solely by magnetic fluctuations, contain actinide or rare earth ions, where the valence one would naively expect both to be identical. Interest- (basically the net charge of the ions) of the electron shells ingly, the magnetic order reappears between P1 and P2, may change continuously or discontinuously as a function coexisting with superconductivity, when a magnetic field is of temperature, external pressure or chemical composition applied [23], suggesting that superconductivity inhibits the [15]. In these materials, the valence electrons do not par- magnetism at zero magnetic field instead of relying on it. ticipate in the bonding of different ions and stay localized. Substitution of Co in CeRh1−xCoxIn5 suppresses the mag- In contrast, the bonding is assumed by electrons in lower netic order at x ≈ 0.8 and superconductivity appears for 0.4 electron shells that form a conduction band. The change of ≤ x ≤ 1. Tc remains at 1.5 K for 0.4 ≤ x ≤ 0.75 and increases the valence of the actinide/rare earth ions is then pro- to 2.2 K for x > 0.75 [24], similar to pressure for CeCu2Si2. moted by the so-called Kondo effect that leads to a partial The application of pressure eliminates the magnetic phase delocalization of the valence electrons into the conduction for increasing x while enlarging the superconducting band [16]. If the strength of the Kondo effect is increased dome, indicating that superconductivity in CeRh1−xCoxIn5 by application of temperature, external pressure, or chemi- exist without being in the vicinity magnetic fluctuations cal composition, this may lead to a macroscopic phase [25]. Identical to CeCu2Si2, the transport measurements transition, where all actinide/rare earth sites collectively agree with the theory for valence-fluctuation-mediated change their valence. If the change of valence at the transi- superconductivity. However, the most promising evidence tion occurs smoothly, it is expected that strong valence for valence fluctuations driving the superconductivity in fluctuations emerge, similar to the magnetic fluctuations CeRhIn5 is a change in the electronic structure at P2 [26] near to a magnetic phase transition. As illustrated in Figure in accord with theory [27]. Finally, a last contender for 2 this may lead to a weak attractive interaction between valence-mediated superconductivity is the material Pu- two conduction electrons, and in turn to unconventional CoGa5 that is isostructural to CeRhIn5. However, the pres- superconductivity [13,17]. ence of plutonium in this compounds makes more detailed investigations complicated, and not much data is currently Here the outstanding scientific problem lies in the iden- available [28]. tification of such a valence transition in the presence of unconventional superconductivity. While a recently devel- While investigations on these materials certainly suggest oped theoretical frame work showed that valence fluctua- valence fluctuation are viable as a candidate for a super- tions should be strong enough to mediate unconventional conducting glue, microscopic evidence remains absent, superconductivity, so far only indirect proof has been and it is easy to identify why. As discussed above a valence demonstrated as we will summarize in the following. The transition corresponds to partial release of a valence archetypal compound that has been investigated in this electron on a rare earth or actinide ion into the valence context is CeCu2Si2. CeCu2Si2 exhibits below a supercon- band. The loss of a valence electron reduces the size of the ducting transition temperature Tc = 0.7 K [18] that shows ion, and consequently a subtle reduction of the size of the a distinct pressure dependence: Tc remains constant up to material will occur at the valence transition. This reduction pressures of roughly 2 GPa, where it increases abruptly to in size can be easily and precisely measured at ambient 2.25 K followed by a slow decrease towards zero [19,20]. pressure. However, conventional types of measurement, By substituting Ge on the Si site the superconducting phase such as capacitive dilatometry, cannot be used at high disintegrates into two distinct domes as shown in Figure 3a pressures, where the valence transitions are expected to [14]. This has led to the view that the SC at ambient pres- occur in the materials discussed above [29]. The goal of sures is due to magnetic fluctuations, whereas the dome this early-career research project was therefore to dem- at high pressures is due to valence fluctuations as corrobo- onstrate a new way to of measuring small changes in the 130

Page 131

lattice parameters that can be employed at high pressure able achievement by itself. We note that a day of neutron while working on the outstanding scientific problem of beam time costs about $10k,andconsequentlythePIhasvalence-fluctuationmediatedsuperconductivity.obtainedintheorderof$400k additional funding to carry out this project. Scientific Approach and Accomplishments As explained and illustrated in Figure 5 we are able to We have used a novel neutron scattering method called pick up the major features of the phase diagram of CeR- Larmor diffraction that may be used in combination with hIn5 such as the antiferromagnetic and superconducting high pressure cells at low temperature. Conventional transition temperatures. More interestingly, the Larmor neutron diffraction methods don’t reach the resolution of diffraction measurements show oscillations below a other methods such as dilatometry for measurements of temperature Tx ≈ 10 K and in vicinity of the pressure P2 at the lattice parameter of a material that are required for which the putative valence transition has been postulated. detecting a subtle valence transition. On the other hand In this pressure and temperature regime corresponds to neutron diffraction is highly suitable for high pressure work the temperature where transport measurements detect because neutrons penetrate deeply into materials be- increased scattering in the electron system, which has cause they are neutral (i.e. they are not decelerated by the been speculated to emerge because of charge/valence charges of the ions and electrons present in any material), fluctuations (cf. Figure 5a) [30]. Finally, at a pressure of 1.9 and thus pressure cells are transparent to them. Larmor GPa, which is the highest pressure at which we were able diffraction combines both worlds by using the small mag- to perform measurements so far, we see a distinct change netic moment that is carried by a neutron and precess a in the low temperature behavior of the lattice parameters magnetic field. As we show in Figure 4a and b, changes of a as explained in detail in Figure 5b. This change is in agree- lattice parameter d can be encoded in this precession, and ment with the expectations for a valence transition, but enhance the resolution in measuring changes of the lattice ultimately more measurements at higher pressures will parameter Δd/d to 10-5 which approaches conventional need to be carried out to confirm this behavior. methods such as dilatometry [29]. For the scope of this project the material CeRhIn5 repre- Impact on National Missions sented the ideal candidate to look for valence-mediated Pressure is an important and well-defined tuning param- superconductivity. Single crystals of very high quality can eter for the study of strongly correlated electron systems be easily synthesized, which preclude problems from disor- and is the ‘cleanest’ way to tune between different quan- der. In contrast, CeCu2Si2 silicon has to be substituted with tum ground states. Neutron scattering on the other hand germanium to tune the material to the putative valence is an excellent microscopic probe that is able to extract transition. Further, in CeCu2Si2 the putative valence tran- information on multiple relevant energy scales from the sition occurs at very high pressures that are currently not bulk of the sample, such as spin fluctuations, crystal field accessible with neutron scattering. In Figure 4c we show excitations and lattice degrees of freedom. A key element Larmor diffraction measurements carried out at ambient of this proposal is to answer an outstanding scientific pressure and compare them to data obtained be means problem while advancing neutron scattering experiments of conventional dilatomtry. This result demonstrates that to higher pressures. This line of inquiry will be of great Larmor diffraction is in excellent agreement with conven- importance for future research on novel electronic materi- tional methods. als broadly, and specifically for the neutron scattering ef- forts funded by the Department of Energy’s Office of Basic Our Larmor diffraction measurements carried out at tem- Energy Science (OBES). As highlighted in last year’s report peratures below 30 K and for various pressures are shown of OBES “neutron scattering is among the most powerful in Figure 5. Before we discuss the data, we want to point tools for characterizing matter to understand and develop out that the measurement of one complete data set at a new materials and chemistries,” and has “applications in single pressure takes between 7-10 days. The only neutron geology, biology, and physics” [31]. To provide US scientists instrument world-wide that is setup to perform Larmor with access to this powerful probe DOE alone operates diffraction measurements on a routine basis is situated at three large neutron user facilities at the Los Alamos and FRM II research reactor in Munich, Germany, and neutron Oak Ridge National Laboratories, with an annual budget of beam time for the measurements carried out during this ~250M$(30early-careerprojecthadtobeobtainedthroughthepro-scientificuserscancarryouttheirresearchusingneutrons.posalsystematMunich,whichisopentotheinternationalDuringthelastdecadeOBESalsoinvestedwellover$1B neutron user community. We obtained a total of 46 days of in the construction of the newest of these three national beam time within less than two years, which is a consider- user facilities, namely the Spallation Neutron Source at 131

Page 132

Oak Ridge. We note, that advanced neutron scattering Figure 2. We illustrate the concept of valence-fluctuation-me- techniques such as Larmor diffraction are currently not diated unconventional superconductivity in a metallic sample available at any of the U.S.’s neutron sources, and building containing actinide or rare earth ions. a In the conduction band electrons can move freely therefore allowing to conduct electric- expertise in such techniques through projects such as this ity. In contrast, the valence electrons on the actinide/rare earth one will be critical to our ability to maintain the science sites are localized (black dotes) and do not move throughout the and technology leadership of the U.S. Finally, we point out material. b As indicated by the blue arrows the valence electrons that discussion with program managers about follow on may hop into and out of the conduction due to valence fluctua- funding for the program of inquiry initiated in this LDRD tions. As indicated by the empty circle, if the valence electron project has already begun. hops into the conduction band, an electron hole carrying a positive charge is generated at the actinide/rare earth ion site. The movable conduction electrons are attracted by this positive charge leading to cloud of conduction electrons shielding the hole. c In case more than one hole exist, the two holes (or shield- ing electron clouds) repel each other. d Energetically it is more favorable to move the two holes in a neighboring configuration, in which the holes can be screened more efficiently. This effec- tively leads to an attractive interaction between to electrons, and would therefore lead to Cooper pairs and superconductivity. Figure 1. The concept of unconventional superconductivity medi- ated by magnetic fluctuations is illustrated. a A schematic phase diagram demonstrating magnetic-fluctuation-mediated super- conductivity as function of temperature T and control parameter t. t can be pressure P, or chemical substitution x. Manipulating t, the material can be smoothly tuned from a nonmagnetic into a magnetic state separated by a zero-temperature magnetic instability. In the nonmagnetic phase the magnetic moments are randomly oriented and not correlated to their neighbors (right inset), whereas in the magnetic phase they are macroscopically Figure 3. The pressure P vs temperature T phase diagrams of ordered (here ferromagnetically, left inset). Approaching the in- the putative valence-fluctuation-mediated superconductors a stability from the nonmagnetic side, magnetic fluctuations arise CeCu2(Si1-xGex)2 and b CeRhIn5 are shown (see text for details). that correlate fluctuating moments over increasing length scales (here denoted by the grey circles in the top inset). Right at the in- stability the fluctuations are strongest, and the correlation length becomes macroscopic, triggering the onset of magnetic order. The fluctuations are also believed to serve as the glue for the Cooper pairs, resulting in a symmetric dome of superconductivity (SC) around the instability (green shade).7-10 b Experimentally determined phase diagram for the compound CePd2Si2 show- ing that this kind of picture has indeed been found for several compounds. Figure 4. The principle behind the Larmor diffraction method is illustrated. a Neutrons exhibit a small magnetic moment (red arrow) that precesses in the presence of a magnetic field (blue arrow) similarly to a gyroscope performing precessions in the gravitational field of earth. b This effect can be used to measure 132

Page 133

changes in the lattice parameter d of a material precisely. Hereby fermion compounds. 1998. NATURE. 394 (6688): 39. the neutrons are scattered of the lattice planes of the material of interest. However, the neutrons incident on the sample (marked 2. Paglione, J., and R. L. Greene. High-Temperature Su- with “in”), and the neutrons scattered off the sample (marked perconductivity in Iron-Based Materials. 2010. Nature with “out”) move through a magnetic field (grey area) in which Physics. 6: 645. the magnetic moments of the neutrons preccess. It turns out that the total number of precessions the neutrons carry out while 3. MONTHOUX, P., A. V. BALATSKY, and D. PINES. TOWARD moving through this setup is directly proportional to the lattice A THEORY OF HIGH-TEMPERATURE SUPERCONDUCTIV- parameter d of the sample. The number of precession is depen- ITY IN THE ANTIFERROMAGNETICALLY CORRELATED dent on d because the path length through the magnetic field CUPRATE OXIDES. 1991. PHYSICAL REVIEW LETTERS. 67 changes as function of d. This setup can measure changes of a (24): 3448. lattice parameter with a precision of 10-5 or better and can be used in combination with pressure cells, because pressure cells 4. Zaanen, J.. A modern, but way too short history of the are transparent for neutrons. C We show the results of our Lar- theory of superconductivity at a high temperature. mor diffraction measurements on the material CeRhIn5 at ambi- 2011. In 100 years of superconductivity. Edited by ent pressure (empty symbols) in comparison with the results by Rochella, H., and P. H. Kes. , p. 10. Boca Raton: Taylor & Takeuchi et al. obtained by conventional dilatometry (solid lines). Francis. 5. Stockert, O., J. Arndt, E. Faulhaber, C. Geibel, H. S. Jeevan, S. Kirchner, M. Loewenhaupt, K. Schmalzl, W. Schmidt, Q. Si, and F. Steglich. Magnetically driven superconductivity in CeCu2Si2. 02. Nature Physics. 7 (2): 119. 6. Dahm, T., V. Hinkov, S. V. Borisenko, A. A. Kordyuk, V. B. Zabolotnyy, J. Fink, B. Buchner, D. J. Scalapino, W. Hanke, and B. Keimer. Strength of the spin-fluctuation- mediated pairing interaction in a high-temperature Figure 5. The results of this project are illustrated. a A modified superconductor. 2009. NATURE PHYSICS. 5 (3): 217. pressure P vs temperature T phase diagram is show. “AF” and “SC” denote the antiferromagnetic and superconducting phases. 7. Kee, H. Y., S. A. Kivelson, and G. Aeppli. Spin-1 neutron The colored dash lines represent the pressures at which we have resonance peak cannot account for electronic anoma- carried out the Larmor diffraction measurements (shown in b). lies in the cuprate superconductors. 2002. PHYSICAL The color scale in the background shows the electrical resistiv- REVIEW LETTERS. 88 (25): -. ity normalized to the electrical resistivity at 5.2 GPa [30], and represents the amount of electron scattering observed in the 8. Eschrig, M.. The effect of collective spin-1 excitations sample as function of pressure. Blue represents lowest scattering on electronic spectra in high-T-c superconductors. and red the highest. The electron scattering is maximal below a 2006. ADVANCES IN PHYSICS. 55 (1-2): 47. temperature Tx ≈ 10 K and around P2 = 2.3 GPa. b The change of the tetragonal lattice parameters a and c of CeRhIn5 is shown 9. Varma, C. M.. Non-Fermi-liquid states and pairing for various pressures and below T = 30 K. As indicated by the instability of a general model of copper oxide metals. purple and light blue dash lines the transition temperatures to 1997. PHYSICAL REVIEW B. 55 (21): 14554. the antiferromagnetic and superconductings state can be seen in our data. Further below Tx,, where electrical resistivity shows 10. Li, Y., V. Baledent, G. Yu, N. Barisic, K. Hradil, R. A. additional scattering, we see oscillations in Delta a/a and Delta Mole, Y. Sidis, P. Steffens, X. Zhao, P. Bourges, and M. c/c that become stronger when approaching P2. Further, at the Greven. Hidden magnetic excitation in the pseudogap pressure P = 1.9 GPa (the highest pressure for which we recorded phase of a high-T-c superconductor. 2010. NATURE. data) the behavior for the c lattice parameter changes signifi- cantly: instead of a slow increase below ≈ 25 K a large decrease 468 (7321): 283. below ≈ 30 K. We note that from the data shown other contribu- 11. Chuang, T. M., M. P. Allan, J. Lee, Y. Xie, N. Ni, S. L. tion such as lattice vibrations were subtracted. Bud’ko, G. S. Boebinger, P. C. Canfield, and J. C. Davis. References Nematic Electronic Structure in the “Parent” State of the Iron-Based Superconductor Ca(Fe1-xCox)(2)As-2.

1. Mathur, N. D., F. M. Grosche, S. R. Julian, I. R. Walker,
2. SCIENCE. 327 (5962): 181. D. M. Freye, R. K. W. Haselwimmer, and G. G. Lonzar- ich. Magnetically mediated superconductivity in heavy
3. Kasahara, S., H. J. Shi, K. Hashimoto, S. Tonegawa, Y. 133

Page 134

Mizukami, T. Shibauchi, K. Sugimoto, T. Fukuda, T. Ter- 22. Knebel, G., D. Aoki, J. P. Brison, and J. Flouquet. The ashima, A. H. Nevidomskyy, and Y. Matsuda. Electronic Quantum Critical Point in CeRhIn5: A Resistivity Study. nematicity above the structural and superconducting 2008. JOURNAL OF THE PHYSICAL SOCIETY OF JAPAN. transition in BaFe2(As1-xPx)(2). 2012. NATURE. 486 77 (11): -. (7403): 382. 23. Park, T., F. Ronning, H. Q. Yuan, M. B. Salamon, R. 13. Onishi, Y., and K. Miyake. Enhanced valence fluctua- Movshovich, J. L. Sarrao, and J. D. Thompson. Hid- tions caused by f-c Coulomb interaction in Ce-based den magnetism and quantum criticality in the heavy heavy electrons: Possible origin of pressure-induced fermion superconductor CeRhIn5. 2006. NATURE. 440 enhancement of superconducting transition tem- (7080): 65. perature in CeCu2Ge2 and related compounds. 2000. 24. Zapf, V. S., E. J. Freeman, E. D. Bauer, J. Petricka, C. JOURNAL OF THE PHYSICAL SOCIETY OF JAPAN. 69 Sirvent, N. A. Frederick, R. P. Dickey, and M. B. Maple. (12): 3955. Coexistence of superconductivity and antiferromagnet- 14. Yuan, H. Q., F. M. Grosche, M. Deppe, C. Geibel, G. ism in CeRh1-xCoxIn5. 2002. PHYSICAL REVIEW B. 65 Sparn, and F. Steglich. Observation of two distinct su- (1): -. perconducting phases in CeCu2Si2. 2003. SCIENCE. 302 25. Jeffries, J. R., N. A. Frederick, E. D. Bauer, H. Kimura, V. (5653): 2104. S. Zapf, K. D. Hof, T. A. Sayles, and M. B. Maple. Su- 15. ROBINSON, J. M.. VALENCE TRANSITIONS AND INTER- perconductivity and non-Fermi liquid behavior near MEDIATE VALENCE STATES IN RARE-EARTH AND AC- antiferromagnetic quantum critical points in CeRh1- TINIDE MATERIALS. 1979. PHYSICS REPORTS-REVIEW xCoxIn5. 2005. PHYSICAL REVIEW B. 72 (2): -. SECTION OF PHYSICS LETTERS. 51 (1): 1. 26. Shishido, H., R. Settai, S. Hashimoto, T. Kubo, H. Hari- 16. Coleman, P.. Heavy Fermions: electrons at the edge ma, and Y. Onukia. de Haas-van Alphen effect in Ce- of magnetism. 2007. In Handbook of magnetism and RhIn5 under pressure. 2005. PHYSICA B-CONDENSED advanced magnetic materials. Edited by Kronmüller, H., MATTER. 359: 184. and S. Parkin. Vol. I, p. 95. published online: John Wiley 27. Watanabe, S., and K. Miyake. Quantum Valence Criti- & Son, Ltd. cality as an Origin of Unconventional Critical Phenom- 17. Holmes, A. T., D. Jaccard, and K. Miyake. Signatures of ena. 2010. PHYSICAL REVIEW LETTERS. 105 (18): -. valence fluctuations in CeCu2Si2 under high pressure. 28. Bauer, E. D., M. M. Altarawneh, P. H. Tobash, K. Gofryk, 2004. PHYSICAL REVIEW B. 69 (2): -. O. E. Ayala-Valenzuela, J. N. Mitchell, R. D. McDonald, 18. STEGLICH, F., J. AARTS, C. D. BREDL, W. LIEKE, D. ME- C. H. Mielke, F. Ronning, J. C. Griveau, E. Colineau, R. SCHEDE, W. FRANZ, and H. SCHAFER. SUPERCONDUC- Eloirdi, R. Caciuffo, B. L. Scott, O. Janka, S. M. Kauzlar- TIVITY IN THE PRESENCE OF STRONG PAULI PARAMAG- ich, and J. D. Thompson. Localized 5f electrons in NETISM - CECU2SI2. 1979. PHYSICAL REVIEW LETTERS. superconducting PuCoIn5: consequences for super- 43 (25): 1892. conductivity in PuCoGa5. 2012. JOURNAL OF PHYSICS- CONDENSED MATTER. 24 (5): -. 19. BELLARBI, B., A. BENOIT, D. JACCARD, J. M. MIGNOT, and H. F. BRAUN. HIGH-PRESSURE VALENCE INSTABIL- 29. Pfleiderer, C., P. Boni, T. Keller, U. K. Rossler, and A. ITY AND TC MAXIMUM IN SUPERCONDUCTING CECU- Rosch. Non-Fermi liquid metal without quantum criti- 2SI2. 1984. PHYSICAL REVIEW B. 30 (3): 1182. cality. 2007. SCIENCE. 316 (5833): 1871. 20. THOMAS, F., J. THOMASSON, C. AYACHE, C. GEIBEL, 30. Park, T., V. A. Sidorov, F. Ronning, J. X. Zhu, Y. Tokiwa, and F. STEGLICH. PRECISE DETERMINATION OF THE H. Lee, E. D. Bauer, R. Movshovich, J. L. Sarrao, and PRESSURE-DEPENDENCE OF T(C) IN THE HEAVY-FER- J. D. Thompson. Isotropic quantum scattering and MION SUPERCONDUCTOR CECU2SI2. 1993. PHYSICA unconventional superconductivity. 2008. NATURE. 456 B-CONDENSED MATTER. 186-88: 303. (7220): 366. 21. Bao, W., P. G. Pagliuso, J. L. Sarrao, J. D. Thompson, 31. Basic Energy Sciences 2011 summary report. 2011. Z. Fisk, J. W. Lynn, and R. W. Erwin. Incommensurate U.S. DOE 2011. Basic Energy Sciences 2011 Summary magnetic structure of CeRhIn5. 2000. PHYSICAL RE- Report. VIEW B. 62 (22): 14621. 134

Page 135

Chemistry and Material Sciences Exploratory Research Continuing Project Understanding Earth’s Deep Water Cycle: Neutron Diffraction and Calorimetric Studies of Hydrous Minerals Hongwu Xu 20110190ER Introduction cal, and biological processes that sustain ecosystems and As part of the global hydrologic cycle, the Earth’s deep influence climate, hazardous events, and related global water cycle plays a significant role in many geological change. The unique neutron and calorimetric capabilities processes, including arc volcanism, deep focus seismic- developed/optimized in this research will have a wide ity, and, in a larger context, the evolution of our planet. range of applications in materials science, physics, chem- To understand Earth’s deep water cycle, it is essential istry, Earth and environmental sciences, and thus will to determine the stability and fate of hydrous minerals, strengthen key capabilities required for future Labora- which occur in hydrated oceanic crust, when the crust is tory mission areas in plutonium science, energy security subducted into the mantle via the mechanism of plate and environmental remediation. Potential future spon- tectonics. The aim of this research is to characterize sors for water/volatile cycle research include NASA, the crystal structures, mechanical properties, thermo- DOE’s Office of Science, and NSF. We will closely engage dynamic stability and phase equilibria of a number of appropriate program managers in the Lab and the DOE important hydrous minerals using neutron and synchro- complex throughout the project. tron x-ray diffraction at simultaneous high-pressure (P) and high-temperature (T) conditions (high P-T neutron Progress diffraction is a unique capability at LANL), together with First, we conducted high-pressure synchrotron infrared high-T solution calorimetry. Specifically, we will exam- (IR) spectroscopic experiments of brucite (Mg(OD)2) and ine three groups of minerals: 1) simple hydroxides; 2) hydrous phase A (Mg7Si2O8(OD)6) using diamond-anvil oceanic crust minerals; and 3) potential hydrous mantle cell (DAC) techniques. Second, we carried out synchro- minerals. In particular, since these minerals contain tron X-ray diffraction (XRD) measurements of Ni(OD)2 large amounts of light elements, particularly hydrogen, and Ni(OH)2 at high pressure (P) and temperature (T) us- neutron diffraction is an ideal tool for such studies. ing multi-anvil techniques. Third, we performed analyses Systematic measurements of these minerals using neu- of high P-T synchrotron XRD and neutron diffraction (col- tron diffraction coupled with our cutting-edge high P-T lected previously) data of jarosite (KFe3(SO4)2(OD)6). technique at LANSCE will yield valuable information such The goals were to study the structures and stability of as hydrogen positions and displacement parameters that these phases at high-P high-T conditions. Hydrous phase cannot be obtained with other methods. The obtained A is a potential host for water in the Earth’s mantle, and results will provide important parameters for models simple hydroxides (brucite and nickel hydroxide) are of Earth’s deep water cycle and thus will shed light on present as component units in the structures of complex mechanisms underlying the storage, fate and dynamics hydrous minerals. Jarosite occurs on Mars, which is con- of water in the Earth’s interior. sidered as a strong evidence for the existence of water (and life) in ancient Mars. Hence, studying the structures Benefit to National Security Missions and stability of these minerals at relevant P-T conditions Successful execution of this project will contribute will provide important insights into the mechanisms of greatly to the fundamental understanding of Earth’s water storage in the deep Earth and Mars (if the same deep cycling of water, CO2 and other volatiles and its plate tectonic mechanism is operating). related geological activities including volcanism and seis- micity. This work is also an essential part of studies of High-P synchrotron IR spectroscopy of brucite and phase the global water cycle which integrates physical, chemi- A: IR spectra of Mg(OD)2 and Mg7Si2O8(OD)6 were col- 135

Page 136

lected from 600 to 8000 cm-1 in frequency at pressures up to 1066 K on polyhalite (K2MgCa2(SO4)4·2H2O), an impor- to 31 GPa. Although both samples are nominally deuterat- tant mineral coexisting with salt crystals in salt reposito- ed phases, their IR spectra indicate that they contain small ries. 2) Two clathrate hydrates containing CO gas molecules amounts of OH, presumably due to the exchange of D by H were synthesized and characterized at high-P low-T condi- from moisture during sample synthesis/handling. Because tions; 3) A piston-cylinder apparatus that can reach 4 GPa IR spectroscopy is more sensitive to OH than OD, the and 2373 K (a new materials synthesis capability at LANL) presence of small amounts of OH results in significant IR was installed. signals arising from OH. Thus the samples can be used for studying both OD and OH vibrational properties. No phase Future Work transition was obvious in brucite on compression; how- During the extension period (FY14) of this project, we will ever, starting from 2.2 GPa, a new IR band (2687 cm–1) as- focus our efforts on neutron diffraction studies of hydrous sociated with OD appeared and its intensity increased with phase A, an important host phase for water in the Earth’s pressure. For Phase A. a phase transition was observed at mantle. A deuterated sample, Mg7Si2O8(OD)6, has been 12 GPa, as evidenced by the slope changes in variations of synthesized specifically for neutron measurements, to OH/OD band frequencies with pressure. avoid large incoherent scattering of H, which would cause high backgrounds of neutron diffraction patterns. We High P-T synchrotron XRD of Ni(OD)2 and Ni(OH)2: The intend to determine the crystal structure, thermo-mechan- reported bulk moduli for nickel hydroxide exhibit a wide ical properties and thermodynamic stability of this phase range, from about 52 to 88 GPa. One possible reason is at high pressure (P) and/or high temperature (T). Neutron the use of Ni(OD)2 versus Ni(OH)2. To determine the pos- diffraction, which detects the nucleus rather than electron sible H/D isotope effect, we carried out synchrotron XRD cloud and can sense light elements, represents a powerful on Ni(OD)2 and Ni(OH)2 at identical P-T conditions. At a tool for locating hydrogen/deuterium positions in hydrous given condition, unit-cell parameters a and c of Ni(OH)2 minerals. With our unique high P-T neutron capability are larger than those of Ni(OD)2. The resulted smaller (up to 10 GPa and 1800 K) at LANSCE, we have success- unit-cell volumes for Ni(OD)2 imply that the structure fully conducted experiments on various materials/miner- may preferentially incorporates D over H at high pressure. als aimed at refining their atomic positions and thermal Furthermore, Ni(OD)2 is only slightly more compressible parameters at high P-T conditions. In addition, we plan to than Ni(OH)2 (primarily along the c-axis). Thus the discrep- perform high P-T synchrotron X-ray diffraction experiments ancies in reported bulk moduli are not due to H/D isotope on this phase using multi-anvil and diamond-anvil cell tech- effect. In fact, for the same Ni(OD)2 sample, we obtained niques, which are complementary to neutron methods and a bulk modulus of 148.4 GPa using DAC synchrotron XRD allow measurements at higher pressures. Lastly, we will and a value of 57.3 GPa with toroidal-anvil-cell neutron dif- complete analyses of the obtained synchrotron infrared fraction. This is likely due to differences in stress state and spectroscopy data of phase A (as well as brucite), summa- diffraction geometry between the two techniques. rize the results, and prepare manuscripts for publication (including those on synchrotron X-ray and neutron diffrac- High P-T synchrotron X-ray/neutron diffraction of jarosite: tion results of brucite, nickel hydroxide and jarosite). The Several sets of measurements were performed to 9 GPa overall goal is, via integrated, multi-technical approaches, and 1273 K to map out jarosite P-T stability regions. For ex- to determine the effects of water on phase relations and ample, at 9.1 GPa, jarosite was stable up to 824 K, but then stability, physical properties and melting temperatures of decomposed partially into yavapaiite, hematite and water related minerals – a critical issue for understanding Earth’s at 839 K. This dehydroxylation temperature is much higher deep volatile cycle. For follow-up projects, we will con- than that at ambient pressure (between 550 and 575 K), tinue to explore the application of high P-T diffraction and suggesting a positive P-T slope for the dehydroxylation spectroscopy methods to studies of minerals (e.g., clay and reaction. At a given pressure, the c dimension of jarosite clathrate hydrate) important to Earth, environmental and expands at a rate more rapidly than that for a. Similar to energy systems, such as enhanced geothermal systems and the behavior at room pressure, this anisotropy is largely salt repositories (such as WIPP). due to the ease of tilting the Fe(O,OD)6 octahedra about the a-axes, which results in rapid increase in the (001) layer Conclusion thickness with temperature. However, the magnitudes of Using neutron/synchrotron x-ray diffraction and solution the thermal expansion are smaller than those at ambient calorimetry, we will determine crystal structures, elastic pressure, apparently due to the confining effect. properties, thermodynamic stability and phase equilibria of a number of oceanic crust and mantle hydrous minerals Other activities: 1) High-T synchrotron XRD was conducted 136

Page 137

at high P-T conditions. In particular, we will characterize hy- drogen bonding in these minerals, which plays a key role in their thermodynamic stability, dehydration processes and phase relations. These results will help construct phase diagrams of hydrous peridotite in MgO-SiO2-H2O and related systems at mantle conditions and will thus shed lights on the mechanisms of volatile circulation through the mantle and the origins of the arc volcanoes and deep focus earthquakes. Publications Xu, H. W., Y. S. Zhao, J. Z. Zhang, Y. J. Wang, D. D. Hickmott, L. L. Daemen, M. A. Hartl, and L. P. Wang. Anisotropic elasticity of jarosite: A high-P synchrotron XRD study. 2010. AMERICAN MINERALOGIST. 95 (1): 19. Xu, H. W., Y. S. Zhao, S. C. Vogel, D. D. Hickmott, L. L. Dae- men, and M. A. Hartl. Thermal expansion and de- composition of jarosite: a high-temperature neutron diffraction study. 2010. PHYSICS AND CHEMISTRY OF MINERALS. 37 (2): 73. Xu, H., D. D. Hickmott, Y. Zhao, J. Zhang, Q. Wei, S. C. Vogel, and L. Wang. High-P/T synchrotron X-ray and neutron diffraction study of carbonate minerals. 2010. GEOCHI- MICA ET COSMOCHIMICA ACTA. 74 (12): A1159. Xu, H., D. Hickmott, J. Zhang, Y. Zhao, S. Vogel, and L. Dae- men. Structure and stability of nickel hydroxide at high T-P conditions. Presented at Goldschmidt 2011. (Prague, Czech Republic, 14-19 Aug. 2011). Xu, H., Y. Zhao, D. D. Hickmott, N. J. Lane, S. C. Vogel, J. Zhang, and L. L. Daemen. High-temperature neutron diffraction study of deuterated brucite. 2013. Physics and Chemistry of Minerals. : 10.1007/s00269. Zhu, J. L., H. W. Xu, J. Z. Zhang, C. Q. Jin, L. P. Wang, and Y. S. Zhao. Thermal equations of state and phase rela- tion of PbTiO3: A high P-T synchrotron x-ray diffraction study. 2011. JOURNAL OF APPLIED PHYSICS. 110 (8): -. Zhu, J. L., S. M. Feng, Q. Q. Liu, J. Z. Zhang, H. W. Xu, Y. C. Li, X. D. Li, J. Liu, Q. Z. Huang, Y. S. Zhao, and C. Q. Jin. Temperature and pressure effects of multiferroic Bi2N- iTiO6 compound. 2013. JOURNAL OF APPLIED PHYSICS. 113 (14): -. Zhu, J. L., Z. J. Lin, J. Z. Zhang, H. W. Xu, S. C. Vogel, C. Q. Jin, and Y. S. Zhao. High pressure neutron and synchrotron X-ray diffraction studies of tetragonal LaFeAsO0.9F0.1. 2012. HIGH PRESSURE RESEARCH. 32 (3): 405. 137

Page 138

Chemistry and Material Sciences Exploratory Research Continuing Project Developing Potentials for Atomistic Modeling of Defect Phenomena at Metal- Ceramic Interfaces Steven M. Valone 20120053ER Introduction last 5 years assumes the availability of atomistic models The goal of this project is to understand defect interac- for complex materials that in fact are not in evidence. tions with metal-ceramic interfaces at an atomistic level Campaigns 1 and 2, BES, Waste Storage programs (Veirs), by means of a new, unified model for both metals and and other LDRD projects have interests in the success of ceramics. Defects e.g., point defects or line defects (dis- this project. If this project is successful, the new model- locations), interacting with complex interfaces, influence ing tool will open new vistas for coupling mechanical and many interfacial properties and the resulting material electronic structure responses of nanoscale materials performance. The response of metal-ceramic interfaces that we cannot attempt now because the atomistic mod- to defects, whether generated from fabrication, cascade els do not have the right physics and the electronic struc- damage from irradiation, or materials deformation pro- ture methods cannot accommodate the time and length cesses, are complicated by the disparity in the materials scales of interfacial interactions and heterogeneity. in intimate contact. However many defects, their move- ments, and mutual interactions can not be observed Progress directly via experiments due to the short-time scales The second stage of the Fragment Hamiltonian (FH) involved. It is then desirable to understand the physical model of Ni metal was revised from the model produced processes and dynamics at the atomistic level via simula- in the first year of the project (2012) because we learned tions such as molecular dynamics (MD) and conventional new properties of the model. That is we came to un- or off-lattice-kinetic Monte-Carlo. All atomistic simula- derstand that the main components of the FH model tions require interatomic potentials to describe the inter- pertained to charge-transfer hopping, rather than intra- actions among atoms, or the bonding. Attaining this goal band hopping in the sense of tight-binding theory. The requires the development of a unified atomistic model “intra-band hopping” energies appear in a different part for both metals and ceramics as a central objective. For of the model than what we had first thought. The Ni mixed bonds environment such as the metal/ceramic paper had to be revised and is now ready for submission. interfaces, it is well known that there are charge transfer The results of the model show that several 3d phases issues that have puzzled the atomistic modeling commu- Ni can be modeled well with the FH model, as can most nity for many years. How to tackle the issue with physical other 3d structures, such as monovacancies and faulted principles, and its potential impact on understanding fcc lattices (ie half-way between fcc and hcp). Lower di- mixed-material interfaces, prompts the writing of this mensional structures (1d and 2d lattices, surfaces, grain ER project. A spin-off of broad importance to materials boundaries) are not as well represented with this first modeling will be high fidelity interatomic potentials that version of the FH model. Postdoc Ghanshyam Pilania describe the metallic, ionic, and covalent bonds in one, was hired in March 2013 from Ramprasad’s group at unified model. The model is to reduce to, or otherwise UConn. Our first pass at expanding the FH model to a approximate, established models for these individual metal/metal-oxide system, namely Be/BeO, was under- classes of materials. taken. Pilania developed the essential database of infor- mation for the Be/BeO composite system from electronic Benefit to National Security Missions structure calculations. Baskes and Liu developed and Atomistic modeling capability is hugely important to applied the methodology to use the electronic structure predictive science capabilities in general. Nearly every data to produce an FH model of the mixed system. This multi-scale modeling proposal around the world for the work is on-going. 138

Page 139

In addition, Pilania was able to identify the lattice param- to a metal/metal-oxide systems, namely Be/BeO and Ni/ eters at which a metal-insulator transition occurred in NiO. The second stage of the model for Ni/NiO is being lattices of 1, 2 and 3 dimensions. As expected, this pat- conducted with the idea of using fits to spectroscopic tern, based on spatial dimensions, is observed in both Be measures, such as XPS, PES, and EELS, in order to devise lattices and BeO lattices and is expected in essentially all a model of the composite system. Our ability to use such crystalline materials. These dimensional dependencies data comes from the “tight-binding-like” interpretation of are important to capture in the model, both for reasons of the FH model. Postdoc Ghanshyam Pilania will develop- fidelity and for validation with electronic properties. Our ing information for both composite systems, and learning first effort to capture the opening and closing of an energy to analyze the results of electronic structure calculations gap in the FH model was restricted to a 1d chain model. to directly measure covalency as an output. The model Two cases were considered. The first case corresponds to derived in this way are to be tested with a 1d chain model. a large lattice spacing and atomic-like energy levels. In this As we refine the details of the FH model, Ben Liu will con- case, the hopping energies are close to zero (atoms are tinue to develop the LAMMPS computer code. The code stuck in a neutral state). There are energy gaps between will be used to estimate defect energies for more complex clusters of energy levels analogous to “HOMO-LUMO” structures (relaxed surfaces, grain boundaries, vacancies) gaps. The second case corresponds to a compressed lat- than what can be estimated analytically. Force fields will tice and bands of energy levels. In this second case, the be added to the code in the coming FY. Invited talks have same hopping energies were allowed to increase to values been accepted at the ACerS and TMS next Winter by both comparable to the gaps in the first case. The gaps closed Valone and Pilania. Also Valone was invited to contribute in a way that is reminiscent to the way a multi-band tight to a book chapter on potential energy surfaces and atom- binding model behaves. istic methods, being edited by colleagues at the University of Az. The FH model will be featured. That task is to be Because we were able to capture some form of energy- completed during the end of the first quarter of FY2014. gap closure in the FH model, Baskes proceeded to develop a model of the basic energy scales in the model so as to Conclusion possess both neighbor-coordination and dimensionality Understanding metal-ceramic interfaces and defect inter- dependencies. For this reason, we are able to represent actions are of great importance to applications of interest: structures of all three dimensionalities in Be. The Ni model radiation damage resistance, mechanical behavior, and still needs to be revised. Pilania was able to devise a novel electronic materials. Understanding the mechanistics method analyze the results of electronic structure calcula- of why some materials can withstand high-level irradiation tions that directly measures covalency as an output. He damage is of great interest for designing better materi- did so by deconstructing the electron density output from als. Metal-ceramic interfaces are of high interest for their an electronic structure calculation according the ansatz of extraordinary mechanical behavior. Finally, metal-ceramic the FH model. The analysis works well for any one materi- layered materials can often have surprising electronic al like Be or BeO. However, we are still experiencing some properties, as for example, the giant room-temperature inconsistency when we compare estimates of covalency magneto-resistance in single-crystal Fe/MgO/Fe magnetic for Be and for Be in BeO. As we have refined the details of tunnel junctions. The outcome of the project will be a new the FH model, Ben Liu has been developing the LAMMPS model that describes the atomistic level for both of these computer code (from SNL) that is serving as the repository disparate materials. for the FH model. The code is being and has been used to estimate defect energies for more complex structures (re- Publications laxed surfaces, grain boundaries, vacancies) than what can Kolluri, K., M. J. Demkowicz, R. G. Hoagland, and X. Y. Liu. be estimated analytically. Valone presented invited talks Behavior of vacancies and interstitials at semicoherent at the Institute for Pure and Applied Mathematics at UCLA interfaces. 2013. Journal of Metals. 65 (3): 374. and at U of Az. Liu, X. Y.. Interphase Defects, Structures, and Phase Stabil- ity. 2013. Journal of Metals. 65 (3): 358. Future Work The first stage of the Fragment Hamiltonian (FH) model of Valone, S. M., K. Muralidharan, and K. Runge. Interatomic Ni metal was revised from the model produced in the first Potentials Including Chemistry. To appear in Multiscale year of the project (2012) because we learned new proper- Paradigms in Integrated Computational Materials Sci- ties of the model. A paper is being prepared at the same ence and Engineering. Edited by Deymier, P. A.. time and is to be submitted in the next month (July 2013). Valone, S. M., S. R. Atlas, and M. I. Baskes. Embedded atom General characteristics of the model will be expanded 139

Page 140

method potentials from the fragment hamiltonian model: Crystalline phases and surfaces of nickel. 2012. LA-UR 2012-01180. Valone, S. M., and S. R. Atlas. Density and spectral-den- sity matrices in atomistic-scale models. 2012. LA- UR-2012-22846. Yadav, S. K., R. Ramprasad, A. Misra, and X. Y. Liu. Core structure and Peierls stress of edge and screw disloca- tions in TiN: A density functional theory study. (submit- ted to Acta Mater.) LA-UR-13-25179. Yadav, S. K., R. Ramprasad, J. Wang, A. Misra, and X. Y. Liu. First-principles study of Cu/TiN and Al/TiN interfaces: weak vs strong interfaces. LA-UR-13-27331 (submitted to Modelling and Simulation in Materials Science and Engineering). Yadav, S. K., X. Y. Liu, J. Wang, R. Ramprasad, A. Misra, and R. G. Hoagland. First-principles density functional theory study of generalized stacking faults in TiN and MgO. submitted to Philosophical Magazine (LA- UR-13-22557). 140

Page 141

Chemistry and Material Sciences Exploratory Research Continuing Project From Waste to Fuels and Feedstocks: Reduction of CO Using Main-Group 2 Catalysts Andrew Sutton 20120197ER Introduction Progress Carbon dioxide (CO2) is the greenhouse gas of greatest A technician has been employed part time and extensive concern in climate change, but it could also be the most screening of Lewis Acids and Lewis Bases has been car- abundant C1 building block on the planet if its capture ried out and this large range has been exposed to CO2 and conversion to a useful feedstock under mild condi- for binding property analysis. The best system is still the tions can be realized. In this proposal, we outline the one discovered with aluminum and phosphines from our pathway by which we believe this conversion can be initial studies and efforts to impart water and air stability accomplished catalytically. By eschewing the use of ex- have been progressing well. pensive and rare precious metals and using main-group (p-block) compounds instead, the process that we are New ligands have been designed and synthesized and proposing will be amenable to future implementation computational analysis has given us confidence these on a global scale. We aim to accomplish this conversion will be robust and catalytically active species. using main-group compounds in the form of frustrated We have also started utilizing bidentate Lewis Bases in Lewis pairs (FLPs, i.e. a Lewis acid/base pair that cannot water and air stable systems and hope these will also form self-adducts) to support the hydrogenation of CO2 show excellent CO2 binding affinities. to yield methanol (MeOH). We will use our experience with the H2 fuel carrier ammonia borane (AB) to design Future Work a regenerable system. Ultimately, we envision a double We have developed several new frustrated Lewis pairs FLP reaction pathway that is capable of the single-stage (FLPs) that we have shown via 13C NMR and IR to bind conversion of a H2/CO2 input stream into MeOH and CO2. These systems need to be further developed to H2O via simultaneous H2 cleavage and CO2 activation. reduce CO2 to methanol using H2, ammonia borane or Throughout this project, we will be advancing scientific other good reducing agent to enable the full sequestra- knowledge on the mechanism of interaction and reac- tion and reduction of CO2 to be performed. We also tion in these systems. need to impart increased aqueous and acid tolerance to our FLP systems. Benefit to National Security Missions Due to concern over CO2 as a greenhouse gas, the DOE Conclusion has taken steps to develop the conversion of CO2 to val- The scale of CO2 production creates significant problems ue-added products such as fuels and feedstocks. In order in capture and sequestration as there is little short-term to do this effectively on a large scale, cost and material economic incentive to do so and the quantities involved availability must be considered. Moving away from pre- are large. The work we propose will address this issue cious metals to main-group elements as catalysts would by developing chemistry to convert CO2 into a value- enable us to develop a catalytic system for CO2 reduc- added molecule, MeOH, which is a C1 feedstock. In the tion which would not only be relatively cost-efficient, but course of this work, we expect to advance our under- also support the DOE & LANL Grand Challenges. Finally, standing of multi-electron reduction chemistry using better understanding of these reaction mechanisms may abundant main group materials. From this knowledge enhance our ability to mitigate other waste streams. base, we will be able to expand our investigations into 141

Page 142

other chemistries such as the activation of small molecules and C-C bond forming reactions. Publications Smythe, N. C., A. D. Sutton, D. A. Dixon, and B. Zelenay. Aluminum Oxide Frustrated Lewis Pairs for CO2 Con- version to Methanol. To appear in TBD. 142

Page 143

Chemistry and Material Sciences Exploratory Research Continuing Project First Principles Many-body Approaches to Strongly Correlated Actinide Metals Jianxin Zhu 20120232ER Introduction Benefit to National Security Missions Since the early days of the Manhattan Project, it has long The development of a new first-principles electronic- been a cherished goal, often thought impossible, to have structure method for strongly correlated actinide materi- an accurate first-principles, parameter-free, capability to als that can accurately calculate effects of electronic calculate and predict, anywhere in the phase diagram, correlations on physical properties at any given tem- the materials properties of plutonium (Pu), which has perature and pressure will have immediate application an extremely anomalous metallurgy. The complexity of in a variety of NW, energy, high-performance (exascale) Pu metals is closely related to the unique position of Pu computing, and IS&T missions at the Laboratory. In the in the last row of periodic table, where the 5f electrons energy arena, for example, this will enable accurate ma- of Pu are borderline between being either itinerant or terials predictions for actinide oxide fuels. Once the ba- localized. This implies the presence of strong electronic sic method is established and validated, future program correlation in these materials. development will be needed to generalize it in order to calculate various specific materials properties of interest. Recent development of the combination of the local For example, a phonon module should be developed so density approximation (LDA) and dynamical mean field that effects of thermal vibrations on the free energy at theory (DMFT), so called the LDA+DMFT method, has finite temperatures can be calculated.The code should advanced our understanding of emergent phenomena also be generalized and exploited as an accurate method like metal-Mott-insulator transitions with structure to calculate phase diagrams of relevant materials. Elastic information in strongly correlated materials. However, constants and mechanical properties could also be calcu- the LDA+DMFT suffers from the weaknesses with the lated. uncertainty of screened local Coulomb interactions and the issue introduced by a double-counting correction. Progress In this project we will develop a first-principles version We have calculated the LDA and relativistic quasi-parti- of DMFT, which will be combined with the quasipar- cle self-energy (GW) renormalized band structure of U, ticle GW theory. GW refers to the theory that predicts Np, Pu, and extended Pu systems. With increasing lattice electronic energy for ground and excited states of constants and partially filled f-orbitals, these systems al- materials. In the latter framework, the screened local low us to understand the correlation physics from itiner- Coulomb interactions are calculated on the fly with no ant to localization limit. With 5f electrons, the spin-orbit adjustable parameters. The development of this theory coupling is non-negligible in these systems and the Fermi will thus form a fundamental basis for new advances in energy is very close to the spin-orbit split. For quasi- our understanding of the superconductivity, magnetic particle correction, we have therefore employed a scalar and heavy fermion behaviors, and many other anoma- relativistic extension of single shot GW approximation. lous properties of strongly correlated d- and f-electron We have also estimated average screened Coulomb materials. It will replace the “empirical” nature of cur- interaction U in the static limit for all systems. In the rent predictions with a new theory that has controllable weakly interacting localized electronic systems, GW self- approximations, and as such can be systematically im- energy is well known to incorporate the dynamic correla- proved. We focus on the application to actinide materi- tion in both long- and short-range limit. With increasing als. This new theoretical technique will position LANL to U, such dynamical quasi-particle correction becomes be a leader in actinides theory in specific, and materials over-burdened in the short range limit and thus calls for theory and simulations in general. 143

Page 144

proper atomic treatment, which is widely addressed by to a superconducting instability. We anticipate predict- various versions of dynamical mean field theory (DMFT). ing the symmetry of the dominant pairing channel(s) and Our first-principles GW calculations thus benchmark cor- classifying the most important Fermi surface sheets for relation strength in the light actinides from itinerant-to- nesting. localized systems including the spin-orbit coupling effects. We have also developed an intermediate Coulomb-U Conclusion coupling model for intermetallic actinides to capture their This project will integrate GW methods with DMFT tech- electronic structure properties affected by electron cor- niques to generate a first-principles, parameter free relations. This model is based on material-specific ab-initio method that can accurately calculate electronic correlation band structure from which correlation effects are com- effects for any material at any temperature and atomic vol- puted via self-consistent GW-based self-energy corrections ume. This code will enable us to confidently predict trends arising from spin fluctuations. in correlation effects systematically for different crystalline phases of elemental actinides as well as for classes of dif- We applied this approach to four isostructural intermetal- ferent actinide compounds. lic actinides PuCoIn5, PuCoGa5, PuRhGa5 belonging to the Pu-115 family, and UCoGa5, a member of the U-115 family. It will provide us with a tool to confidently understand The 115 families share the property of spin-orbit split den- many anomalous properties in actinide metals. It will pro- sity of states enabling substantial spin fluctuations around vide a general-purpose electronic structure method that 0.5 eV, whose feedback effect on the electronic structure can also be applied to related correlated transition metal creates mass renormalization and electronic ‘hot spots’, and rare-earth compounds. i.e., regions of large spectral weight in momentum-energy space. Publications Ahmed, T., R. C. Albers, A. V. Balatsky, C. Friedrich, and J. A detailed comparison was provided for the angle-resolved Zhu. GW quasiparticle calculations with spin-orbit cou- and angle-integrated photoemission spectra and de Haas- pling for the light actinides. 2013. arXiv.org. van Alphen experimental data as available. The results suggested that this class of actinides is adequately de- Das, T.. Interaction induced staggered spin-orbit order in scribed by the intermediate Coulomb interaction regime, two-dimensional electron gas. 2012. Physical Review where both itinerant and incoherent features coexist in the Letters. 109: 246406. electronic structure. Das, T.. Staggered spin-orbit order: A new paradigm of bro- ken symmetry phase of matter. 2013. Journal of Super- Future Work conductivity and Novel Magnetism. 26: 1673. In the next fiscal year, we will continue applying the Das, T., J. Zhu, T. Durakiewicz, J. J. Joyce, and M. J. Graf. Ma- GW+SO method to investigate the electronic structures of terials specific electronic correlation effects and spec- many heavy fermion systems including some new actinide- tral weight `hot spots’ in inter metallic actinides. 2012. based topological insulators, e.g., PuSb. We are particularly In 2012 MRS Spring Meeting. (San Francisco, 9-13 interested in looking into the bulk energy gap in these April, 2012). Vol. 1444, p. 169. Cambridge: Cambridge systems. We will start exploring the possibility of calcu- University Press. lating the total energy within the GW Green’s function formalism, which is important to the study of equilibrium Das, T., J. Zhu, and M. J. Graf. Spin fluctuations and the volume of actinides. Therefore, our short term priority is peak-dip-hump feature in the photoemission spec- to develop this methodology which can be built upon the trum of actinides. 2012. Physical Review Letters. 108: 017001. existing scalar relativistic GW code. Das, T., J. Zhu, and M. J. Graf. The electronic structure and We plan to integrate our SO+GW method in a self-consis- self-consistent spin-fluctuation corrections of inter tent GW+DMFT(CTQMC) formalism which will not only metallic actinides. 2013. Journal of Materials Research. allow us to study the strongly correlated systems or metal- 28: 659. to-insulator(MIT) transitions from first-principles, but also enable us to understand electronic structure associated Das, T., T. Durakiewicz, J. Zhu, J. J. Joyce, J. L. Sarrao, and M. J. Graf. Imaging the formation of high-energy disper- more complicated Kondo like physics in actinide systems. sion anomaly in the actinide UCoGa5. 2013. Physical In parallel, we plan to extend the GW-based self-energy Review X. 2: 041012. method based on spin fluctuations necessary for identify- ing the electronic hot spots at the Fermi surface that lead 144

Page 145

Nica, E. M., K. Ingersent, J. Zhu, and Q. Si. Quantum criti- cal Kondo destruction in the Bose-Fermi Kondo model with a local transverse field. 2013. Physical Review B. 88: 014414. Ronning, F., J. Zhu, M. J. Graf, R. C. Albers, H. Rhee, and W. E. Pickett. Superconducting gap structure of the 115’s revisited. 2012. Journal of Physics: Condensed Matter. 24: 294206. Svane, A., R. C. Albers, N. E. Christensen, M. van Schilf- gaarde, A. N. Chantis, and J. Zhu. The electronic cor- relation strength of Pu. 2013. Physical Review B. 87: 045109. Zhu, L., and J. Zhu. Singularity in self-energy and compos- ite fermion excitations of interacting electrons. 2013. Physical Review B. 87: 085120. 145

Page 146

Chemistry and Material Sciences Exploratory Research Continuing Project Lanthanum Bromide Glass Ceramics for Gamma-Ray Spectroscopy Markus P. Hehlen 20120246ER Introduction the single crystal-performance correlation.The use of Our goal is to demonstrate a novel rugged lantha- scalable synthesis techniques that do not require long num bromide glass-ceramic scintillator for gamma-ray time periods and expensive equipment suggests that spectroscopy. This is important because the best and these glass ceramic scintillators can be produced at costs most desired gamma-ray scintillator available today– at least 20× lower than respective crystals. At the same cerium-doped lanthanum bromide (LaBr3:Ce3+) – is an time, the high-Z and high light output of these materials extremely hygroscopic crystal that is difficult to grow will preserve good performance. This is of interest to our and a challenge to handle. The resulting high cost, small homeland security, nuclear nonproliferation, and space size, and few suppliers limit its use to a small number systems customers. of specialized detectors. What is urgently needed is a gamma-ray scintillator with LaBr3:Ce3+ performance Progress that is chemically inert, mechanically rugged, and that The main goal of the project is to create for the first time can be manufactured economically in large size. If suc- a material that contains nano-particles of a high-perfor- cessful, we will be the first to demonstrate such a mate- mance halide scintillator (such as LaBr3:Ce) dispersed in rial. This would be a breakthrough in scintillator research a chemically inert silica matrix. Our baseline approach and a game-changer for a broad range of customers was to utilize a non-hydrolytic sol-gel (NHSG) synthesis who rely on gamma-ray spectroscopy in a wide range of to form the silica matrix in the presence of dissolved homeland security, non-proliferation, and intelligence LaBr3:Ce, followed by heat treatment to induce the applications.Our primary research objective is to fabri- precipitation of LaBr3:Ce nano-crystals. NHSG synthesis cate a chemically inert and mechanically rugged oxide is attractive because it allows for the solution-based glass matrix in which we grow in situ a high density of formation of a silica-based glass at room temperature LaBr3:Ce3+ nano-crystals. The resulting scintillating glass under the exclusion of water, which would readily react ceramic will be optically transparent and encapsulate in undesired ways with LaBr3:Ce. The focus in the first the hygroscopic LaBr3:Ce3+.Such a material is novel and half of FY13 was on completing the development of the does not currently exist. Its fabrication requires special NHSG process. We successfully fabricated transparent synthesis methods because bromides react readily with silica-gels with this method. A setback was encountered water and oxides in undesired ways. We will use non- when we observed that the solubility for LaBr3:Ce in the hydrolytic sol-gel (NHSG) processing as an anhydrous, starting solutions was insufficient. This initiated an ex- low-temperature, and scalable synthesis route to this tensive experimental campaign that exhaustively tested new class of nano-composite materials. various starting reactants, the process temperature, and several additional solvents for their ability to dissolve Benefit to National Security Missions LaBr3:Ce. We concluded that the LaBr3:Ce solubility Scintillators are currently used in numerous types of was too low in any of the NHSG chemistry to obtain the radiation detection systems for basic science, medi- required large mass fraction of LaBr3:Ce needed in the cal imaging, non-destructive evaluation, and radiation solution for the final application. detection. This projects seeks to enable a breakthrough in scintillators that would directly benefit all of these This extensive sol-gel synthesis effort led us to discover areas. Our prime goal is to break the price-performance an alternative sol-gel approach to a nano-structured correlation in radiation detection materials by breaking scintillator. LSO:Ce is a bright gamma-ray scintillator with 146

Page 147

an energy resolution comparable to NaI:Tl but 6 times and measure a gamma-ray spectrum determine the energy shorter decay time. Being able to produce it via a low-cost resolution of this composite material. This effort is guided sol-gel process would be very attractive for a number of by a diverse set of optical and structural characterization applications. We found that an LSO:Ce based nano-struc- methods. tured glass ceramic can be fabricated with a sol-gel process that uses regular siloxane chemistry. Our first experiments Both these development efforts will allow us to obtain the yielded relatively transparent gels with the required high first nano-structured glass ceramic samples that may offer concentration of the lutetium precursor. This process is a lower cost way to high resolution scintillators compared currently being systematically optimized with the goal of to bulk crystals. We will publish this work in peer-reviewed scaling it up to larger sample sizes (several cubic centime- papers and, if possible, present at an international confer- ters) in FY14 for subsequent measurement of scintillation ence (ICL’2014). photon yield and energy resolution. Conclusion We have also initiated scintillator syntheses using meso- We expect to fabricate a transparent silica glass ceramic porous silica. These materials are silica-glass monoliths containing a 50-60% volume fraction of <10 nm diameter that contain a network of open pores with pore sizes of LaBr3:Ce3+ nano-crystals. This nano-composite mate- <10 nm and are therefore optically transparent. They lend rial will have scintillation performance that rivals that of themselves to infusion of the pores with scintillator materi- LaBr3:Ce3+ single-crystal. Success will be achieved when als to form a highly loaded nano-structured scintillator. we have demonstrated (1) an energy resolution of <3.5% This approach has not been reported before. We have suc- at 662 keV, (2) a scattering length of ≥10 cm at the 380 nm cessfully filled a large fraction of the available pore volume scintillation wavelength, and (3) stability against hydrolysis (28 vol% for Vycor) with LaBr3:Ce and SrI2:Eu (two highly in ambient air over a period of 1 month. Achieving this efficient scintillators) using immersion of the mesoporous would constitute a breakthrough in scintillator research silica in the respective halide melt. The resulting compos- and would result in publications/presentations, patents, ites exhibit bright photo- and radioluminescence, which follow-on funding, and interest by commercial companies. are pre-requisites for efficient gamma-ray scintillation. This development currently focuses on developing the process Publications parameters to obtain composites with high optical trans- Blair, M. W., M. Fasoli, S. C. Tornga, A. Vedda, N. A. Smith, parency. B. L. Bennett, M. P. Hehlen, and R. E. Muenchausen. Nanophosphor GdOBr:Ce via combustion synthesis: Our work in FY13 has also resulted in two accepted publi- luminescence results. 2013. PHYSICA STATUS SOLIDI C: cations in the Journal of Luminescence, an accepted publi- CURRENT TOPICS IN SOLID STATE PHYSICS, VOL 10, NO cation in Optical Materials, and an invited presentation at 2. 10 (2): 227. the Phosphor Global Summit. Hehlen, M. P.. Short Course: Fundamentals of rare-earth spectroscopy. 2013. In Phosphor Global Summit 2013. Future Work (New Orleans, 18-20 March 2013). , p. 1. Falmouth, The following goals and associated tasks are scheduled for ME: Smithers. next fiscal year. Hehlen, M. P., M. G. Brik, and K. W. Kramer. 50th anniversa- Sol-gel process development ry of the Judd-Ofelt theory: An experimentalist’s view of the formalism and its application. 2013. JOURNAL The sol-gel process development for LSO scintillator will OF LUMINESCENCE. 136: 221. be completed. The goal is to produce a nano-structured scintillator sample of sufficient size to measure a gamma- Hehlen, M. P., N. A. Smith, M. W. Blair, A. Li, S. Stange, R. ray spectrum and determine the energy resolution of D. Gilbertson, E. A. McKigney, and R. E. Muenchausen. this composite material. This involves optimization of the Nanocomposite scintillators. Nanocomposite, ceramic sol-gel chemistry as well as the parameters for the heat and thin-film scintillators. Edited by Nikl, M.. treatment of the gel. This effort is guided by a diverse set Lakshminarayana, G., E. M. Weis, A. C. Lira, U. Caldino, D. of optical and structural characterization methods. J. Williams, and M. P. Hehlen. Cross Relaxation in rare- earth-doped oxyfluoride glasses. 2013. JOURNAL OF Infused meso-porous silica LUMINESCENCE. 139: 132. Encouraged by our first results in FY13, we will pursue this alternate method of fabricating a nano-structured scintil- Lakshminarayana, G., E. M. Weis, B. L. Bennett, A. Labouri- lator. We will tune the process to maximize transmission au, D. J. Williams, J. G. Duque, M. Sheik-Bahae, and M. 147

Page 148

P. Hehlen. Structural, thermal, and luminescence prop- erties of cerium-fluoride-rich oxyfluoride glasses. 2012. OPTICAL MATERIALS. 35 (2): 117. Li, A., N. Smith, M. P. Hehlen, E. A. McKigney, and R. Gard- ner. Light yield measurement method for milled nano- size inorganic crystals. 2012. Applied Radiation and Isotopes. 70: 1219. Vasudevan, K. V., W. L. Boncher, N. A. Smith, M. W. Blair, B. L. Scott, B. L. Bennett, M. P. Hehlen, R. E. Muenchau- sen, and J. C. Gordon. Nitrile-supported coordination polymers of Cerium (III) Bromide. To appear in Inor- ganic Chemistry. 148

Page 149

Chemistry and Material Sciences Exploratory Research Continuing Project Are Nanoscale Foams Radiations Resistant? Jose A. Caro 20120250ER Introduction in the functionality of nano-structured materials. There The search in recent years of materials with improved is also a close connection to MaRIE, which has a focus radiations response has been based on engineering on materials in extreme irradiation environments; LANL nanoscale features that act as efficient sinks for irradia- will benefit from supporting research that explores the tion created defects, such as the Nanostructured Ferritic fundamental science of radiation tolerance in materi- Alloys (NFA), or Oxide Dispersion Strengthened (ODS) als. Developing a basic understanding in the area of this steels. Similarly, the search of materials with improved project will place us in a strategic position to seek future mechanical properties (mainly strength) has been based funding from the DOE Office of Nuclear Energy and Of- on engineering nanoscale features that inhibit disloca- fice of Basic Energy Science. tion motion over long distances, typically by decreasing grain size to the nm range. These strategies have given Progress impressive results by increasing radiation tolerance and In previous work, we studied the response of nano- hardness compared to coarse grain counterparts. porous gold (np-Au) foams under irradiation at room temperature and different dose-rates [E. Fu, et al. APL By shifting from internal interfaces in bulk materials to 101 (2012) 191607]. Our experimental findings show surface derived phenomena in nanoporous materials, that 400 keV Ne++ ion irradiation for a total dose of 1 a field that has not been explored so far, a new under- dpa leads to the formation of Stacking Fault Tetrahedra standing of radiation tolerance and strengthening will be (SFTs) at high and intermediate dose-rate, while no SFTs obtained. Foams with filament and porous sizes in the are formed at low dose-rate. An atomic-view of the range of nanometers are ideal systems to study these ef- process based on Molecular Dynamics simulations (MD) fects and could provide insight on designing the ultimate shows that vacancy migration distance and nanofoams material regarding perfect sink strength offered by the ligament size play a key role in explaining the dose-rate surface of the pores, and ultimate strength offered by dependent defect accumulation. dislocation free ligaments. In the past months, we continued this study by inves- The goal of this project is to investigate the role of sur- tigating the mechanical response of nanofoams under face-to-volume ratio on the irradiation and mechanical irradiation. These investigations have two parts: 1) properties of nanoscale foams. We will prove that high computer simulations studies in which we observe that surface density nano-foams can be tailored to become nanofoams soften under irradiation and 2) experimental simultaneously radiation tolerant and ultra-strong. indentation studies which are in good agreement with this prediction. Benefit to National Security Missions This work will lead to improved understanding and A paper was published on the deformation mechanism control of defects and interfaces, which is a critical of the ligaments that forms the nanofoam [L. Zepeda- capability for the Laboratory’s mission in energy secu- Ruiz, et al., APL (2013) in press]. MD simulations of rity, specifically in the area of where defects and inter- Au nanopillars under compression in the unirradiated faces control phenomena such as radiation hardness of (defect-free) and irradiated condition (with a SFT at the materials in nuclear reactors. It will also provide insight center of the nanopillar) show that SFTs play an im- into how interfaces and defects play an important role portant role in the deformation. Upon applied stress, 149

Page 150

plastic deformation originates at the SFT, which becomes Publications an almost inexhaustible source of dislocations. These Caro, M., E. G. Fu, L. Zepeda-Ruiz, W. Mook, Y. Q. Wang, J. results indicate that the presence of SFTs in Au ligaments K. Baldwin, C. Sheehan, E. Bringa, M. Nastasi, and A. after irradiation will induce a decrease of yield stress under Caro. Radiation response of nano-scale gold foams. compression, reducing its yield strength. Invited presentation at 8th Pacific Rim International Congress for Advanced Materials and Processing (PRICM-8). (Waikoloa, HI, 4-9 Aug., 2013). np-Au deformation behavior under compression was stud- ied before and after ion irradiation at room temperature. Caro, M., E. G. Fu, L. Zepeda-Ruiz, Y. Q. Wang, J. K. Baldwin, Nanoporous gold specimens were prepared by chemically C. Sheehan, E. Bringa, M. Nastasi, and A. Caro. Radia- dealloying in nitric acid ~ 100 micron thick Au30/Ag70 thin tion effects in nanoporous gold. Presented at TMS An- films deposited on a multilayered substrate. Preliminary nual Meeting Symp. on Functional Nanomaterials. (San characterization and nanoindentation test results obtained Antonio, TX, 3-7 Mar., 2013). as a function of etching time and irradiation conditions Caro, M., E. G. Fu, L. Zepeda-Ruiz, Y. Q. Wang, K. Baldwin, confirm that: 1) np-Au hardness decreases as ligament E. Bringa, M. Nastasi, and A. Caro. Characterization of size increases (Hall Petch effect) and 2) np-Au hardness a window of radiation resistance in nanoporous gold. decreases under irradiation indicating that metallic foams Presented at 2012 MRS Fall Meeting Symp. MM, Ma- soften under irradiation in very good agreement with pre- terials under Extreme Environments. (Boston, MA, 25 vious computational simulations. Nov. - 30 Dec., 2012). We presented our work in two conferences (MRS 2012, Caro, M., W. M. Mook, E. G. Fu, Y. Q. Wang, J. K. Baldwin, C. and TMS 2013) and have three papers in preparation. Sheehan, L. Zepeda-Ruiz, E. Bringa, and M. Nastasi. Ra- diation effects on mechanical properties of nanoscale gold foams. Presented at CINT Workshop . (Santa Fe, Future Work NM, 24-26 Sep., 2013). On the experimental side we will irradiate foam samples with different projectiles to explore the dependence of Farkas, D., A. Caro, E. Bringa, and D. Crowson. Mechanical damage on the recoil spectrum. This variable is the last response of nanoporous gold. 2013. Acta Materialia. we proposed to explore. On the irradiated samples we will 61: 3249. perform high resolution microscopy to determine the dam- Fu, E. G., L. Zepeda-Ruiz, M. Caro, Y. Q. Wang, E. Bringa, J. age and nano indentation to determine the hardening. Baldwin, M. Nastasi, and A. Caro. Temperature depen- On the computational side, we will perform large scale dent radiation response of nanoporous (np) Au foams. simulations of damage on foams containing stacking fault Presented at Computer Simulation of Radiation Effects tetrahedra to confirm the experimental findings and pro- in Solids COSIRES 2012. (Santa Fe, NM, 24-29 June, vide a quantitative explanation. 2012). We will write and publish the work already done on me- Fu, E., M. Caro, L. A. Zepeda-Ruiz, Y. Q. Wang, K. Baldwin, E. Bringa, M. Nastasi, and A. Caro. Surface effects on chanical properties of irradiated foams, and on the recoil the radiation response of nanoporous Au foams. 2012. spectrum dependence of damage. Applied Physics Letters. 101: 191607. Conclusion Fu, E., M. Caro, Y. Q. Wang, M. Nastasi, L. Zepeda-Ruiz, E. This work will break new ground in understanding the Bringa, K. Baldwin, and A. Caro. Nanofoams Response to Radiation Damage. Invited presentation at CAARI basic physics of material response under radiation condi- 2012, 22nd Int. Conf. on the Application of Accelera- tions where the architectural dimensions of the material tors in Research and Industry. (Fort Worth, TX, USA, are smaller than the typical size of the collision event that 5-10 August 2012). characterize the radiation damage. It will lead to improved understanding and control of defects and interfaces, spe- Fu, E., M. Caro, Y. Q. Wang, M. Nastasi, and A. Caro. Irradia- cifically in the area of where defects and interfaces control tion Response and Stability of Nanoporous Materials. phenomena such as radiation hardness of materials in Presented at 18th International Conference on Ion nuclear reactors. It will also provide insight into how inter- Beam Modification of Materials (IBMM 2012). (Qing- faces and defects play an important role in the functional- dao, China, September 2-7 2012). ity of nano-structured materials. Zepeda-Ruiz, L., E. Martinez, M. Caro, E. G. Fu, and A. Caro. Deformation mechanisms of irradiated metallic nano- foams. 2013. Applied Physics Letters. 103: 31909. 150

Page 151

Chemistry and Material Sciences Exploratory Research Continuing Project A Novel Mo-99 Separation Process Designed for Next Generation Medical Isotope Production Iain May 20120255ER Introduction ing a US Mo-99 domestic supply thus addresses two na- Mo-99 is a radioactive isotope that decays to Tc-99m, tional security needs, secure medical isotope production the most important radioisotope in nuclear medicine. and lowered proliferation concerns through elimination Two front runner technologies for the domestic produc- of HEU use. By providing a separation science solution to tion of Mo-99 are based on the fission of U-235 in mildly Mo-99 recovery from two of the front runner technolo- acidic low enriched uranium (LEU) solutions, in either gies for US based Mo-99 production LANL will be posi- nitrate or sulfate media. Both require the extraction of tioned to support the supply of Mo-99 in this country for trace amounts of Mo-99 (low mass of material but high decades to come, of direct interest to the NA21 Global specific radioactivity) from a vast excess of uranium. The Threat Reduction Initiative. current most technically advanced chemical process for this separation involves two inorganic columns designed Progress to recover a crude Mo-99 product and a combined In our original proposal the end of FY 2013 milestone precipitation and solvent extraction process to purify the was to ‘prove efficient separation of Mo-99 from ura- LEU fuel for recycle. nium in nitrate and/or sulfate media.’ This milestone has been met, we have performed the experiments that We proposed to develop a novel separation process that indicate that medical isotope Mo-99 can be successfully will allow efficient Mo-99 recovery after production, recovered from a vast excess of irradiated Low Enriched recycle of LEU fuel and the generation of a waste stream Uranium (LEU) using our novel flow sheet. This technolo- that can readily be converted to low level waste (LLW). gy can potentially be applied to the separation of Mo-99 The flowsheet we envision comprises steps that are generated as a fission product from LEU solution fuels. known individually but have not previously been inte- Using Mo-99 milked from a Tc-99m generator we proved grated into an overall process. We will evaluate special- ‘proof of concept’ that our novel separation technology ized apparatus for several of these individual operations, could be applied to the separation of Mo-99 from a mild- including utilizing automation where possible. Our ob- ly acidic uranium nitrate solution. The next step was to jective will be to prove that this “inventive application” prove that this separation could still be achieved in the of apparatus and individual chemical separations into an presence of the full suite of fission products generated integrated process flow can successfully separate Mo-99, during irradiation of a Low Enriched Uranium (LEU) solu- at the bench top scale, from LEU solutions irradiated at tion. This required the development of a new sample LANSCE (Los Alamos Neutron Science Center). Irradiated irradiation capability at Target 4 LANSCE, developed in solution tests will be used to evaluate the best chemical part for this project. Target 4 represents a new sample processing conditions for efficient Mo-99 recovery and irradiation capability at LANSCE, the contribution made LEU fuel recycle. through this proposal towards this capability develop- ment directly supported the collection of technical data Benefit to National Security Missions for this project - no augmentation. The US requires the domestic production of Mo-99 from Low Enriched Uranium (LEU), both to ensure continu- The Mo-99 generated in the irradiated uranium nitrate ous supply of this important medical isotope and to solution was successfully separated from the vast excess eliminate dependence on the international use of High of low enriched uranium. While the % recovery of Mo-99 Enriched Uranium (HEU) for Mo-99 production. Develop- was not as high as we were aiming (> 85 %), further pro- 151

Page 152

cess optimization allowed us to get very close to achieving fuel recycle. our target through a subsequent Mo-99/depleted uranium trial. Evaluate Mo-99 Recovery from LEU Sulfate Fuel We plan experiments to evaluate the possibility of our While the major technical development FY13 was the novel flow sheet being adapted to the successful recovery Target 4 sample irradiation capability, significant advances of Mo-99 from an irradiated LEU solution in dilute sulfate in analytical chemistry procedures were also made in sup- media. port of the separation chemistry. This included developing a spectroscopic uranium analysis capability that allowed Communication of Results for quick, and accurate, analysis of uranium concentration. We have communicated results obtained in this project to Also, a literature analytical procedure for acidity analysis the NA-21 Global Threat Reduction Initiative, who could was refined and adapted for our specific application. provide follow up funding to apply our research to com- We had previously submitted an Idea to LANL Tech Transfer mercial Mo-99 production programs. Research performed based on our novel separation process and, based on FY13 in this project could also find application in Spent Nuclear results we have initiated the process for submission of Fuel Reprocessing, and the automated chemical processing patent protection. Once our intellectual property has been applied here could be applicable to many chemistry proj- protected we will seek to publish our work in high impact ects. Therefore, once patent protection is in place we will chemical science/technology peer reviewed journals. seek to present our results to a wider audience through During FY13 we took the decision to focus solely on recov- conference presentations and peer reviewed publications. ery of Mo-99 from irradiated LEU nitrate solution fuels, and we did not continue pursuing research into LEU sulfate Conclusion solutions. In the summer of FY12 initial testing with our Mo-99 decays to Tc-99m, the most commonly utilized ra- newly installed semi-automated separation equipment dioisotope in nuclear medicine. 50,000 procedures are per- revealed complexities in the uranium sulfate chemistry formed every day here in the US with Mo-99, but the short that were not present with the uranium nitrate chemistry. radioactive half life of this isotope leads to rapid shortages When writing the original proposal we acknowledged that when there are interruptions to production. Currently the the sulfate chemistry could be challenging, but in FY14 we US is entirely dependent on increasingly unreliable foreign plan to return to uranium sulfate fuel to evaluate the feasi- suppliers and there is an urgent need to develop domestic bility of applying our separation technology for this fuel. production and purification processes. Our goal will be to develop a separation process for Mo-99 that will comple- Future Work ment Mo-99 production technologies based on U-235 Patent Protection fission that are being developed here in the US with low We plan to submit a patent, protecting the intellectual enriched uranium. property associated with our novel technology for medi- cal isotope Mo-99 recovery from irradiated Low Enriched Uranium (LEU). Advances in Flow Sheet Separation Chemistry We want to confirm that we can recover >85 % Mo-99 from an irradiated LEU solution in dilute nitric acid, while obtaining near quantitative recovery of the uranium for recycle. This will involve additional flowsheet chemistry refinements. Secondly, we wish to turn our attention to more careful control of other fission product contaminants of the Mo-99 product. Of most relevance is I-131, an iso- tope that partitions between numerous product and waste streams in Mo-99 recovery processes from enriched urani- um targets. We have ideas for chemical control that could be very efficient, when coupled with our novel separation technology. Finally, we will evaluate additional purification steps for the LEU product post-recovery of Mo-99. Our FY13 results indicate that it should be possible to obtain a higher purity LEU product, enhancing the attractiveness of 152

Page 153

Chemistry and Material Sciences Exploratory Research Continuing Project Ultrafast Spectro-microscopy for Nanoscale Magnetic Domain Imaging Richard L. Sandberg 20120278ER Introduction nanoscale imaging of heterogeneous complex materials Our goal is to develop an ultrafast soft X-ray coherent at LANL and worldwide is envisioned with this micro- diffractive imaging (CDI) microscope for magnetic mate- scope. rials. Our novel design will use high harmonic generation (HHG) as a tabletop coherent soft X-ray (SXR) source. Benefit to National Security Missions Complete characterization of the magnetic inhomoge- The proposed integration of “table-top” ultrafast magne- neity and its non-equilibrium dynamics holds the key to-optical X-ray spectroscopy and coherent X-ray imaging to understanding of the exotic properties of complex will provide LANL with a novel capability to investigate magnetic and other nanostructured materials. Tools the dynamics of the magnetic ordering in a broad class that image element-specific magnetization dynamics at of materials at the fundamental time and spatial scales the spatial and temporal scales of electron-spin-lattice and place LANL at the forefront of this field. This project interactions (fs-to-ps) are therefore essential for further will also enable an important characterization capabil- progress in complex material science. ity for MaRIE, LANL’s future signature facility, since the ability to unravel functionality in novel materials with While time-resolved X-ray magnetic linear and circular coherent X-rays is an important competence underpin- dichroism microscopies(tr-XMLD/tr-XMCD) have proven ning MaRIE’s M4 facility. This work will directly address to be powerful techniques to unveil magnetic properties the LDRD Grand Challenge in Materials. In the long term, of a broad class of materials, they have relied on larger our work will be directly applicable to exploration of spin synchrotron X-ray sources and had a limited temporal ordering in actinide and other mission-relevant materi- resolution (>ps), far from the fundamental timescales of als, which aid in determine material functionality. Fur- the underpinning physical processes. thermore, development of ultrafast magnetic CXDI will introduce a unique capability to the LANL-based CMIME- Despite recent progress in the development of femto- EFRC and CINT and will attract more users to this facility. second X-ray imaging techniques, ultrafast time-resolved imaging of magnetic material structure has never been Progress demonstrated to our knowledge. The magnetic x-ray CDI The goals of the first year and a half of this project fo- (mCXDI) microscope proposed here will be capable of cused on building the table-top magnetic coherent X-ray resolving ultrafast processes such as charge relaxation diffractive imaging (mCXDI) system, optimizing the X-ray and spin ordering at the nanometer (sub-50 nm) spatial magnetic linear dichroism (XMLD) signal spectroscopi- and femtosecond (<30 fs) time scales. This technique cally, and completing the first static imaging studies of will fill the gap between electron, scanning probe, and non-magnetic (initially) and then magnetic samples. At optical photon imaging technologies and will provide the time of writing this report in June – about one and an unprecedented combination of elemental specificity, a half years through this project – we are slightly behind sensitivity to ferromagnetic and anti-ferromagnetic spin in achieving all of these goals due to several challenges ordering, and the ability to probe material dynamics at we have had. However, we have made good progress the fundamental space and timescales. recently and believe we can still achieve all of the goals of the project in the next year and one half. The development of time-resolved X-ray dynamic imag- ing is still at the early demonstration phase and signifi- Last year we highlighted the successful design and cant impact for ongoing and future LANL programs in 153

Page 154

integration of the mCXDI chamber into the existing time- least two ways: 1st, these XAFS studies will tell us about resolved, spectroscopic high harmonic generation (HHG) the effect of strain-inducing elements on the local atomic system. We also mentioned initial studies that were per- structure of the materials, a critical direction for the future formed on an antiferromagnetic strontium-doped manga- of multiferroic materials like BFO, and 2nd, these initial nite film (SrMnO3). Those studies lead us to an upgrade studies open the door to pursuing coherent diffractive im- in the laser system that has greatly increased its flux and aging at the SSRL. To this effect, a user proposal has been usefulness. Furthermore, we developed a robust broad- submitted to the SSRL for coherent scattering and imaging band phase retrieval algorithm for inverting the collected studies of BFO and other multiferroic materials in collabo- diffraction patterns into high resolution images. We have ration with Alex Reid of SLAC. begun efforts to use this algorithm on actual soft x-ray (SXR) diffraction patterns. Future Work Our goals for the next FY will be to conclude and publish With the fully operational mCXDI chamber, we have static imaging demonstrations, conduct time resolved conducted many studies on test patterns to optimize the studies of magnetic materials, and conduct complemen- coherence and flux. We are proud to report that the soft tary static studies of antiferromagnetic (AFM) materials at x-ray flux available from the upgraded laser setup appears synchrotrons. First, throughout the rest of this FY and in to have approximately an order of magnitude higher flux the beginning of FY14, we plan on finishing the static imag- at the magnetically important iron M-edge (55 eV) than ing experiments already begun on three samples: static previous sources used for such efforts. test patterns, absorption contrast gold/chromium patterns, and thin film bismuth ferrite, BiFeO3 (BFO). Imaging the During this last year we have also designed and purchased antiferromagnetic domain structure of a material like BFO multilayer soft x-ray mirrors that have made our chamber will be a first on either a tabletop or synchrotron x-ray more versatile. However, the coherence of the source has source. These results will be finished and written up in the been limited due to challenges in alignment and vibration next few months. Additional imaging of static AFM materi- in the HHG source that we did not discover and correct un- als such as SrMnO3 (SMO) or other xMnO3 materials will til recently. These vibrations had prevented reconstruction also be pursued. of high quality, high resolution (sub-100 nm) SXR images of the test samples. Beside test objects, a gold-chromium The bulk of the efforts during the last FY of this project will test sample was designed and lithographically patterned be dedicated to time-resolved imaging of novel magnetic that will allow us to demonstrate absorption contrast di- materials. We will begin with laser driven demagnetization chroism imaging. This type of imaging will allow elemental studies of BFO or SMO. We stress that such studies have specific imaging of nanometer scale samples – a poten- not been demonstrated in a time resolved fashion on any tial first for a tabletop source. We are currently finishing existing X-ray source. As a complementary effort to the ta- the collection of this data and expect to have two papers bletop source, we have also submitted a user proposal for submitted by the end of this FY. Furthermore, these initial beamtime at the SLAC SSRL synchrotron. During FY14, we studies were highlighted in an invited talk at a Special Sym- have planned a series of coherent scattering and imaging posium on coherent diffractive imaging at the TMS 2013 experiments on AFM materials that will avail of the higher Conference. flux and contrast at the L-edges (~700 eV) not currently accessible on our tabletop source. These first experiments, Additionally, this year we have identified and designed a combined with tabletop results, will be combined into a transmissive magnetic sample geometry under a CINT User user proposal for time resolved studies at the LCLS X-ray Proposal with Quanxi Jia. These samples are currently be- Free Electron Laser (XFEL) at SLAC. ing manufactured and include a thin layer of multiferroic and antiferromagnetic (AFM) bismuth ferrite (BiFeO3 or Conclusion BFO). In the coming weeks, we expect to begin studies on Our proposed tabletop magnetic coherent x-ray diffractive these samples in order to determine their optimal contrast imaging (mCXDI) microscope will fill the urgent need of and magnetic domain structure. These samples will also ultrafast and nanoscale probes to address several materi- be an excellent candidate for time-resolved imaging in als challenges, especially novel magnetic materials that FY14. Additionally, in collaboration with Steve Conradson will have a large impact on wide ranging technology from (MST-8), we have begun x-ray absorption fine structure information storage to energy applications. Therefore, we (XAFS) studies at the SLAC SSRL synchrotron on BFO sam- expect this work to have significant impact in both materi- ples doped with strain inducing SmO3 nanopillars. These als science and ultrafast nano-probe development both studies will aid our efforts at imaging AFM domains in at 154

Page 155

for the LANL and worldwide community. Additionally, this work will aid in providing a stepping stone capabil- ity towards understanding materials under extremes – a category which has critical and cross cutting applications to our energy security. Publications Barber, J. L., R. L. Sandberg, C. W. Barnes, and R. L. Shef- field. Diffractive Imaging at Large Fresnel number: The Challenge of Dynamic Mesoscale Imaging with Hard X rays. Phys. Rev. B. Sandberg, R. L., Z. F. Huang, R. Xu, J. A. Rodriguez, and J. W. Miao. Studies of Materials at the Nanometer Scale Us- ing Coherent X-Ray Diffraction Imaging. 2013. JOM. 65 (9): 1208. 155

Page 156

Chemistry and Material Sciences Exploratory Research Continuing Project Unlocking Plasmons in Graphene Kirill A. Velizhanin 20120300ER Introduction technologies focused on understanding/manipulating A plasmon is the collective movement of electrons in optical plasmons for new applications. In particular, this a conducting material, i.e., a metal. Controlling the work will bring new capabilities to CINT that may be movement of plasmons can have many potential appli- made available to new users, which directly supports cations in fields such as telecomunications, sensing and our DOE-BES mission. Ultimately, the work proposed renewable energy. Graphene, which is a newly isolated here could lead to breakthrough capabilities for optical form of carbon that is only a one atom thick sheet, holds switching, signal processing and active metamaterial promise to become a useful material for such plasmonic applications that would be of interest to a variety of applications due to its high electrical conductivity and its sponsors including BES, as well as lead to new intellec- unique electronic and mechanical properties. However, tual property. systematic studies of the plasmons in graphene have not been performed to date. Thus, the goal of our project is Progress to provide a fundamental understanding of the plas- Surface plasmons on a single-layer graphene reside in mon’s in graphene using a combined theoretical and ex- the infrared range of the spectrum, depending on the perimental approach to “unlock” its potential for future exact amount of doping. The success of the experimen- applications. We will experimentally measure several tal component of this ER project relies heavily on the critical parameters (e.g., plasmon dispersion relation and availability of high-quality graphene samples with the lifetime) under a variety of conditions. These param- ability to control the amount of doping. An experimental eters will be used to theoretically predict the behavior of postdoc, Akhilesh Singh, has been hired (started March plasmons in graphene. 2013) to accomplish the graphene synthesis, fabrication of graphene-based devices and their optical character- With this knowledge, we will then attempt to control the ization. To date, Dr. Singh has been able to accomplish plasmons in graphene through the addition of semi- the chemical vapor deposition (CVD) synthesis of good conductor or metal nanoparticles onto the graphene quality graphene samples with a few 100 µm continuous surface, as well as by patterning or functionalizing the single layer flakes. This has been confirmed with optical substrate supporting the graphene film. These features microscope and Raman measurements. This contributes will provide a rudimentary means for spatial confine- to the Lab (and CINT) capabilities in synthesis of novel ment and directing of plasmons in graphene. Direct vi- materials with unique properties. sualization of graphene plasmons will be performed with advanced optical characterization tools. The detailed Defects/functionalization have to be introduced to understanding of plasmons in graphene gained from this graphene in order to couple the external optical long- work is expected to allow us to predict how to optimize wavelength excitation to short-wavelength graphene graphene-based devices for a variety of applications. plasmons. We have continued the studies of the interac- tions between graphene surface and adsorbate/ligand Benefit to National Security Missions molecules. In particular, D. Yarotski has applied scanning The work will directly support a LANL Materials Grand tunneling microscopy to determine the quality of the Challenge directed at “Control of electronic or photonic CVD-grown graphene samples. The images reveal atomi- functionality…”. Further, this work supports ongoing ef- cally flat graphene surface that exhibits periodic Moiré forts in T-Division and the Center for Integrated Nano- patterns. These patterns confirm the presence of single 156

Page 157

layer graphene and are induced by the lattice constant mis- well as thorough theoretical analysis of efficiency of plas- match between graphene and a substrate. We are current- mon excitation and propagation in graphene put us well on ly working on mapping the spatial distribution of fluorine our way to study, analyze and model plasmonic response molecules bonded to graphene as a function of doping flux of graphene-based devices. and exposure time. These results will be used to optimize graphene functionalization approaches and create co- Future Work valently bonded fluorine ‘mediators’ between graphene Both theoretical and experimental efforts during the next and other nano-to-microscale adsorbants relevant to its fiscal year will be focused on increasing the efficiency of electro-optical applications. plasmon propagation and plasmon excitation in graphene. The former task will incorporate the fabrication high- The optical characterization of graphene samples and the mobility graphene samples (using both chemical vapour interferometric time-resolved measurements of plasmons deposition and exfoliation techniques). On the theory side, propagation in graphene require new experimental tools various mechanism of plasmon energy losses will by stud- in LANL and CINT labs. A high-end infrared optical micro- ied (e.g., defect scattering, phonons), so that the sample scope and spectrometer were set up in the ultrafast optical fabriction can be guided. spectroscopy lab at CINT by A. Efimov. The spectrometer features step-scan capability so the system will be used in The latter task, i.e., efficient plasmon excitation in gra- the interferometric experiments on plasmon propagation. phene, will involve theoretical (and, subsequently, experi- Preliminary experiments indicate superior signal-to-noise mental) design of defects in graphene in order to achieve and drift resistance as compared to the previous systems the efficient coupling between laser beam and plasmons. used. Semiconductor quantum dots are easy to use in a laborato- ry, but they do not provide very efficient photon-plasmon Electrical gate-dependent measurements on the gated gra- coupling. phene samples were performed using a microscope probe station from Agilent. It was found that presently all the Our preliminary theoretical analysis shows that in-gra- samples tested posses a substantial level of p-doping po- phene voids (holes, tranches) can drastically increase the tentially indicating large levels of defects and adsorbates. efficiency of this coupling. During the next year, theory will Vacuum measurements are ongoing. focus on accurate quantitative studies and optimization of plasmon excitation by in-graphene voids. Fabrication On the theory side, we have continued investigating the of graphene samples with incorporated voids (made by properties of plasmons in graphene and their dependence means of lithography or focused ion beam etching), opti- on such graphene parameters as doping level and concen- cal characterization of such samples and observation of tration of defects. Specifically, K. Velizhanin has published plasmon propagation in such structures will constitute the a paper on the interaction of charge carrier with defects/ experimental efforts. adsorbates on graphene. An interaction of graphene plasmons with semiconductor quantum dots have been Conclusion studied by us recently. We expect to directly measure, for the first time, the criti- cal parameters needed to describe the collective move- In particular, we have demonstrated the strong enhance- ment of electrons, i.e., plasmons, in graphene. We will also ment of resonance energy transfer between quantum dots study the behavior of plasmons in graphene in the pres- in the presence of graphene as compared to the standard ence of controlled defects, such as nanoparticles, as well Forster resonance energy transfer. This finding might as under applied external electrical fields. We will use this become of use in applications of hybrid graphene-based knowledge to optimize predictive tools needed to theoreti- materials in photovoltaics. The theory paper on strong cally describe plasmons in graphene. We expect that this interaction of excitons in semiconductor nanostructures predictive capability will allow us to design new graphene- (e.g., quantum wells) and plasmons in graphene is cur- based devices for practical applications that may lead to rently in preparation. new break throughs in fields such as telecommuication, photovoltaics and sensing. Overall we feel that the progress within experimental and theoretical components of the project is very good. Publications The recent additions of such capabilities to the project as Solenov, D., and K. Velizhanin. Adsorbate Transport on Gra- successful CVD graphene growth, optical microscope/spec- phene by Electromigration. 2012. Phys. Rev. Lett.. 109 trometer and scanning tunneling microscope studies, as 157

Page 158

(9): 095504. Velizhanin, K. A., and T. V. Shahbazyan. Long-range plas- mon-assisted energy transfer over doped graphene. 2012. Phys. Rev. B. 86: 245432. Velizhanin, K., and A. Efimov. Probing plasmons in gra- phene by resonance energy transfer. 2011. Phys. Rev. B. 84 (8): 085401. 158

Page 159

Chemistry and Material Sciences Exploratory Research Continuing Project Plasmon-exciton Interactions in Single-wall Carbon Nanotube – Metal Nanostructure Complexes Han Htoon 20120330ER Introduction This project, therefore, directly addresses a LANL 2011 Single-wall carbon nanotubes (SWNTs), a graphene sheet Material Grand Challenge, “Understanding and Control rolled up to form a tube of ~1 nm in diameter and tens of Emergent Functionality Using Extrinsic Techniques.” of microns in length, represent a nanoscale laboratory Furthermore the SWNT-metal nanostructure complexes for exploring emerging quantum phenomena in an ideal could also have potential to bring transformational one dimensional (1D) system. Intense research efforts breakthrough in light emitting as well as light harvesting focused on the optical properties of SWNTs in the past applications. This project therefore addresses “Energy decade have revealed novel phenomena such as the and Earth System” grand challenge, “Concepts and Ma- Aharanov-Bohm effect, violation of the Condon approxi- terials for Clean Energy. mation, and coherent exciton energy transfer. Here, aim- ing toward establishing a new branch in this very active Progress research field, we propose to investigate the interaction Fabrication of SWNT-metal nanostructure complexes of two different fundamental optical excitations, namely, The team designed several metal-dielectric- metal 1D excitons of SWNTs (electron-hole pairs) and plasmons (MDM) plasmonic cavity arrays with resonances aligned (collective excitations of charge density waves supported to the emission of SWNTs, and developed an e-beam by metallic nanostructures). Although plasmon-exciton lithography process for fabrication of these structures interactions has been investigated intensely in nano- on transparent quartz substrates. We also successfully crystals, quantum dots, molecules and quantum wells, incorporated nanocrystal quantum dots into MDM cavi- so far very few studies have been conducted on SWNTs. ties. Collaborations are ongoing with C-PCS to adapt this This study possesses tremendous potential for revealing scheme for incorporation of polymer wrapped SWNT never before seen phenomena that lie at the interface into MDM cavities. We expect to achieve SWNT-MDM of classical electrodynamics and quantum mechanics. coupled structures in a few months. The non-linear Fano effect (constructive and destruc- tive interference of a narrow discrete resonance with a Acquisition of superconducting nanowire single photon broad spectral line) and exciton induced transparency detection system (SNSPD) (an analog of electromagnetically induced transparency) With the financial support of CINT, we are acquiring represent two examples of the theoretically predicted the SNSPD system capable of performing time tagged phenomena we expect to encounter in this study. In ad- time correlated photon counting experiments on large dition, as this investigation is reaching into unexplored diameter SWNT emitting at wavelengths beyond 1 µm areas, we also expect the emergence of completely (cost: ~180K). This instrument will allow us to investi- unpredicted phenomena. gate plasmon exciton interactions at 1.3 and 1.5 micron telecommunication wavelength regimes. Benefit to National Security Missions The novel physical phenomena we aim to explore in this Studies of plasmon induced localization project will pave the way to tailor not only the spon- A publication is underway to report on the observa- taneous optical excitation and emission processes but tion of plasmon induced location of excitons in SWNTs also the coherent dynamical behaviors of SWNT-metal coupled to the atomically smooth gold pyramids that is nanostructure complexes for applications ranging from reported in the last year progress reports. bio/chemo sensing to quantum information processing. 159

Page 160

The team is also finishing up the preparation of two theo- applications ranging from solid state lighting to quantum retical papers explaining this plasmon induced exciton communication as they provide handles to manipulate not localization effect. only spontaneous emission rates but also quantum coher- ent dynamics and photon emission statistics. Investigation of Plasmonic effects on exciton diffusion, recombination dynamics and photon emission statistics So far very few studies on plasmon-exciton interaction The team is conducted PL imaging, time resolved PL and have been conducted on single walled carbon nanotubes photon-correlation spectroscopy studies on individual (SWNTs). Because of the SWNT’s unique one dimensional stand-alone SWNT as well as SWNT-metal nanostructures electronic structure, the manifestation of this interac- complexes to understand the competing role of exciton tion is bound to be different at the fundamental level diffusion and Auger Recombination in defining photon and therefore holds great potential for the emergence of emission statistics of SWNTs. These studies revealed that unexpected phenomena. To this end here we propose to (1) the efficiency of Auger recombination decreases with investigate plasmon-exciton interactions in high quality, the increase of exciton diffusion length, (2) a strong photon well-defined SWNTs. In this study, we will first develop two anti-bunching can only be achieved in short SWNTs in novel approaches- plasmon induced trapping and chemical which the exciton diffusion is limited by the non-radiative assembly- to controllably couple SWNTs to variety of me- recombination at the end of the SWNTs and (3) coupling tallic nanostructures. We will then apply a suite of advance to a metallic nanostructure can lead to increase of multi- single nanostructure optical spectroscopy approaches to exciton emission. A model was successfully developed to explore the effect of plasmon on diffusion localization and describe the interplay of exciton-exciton Auger annihila- recombination dynamic of 1D excitons. In addition we tion and one dimensional exciton diffusion to explain the will also exploit the strong transition dipole moments of first two findings. These results were presented at the APS SWNT’s 1D exciton to explore phenomena of strong inter- March meeting in 2013 and WONTON. action regimes such as nonlinear Fano effect and exciton induced transparency. We will also apply hybrid quantum- A new single nanotube imaging microscope was assembled mechanical-semiclassical approaches to model the exciton- in the last year. The capability of the microscope was plasmon dynamic in 1 D systems. expanded to do simultaneous “two-color” imaging on InGaAs (For NIR emission ) and EM-CCD (for visible emis- Conclusion sion) cameras with wide area or confocal excitation of Through this project, we expect to attain the capability to nanotubes and to obtain single-tube and even single-site manipulate various aspects of light-SWNT interactions that spectra along the length of an individual SWNT. Significant hold the key to the realization of a variety of technological progresses were also made towards characterizing the applications. Specifically we expect to attain the capability emission characteristics of oxygen doped SWNTs. These to tailor both spontaneous optical excitation and emis- will serve as a baseline for probing interaction of dopant sion processes. This capability will bring transformational optical behaviors with plasmonic structures. In particular, breakthroughs in the utilization of SWNTs in photonic, it will be interesting to compare spectral response of exci- chemo/bio sensing and light-harvesting applications. ton localized at dopant sites with respect to that of exciton Furthermore, this study is also expected to shed new light localized at plasmonic hotspots. on the manipulation of coherent dynamics and photon emission statistics of SWNTs and lay the foundation for Future Work their applications in quantum information processing and Interactions of two different fundamental optical excita- communication applications. tions often lead to the emergence of new phenomena that cannot be observed in neither of the original excita- Publications tions. The interaction between the optical excitation of a Cherqui, C.. Plasmon assisted polaron effects in carbon quantum emitter (exciton) and a propagating excitation nanotubes. Student Award Presentation . Invited pre- of charge density waves on the surface of a conductor sentation at TSRC Workshop on Nanomaterials: Theory (plasmon), represents one such instance. Phenomena and Computation. (Telluride, July 16-20, 2012). that lie at the interface of classical electrodynamics and Cherqui, C., D. H. Dunlap, and A. Piryatinski. Diffusion of quantum mechanics such as enhancement of emission Plasmon-Exciton Polaron. Presented at APS March rates and exciton Rabi splitting have emerged in the study Meeting. (Boston, Feg 27- March 2, 2012). of this interaction in variety of nanostructures. In addi- tion to being important for fundamental reason, these Crochet, J. J.. Exciton transport and manipulation in col- phenomena exhibit tremendous potential for technological loidal semiconducting carbon nanotubes. Invited pre- 160

Page 161

sentation at 5th Workshop on Nanotube Optics and Nanospectroscopy. (Santa Fe, 16-20 June, 2013). Doorn, S. K.. Surface Chemistry of Carbon Nanotubes: Photoluminescence Probes at the Ensemble, Single- Tube, and Single-Site Levels. Invited presentation at Telluride Science Research Center Workshop on The Chemistry and Physics of Defects in Carbon Nanotubes. (Telluride, CO, 8-12 July, 2013). Doorn, S. K.. Exciton transport and manipulation in col- loidal semiconducting carbon nanotubes. Invited pre- sentation at 223rd Electrochemical Society Meeting. (Toronto, Canada, 12-16 May, 2013). Doorn, S. K.. Surface chemistry of carbon nanotubes: En- abling novel composite materials and studies of dop- ant chemistry at the 1-D interface. Invited presentation at Lehigh University, Physics Colloquium. (Lehigh Uni- versity, 3 Oct, 2013). Ma, X., J. G. Duque, J. J. Crochet, S. K. Doorn, and H. Htoon. Photon Statistics of Single Carbon Nanotubes at Room Temperature. 2013. In American Physical Society, March Meeting, . (Blatimore, MD, 18-22 March, 2013). Vol. 58, p. 1. Baltimore, MD: American Physical Society. Ma, X., J. G. Duque, J. J. Crochet, S. K. Doorn, and H. Htoon. Influences of exciton diffusion and exciton-exciton annihilation on photon emission statistics of carbon nanotubes. Presented at 5th Workshop on Nanotube Optics and Nanospectroscopy. (Santa Fe, June 16-20). Piryantinski, A.. Localized Surface-Plasmon Effect on Ex- citon Transport and Radiative Properties in 1D Nano- structures. Invited presentation at TSRC Workshop on Nonequilibrium Phenomena, Nonadiabatic Dynamics and Spectroscopy. (Telluride, CO, 21-25 July, 2013). Piryantinski, A.. Surface-Plasmon Assisted Charge Carrier and Exciton Transport in 1D Nanostructures. Invited presentation at Seminar, Department of Physics, Uni- versity of Wyoming. (Laramine, 26 April, 2013). Piryantinski, A., C. Chirque, and D. Dunlap. Surface-Plas- mon Assisted Charge Carrier and Exciton Transport in 1D Nanostructures. Presented at 5th Workshop on Nanotube Optics and Nanospectroscopy. (Santa Fe, 16- 20 June, 2013). Piryatinsko, A.. Surface-Plasmon Assisted Charge Carrier and Exciton Transport in 1D Nanostructures. Invited presentation at TSRC Workshop on Nanomaterials: Theory and Computation . (Telluride, 16-20 July, 2012). 161

Page 162

Chemistry and Material Sciences Exploratory Research Continuing Project Novel Inverted Nanoshells for Multimodal Diagnostic Imaging and Cancer Therapy Jennifer A. Hollingsworth 20120365ER Introduction (ensemble to single-particle, static to real-time), sig- The overarching goal is to design and implement a new nificantly improve our understanding of interactions at nanoparticle-based platform for combined multimodal the nano-bio interface, and provide proof-of-concept imaging and cancer therapy. The approach is to encapsu- for a new approach to realizing simultaneous imaging/ late within a metal shell a semiconductor and/or lantha- therapy. nide (Ln)-comprising core that affords both fluorescence and magnetic resonance (MR) contrast capabilities. The Combined, this agenda advances basic understanding metal shell also serves dual duty, simultaneously en- of materials and basic health research at a nationally hancing fluorescence originating in the core and generat- competitive level. It capitalizes on, advances, and unifies ing heat upon optical activation for photothermal cancer three LANL technologies: patented giant quantum dots, therapy. In contrast with other nanoparticle-based patented 3D particle tracking, and human tissue models. approaches that aim to couple diagnostics with therapy, The work will be of strong interest to DOE Office of Sci- this approach affords greater simplicity, smaller/tunable ence with respect to new optical materials design/devel- particle sizes, enhanced optical performance, and bio- opment, with applications beyond biosciences to effi- compatibility. Our design strategy is guided by modeling cient lighting, light-based communications technologies, studies, which revealed the transformative potential of low-threshold lasing, single-photon sources for quantum “inverted” nanoshell (INS) structures for enhancing a cryptography, etc., as they incorporate additional func- fluorophore’s emission. The inverted geometry places tionality, while solving the remaining outstanding flaws the emitter at the center of a metal shell, where the of more conventional emitters. It will also be directly optical intensity enhancement is strong and uniform. In relevant to the needs of DHHS agencies, even beyond contrast, the field enhancement outside a metal shell the direct goal of advancing nanoparticles for combined structure, though very large at the surface of the metal, imaging/therapy, where predicting and controlling drops off quickly and non-uniformly with distance from nanoparticle mediated activity at the molecular, cellular the metal. Here, we take advantage of the internal field and tissue levels has implications for a range of future enhancements by placing our fluorophores on the inside applications from nano-enabled bio/chem-sensors to of a metal nanoshell, advancing the state-of-the-art in non-invasive drug delivery and “battle suit fluorophore probe design, while at the same time com- medicine.” Also along these lines, the new dual-purpose bining this concept with an additional imaging function- nanomaterials developed here can find implementation ality (MR) and a therapy functionality (photothermal). beyond cancer, e.g., in the targeted treatment of non- We will demonstrate proof-of-concept deep-tissue imag- cancerous diseased tissue. ing and selective hyperthermal cancer-cell ablation using peptide-INS conjugates. Progress INS synthesis Benefit to National Security Missions While silica (SiO2) shell growth onto giant quantum In addressing our overarching goal to develop and test dot (gQD) emitters has proven successful for a range novel inverted nanoshell nanoparticles for coupled of SiO2 thicknesses (10 nm initially and more recently multimodal imaging and photothermal cancer therapy, 13, 15, and 20 nm), successful addition of the targeted we will concurrently build new capabilities in optical gold (Au) shell has been more elusive. We define suc- nanomaterials and advanced imaging in complex media cess as the ability to apply a uniform/thin Au shell to all 162

Page 163

nanoparticles without emitter quenching. We continue to N-hydroxysuccinimide ester intermediate) to Neutravidin. work out the required chemistry for controlled Au shell Gels confirmed gQD-Neutravidin coupling (1% agarose gel, growth onto these relatively large gQD/SiO2 nanocrystals, 120 V, 60 min). Following conjugation to the cell-targeting but of primary concern is the tendency of the Au reduction agent, cell-binding specificity was determined – targeted- step to afford nearly complete suppression of emission to-untargeted ratio of 4-to-1. With this level of specificity, even without complete Au shell coverage on the nanocrys- we were able to conduct extensive 3D particle-tracking/ tals. Extensive characterization by transmission electron cell-association experiments, obtaining many 10’s of microscopy (TEM) at different stages of Au reduction and successful tracks affording an average tracking duration growth has allowed us to confirm that the source of gQD 20-fold higher than commercial QDs, dramatically improv- quenching is attachment of Au ions at the gQD surface ing our chances for successfully observing QD (or INS)-cell and nucleation of Au nanoparticles directly onto the gQD. interactions. To further improve cell targeting, we are The ability of the Au ions to penetrate the SiO2 shell likely improving our conjugation chemistry for enhanced cell results from the porosity of this amorphous layer, as indi- specificity (to equal commercial QDs’ ~10-to-1), replac- cated by recent literature suggesting that the SiO2 is po- ing Neutravidin with Streptavidin (promising: significantly rous to ions/small molecules. We have attempted several improved gel-filtration chromatograms). approaches to limit/prevent the transport of Au ions to the gQD surface: (1) “seal” the SiO2 against ion penetration by We have also made progress in developing suitable cell- (a) addition/optimization of multiple layers of polyelectro- matrix composites that afford sufficiently low viscosity to lyte – recent improvements affording highly charged/stable observe QD mobility combined with retention of good cell layers (zeta potential ~±40) or (b) growth of a crystalline growth. Specifically, we have evaluated the penetration oxide (ZnGa2O4) onto the gQD surface (completed–await- efficiency of our gQDs (INS surrogates) through decreasing ing characterization) or onto the SiO2 surface, and (2) concentrations of matrigel: 9.5 - 3 ng/mL, finding lower increase the SiO2 thickness (>10 nm). The latter approach concentrations (3 – 5 ng/mL) to be more favorable for gQD combined with highly ionic polyelectrolyte coverage has penetration. 3T3 cells were grown in the optimal matrigel recently afforded Au-coated gQDs that retain much of their suspension, and gQDs evaluated for penetration-efficiency emission despite exhibiting almost exclusively Au-absorp- and non-specific binding/phagocytosis by cells. We are tion features. We are awaiting TEM analysis to confirm the currently assessing non-specific gQD-cell binding in this structure affording this very promising result. environment and will proceed with specific-binding experi- ments. gQDs were evaluated for cytotoxicity and found to Spectroscopic evaluations be non-cytotoxic in the relevant concentration range – im- We have quantified brightness per particle (BPP) using portant for distinguishing future “intentional” cell toxicity. fluorescence correlation spectroscopy (FCS) and developed BPP curves as a function of laser power for four different Future Work laser sources. The lasers thus far investigated include 405 Guided by theory/optical-spectroscopy characterizations, and 470 nm pulsed, and 488 and 532 nm continuous wave we will synthesize and study enhancement and/or damp- (CW). The tracking studies discussed below were primar- ing effects as a function of core size, spacer-layer thickness, ily obtained using a 470 nm pulsed laser, with laser power and metal-shell thickness toward optimized structures. As optimized to achieve maximum BPP. It was subsequently the emitting “core” material, we will utilize two types of determined that by using a CW source we could avoid QD nanocrystals: our unique “giant” (g-QDs) (focus to date) saturation, allowing continuously increasing QD BPPs with and either lanthanide or manganese-doped QDs. We now increasing laser power, even for the 532 nm CW source have two g-QDs to explore: CdSe/CdS g-QDs that can be despite the minimal gQD absorption cross-section at tuned to emit in the far-red and InP/CdS g-QDs that emit this wavelength. Using CW sources, we expect to further in the near-infrared. Both are appropriate wavelengths for improve the already dramatically increased tracking times deep-tissue imaging. Time permitting, we will incorporate (below), and, significantly, use of “redder” 532-nm source a paramagnetic material to add MR-imaging functionality. will reduce background emission/unintended cell damage. To assess the significance of any ensemble-level absorp- Cell targeting/tissue mimics tion/emission enhancements observed in Task 1 for Focusing on peptide-conjugation chemistry, the gQDs were deep-tissue fluorescence imaging, we aim to obtain 1 functionalized with bis(DHLA)-PEG-COOH ligands (provided and 2-photon action cross sections for individual particles by NRL colleagues). Carboxyl termination was used to using fluorescence correlation spectroscopy (FCS) and facilitate 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide time-correlated single-photon counting (TCSPC) measure- (EDC)-mediated coupling (semi-stable amine-reactive ments, where “action” cross sections effectively include 163

Page 164

both quantum yield and absorption characteristics of the National Laboratory Laboratory Directed Research and material, which are both critical for effective detection Development Day. (Santa Fe, October 23, 2012). through distance (i.e., a biological tissue). In all cases, we Keller, A. M., Y. Ghosh, M. E. Phipps, D. S. Lidke, B. S. Wil- will compare the action-cross-sections for both the emit- son, J. A. Hollingsworth, and J. H. Werner. 3-dimen- ting core and the INS form. The most promising candidates sional tracking of blinking-suppressed quantum dots will be incorporated into “tissue mimics” to investigate in live cells. 2013. In Biophysical Society 57th Annual imaging in a more realistic environment (tissue complexity Meeting. (Philadelphia , February 2-6, 2013). , p. 1. will be tuned) and as a function of depth/imaging mode Conference Abstracts: Biophysical Society. (1- or 2-photon). Keller, A. M., Y. Ghosh, M. S. DeVore, M. E. Phipps, M. H. Stewart, D. S. Lidke, B. S. Wilson, J. A. Hollingsworth, Toward our goal to establish proof-of-concept cancer-cell and J. H. Werner. 3-Dimensional Tracking of Non-blink- targeting/photothermal therapy, we have worked out the ing ‘Giant’ Quantum Dots in Live Cells. Nature Nano- required peptide conjugation chemistry that will enable technology. the INS to selectively target and eliminate cancer cells in a matrix of normal cells. We now aim to establish this Park, Y. S., Y. Ghosh, P. Xu, N. H. Mack, H. L. Wang, J. A. capability for mixtures of cancerous and normal cells em- Hollingsworth, and H. Htoon. Single-nanocrystal pho- bedded in a collagen-matrix (co-culture tissue-like matrix), toluminescence spectroscopy studies of plasmon-mul- then transitioning to re-constructed cancerous tissue. tiexciton interactions at low temperature. 2013. The Journal of Physical Chemistry Letters. 4: 1465. Conclusion We anticipate three advances impacting both basic and applied biosciences. First, we will establish materials-de- sign principles for advanced multimodal-imaging/therapy nanoparticles. The approach is flexible and can be applied to different imaging agents, e.g., as new, “greener” NQDs are developed, these can be incorporated into the design motif, or different MR contrast agents could be use. We will establish a dual-imaging capability for deep-tissue studies, including a novel capability for real-time tracking of nanoparticle-based cell targeting. Lastly, we will estab- lish the materials and photoactivation requirements for se- lective cancer cell hyperthermal ablation that is applicable to multiple cancer types and other diseases. Publications Hollingsworth, J. A.. Heterostructured Quantum Dots: From Molecular Probes Toward Cancer Therapy. Invit- ed presentation at Nanoparticle Synthesis and Applica- tions Cancer Imaging and Treatment Symposium . (Uni- versity of New Mexico, Albuquerque, 16 Aug. 2013). Hollingsworth, J. A., J. H. Werner, H. Htoon, A. Piryatin- ski, R. Schaller, Y. Ghosh, A. M. Dennis, A. M. Keller, B. Mangum, and D. C. Hannah. Non-blinking “giant” quantum dots: Ideal probes for real-time three-di- mensional tracking. 2013. In CLEO:2013. (San Jose, CA, 9-14 June 2013). Vol. CLEO:QELS Fundamental Sci, p. Biophotonics (QTu1P). OSA Optics InfoBase: The Opti- cal Society. Hollingsworth, J. A., J. Werner, R. Iyer, N. Karan, A. Keller, A. Nagy, Y. Ghosh, and J. L. Casson. Novel inverted nanoshells for multimodal diagnostic imaging and cancer therapy. Invited presentation at Los Alamos 164

Page 165

Chemistry and Material Sciences Exploratory Research Continuing Project Forensic Archaeology of a Manhattan Project Era Nuclear Site Warren J. Oldham 20120459ER Introduction dation to position the laboratory as a credible national The goal of this research is to develop accurate and resource to support these expanded program require- precise methods for characterizing nuclear sites, includ- ments. ing those that have undergone demolition and decon- tamination efforts intended to mask nuclear prolifera- Progress tion activity. This will be achieved by reconstructing the The ubiquitous and persistent background resulting from weapons production activities that occurred during the fallout during the era of atmospheric nuclear testing historic Manhattan Project at the plutonium processing complicates analytical detection of low-level emissions facility previously located near Ashley Pond in downtown resulting from modern nuclear activities. This project Los Alamos. This represents a significant experimental aims to develop a novel sampling and analysis strategy challenge since environmental samples collected near to distinguish local signatures from global fallout based this location inevitably contain mixtures of plutonium on improved understanding of actinide geochemistry that originate from a variety sources and were deposited and environmental behavior. Our work has shown that at various times between 1944 and 2011. We propose rusty metal adsorbs trace actinides in the environ- here to make use of this unique site to develop a system- ment through natural weathering processes to facilitate atic approach for reconstructing nuclear process activi- ultra-low plutonium detection using a simple and robust ties, even when complicated isotopic compositions exist. analytical method. We hypothesized that if industrial Specific scientific goals include: Developing methods for effluents are more easily mobilized than global fallout, identifying trace levels of radionuclides that are associ- then analysis of the surface oxide coating of discarded ated single source terms in areas where large regional metal objects could reveal a local source-term of inter- contributions of actinides and fission products are pres- est. ent. Demonstrating for the first time that relationships exist between actinides and stable isotopes that contain During FY13 this idea was directly tested using a “case unique insight associated with process history and intent study” approach focused on the former site of original of the nuclear material. Conducting carefully designed Manhattan Project Era laboratories near Ashley Pond. experiments that advance current methods for age dat- This site included the world’s first plutonium processing ing and accounting for global fallout contributions to and fabrication facility that operated in support of the radioanalytical measurements. war effort from 1943 to 1945. Although this area was ultimately remediated for unrestricted public access, we Benefit to National Security Missions suspected that the ultra-sensitive nature of the iron-ox- Analytical methods that enhance the detection and ide adsorption process might be used to detect faint plu- isotopic characterization of trace-level plutonium in the tonium signatures from the Manhattan Project. We’ve environment support LANL’s core mission in nuclear now shown that rusty bottle caps (contemporary origin) nonproliferation and treaty verification. In the future the collected within the boundaries of the former TA-1 site, laboratory may be tasked with monitoring compliance reveal ultra-low 240Pu/239Pu atom ratios representing with additional international agreements including the Pu from the earliest years of the U.S. nuclear weapons Fissile Materials Cut-off Treaty and the Comprehensive program. Although local sample concentrations were not Test Ban Treaty. The analytical and evaluation techniques appreciably elevated compared to regional background pioneered in this research will provide a technical foun- examples, isotopic measurements (240Pu/239Pu) for TA-1 165

Page 166

samples varied from 0.0036 to 0.050 compared to 0.168 excellent purity using high performance liquid chromatog- ± 0.012 for regional collections. The results of this study raphy. The 151Sm and 155Eu analytes were quantified using suggest that the novel surface oxide sampling strategy can liquid scintillation counting and γ-spectroscopy, respectful- be used to selectively identify local nuclear signatures even ly. Overall, the procedure represents a robust and accurate when only trace-level contamination is present. The work method for evaluating abundances of 151Sm and 155Eu in was presented at the most recent LDRD Day at Buffalo environmental samples. Thunder Casino and formed the basis of a PADSTE Science Highlight (2-20-2013). The work was also a featured high- Future Work light on the LDRD Homepage. This project aims to develop novel analytical methods to improve detection and interpretation of signatures associ- The results of the “rusty bottle cap” study have changed ated with nuclear activities. The effort is proceeding along our thinking about what is central to faint signature detec- three main thrust areas: (1) Trace-level measurements of tion. We believe that relative mobility is the key concept actinides in the environment associated with discarded for environmental monitoring efforts. Our work suggests rusty iron relics (e.g. nails, bottle caps, etc.). During FY14 that the chemical form of industrial contaminants com- measurements of Am-241 and Np-237 compared to Pu-239 pared to nuclear fallout will ultimately control relative and Pu-240 in collections of this type will be completed to environmental mobility. To rigorously test this idea, we aim determine relative environmental mobility of the different to conduct careful laboratory based experiments com- actinide elements. The presence of these anthropogenic paring aqueous mobility of plutonium contained within radionuclides within the surface oxide coating of rusty these two distinct sample types. While well-characterized artifacts implies aqueous transport from the local environ- standard reference materials containing industrial contami- ment. The relative concentrations and isotopics measured nation are available from NIST (e.g. SRM 4353 Rocky Flats for rusty metal samples will be compared with adjacent Soil), analogous nuclear fallout standards are not readily soils to characterize relative mobility, (2) Radioanalytical available. To solve this problem, we have identified SRM measurement of Sm-151, combined with Pu isotopics, 1648a (Urban Particulate) as a new, previously unrecog- is being developed to correct for global fallout in mixed nized source of atmospheric fallout from nuclear weapons samples that also contain a local signature of interest. testing. The material was originally collected by NIST using During FY14, we will validate the radioanalytical methods an industrial scale baghouse during 1976 and 1977 in St. developed in FY12/13 to measure Sm-151. A significant Louis, MO and inadvertently contains significant concen- challenge will be measurement of trace Sm-151 derived tration of nuclear fallout from Chinese and French nuclear from nuclear fallout contained within SRM 1648A. (3) A tests. Plutonium and americium concentrations as well as new capability to simultaneously measure Np-237 and isotopic compositions were determined during FY13. These Pu is being developed. During FY14 we will measure the measurements will form the basis for comparisons with relative ratio of Np-237 and Pu-239 contained in samples natural matrix standards containing plutonium contamina- collected from TA-18, which formerly contained various tion derived from industrial releases. The analytical study critical assemblies used in weapons development work. of SRM 1648a was presented at the American Chemical During the time that these machines were operated, U-238 Society meeting in New Orleans, LA within a dedicated ses- underwent neutron capture reactions to simultaneously sion on Nuclear Forensics. form U-237 and U-239. Today these short-lived isotopes will be detectable as Np-237 and Pu-239, respectively. The Developing methods for quantitatively evaluating long- relative ratio of these analytes could provide forensic in- lived lanthanide fission products, i.e. 151Sm and 155Eu, from formation on the nature of the original activities that were environmental samples offers opportunity to advance conducted at TA-18. theories regarding the fate-and-transport of electroposi- tive and radioactive metals in environmental systems. Conclusion While this method can be applied in analysis of nuclear The research will (1) develop methods for distinguish- waste and assessment of environmental contamination ing trace levels of radionuclides that are associated with sites, its application toward nuclear forensics is of par- nuclear operations from regional and global backgrounds, ticular importance given that 151Sm and 155Eu (t½ = 90 and (2) identify signatures and observables between radio- 4.75 y, respectively) can be used to assess fission events nuclide and stable element emissions, and (3) validate that happened long ago. Over the last two years we have methods for identifying proliferation activity in the face of developed a chemical purification process that provided active deception. highly pure lanthanide concentrates from which individual fractions of 151Sm and 155Eu were isolated in high yield and 166

Page 167

Publications Oldham, W. J., B. S. Matteson, J. L. Miller, C. T. Lake, and M. Attrep. Rusty nails adsorb trace plutonium from the environment. 2013. JOURNAL OF RADIOANALYTICAL AND NUCLEAR CHEMISTRY. 296 (2): 889. 167

Page 168

Chemistry and Material Sciences Exploratory Research Continuing Project Exploiting Metal/Organic Interfaces as Potential Bulk Heterojuntions: Unlocking the Efficiency of Organic Photovoltaics Hsing-Lin Wang 20120464ER Introduction Combining the above approach allows solving critical is- In this project, we adopt an unorthodox approach to sues that hampering efficiencies of organic photovoltaic develop organic photovoltaic devices with improved ef- devices. ficiency and reproducibility. We will focus our efforts on the following: Benefit to National Security Missions This research allows better understanding of fundamen- • Developing a unique donor-acceptor interface, bulk tal mechanisms for self-assembly and a viable approach heterojunction (BHJ), using nanoporous metals to to fabricating nano-assemblies with complex structures improve charge separation and collection efficiency. and functionality. The proposed study addresses funda- We will synthesized nanoporous metal with control mental challenges at the interface between chemistry, pore size ranges from 10-50 nm, matching the exci- materials, and nanotechnology. Success will have wide- ton diffusion length in the polymer systems, ~ 5-15 spread impact on the fabrication of photovoltaic devices. nm, thus ensuring minimal exciton diffusion losses. The proposed work will strengthen our capability in ad- dressing laboratory missions, particularly in the areas of • Surface functionalization of nanoporous metals by materials development and energy security. Understand- covalently attaching electron acceptors onto the ing the underpinning mechanism that dominate charge metal surface, allowing instaneous electron transfer separation and charge collection efficiencies will form and collection on the metal electrode. This novel the foundation for developing next generation solar cells approach to the best of our knowledge has never with unprecedented high efficiency and support new attempted by any other research group. program development in Office of Science and DARPA. • Enhancing light absorption by increasing light trap- ping within the active polymer, as well as synthesiz- Progress ing novel low-band gap polymers with absorption Our goals in this project are (a) enhance light absorp- that covers the solar emission from UV to near IR. tion due to increased light scattering pathways from the We plan to develop synthesis of conjugated poly- porous metal into the active later and by synthesizing mers with low band gap molecules in the backbone novel low-band gap polymers (b) enhancing light absorp- structures, and pentathiophene as side chains. Our tion by synthesizing novel low-band gap polymers and by design entails extension of polymer absorption to increasing the thickness of organic film and (c) enhanc- encompass the spectra of UV, visible, and near in- ing hole-transport characteristics by synthesizing crystal- frared lights, allowing overlaps portions of the solar line polymers. We have made significant progress in all emission that were previously inaccessible. fronts of the project. • Enhancing hole-transport by design synthesis of con- jugated polymer thin films with improved crystallin- Polymer synthesis ity. We will extend our synthesis to a variety of semi- For the past three 12 months, we have demonstrated rigid side chain structures that are likely to couple synthesis and characterization of a series of two di- with each other and form highly ordered crystalline mensional polythiophene-based conjugated polymers; thin films. Higher crystallinity leads to higher carrier alternative copolymers consist of benzo[1,2-b:3,4-b] mobility, minimize the possibility of exciton recom- bithiophene-bithiophene (donor-donor) and benzo[1,2- bination. b:3,4-b]bithiophene-benzothiadiazle (donor-acceptor) as 168

Page 169

building blocks for the conjugated polymer main chain. We an optimized film morphology. The D/A weight ratio was observe a 0.4 eV difference for the band gap and LUMO tuned from 1:0.5 up to 1:3.0 to find the optimized compo- level between the donor-acceptor and donor-donor type sition for efficient charge separation and transportation. polymers; however, the HOMO level for the two polymers As a result, the best performing device (D:A=1:1.5) showed remained unchanged (~ -5.4 eV), suggesting that the an average power conversion efficiency (PCE) around HOMO level was essentially dominated by the benzo[1,2- 3.30% and peak PCE~3.68% with open circuit voltage b:3,4-b]dithiophene (BDT) moieties and the LUMO level (Voc)0.804V short circuit current density (Jsc)7.24 mA/ and band gap are determined by the nature of the co- cm2, fill factor62.0%. Another polymer system processed monomer. Overall, these 2-D conjugated polymers integrat- under the same condition reveals a superior PCE4.25%, ing low band gap molecules in the backbone structures, which is due to the high degree of crystallization as mani- and oligothiophene as side chains. Such design entails fested by the UV-Vis absorption spectrum of polymer extension of polymer absorption to encompass the spectra solutions. Of particular importance is the enhanced absor- of UV, visible, and near infrared lights, allowing overlaps bance peak at ~600 nm which implies the increase in chain portions of the solar emission that were previously inac- packing and crystallinity. The temperature-dependent cessible. photoluminescence spectra show a gradual decrease of the emission peak at 620nm, further validate the crystal- Fabrication of porous electrodes line nature of the polymer. We have focused our effort to synthesize nanoporous metal based photovoltaics. Future Work • We will perform synthesis and characterization of poly- We have synthesized various nanoporous metals includ- thiophene derivatives with optimum absorption that ing Au, Ag, Ni, and Ti. Nanoporous gold have high work covers wavelength from 300-800 nm. function of 5 eV and small pore size (5-20 nm), therefore • Preparation of different Nanoporous Metals including resulting in low efficiency. We have recently synthesized Titanium, Aluminum and Silver that have lower work nanoporous Titanium by dealloying a Ti/Al alloy. Nanopo- function with appropriate pore size. rous Ti have bigger pore size (60-100 nm) and a low work function of 4.1 eV. Organic polymers easily infiltrate into • Infiltrate them with functionalized C-60 using vapor the pores and the nanoporous Ti acts as a superior elec- deposition and solution-based techniques. tron acceptor. Initial photovoltaics were prepared using a • Demonstrate a functional OPV standard donor/acceptor pair e.g. P3HT and PCBM. Initial results reveal that Nanoporous Ti based solar cells have Conclusion better performance than planar Ti. However these cells We expect to demonstrate fabrication of bulk heterojunc- exhibit low shunt resistance, which leads to shorting of de- tions (BHJs) with extremely high surface area, controlled vices. Nanoporous Titanium (have been reported for first pore size and morphology to maximize charge separation time) will have numerous other applications including Dye and collection efficiencies. Success of this work will allow sensitized solar cells. We are currently focusing our efforts development of next generation OPV devices with high ef- on optimization of nanoporous based photovoltaics by (a) ficiencies that rival the mono- and polycrystalline Si panels depositing an oxide layer to prevent shorting (b) lowering with efficiency ranges from 12-18%. Significantly lower the work function of materials. the price for manufacturing and installation, meanwhile reduce the unit price of solar power. This will significantly OPV devices shorten the time frame originally set by the DOE SunShot We then use these polymers to fabricate organic photo- Initiative to achieve cost competitive solar energy by 2020. voltaic (OPV) devices. The device structure comprise of multilayered ITO as transparent electrode, MoO3 is the hole transporting layer, polythiophene derivative and Publications C60 mixture as electron donor and acceptor, and Ca/Al as Kang, L. L., P. Xu, B. Zhang, H. H. Tsai, X. J. Han, and H. L. another electrode. Polymer solar cells are fabricated using Wang. Laser wavelength- and power-dependent plas- alternating copolymers blended with Phenyl-C71-butyric mon-driven chemical reactions monitored using single particle surface enhanced Raman spectroscopy. 2013. acid methyl ester (PC71BM) to form donor (D) /acceptor CHEMICAL COMMUNICATIONS. 49 (33): 3389. (A) bulk-heterojunction. Our polymers with low band gap and high degree of crystallinity are used for initial optimi- Kuo, C. Y., Y. C. Huang, C. Y. Hsiow, Y. W. Yang, C. I. Huang, S. zation. We spun cast polymer/PC71BM blend solution in P. Rwei, H. L. Wang, and L. Wang. Effect Side-Chain Ar- high boiling point 1,2-dichlorobenzene (oDCB) to achieve chitecture on the Optical and Crystalline Properties of 169

Page 170

Two-Dimensional Polythiophenes. 2013. MACROMOL- ECULES. 46 (15): 5985. Li, Q., P. Xu, B. Zhang, G. Wu, H. T. Zhao, E. G. Fu, and H. L. Wang. Self-supported Pt nanoclusters via galvanic re- placement from Cu2O nanocubes as efficient electro- catalysts. 2013. NANOSCALE. 5 (16): 7397. Park, Y. I., B. Q. Zhang, C. Y. Kuo, J. S. Martinez, J. Park, S. Mallapragada, and H. L. Wang. Stimuli-Responsive Poly-N-isopropylacrylarnide: Phenylene Vinylene Oligo- mer Conjugate. 2013. JOURNAL OF PHYSICAL CHEMIS- TRY C. 117 (15): 7757. Park, Y. S., Y. F. Chen, Y. Ghosh, A. Pirytinski, P. Xu, N. H. Mack, H. L. Wang, V. I. Klimov, J. A. Hollingsworth, and H. Htoon. Strong Photon Bunching in Individual Nano- crystal Quantum Dots Coupled to Rough Silver Film. 2012. In 2012 CONFERENCE ON LASERS AND ELECTRO- OPTICS (CLEO). , p. -. Park, Y. S., Y. Ghosh, Y. Chen, A. Piryatinski, P. Xu, N. H. Mack, H. L. Wang, V. I. Klimov, J. A. Hollingsworth, and H. Htoon. Super-Poissonian Statistics of Photon Emis- sion from Single CdSe-CdS Core-Shell Nanocrystals Coupled to Metal Nanostructures. 2013. PHYSICAL RE- VIEW LETTERS. 110 (11): -. Sasaki, D. Y., N. Zawada, S. F. Gilmore, P. Narasimmaraj, M. A. A. Sanchez, J. C. Stachowiak, C. C. Hayden, H. L. Wang, A. N. Parikh, and A. P. Shreve. Lipid Membrane Domains for the Selective Adsorption and Surface Pat- terning of Conjugated Polyelectrolytes. 2013. LANG- MUIR. 29 (17): 5214. 170

Page 171

Chemistry and Material Sciences Exploratory Research Continuing Project 3-Dimensional Characterization of Nuclear Fuels: Microstructural Evolution under Representative Temperature and Thermal Gradients Donald W. Brown 20130232ER Introduction in that it provides data that will be used to validate com- Recently, interest in nuclear energy in the Unites States putational models which are currently being developed has surged with increased efforts to extend the lifetime as part of DOE NE. It will advance the field of ceramic of the current fleet of reactors and to develop advanced nuclear fuels specifically, and ceramics in general. The reactors which operate at higher temperatures and use development of the technique is directly relevant to the fuel to higher burn-up. Safety margins and predictions of Materials and Radiation Interaction in Extremes (MaRIE) the engineering performance of nuclear reactor fuel rely project in that one would imagine every sample run in on modeling codes used to predict dimensional change, the multi-probe diagnostic hall (MPDF) first being char- stress state and fission gas release as a function of burn- acterized with the high energy diffraction microscopy up and temperature history. Thermal conductivity is (HEDM) developed herein. This technique development the single most important material parameter in these is particularly relevant to the nuclear stockpile mission codes and is heavily dependent on microstructure e.g. in that it will focus on specific advances necessary to grain morphology, porosity, etc. To date, empirical and study high-Z number materials with x-ray diffraction and phenomenological descriptions of the evolution of the tomography. microstructure of fuel materials are used to calculate the thermal conductivity, but this leaves huge uncertainties Progress and necessarily results in large safety margins and inher- In FY13 we wrote proposals for and were awarded two ently inefficient use of resources as well as increased rounds of beamtime on beamline 1-ID at the Advanced production of waste. Photon Source (APS) to complete High energy diffraction microscopy on ceramic uranium dioxide (UO2) and me- The goal of the proposed research is to advance the un- tallic uranium (U) samples. The experiments were run in derstanding of the microstructural evolution of nuclear October of 13 and April of 14. Another proposal will be fuel materials under conditions simulating those per- submitted to the APS for the July 8th deadline. Also, we ceived during manufacture and in-pile. Specifically, we wrote a proposal and received beamtime for tomograph- will monitor grain morphology, pore migration, chemical ic studies of sintered UO2 samples on beamline 10BM segregation, and fission gas transport in nuclear fuels at the APS. Finally, we have prepared and submitted a under processing temperatures (~1650ºC) and under proposal for Small Angle Neutron Scattering beamtime. service conditions (themal gradient of 1000ºC/5mm). (HEDM) experiments on 1ID on UO2+x pellets with Toward this end, we will develop high energy synchro- varying O content and, in a burgeoning collaboration tron x-ray techniques for characterization of fuel under with Idaho National Lab, metallic U foils were examined. representative temperature conditions. These tech- Near field measurements, which map grains by orienta- niques provide a non-destructive means to monitor the tion were completed on three samples with different evolution of fuel microstructure under representative O content (2 off stoiciometric) and 2 metallic uranium conditions. The results will be used by researchers across foil samples. Simultaneously, we completed tomogra- the nation working on multi-length scale modeling of phy on the same samples. Finally, looking forward to ceramic nuclear fuels. in-situ measurements, far field measurements, which can determine orientation and strain in grains, were also Benefit to National Security Missions completed. The work is directly relevant to DOE NE (nuclear energy) 171

Page 172

Analysis of the HEDM data has lagged because we have We will understand preferential segregation of fission struggled to hire the postdoc/student for this work and to products to specific grain boundary types. This data will get a contract in place with Livermore, who has the analy- be used by the fuel modeling community represented, for sis capability in house. The data has been examined to the instance by the CASL (Consortium for Advanced Simula- point that we are sure the analysis can be completed. A tion of Light Water Reactors) at LANL, to both drive model new postdoc has accepted the position and will start July development and validate existing models. 15th. The contract with Livermore is now in place. This will allow analysis of the HEDM data to commence. Publications Brown, D. W., L. Balogh, D. Byler, C. M. Hefferan, J. F. Hunt- Preliminary analysis of the tomography data has been er, P. Henesei, S. F. Li, J. Lind, S. R. Niezgoda, and R. M. completed and is very promising. By including LANL in- Suter. Demonstration of near field high energy x-ray house expertise (James Hunter) and aggressively pursuing diffraction microscopy on high-Z ceramic nuclear fuel real time data analysis, we were able to improve upon the material. . To appear in MECA SENS (Approximately optimization of the instrument setup. It appears we have Mechanical Evaluation of Stress With Neutron and X- ray Scattering Techniques). (Sydney, Australia, 10-12 achieved sub-micron resolution on both UO2 and metallic Sept 213). U samples. The tomographic experiments on sintered and quenched UO2 cylinders have been completed and analysis is ad- vanced. We are currently preparing a manuscript including this data and previous tomographic measurements sup- ported by earlier LDRD projects. The title will be ap- proximately: “The Use of High Energy Xray Tomography to Characterize Porosity in Ceramic Uranium Dioxide Nuclear Fuel.” Don Brown (PI) has been invited to give a presentation at the Mechanical Evaluation of Stress Using Neutron and Synchrotron Techniques (MECA SENS), in Australia in Sep- tember, in which this work will be highlighted. Future Work • Prepare samples in ex-situ thermal gradient • Complete Small Angle Neutron Scattering (SANS) mea- surements during sintering of UO2 samples. • March 14, Prepare ion bombardment assisted deposi- tion (IBAD) samples with Krypton or Xenon bubbles. • Complete High Energy Diffraction Microscopy (HEDM) experiments on heat treated samples. • Complete Small Angle Xray Scattering (SAXS) experi- ments on IBAD samples. • Present HEDM results at topical conference and/or publish results in peer reviewed journal. • Present/publish SANS and SAXS data. Conclusion This experimental work will provided the nuclear fuel theoretical community with unprecedented information concerning the microstructural evolution of fuels during usage. We will not only recognize that the grain structure of a ceramic fuel evolves during use, but will also deter- mine kinetics and mechanisms for said motion of the grain. 172

Page 173

Chemistry and Material Sciences Exploratory Research Continuing Project Very Low Temperature Scanning Point Contact Spectroscopy Investigation of Inhomogeneous States on the Nano-scale Roman Movshovich 20130285ER Introduction Benefit to National Security Missions We will combine our expertise in several Scanning Probe This project addresses the Grand Challenge: “Materials: Microscopy (SPM) and Spectroscopy (SPS) techniques Discovery Science to Strategic Applications”. It creates and a dilution refrigerator instrumented with a 14 Tesla new measurement and modeling capabilities in materi- magnet to develop a novel Low Temperature Scanning als science for superconductivity and f-electron physics. Point Contact Spectroscopy (LTSPCS) apparatus. Point The VLTSPCS will ensure LANL’s position of leadership Contact Spectroscopy (PCS) has been used extensively in the area of correlated electron research. It will be for investigating homogeneous superconducting (SC) directly applicable to investigations of heavy fermion states in zero magnetic field. Our LTSPCS apparatus physics, an area of traditionalyl strong interest at LANL will go well beyond this: a sub-nm resolution of the due to its focus on f-electron elements, especially the ac- PCS combined with the scanning capability and in-situ tinide series. Overall, this project supports both themes control of the force between the PCS tip and the sample, of Emergent Phenomena and Defects and Interfaces, combined with high magnetic field, and very low tem- two of the three areas of the Materials Strategy at LANL. peratures environment, will allow us to explore the spa- The VLTSPCS will open up avenues for new program tially inhomogeneous states in an unprecedented phase development based on the multi-component suite of the space. Within this proposal we will address some of Scanned Probes. John Sarrao, former Office of Science the hottest topics in correlated electron physics, among Programs Director and MaRIE Capture Manager, is cogni- them (1) spatial variation of the energy gap in the postu- zant of the proposed research. lated Fulde-Ferrell-Larkin-Ovchinnikov superconducting state in CeCoIn5, and (2) direct visualization of the high The experimental part of this project cannot be success- field vortex structures, up till now studied only by Small ful without advances in the theoretical understanding of Angle Neutron Scattering (SANS) techniques, in e.g. Ce- the details of the FFLO state (especially at the surface), CoIn5 and UPt3. Our LTSPCS apparatus will be developed and a substantial increase in our capabilities to model on the basis of the combination of techniques and com- point contact spectroscopy of heavy-fermion supercon- ponents adapted from a number of different scanning ductors. The developed modeling capabilities will allow probes, such as a unique all metal tip cantilever (based the interpretation of measurements within this project, on cantilevers used for Scanning Tunneling Microscopy and apply to other correlated electron systems as well. (STM)), and fiberoptic interferrometery employed in conventional Magnetic Force Microscopy (MFM) and Progress Atomic Force Microscopy (AFM). The experimental part We have designed the stage for the dilution refrigerator. of this project will be complimented by advances in the The stage drawings have been submitted to the shop. theoretical understanding of the details of the FFLO state We have placed an order for the digital controller that (especially at the surface), and a substantial increase in will be used in the SPCS apparatus. We have initiated our capabilities to model point contact spectroscopy of a leak check, thermometry calibration, and cool-down heavy-fermion superconductors. The unique combina- procedures for operational test of the dilution refrigera- tion of achieved capabilities will open a new era of cor- tor with a new 14 Tesla magnet. The 1K pot fill line of related electron research at very low temperatures and the dilution refrigerator is being rebuilt. high fields, and help assure the preeminent position of We derived an analytic formulation of the tunneling con- LANL in this field. ductance in the normal state based on our multi-channel 173

Page 174

point-contact tunneling model for heavy-fermion materi- Publications als. This new formulation provides a great simplification Chen, H., Y. Y. Tai, C. S. Ting, M. J. Graf, J. Dai, and J. X. Zhu. over previous formulations and allows us to better under- Disorder effects in multiorbital s±-wave superconductors: stand qualitatively the normal state properties of electrons Implication for Zn-doped BaFe2As2 compounds. Phys. Rev. tunneling from a metallic tip into a heavy-fermion metal X. beyond numerical results. In addition, we have begun with the construction of a tunneling Hamiltonian that incorpo- rates the magnetic field and other ordered states, such as a spin-density wave. The effects of both magnetic field and spin-density wave in the superconducting phase need to be accounted for when analyzing SPCS measurements in the suspected FFLO phase of CeCoIn5 at low temperatures and high magnetic fields, because of the NMR observation of antiferromagnetic order in the same part of the phase diagram. Jeehoon Kim will be leaving the lab by August 1. We have initiated a search for his replacement. Neliza Leon-Brito from the Alex DeLosanne’s group at the UT Austin has ar- rived to LANL for training and experimental collaboration with Jeehoon while he is still here. Future Work Within the experimental part of our project, we will build the Very Low Temperature Scanning Point Spectroscopy (VLTSPCS) stage and combine it with a dilution refrigerator and 14T magnet. We will demonstrate the scanning capa- bility of our device on NbSe2; image the superconducting vortex (spatial variation of the energy gap) and vortex lat- tice with the SPCS apparatus. Within the theoretical part of our project, we will develop the model for the FFLO state in the presence of localized states at the surface and calculate the conductance in the vortex state for fields at arbitrary angle relative to the surface. Conclusion We will develop new experimental capability of Very Low Temperature Scanning Point Spectroscopy, together with theoretical modeling support, to image the superconduct- ing energy gap on the sub-nm scale, at temperatures down to few tens of mK, and fields up to 14 T. Visualization of the spatial variation of the superconducting energy gap in CeCoIn5 will settle the controversy of whether the High Field Superconducting (HFSC) phase is a realization of the FFLO state, predicted theoretically in 1960’s, or an exotic field-induced magnetic state. We will directly image the vortex lattice of CeCoIn5 and explore its rich and exotic phase diagram. 174

Page 175

Chemistry and Material Sciences Exploratory Research Continuing Project Excited State Quantum Interactions in Carbon Nanotubes Stephen K. Doorn 20130309ER Introduction Each of these focal areas represents pioneering studies Our recent first-time demonstration of quantum interfer- that will open significant new areas of CNT physics for ence and violation of the Condon further investigation. No other molecular or condensed- approximation in carbon nanotubes (CNTs) overturns matter system offers the ability to study non-Condon long-held assumptions regarding their optical response and interference behaviors in such a controlled and and opens the way for new quantum physics to be systematic way and the information gained here will explored in these systems. While these fundamental impact understanding of photophysical behaviors of behaviors are fascinating by themselves, they have the broad classes of optical materials and serve as a bridge potential to alter relaxation pathways and dynamics between molecular and condensed matter systems. underlying anticipated applications in photonics, opto- electronics and energy harvesting. It is therefore impor- Benefit to National Security Missions tant to understand both the origins and consequences Results of this project will have a direct bearing on of these phenomena in CNTs. Our goal is to understand developing optical and electronic properties of carbon how the interplay of electronic and vibrational structure nanotubes for photonic and energy harvesting applica- in CNTs gives rise to non-Condon and interference be- tions and may also contribute to sensing and spectral haviors and determine what consequences they hold for tagging applications. As a result this work will have direct observed photophysical responses. We focus on three relevance to the mission of the DOE-BES funded Center objectives: for Integrated Nanotechnologies and the potential appli- cations will be of interest to agencies including NIH, DOE,

1. Non-Condon Behaviors will be probed for multiple DHS, and DOD with potential impact on threat reduction nanotube structures, vibrational modes, and excited and renewable energy missions. Our effort will also drive states using Raman and modeled with quantum development and understanding of fundamental surface chemical theory to gain an understanding of their chemistry of low-dimensional materials. Additionally, origins. Dynamic perturbation of CNT structure will the quantum behaviors we will study are found in broad further inform on their origins and reveal routes to classes of materials, but only carbon nanotubes allow control the effect. the detailed study that will be pursued in this effort. As a result, the nanotubes serve as ideal model systems for
2. Dynamic Control of Quantum Interference by strain obtaining a basic understanding of several novel quan- tuning of the electronic energy spacings that deter- tum behaviors. As such, their study supports DOE goals mine interference will allow us to sample all pos- for expanding fundamental understanding of functional sible interference behaviors while using interference nanomaterials. response to probe consequences for underlying photophysical behaviors. Progress
3. Relaxation Pathways and Dynamics will be probed We report progress in 6 separate areas related to the using time-resolved spectroscopies to learn the fate goals of the project of optical excitations as determined by phonon-me- Non-Condon Behavior in Armchair Carbon Nanotubes: diated electronic state interactions. We have successfully obtained Raman excitation profiles (REPs) for the (5,5), (6,6), and (7,7) chiralities (so-called armchair metallic structures) of nanotube. Material 175

Page 176

enriched in these structures was obtained via two separate son shows the 2D behavior arises from an LO phonon and enrichment methods and from two separate nanotube allows us to confirm our understanding of phonon disper- source materials. The strong asymmetry in the REPs that sion in nanotubes. Our next step is to apply similar com- was previously observed in semiconducting tubes was parisons for all the tubesfor which we have overlapping found to continue for these metallic structures. The results theory and experimental data. suggest a continuation of the trend that asymmetry in- creases with tube diameter and towards large chiral angles Sample Development for Strain-Based Measurements and further indicates the behavior is intrinsic and general We require increased amounts of enriched material for for all tube types. The results have been submitted to Phys. our strain-based measurements and are applying a new Rev. B and the manuscript is under revision after a first two-phase extraction method for separations that depends round of reviews. on tuning relative degrees of hydrophobicity/-philicity be- tween different structures. We are already showing good Frequency Behavior of the TO Phonon in Armchair Nano- results for isolation of (6,5) tubes and are progressing to tubes isolation of other structures. The strain measurements also Our work on enriched samples of single armchair struc- require a composite matrix that will transfer applied strain tures has allowed us to obtain the first ever detailed to the constituent nanotubes. We have evaluated PDMS (a frequency values for the TO phonon mode over 6 differ- silane polymer), which may be suitable but requires tricky ent structures. Frequencies show a sharp decrease to processing conditions. We are currently looking into tri- smaller diameter tubes, but can be modeled by a strongly block copolymers as an alternative. upshifted response similar to the TO mode of semiconduct- ing tubes. By also probing the response of the frequencies Pump-Probe Capability for Evaluating Dynamics Response to electrochemical doping we are able to understand the to Quantum Interference TO behavior as arising from the effects of rehybridization We are nearing completion of building femtosecond pump- induced by strong tube curvature. Upshifted frequencies probe capability to evaluate excited state dynamics in result from an anomalie in the phonon dispersion. Our single-chirality samples. In particular we are interested in results resolve a controversy in the theoretical community measuring those in the (10,5) tube, which is anticipated to over how to describe electonr-phonon coupling in metallic reveal new quantum interference behaviors. This structure nanotubes. requires a pump pulse capable of being tuned to 375 nm. We are currently waiting on a chirped mirror set used to Transition and Phonon Mode Dependence of Non-Condon shorten pulses to an anticipated 5 femtoseconds. Once Effects incorporated in the instrument, we will have tunable 5 fs We have now obtained REPs for the E11 thru E44 (first pulses available from 335-430 nm and from 550 to 900 nm. through fourth exciton transitions) of a wide range of semiconducting tube structures. Additionally, we have Future Work this data for the LO, TO, and radial breathing mode (RBM) Critical to understanding the origins of the non-Condon phonon modes. In all cases except for one, non-Condon- effects is determining their dependence on nanotube induced asymmetry is observed. For the RBM, surprisingly, structure. We expect the effect is related to the curvature we observe no asymmetry. For the LO and TO, we find of the related band structure, which is a strong function of that as one goes from E11 to E22 excitation the asymmetry excitation energy. To confirm this expectation, our inves- increases significantly, but shows no particular trend on tigations will probe the behavior for the first four bright going to higher-energy excitations. The results show that exciton states, using the nanotube G-band vibration as we do need to consider how extrinsic effects may impact a probe. We will also evaluate the behavior for metallic the non-Condon behavior. Additionally, we have begun structures, in order to both extend the generality of the quantum chemical modeling to evaluate where differences behavior beyond semiconducting structures and to evalu- in non-Condon response for the different phonon modes ate how the effect may be linked to exciton formation or to originates. The experimental part has been written up for non-adiabatic behaviors inherent to other electron-phonon publication, which will wait on the theory results. coupling phenomena in metallic structures. Probing Double Resonance Raman Response Equally important is determining how vibrational mode We have completed theoretical modeling of the double and different coupling mechanisms impact the non- resonance (2D) Raman response for several nanotube Condon response by evaluating the effect over additional chiralities and have performed a detailed comparison to modes (not just G-band). Analysis of excitation behavior experimental results on the (9,1) structure. The compari- of the nanotube radial breathing mode (RBM), 2D, and 176

Page 177

M modes will be our targets. The resonance behaviors for these vary from direct exciton resonances to double resonance processes that link the nanotube electronic and phonon densities of states. Experimental data will be ana- lyzed and modeled within a 3D evaluation of these state behaviors. The result should help address current con- troversy about so-called internal and external resonance processes in these materials and will provide further insite into how non-Condon effects arise. In anticipation of an emphasis on dynamic tuning of inter- ference and non-Condon behaviors for follow-on years, we will also focus effort on development of appropriate sample matrices to enable our envisioned strain-tuning measurements. Finally, on the theory side of our effort, we will extend our existing modeling of non-Condon effects on the G-plus mode to also include modeling of expected behavior for G-minus and RBM vibrations and establish a foundation for also modeling strain effects. Conclusion Our effort will lead to an understanding of how interac- tions between electronic and vibrational structure in carbon nanotubes (CNTs) leads to quantum interference behaviors and gives rise to breakdown in fundamental descriptions of quantum interactions. These will be pio- neering studies in an emerging area of CNT photophysics. The result will allow us to determine what consequences these behaviors have for nanotube optical responses that underlie photonics, optoelectronics, and energy harvesting applications. Ultimately, by revealing the origins and con- sequences of these new CNT quantum phenomena we will also generate the principles for how we may control them for enabling applications. Publications Berciaud, S., X. Li, H. Htoon, L. E. Brus, S. K. Doorn, and T. F. Heinz. Intrinsic line shape of the Raman 2D mode in freestanding graphene monolayers. 2013. Nano Let- ters. 13: 3517. Duque, J. G., E. H. Haroz, H. Telg, J. J. Crochet, J. Simpson, A. R. Hight Walker, X. Tu, M. Zheng, J. Kono, and S. K. Doorn. Asymmetric excitation profiles and non-Condon effects in the resonance Raman respons of armchair carbon nanotubes. To appear in Physical Review B. Haroz, E. H., J. G. Duque, X. Tu, M. Zheng, A. R. Hight Walk- er, R. H. Hauge, S. K. Doorn, and J. Kono. Fundamental optical processes in armchair carbon nanotubes. 2013. Nanoscale. 5: 1411. 177

Page 178

Chemistry and Material Sciences Exploratory Research Continuing Project “Upscaling” Nanoscale Thermoelectrics: The Meso-macroscale Design Challenge for Real-World Energy Needs Jennifer A. Hollingsworth 20130350ER Introduction Benefit to National Security Missions The overarching goal is to address the challenge of scale Successful completion of the proposed objectives will that is uniquely relevant to nano-enabled thermoelec- provide advances in synthesis, characterization trics (TEs). Nanoscale phenomena can dramatically and theory that will lay the foundation for success- alter and improve the fundamental properties of semi- ful, functional translation of nanoscale phenomena to conductor materials for applications in TEs, such as the the macroscale, as well as understanding how proper- electronic density-of-states (DOS) and charge carrier or ties scale with changes in length scale and complexity. phonon mobilities and interactions. However, in contrast More specifically, we will also develop novel functional with other potential nano-based technologies (light- composites comprising NW networks in aerogels for ing, sensing, energy harvesting/storage), TEs requires applications in thermoelectrics. Further, we will estab- bulk-scale assemblies to achieve high efficiencies at the lish new non-contact characterization and modeling device level (cm rather than nm-μm scale). This require- tools for addressing transport processes from simple to ment remains a key challenge to translating properties branched NWs and their composites. This aspect of the efficiencies at the nanoscale to practical efficiencies in work contributes directly to DOE Basic Science initiatives applications. We aim to bridge this literal, physical gap and is relevant to the Scientific Discovery and Innovation in nanomaterials integration by developing new syn- Mission in the Basic Understanding of Materials. thetic methodologies for interconnecting nanomaterial building blocks into functional networks and compos- The work also fundamentally addresses challenges faced ite structures. We will also address an underpinning by the implementation of nano-enabled thermoelectrics “knowledge gap” in terms of understanding how the key in real-world device architectures, where success would transport properties—electrical and thermal—change have a clear impact on Mission Relevance in Energy Se- across the extremity of length scales from nano/meso- curity (Renewable Energy) and the Environment (Climate to macroscale. Though the problem of scale is distinctly and Energy Impact and even “waste” management), as paramount for TEs, the synthetic strategies and new theremoelectrics tackles both waste heat utilization and principles of multi-scale transport will be applicable to a heat management issues. wide range of nano-enabled technologies. To these ends, we will pursue the following three objectives: Lastly, the new composite materials will be tested and operated at temperature extremes (cold and hot) and • Assessing and optimizing nanoscale TE functionality entail conversion of energy types (e.g., heat to electric- by way of baseline studies in single-NW systems of ity) and energy translation across interfaces, with direct controlled size and surface roughness. relevance to complex materials design and characteriza- tion for MaRIE. In addition, the technology developed • Understanding and controlling meso-macroscale will support longer term development of small-size, TE properties of novel branched/networked NW rapid-response temperature probes for extreme environ- constructs. ments, e.g., explosives. • Assessing the viability of aerogels as a NW TEs ma- trix. Progress Significant progress in both theory and experiment are described below for each objective. 178

Page 179

Despite a late start by the primary synthetic chemist on key question of how branch points affect carrier/phonon the effort (Dr. Mishra: mid-April arrival), we have already transport. Significantly, members of this ER team have synthesized several of the test materials for project Objec- already demonstrated theoretically a key premise of our tive 1 – ultra-narrow and thicker (~40 nm) diameter PbS proposed work. They have developed a model that maps and PbSe nanowires and surface-roughened nanowires. a nanowire network on a two-port network used in the We have added to our “test systems” with the inclusion theory of electrical circuits to describe charge-carrier/heat of PbSe/PbS tetrapods, which afford in a single nano-con- transport in interconnected systems. This model has been struct multiple branch points and, thereby, an additional implemented into an efficient computational code and option for testing our hypotheses pertaining to branching- allows us to perform simulations at the device-modeling effects on thermoelectric properties, especially electrical/ level. The simulations have been performed for Bi2Te3 thermal transport. The tetrapods also inherently contain nanowire-networks and have demonstrated that a network an internal interface that may further serve to modify structure enhances thermoelectric figure-of-merit, ZT. Ad- carrier/thermal transport as a function of its crystalline ditionally, they have developed a number of numerical mi- quality (expected to be quasi-epitaxial) and the size of the croscopic techniques for modeling thermoelectric proper- tetrapod arm-lengths/diameters. The tetrapod syntheses ties of individual nanowires and the effect of junctions for entail first synthesizing CdSe/CdS tetrapods then chemical- branched nanowires. These techniques have been applied ly converting these by cation exchange reactions into the to Bi2Te3 materials and successfully benchmarked against desired PbSe/PbS structures. We have to date prepared 6 published data. They have obtained data that has been nm-diameter/35 nm-arm length and 12 nm-diameter/30 used as input parameters to the network-simulation code. nm-arm length CdSe/CdS tetrapods, which have been con- A manuscript is in preparation (O. Roslyak, A. Piryatinski verted to their PbSe/PbS counterparts. “Enhanced thermoelectric properties of semiconductor nanowire networks” Nano Letters (in prep), and two oral We have also begun efforts to synthesize Bi2Te3 nanow- presentations have been given (O. Roslyak, A. Piryatinski; ires. Our goal is to synthesize controlled-diameter nanow- “Thermoelectric effects in disordered branched nano- ires (~5 and ~40 nm) by solution-liquid-solid (SLS) growth, wires”, APS Annual Meeting, 2013; O. Roslyak “Enhanced as this method affords facile control over nanowire di- thermoelectric effect of branched nano-wire trees.” Invited ameter by simple manipulation of the size of the metal lecture, Fordham University). nanoparticles used to “catalyze” nanowire nucleation/ layer-by-layer growth/elongation. This approach for Bi2Te3 We have successfully incorporated semiconductor nanow- is not yet known in the literature. To this end, we are at- ires into an aerogel matrix. Here, we employed “model” tempting to modify or develop new synthetic procedures nanowires comprising CdSe, as this material was immedi- for “Bi2Te3 quantum dots” for adaptation to SLS growth. ately available. We successively increased the concentra- Specifically, SLS requires that Bi2Te3 is formed by direct tion of nanowires embedded within the porous aerogel reaction of Bi and Te molecular precursors, rather than re- matrix. We have yet to hit a synthetic limit in this process, action through a Bi (or Te) crystalline intermediate (litera- but are challenged by solvent compatibility issues (differ- ture approaches). We have to-date performed numerous ences in preferred solvent for the nanowires compared reactions between different Bi/Te molecular precursors in to the aerogel formation chemistry). Nevertheless, com- a range of solvents, mediated by various surfactants over posites-to-date are dark in color (nanowires) but remain several temperatures. These have been characterized by uniform in appearance, translucent and structurally sound. powder x-ray diffraction (XRD) and transmission electron We will now extend to thermoelectric-relevant nanow- microscopy (TEM). While XRD shows we are creating ires using modified nano-surface chemistry for enhanced crystalline Bi2Te3, morphology remains non-optimal and/ chemical-compatibility. or the nanoparticles are aggregated. With a satisfactory approach in hand, we will attempt SLS Bi2Te3-nanowire Future Work growth. Nanoscale test systems Using solution-phase methods (low-cost and scalable), We have identified a “test-bed” device-design to use we will continue to synthesize NWs that allow us to test for comparative studies of the different thermoelectric nanoscale effects on key TE parameters: the electronic DOS nanomaterials: reported in Wang et al. Nano Lett. 8, 2008, – i.e., ultranarrow sub-10 nm NW diameter versus tens of 2283. We will fabricate/test similar devices in CINT using nms, and carrier/phonon transport – i.e., smooth versus our nanowire/tetrapod series. nanoroughened wire surfaces. Additional NW structural/ compositional characteristics can be modified to achieve Theoretical efforts have begun in earnest to address the more complete optimization of TE properties, but our aim 179

Page 180

here is instead to investigate the translation of nanoscale Publications effects across vast length scales. For this reason, we focus Roslyak, O.. Enhanced thermoelectric effect of branched our studies on NW size (quantum confinement) and sur- nano-wire trees. Invited presentation at Fordham Uni- face properties. versity Seminar. (New York, 12 May 2013). Roslyak, O., and A. Piryatinski. Thermoelectric effects in Meso-macroscale integration/exploration disordered branched nano-wires. 2013. In American We will develop new synthetic methods for interconnect- Physical Society Annual Meeting 2013. (Baltimore, ing NWs into branched/hyperbranched structures as de- MD, 18-22 March 2013). , p. BAPS.2013.MAR.W12.14. signed model systems for understanding scaling effects on College Park, MD: Bulletin of the American Physical TE properties. By linking NWs together into progressively Society. more complex assemblies, we will address how branch points affect carrier/phonon transport and whether these Roslyak, O., and A. Piryatinski. Enhanced thermoelectric properties of semiconductor nanowire networks. Nano processes can be controlled by tuning the NW-NW connec- Letters. tion (size, composition or crystal-structure mismatch) and concentration of branch sites. Experimental and theoreti- cal characterizations on individual or small collections of NW branch points will encompass mesoscale processes. Going beyond this, we will further increase NW network complexity (via extreme hyperbranching or gelation condensation chemistry) to access transport processes in complex interconnected NW networks and to investigate non-equilibrium phonon dynamics that may result from bulk disorder & fractional dimensionality. Aerogel utility Aerogels are characterized by extremely low thermal con- ductivity and continuous porosity. They can be molded into pellets and machined. We will optimize NW-aerogel com- posite chemistry. We aim to assess the utility of aerogels as scalable matrices for bulk-scale NW TEs by establishing basic transport properties of NWs-in-aerogels from single/ branched aerogel-embedded NWs to NW networks in an aerogel framework. Conclusion The technical goals that will be pursued are: • Using solution-phase methods, we will synthesize NWs that allow us to test nanoscale effects on key TE pa- rameters: the electronic DOS – i.e., ultranarrow sub-10 nm NW diameters, and carrier/phonon transport – i.e., smooth versus nanoroughened wire surfaces. • We will develop new synthetic methods for intercon- necting NWs (from Obj 1.) into branched and hyper- branched structures as designed model systems for understanding/controlling scaling effects on TE proper- ties. • We will develop methods for embedding NWs into aerogels and assess the utility of aerogels as scalable matrices to realize bulk-scale NW TEs. 180

Page 181

Chemistry and Material Sciences Exploratory Research Continuing Project Threat Reduction via Nanomaterials: Engineering a Novel SERS Plat-form for Chemical Detection Nathan H. Mack 20130416ER Introduction interest in the detection of chemical species. Surface enhanced Raman spectroscopy (SERS) is an established technique for boosting the intrinsic Raman Benefit to National Security Missions signal from a molecular compound by binding it to an This work will have two primary areas of mission rele- appropriately designed metal surface. It has the capabil- vance which can be extended to any agency interested in ity for not only ultra-sensitive detection levels, but also the ultra-sensitive detection and identification of chemi- provides a molecular fingerprint spectrum that can be cal species. First, we hope to gain a better understand- used for compound identification. For nearly a decade ing of how these novel materials interact with various now, this technique has been studied as a potential threat agents (and their simulants). This will include break through approach to sensing threat agents (e.g. how these chemical species bind to metal surfaces, what explosives, chemical weapons, and toxic chemicals). their relative affinities are for these surfaces, how these Unfortunately, the processes required to make an effec- affinities can be enhanced, and what optical signatures tive SERS metal substrate required expensive lithogra- can be exploited for chemical identification. All of this phy techniques and were often irreproducible, thereby falls under the fundamental basic understanding of limiting it’s use to carefully controlled laboratory experi- materials. ments. Second, this work will have national security and threat Recent developments at LANL have produced a polymer- reduction mission relevance in that we hope to demon- based metal-nanocomposite that can act as a highly en- strate real world detection and identification of various hancing SERS substrate. We have found it to be sensitive chemical threat agents. This could be applied to portal to a variety of organic compounds down to 10^-12 molar monitors, airport screening, first responder devices, and concentrations. Additionally, it can be produced at a be applied to war-fighter support. Additionally, chemi- fraction of the cost and has superior physical and me- cal detection and identification will have intelligence and chanical properties as compared to traditional SERS sub- reconnaissance applications. strates. This initial work hints that these materials may be ideally suited as a SERS based detection platform as Progress the material can be produced in flexible thin film formats One of the greatest challenges to realizing SERS based that opens up a wide array of potential detection ap- sensing is the reproducible production of highly enhanc- proaches. In this work, we will exploit these enhanced ing structured metal surfaces at a realistic cost. In order properties to develop a field-able sensing platform for to move our technology beyond TRL1 as well as move threat agent detection. We will determine the limits of towards a more portable form factor (i.e. a test stick or detection for a variety of explosive residue and chemical cartridge based substrate), we set out to develop a set agent simulants, develop new process techniques to en- of polymer process rules that define how to produce able high throughput substrate production, and design reliable SERS substrates. Initial batches of polymer films a proof of concept sampling system that will be suitable were producing SERS sensing substrates that perform, for portable detection applications. however, they exhibit incomplete coverage of the poly- mer surface with structured metal, leading to inhomoge- Successful completion of this work will bring SERS to neous SERS signals. We altered the process conditions the forefront of detection technologies and have wide to include both solvent removal, acid doping, and me- applications among DOD, DHS, and other agencies with 181

Page 182

chanical pressing steps, as well as empirically adjusted rela- representative spectra. We hope to submit our purchase tive concentrations of the Ag metal precursors to result in request within the month and have all acquisitions com- a process that yields films with >80% metal coverage. This pleted by the end of this FY. coverage is high enough such that any voids in the metal film are smaller than the spotsize of our Raman laser, and Future Work thus insignificantly affect the resulting SERS signal. In the first year of this project we will have run our nano- composite through a variety of sensitivity tests to deter- With these films in hand, we are now starting to screen for mine the lower limits of detection to several threat agents. potential threat agents of interest, starting with simple or- In the next FY, this work will be extended to include explo- ganophosphate solutions. Initially, these solutions are rela- sives residues and explosive precursors to test for potential tively high in concentration (millimolar); ultimate limits of applications in detection of energetic materials. Addition- detection will be resolved once we have a more complete ally, we will explore surface modification of the sensor sensing matrix indicating what agents yield the greatest surface to affect enhancements in sensitivity. We intend SERS signal. We expect to have this matrix completed for to explore self assembled monolayers to vary the sensor’s CW simulants by the end of the FY, with similar matrices hydrophobicity and surface charge, as well as ligand based for home made explosives and energetic materials early in molecules to serve as capture moieties for binding analytes the next FY. We have contacted WX personnel to assist in of interest. providing energetic materials for testing. On the engineering side of this project, we have devel- In parallel, we have been working with ISR-5 to explore oped a reproducible method for batch production of our new SERS sensing form factors that take advantage of that sensor materials. In the next FY, we will work to develop groups automated production facilities. This approach a new polymer processing approach (based on our batch would automate the polymer processing steps in order synthesis approach) that will enable continuous fabrication to 1) improve batch reproducibility and reliability and 2) of ribbons of the nano-composite. These ribbon materials enable precise spatial control over polymer deposition by will be verified for sensing capabilities and be used as one extruding the polyaniline (PANI) from a fine needle prior to component in our proposed sampling scheme. We will solvent removal. This will result in patterned and layered study the mechanical properties of laminating this material PANI structures that can then be subsequently metalized. to a flexible permeable membrane, which will serve to pro- Our hope is that we can use these 3d hierarchical struc- tect the sensing surface and act as a sample compartment tures to further enhanced analyte capture (and retention). for analytes of interest. Key metrics in this process are the polymer gelation time, as all patterning must be performed prior to the polymer Our new portable spectrometer will come online at the setting as a gel, which could affect subsequent metalliza- beginning of the next FY. Work in this area will focus tion processes. on fabricating any necessary components to allow us to rapidly analyze test samples for demonstration purposes. Finally, we have been investigating the purchase of a por- Lower limits of detection will be determined using both table raman spectrometer to be used for demonstration model analytes as well as analytes from our previously of real world SERS analyte detection. Typical SERS work in determined detection matrix. the literature is performed on a highly tuned spectrometer that requires large lasers, microscopes, optics, etc. How- Success in these three areas will enable use to gain valu- ever, translating results on an expensive research grade able information on analyte sensor affinities and spectral spectrometer over to a realistically portable spectrometer responses as well as perform real world demonstrations of often results in failure. We had one vendor visit our labs threat agent detection to sponsors for follow-on funding in to demonstrate their portable spectrometer that can be out years. battery powered, and is approximately the size of a carry on luggage bag. This spectrometer performed well as Conclusion compared to data taken on an identical sample using our In this work, we seek to develop our novel nano-composite research grade Raman system. This spectrometer also has into a field-able sensing platform for threat agent detec- dual wavelength capabilities, which will enable a wider tion. We will determine this material’s limits of detection range of analyte detection through the elimination of for a variety of explosive and chemical agent simulants and fluorescence in troublesome samples. This spectrometer is then use this data to guide the development of a proof of our top choice, but we have sent several samples for dem- concept sampling system that will be suitable for portable onstration at several vendors’ labs and are awaiting their 182

Page 183

detection applications. The engineering component of this work will involve developing new process techniques to enable continuous production of the polymer-based mate- rial to allow for high throughput substrate production at previously unrealized low costs. Publications Kang, L. L., P. Xu, B. Zhang, H. H. Tsai, X. J. Han, and H. L. Wang. Laser wavelength- and power-dependent plas- mon-driven chemical reactions monitored using single particle surface enhanced Raman spectroscopy. 2013. CHEMICAL COMMUNICATIONS. 49 (33): 3389. Kang, L. L., P. Xu, D. T. Chen, B. Zhang, Y. C. Du, X. J. Han, Q. Li, and H. L. Wang. Amino Acid-Assisted Synthesis of Hierarchical Silver Microspheres for Single Particle Sur- face-Enhanced Raman Spectroscopy. 2013. JOURNAL OF PHYSICAL CHEMISTRY C. 117 (19): 10007. Li, S. W., P. Xu, Z. Q. Ren, B. Zhang, Y. C. Du, X. J. Han, N. H. Mack, and H. L. Wang. Fabrication of Thorny Au Nanostructures on Polyaniline Surfaces for Sensitive Surface-Enhanced Raman Spectroscopy. 2013. ACS AP- PLIED MATERIALS & INTERFACES. 5 (1): 49. Park, Y. S., Y. Ghosh, P. Xu, N. H. Mack, H. L. Wang, J. A. Hol- lingsworth, and H. Htoon. Single-Nanocrystal Photolu- minescence Spectroscopy Studies of Plasmon-Multiex- citon Interactions at Low Temperature. 2013. JOURNAL OF PHYSICAL CHEMISTRY LETTERS. 4 (9): 1465. Park, Y. S., Y. Ghosh, Y. Chen, A. Piryatinski, P. Xu, N. H. Mack, H. L. Wang, V. I. Klimov, J. A. Hollingsworth, and H. Htoon. Super-Poissonian Statistics of Photon Emis- sion from Single CdSe-CdS Core-Shell Nanocrystals Coupled to Metal Nanostructures. 2013. PHYSICAL RE- VIEW LETTERS. 110 (11): -. 183

Page 184

Chemistry and Material Sciences Exploratory Research Continuing Project Accurate Interfacial Structures for Atomistic Simulations: Minimizing the Grand-canonical Free Energy Danny Perez 20130517ER Introduction Benefit to National Security Missions To carry out effective simulations at the atomic scale, it is Atomistic computer simulations play an increasingly crucial to know the details of the atomistic structures of important role in the prediction of properties of mate- materials. Often, obtaining that information experimen- rials, or in the interpretation of experimental results. tally is impractical so simple rules of thumb or geomet- However, to be effective, these simulations need ad- ric constructions are used to fill in the missing details. equate starting points that properly represent the most Sometimes, limited searches are used to explore the probable atomic-scale structure of the material. This neighborhood of these guesses to assess their thermo- is crucial for a number of the lab’s missions, specially dynamic stability, which is often a good metric of their those dealing with materials in extreme conditions. For practical relevance. However, due to the large size of the example, we recently found that adding a small propor- accessible configuration space, unconstrained searches tion of interstitials to a grain boundary in tungsten (a are scarcely attempted. This problem is even more leading candidate for the first wall of fusion reactors) can severe when the number of atoms in the structures dramatically lower the stress at which the boundaries of interest is unknown. This is a common case when slide, which could lead to creep, but also enhance the defects are involved. This inability to thoroughly search ability of the materials to heal radiation-induced defects. the space of possible configurations severely limits the To quantify these effects, it is paramount to first identify accuracy of atomistic simulations. In this project, we the structures that are thermodynamically relevant (i.e., propose to develop efficient sampling algorithms that who are the most stable). These same challenges exist would allow to systematically and autonomously search everywhere atomistic simulations of materials are used. for the most relevant structures. Our work will be based Our proposed methodology will assist in the sampling on a grand-canonical approach where the number of of the possible structures and greatly help in the identi- atoms in the simulation cell is not fixed, but can freely fication of the relevant ones. This has the potentials to evolve according to its own dynamics, i.e., atoms will be significantly increase the power of atomistic simulations able to dynamically come in and out of the simulation as across a very wide range of problems of direct relevance time passes. to the lab’s missions, for example to investigate the properties of cladding materials in nuclear reactors, or to This framework will be coupled to state-of-the-art sam- be able to predict the response of polycrystalline materi- pling algorithms that will allow an efficient search of this als to shock, to name only a few examples. This capabil- huge grand-canonical (where the number of atoms is ity directly supports mission needs for DOE Office of also free to vary) space. Our method will first be demon- Science, Nuclear Energy, and MaRIE while enhancing our strated on problems related to grain boundary and inter- basic understanding of materials mission. faces in materials, a very important class of two- dimen- sional defects that often significantly affect mechanical Progress properties. Our approach could have a profound impact The last year was dedicated to assessing the perfor- in the field by leveraging modern simulation techniques mance of our initial simulation code and to the analysis to turn the “art” of predicting the atomic structure of of the most promising sampling algorithms that could materials into a systematic search process. This could be leveraged to ensure a thorough exploration of the dramatically improve the accuracy and predictive power possible atomistic configurations of our system. We have of atomistic simulations. tested approaches based on meta-dynamics, a well- 184

Page 185

known sampling algorithm that learns about the features a comparison with extensive accelerated molecular dy- of the system as the simulation proceeds in order to design namics simulations, we have showed that our method can a bias function that would make sampling more efficient by quickly and reliably identify highly non-trivial low-energy promoting the exploration of unlikely regions in the space configurations, such as a new type of interstitial clusters of configurations that can act as pathways to other isolated that consist of delocalized defects in nanoscale triangles regions. We have shown that this approach performs well located in the basal planes of Zirconium. Interestingly, this for grain boundaries, and we observe that we are able defect, which was previously largely overlooked, appears to add and remove whole layers of the boundary freely, to be very stable, and could therefore be very relevant in which indicates very good sampling efficiency. Our results applications. We have also demonstrated that different demonstrate that for certain boundaries, local reorgani- interatomic potentials give drastically different predictions zations can lead to relatively small changes in formation of the clustering of small vacancy clusters. This information energy as more atoms are added or removed from the will be leveraged by creep modeling activities carried out boundary. in T and MST divisions, as part of the CASL Nuclear Energy Hub. We have also investigated and adopted a novel sampling algorithm, so-called statistical temperature sampling. This Future Work algorithm allows us to go beyond the initial expectations of During the second FY, we will pursue the derivation and the proposal in that it enables correct finite-temperature implementation of our statistical temperature sampling sampling, in contrast to zero-temperature sampling as we scheme that will increase the efficiency of the simulation initially expected (i.e., we can now identify structures that by promoting the exploration of different configurations, are the most stable at finite temperature, and not only at with varying numbers of atoms and energies. We expect the absolute zero). This will allow us to investigate thermo- that this approach will lead to very efficient sampling, dynamic effects, such as interface phase transitions, while while allowing for a rescaling of the simulation results that greatly enhancing the sampling efficiency compared to will allow us to make predictions over a range of chemical direct simulation techniques. This method will also allow potentials and temperatures from a single simulation. This us to obtain results for a wide range of temperatures from will go beyond what we expected to be able to achieve a single simulation. We have implemented the method in during the course of this project, as it will open the door our simulation code and demonstrated that it is able to to the study of finite temperature effects, such as phase quickly recover the crystalline minimum energy state, even transitions, on boundaries. This will be a unique method when starting from a completely disordered configuration. that will greatly expand our current capabilities. Note that This is highly suggestive that a similar performance will be the same algorithm can be run both into exploration and achieved in the case of interfaces. We have begun deriving accurate sampling mode, which allows for exploration of the grand-canonical variant of this method that will allow the possible configurations at a modest cost, and for accu- for correct sampling in the grand canonical ensemble, i.e., rate thermodynamics with a larger investment in computer where the number of atoms can freely change during the time. course of the simulation. This combined approach will lead to very efficient sampling of different configurations with We will validate and thoroughly test the algorithm. Our varying numbers of atoms and energies, which will maxi- main model system at this stage will be a family of grain mize our chances of identifying the most relevant ones for boundaries in copper. We will also quantify the landscape matching experimental conditions. Further, as part of our of defective boundaries and qualitatively assess their work, we have discovered links between the meta-dynam- thermodynamic relevance in different conditions. This will ics and statistical temperature approaches that went unno- inform us on the behavior of boundary upon injection of ticed up to now. We are now developing a unified picture defects, a topic of great relevance to understand radiation of these different methods, which will be very important damage, for example. to the field in terms of assessing, comparing, and improv- ing methods. A manuscript detailing our findings is now in Conclusion preparation. We will develop sampling algorithms based on an innova- tive grand-canonical approach. These methods will be used Finally, we have applied our technique to the investigation to explore the landscape of possible atomic-scale struc- of the configuration of point defect clusters in Zirconium, tures of materials. These techniques will allow research- which are of relevance to understanding radiation-induced ers to systematically, simply, and efficiently search for the creep (the slow plastic deformation of materials when ex- most relevant structures, instead on relying to informed posed to radiation) in the cladding of nuclear fuel. Through 185

Page 186

guesses or searches in narrow parameter spaces. These can then be used as the starting points of very high quality atomistic simulations that aim at predicting the behavior of materials. This new capability will first be applied to the exploration of the structure of boundaries and interfaces in materials. Figure 1. Illustration of a grand-canonical simulation of an interface. Green atoms are fixed and impose a certain kind of boundary. Red atoms are “real”, or “normal”, atoms, while blue atoms are “virtual”, i.e., they are only partially present in the simulation. Whether a certain atom is real or virtual evolves naturally during the simulation. 186

Page 187

Chemistry and Material Sciences Exploratory Research Continuing Project Understanding and Controlling Magneto-electric Coupling in Multiferroic Materials Dmitry A. Yarotski 20130525ER Introduction tiferroic materials is poised to make a broad impact on The last few decades have seen the discovery of novel condensed matter physics and will open new directions classes of correlated electron materials with exotic prop- in complex materials research. erties that few would have imagined. Multiferroic mate- rials stand out among them, due to the ability to change Benefit to National Security Missions their magnetization and/or polarization state with either Our work directly addresses the Grand Scientific Chal- electric or magnetic fields. This has the potential to revo- lenges identified in the Basic Energy Sciences Advisory lutionize future energy, sensing and information technol- Committee (BESAC) report, which are central to DOE’s ogies, as multiferroic circuits will combine the low power missions in energy, science, and security in general, and consumption and speed of field-effect devices with the to the LANL Materials Grand Challenge in particular. The permanence of magnetic elements. The remarkable proposed integration of material synthesis, ultrafast opti- properties of multiferroics emerge from strong cou- cal techniques and forefront condensed matter theory pling between coexisting electric and magnetic orders. represents the LANL Materials Strategy and should pro- Despite recent progress in the synthesis and character- vide LANL, as well as CINT, with the capability to investi- ization of single-phase and heterostructured multiferroic gate emergent properties of complex materials through materials, the existing understanding of magnetoelectric observation of the dynamical behavior of relevant order (ME) coupling mechanisms is still controversial, and the parameters and through selective excitation of the low- dynamical properties of multiferroics remain practically energy modes responsible for material functionality. Our unexplored. In particular, the role of low-energy excita- thrust to interface materials science with ultrafast THz, tions (phonons, magnons, etc.) in the emergence of optical and X-Ray coherent photon probes represents multiferroic functionality has not been clarified. an essential element in the MaRIE strategy that con- nects the M4 facility to the Multi-Probe Diagnostic Hall. Our ultimate goal is to leverage the unmatched LANL Proposed experiments will provide critical understand- integration of material synthesis, advanced ultrafast ing of the mechanisms of magnetoelectric coupling and optical techniques and forefront solid-state theory to thus enable the design and synthesis of new multiferroic test the dynamic limits of ME phenomena and reveal the materials with controlled functionalities. Materials with mechanisms governing ME coupling in representative tunable and novel functionality are an enabling com- multiferroic materials. In pursuit of this goal, we will use ponent in the development of next-generation devices intense THz pulses to directly excite low-energy modes for sensing, information storage, and spintronics ap- and investigate their effect on magnetic and electric plications. We believe that our integrated capabilities in orders in each material. complex material design, synthesis and characterization will be of great interest to multiple sponsors, including A unique set of time-resolved optical and soft X-Ray DOE-BES, DOD, IC, and industry. probes will allow unambiguous separation of spin, charge and lattice dynamics, and in conjunction with Progress theoretical modeling will unveil the microscopic origin of Using our existing theoretical and experimental exper- ME coupling in multiferroics. This integration of “pas- tise in the physics of complex materials, we have made sive” and “active” broadband ultrafast optical methods substantial progress in the first year of this project. In with forefront modeling expertise is a unique capability particular, our experimental efforts have focused on at LANL. The application of this capability to study mul- 187

Page 188

elucidating the dynamic coupling between magnetic and tion, novel composite materials were fabricated where electric orders in a number of multiferroic materials by components are interfaced in three dimensions, unlike cur- means of broadband ultrafast spectroscopy. We have rently used one-dimensional layered structures. Such 3D used optical pump – Terahertz (THz) probe experiments to composite films are expected to improve magneto-electric investigate the dependence of magneto-electric response coupling due to better control of strain fields and proximity of Eu0.75Y0.25MnO3 multiferroic compound on tempera- effects. We will systematically study the effects of structure ture. These studies revealed drastic changes in the relax- and composition on the performance of these materials ation dynamics of photo-induced charge carriers when using conventional techniques and newly developed THz material passes through antiferromagnetic phase transition capability. near Neel temperature of 45K. Onset of magnetic order enhances the energy dissipation from hot carrier to other Experimental advances were matched by theory devel- degrees of freedom as indicated by reduction in relaxation opment aimed at explaining the data and directing the timescales from more than 1 nanosecond at room temper- experiment to discovering the mechanisms underpinning ature to ~500 ps below Neel temperature. This behavior multiferroic response. Specifically, we have carried out clearly reveals strong interaction between the electronic first-principles calculations of electronic/magnetic struc- and spin excitations, and should provide more quantitative ture and optical properties of bulk materials mentioned information about the underpinning mechanisms when above as well as new Bi2FeMnO6 crystals. The reveled ongoing theoretical simulations are completed. magnetic structure and its implications on ferroelectric/ lattice order helped to identify the signatures of multiple- One of the goals of our project is to reveal the role of order interactions that should be observable with ultrafast the low energy excitations in an emergence of magneto- THz, optical and X-Ray probes. Moreover, we have shown electric coupling. To address this goal, we have established that thinner films have larger magnetic moment than collaboration with Richard Averitt from Boston University. thicker ones thus indicating an importance of the strain We have used intense THz source available in his labora- effects. These results will provide a solid foundation for tory to selectively excite phonon/electromagnon modes in modeling more complex heterostructured materials where Eu0.75Y0.25MnO3 and TbMn2O5 multiferroic crystals, and the interfaces are used to control strain state and order to probe the ensuing time-averaged effects on magnetic/ coupling. In addition, the theory will focus more on mod- electric orders and coupling between them. However, eling spin response in X-Ray region to substantiate novel no discernible nonlinear responses were observed in the ultrafast coherent X-Ray probes of time-resolved magnetic THz electric fields up to 200 kV/cm. This result does not phenomena in multiferroic materials. imply the weakness of nonlinear effects in THz range but is simply the consequence of the limitations of an experi- Future Work mental apparatus where the changes in the short THz In the next year, we will base on our initial success and pulse are averaged over long periods of time. Dr. Averitt is continue the synergistic theory-synthesis-characterization currently working on introducing high temporal resolution effort aimed at elucidating the physics of multiferroic in the system, and we are already scheduled to conduct materials. We will complete the construction of intense time-resolved experiments that will allow direct observa- THz source which will provide sub-picosecond THz pulses tion of the phonon/electromagnom dynamics. Simultane- spanning 0.5-5 THz with electric fields up to 1 MV/cm. This ously, an experience gained during this work allowed us system will be coupled with THz/second harmonic/Kerr to accelerate the construction of intense THz source at rotation probes for the ultrafast dynamics of electric and LANL which will be completed by the end of this year. This magnetic orders. We have already demonstrated strong new capability will provide access to sub-picosecond THz interaction between the electronic and spin excitations in pulses spanning 0.5-5 THz with electric fields up to 1 MV/ representative Eu0.75Y0.25MnO3 compound using optical cm. Moreover, this system will be coupled with second pump-THz probe technique. Next, we will apply the new harmonic/Kerr rotation probes that will enable indepen- THz capability to this and TbMn2O5 materials to directly dent monitoring of the ultrafast dynamics of electric and excite electromagnon modes and reveal the mechanisms magnetic orders. underpinning spin-polarization coupling. Simultaneously, we plan to use ultrashort optical pulses to excite the elec- We have also been successful in synthesizing vertically tronic subsystem in Eu0.75Y0.25MnO3, and TbMn2O5, and aligned nanocomposite multiferroic films with nanomaze probe the timescales of spin order switching with time- structure. Our results suggest that nanomaze architecture resolved x-ray magnetic linear spectroscopy at Mn3+ edge. consist of epitaxial phases and presents a novel platform Similarly, we will investigate magnetic and electric order for creating tunable multiferroic functionalities. In addi- dynamics in BFO/LCMO bilayer films, which have shown 188

Page 189

a promise for building ‘artificial’ multiferroics. Moreover, recent studies have clearly shown that improved multifer- roic functionalities can be achieved by creating 3D com- posite films. We will systematically synthesize a variety of different vertically aligned nanocomposite films with dif- ferent chemical compositions, and apply our unique suite of ultrafast techniques guided by theoretical modeling to establish the process-structure-functionality relationship in the 3D composite BFO/LCMO thin films. Further theory de- velopment will aim at modeling the influence of strain on the magnetism in these compounds in a hope to uncover the mechanism for the enhancement of the magnetic mo- ment in the thinner films accompanied by the calculations of electronic, magnetic, and optical properties. Conclusion We expect to provide a better understanding of the dy- namics and origins of magnetoelectric (ME) coupling in multiferroic materials. We will answer three longstanding questions in the basic and applied science of ME sys- tems: 1) How fast can the magnetization/polarization be switched, and how can this be improved? 2) What low-en- ergy excitations are responsible for the strong interactions between electric and magnetic orders? 3) How can we manipulate these excitations to enhance the ME coupling and design better multiferroics? This will enable develop- ment of the basic principles of ME material design, with a significant impact on energy and information technologies. Publications Aguilar, R. Valdes, Y. M. Sheu, A. J. Taylor, R. P. Prasanku- mar, and D. A . Yarotski. Time resolved terahertz and second harmonic investigations in multiferroic RMnO3 and RMn2O5. Presented at 2013 March Meeting of The American Physical Society. (Baltimore, MD, 18-22 March, 2013). Chen, A. P., H. H. Zhou, Z. X. Bi, Y. Y. Zhu, Z. P. Luo, A. Bayraktaroglu, J. Phillips, E. M. Choi, J. L. MacManus- Driscoll, S. J. Pennycook, J. Narayan, Q. X. Jia, X. H. Zhang, and H. Y. Wang. A New Class of Room-Tem- perature Multiferroic Thin Films with Bismuth-Based Supercell Structure. 2013. ADVANCED MATERIALS. 25 (7): 1028. Chen, A. P., W. R. Zhang, F. Khatkatay, Q. Su, C. F. Tsai, L. Chen, Q. X. Jia, J. L. MacManus-Driscoll, and H. Wang. Magnetotransport properties of quasi-one-dimension- ally channeled vertically aligned heteroepitaxial nano- mazes. 2013. APPLIED PHYSICS LETTERS. 102 (9): -. Chen, A. P., Z. X. Bi, Q. X. Jia, and H. Y. Wang. Microstruc- ture, vertical strain control and tunable functionalities in self-assembled, vertically aligned nanocomposite thin films. 2013. ACTA MATERIALIA. 61 (8): 2783. 189

Page 190

Chemistry and Material Sciences Exploratory Research Continuing Project Phase Transitions at Extremes: Emergence of Topological Defects Vivien Zapf 20130601ER Introduction ces and other defects in phase transitions out of equi- Ever since an early metal worker plunged hot iron into librium. This is known as the Kibble Zurek mechanism. water and discovered quench-induced hardening, the We will focus specifically on magnetism, thus enhancing importance of speed in phase transitions was clear. A fundamental understanding of materials. These top- phase transitions is e.g. the process of transforming from ics fall directly within the purview of the DOE Office of solid to liquid or from a non-magnetic to a magnetic Science. However this theory applies to all phase transi- state. Phase transitions traversed at extreme speeds are tions that are traversed rapidly and out of equlibrium. central to many LANL missions. First order transitions We will be testing the central quantitative prediction of (which are abrupt and have latent heat,) are well under- the Kibble Zurek mechanism, and thus we will be making stood and phenomena such as nucleation and anneal- contributions to other fields in which out-of-equilbrium ing are the basis of vast industries. Second order phase phase transitions are studied including ferroelectrics, transitions at extreme speeds (which are continuous and charge-density waves, superconductors, cold atoms, and have no latent heat) can also produce profound effects even cosmology. The Kibble Zurek Mechanism was first on materials. However, our understanding of second proposed in cosmology to explailn the formation of the order transitions is incomplete. The Kibble-Zurek Mecha- early universe. nism (KZM) was developed starting in 1985, largely at LANL, and describes how rapid quenching can “freeze in” It applies to information science and technolgy because fluctuations in second order phase transitions, creating topological defects are a new frontier for quantum com- defects such as domain walls, vortices, etc. These defects puting, and even in a classical context can be used to are long-lived fossils of transition dynamics, revealing store and manipulate date. Understanding how to create nature of the fluctuations and the symmetry breaking and manipulate vortices, both wanted and unwanted process. KZM applies to second order phase transitions vortices, is thus important for fields directly impacting of interest to LANL and DOE from cosmology to magnets information technology including spintronics, quantum to superconductors. computing, and computer memories. Recent breakthroughs allow us to test KZM in materials Progress with practical importance. We will study KZM in magnets Success in exploring the Kibble-Zurek mechanism (KZM) for the first time in theory and experiment. We innovate as outlined in our proposal requires four ingredients: by using magnetic fields to induce phase transitions with appropriate compounds, variable-pulse speed magnetic up to eleven orders of magnitude of quench rate. Thus, fields, the ability to measure the resultant topological for the first time, we can test key quantitative prediction defects, and detailed modeling of the compounds. We of KZM: the dependence of defect density on quench describe outline the progress in these four areas below. rate. We build on our expertise in a new class of mag- nets that have the right symmetry to allow spin vortices Appropriate compounds and other topological defects, testing KZM in both ther- We previously suggested NiCl2-4SC(NH2)2 as an excel- mal and quantum phase transitions. lent candidate compound, and we are conducting exper- iments now. However, we are also exploring the physical Benefit to National Security Missions properties of several new magnetic compounds that are This project seeks to understand the formation of vorti- candidates for observing the Kibble-Zurek mechanism. 190

Page 191

These include CoCl2-4SC(NH2)2, Ca3CoMnO6, REInO3 (RE class as Ising model in one dimension. Zurek has also co- = rare earth), Ho2CuTiO6 and Er2CuTiO6, and Ni3TeO6. authored a review paper with del Campo and Kibble on the Three papers are in progress on this preliminary character- interplay between inhomogeneity of the system and KZM. ization and Zapf and Jaime are finalizing a Rev. Mod. Phys. on quantum magnets exhibiting Bose-Einstein Condensa- Future Work tion, which are the prime candidates for KZM. In Fiscal year 2014 we will complete our detailed theory of how the Kibble Zurek Mechanism applies to XY antiferro- Variable pulse speeds magnetism for quantum phase transitions in two and three We have developed new pulse algorithms for our unique dimensions, and the physical properties being studied 60 T shaped-pulse magnet to achieve the widest possible experimentally in this ER (see below). Particular attention range of pulse speeds, and tested these on Ca3CoMnO6. will be paid to the quantum aspects of the dynamics of This magnet will be the main magnet in which the final phase transitions. tests of the Kibble-Zurek mechanism will be conducted. On the experimental side we will complete the develop- Measurements of the KZM effect ment of a magnetostriction measurement method that is Our direct measurements of the KZM effect are stalled for compatible with our samples (an unexpected addition to the moment because our optical fiber grating magneto- our work). Then we will measure the magnetostriction, the striction method is not effective at measuring the magne- dielectric constant and the ac susceptibility measurements tostriction in the samples most relevant to this ER due to in at variable pulsed speeds for the compound NiCl2- mismatched thermal expansion. We are also learning that 4SC(NH2)2, and then for analog compounds that order magnetostriction measurements transverse to the applied magnetically even in zero field, allowing us to pulse the magnetic field are optimal for testing KZM and the optical magnetic field first and then measure afterwards in zero fiber system does not have the ability to measure trans- field. We will continue exploring new antiferromagnets verse magnetostriction in many of our magnets. Therefore that are candidates for our KZM investigations due to their we are taking a different tack and developing a method XY antiferromagnetism and magneto-electric coupling. of measuring magnetostriction that uses a piezo-electric cantilever to track the length changes of the sample. This Conclusion method will be more robust for a wide range of samples, We expect to test the central prediction of the Kibble Zurek and can be oriented in any direction relative to the applied Mechanism that topological defects, like vortices, can be magnetic field. Initial results strongly suggest that it is at induced by rapidly quenching through 2nd order phase least as sensitive as the optical fiber system, and the first transitions. We will study magnetism and perform the first measurements on our candidate systems are commencing detailed quantitative test of the power-law scaling of vor- this summer and fall. tex density with quench speed. We will develop the theory for magnets and quantum magnets, and perform experi- Theoretical Modeling ments in the unpredented range of magnetic field quench We have investigated formation of topological defects fol- speeds at the National High Magnetic Field Laboratory. lowing a phase transition in several models relevant to the ongoing experiments. These include formation of topologi- Publications cal defects in a quench from a Mott insulator to superfluid, Campo, A. Del, and W. Zurek. Universality of phase transi- which is most relevant to NiCl2-4SC(NH2)2. Two papers on tion dynamics: topological defects from symmetry this subject have been published. We have also developed breaking. To appear in 80th Birthday Symposium for a simple model that accounts for the winding number in Tom Kibble. two-dimensional systems through the correlations be- tween pairs of topological topological defects. This has led Campo, A. del, T. Kibble, and W. Zurek. Causality and non- to re-interpretations of the data of the experiments on equilibrium second-order phase transitions in inhomo- tunnel Josephson junctions (and brought them into agree- geneous systems. 2013. J. Phys. Condensed Matter. 25: ment with the Kibble-Zurek mechanism). Furthermore, 404210 . Zurek and del Campo are involved in a paper that tests Matlashov, A. N., L. J. Schultz, M. A. Espy, R. H. Kraus, I. M. KZM in ion traps, where chains of ions (“Coulomb crystals”) Savukov, P. L. Volegov, and C. J. Wurden. SQUIDs vs. can buckle (creating a “zigzag”) as a result of a change of Induction Coils for Ultra-Low Field Nuclear Magnetic the external potential. Topological defects (places where Resonance: Experimental and Simulation Comparison. a “zig” is followed by another “zig” rather than a “zag”) 2011. IEEE TRANSACTIONS ON APPLIED SUPERCON- form in a phase transition that is in the same universality DUCTIVITY. 21 (3): 465. 191

Page 192

Mun, E. D., M. Frontzek, A. Podlesnyak, G. Ehlers, S. Barilo, S. V. Shiryaev, and V. S. Zapf. High magnetic field evolu- tion of ferroelectricity in CuCrO2. Physical Review B. Oh, Y. S., S. Artyukhin, J. J. Yang, V. S. Zapf, J. W. Kim, D. Vanderbilt, and S. W. Cheong. Colossal magnetoelec- tricity in a collinear antiferromagnet. Nature Commu- nications. Pyka, K., J. Keller, H. L. Partner, R. Nigmatullin, T. Burger- meister, D. M. Meier, K. Kuhlmann, A. Retzker, M. B. Plenio, W. H. Zurek, A. del Campo, and T. E. Mehl- staubler. Topological defect formation and spontane- ous symmetry breaking in ion Coulomb crystals. 2013. NATURE COMMUNICATIONS. 4: -. Torrontegui, E., S. Ibanez, S. Martinez-Garaot, M. Modug- no, A. del Campo, D. Guery-Odelin, A. Ruschhaupt, X. Chen, and J. G. Muga. Shortcuts to Adiabaticity. 2013. ADVANCES IN ATOMIC, MOLECULAR, AND OPTICAL PHYSICS, VOL 62. 62: 117. Tsyrulin, N., C. D. Batista, V. S. Zapf, M. Jaime, B. R. Hansen, C. Niedermayer, K. C. Rule, K. Habicht, K. Prokes, K. Kiefer, E. Ressouche, A. Paduan, and M. Kenzelmann. Neutron study of the magnetism in NiCl2 center dot 4SC(NH2)(2). 2013. JOURNAL OF PHYSICS-CONDENSED MATTER. 25 (21): -. Wulf, E., D. Huvonen, J. W. Kim, A. Paduan-Filho, E. Res- souche, S. Gvasaliya, V. S. Zapf, and A. Zheludev. Criti- cality in a disordered quantum antiferromagnet by neutron diffraction. Physical Review B. Zapf, V. S., M. Jaime, and C. D. Batista. Bose-Einstein Con- densation in Quantum Magnets. Reviews of Modern Physics. Zurek, W. H.. Topological relics of symmetry breaking: winding numbers and scaling tilts from random vor- tex–antivortex pairs. 2013. J. Phys.: Condens. Matter. 25: 404209. 192

Page 193

Chemistry and Material Sciences Exploratory Research Continuing Project Redox active Catalysts for C-C Coupling Reactions Relevant to Renewable Energy John C. Gordon 20130672ER Introduction Progress An effort to upgrade derivatives of glycerol (essentially a The concomitant rise in global energy demands and waste molecule produced on a huge scale as a result of growing concerns over climate change have spurred biodiesel production) for the synthesis of fuels precur- interest in the development of alternative energy sors and higher value chemicals, and to do it using well- sources as well as less expensive synthetic methods for defined, inexpensive catalysts based on earth abundant high-value chemical feedstocks. The development of metals is a desirable goal. The research outlined herein novel methods for biomass derived molecule conversion targets the development of inexpensive catalysts, com- is a burgeoning area. We are specifically interested in prised of earth abundant metals, which are capable of approaches for the reductive coupling of three-carbon such biomass derived molecule upgrading. containing ketonic molecules derived from glycerol into six (or higher)-carbon containing molecules as this rep- Precious metals are commonly used in catalysis due to resents a potential route for both fuel and fine-chemical their ability to promote two-electron processes, such as synthesis from readily-available and inexpensive precur- oxidative addition reactions. In contrast however, 1st- sors. row metals tend to undergo one-electron redox changes or promote radical reactions, which has limited their Our approach to the homocoupling of glycerol deriva- more general use in catalytic chemistries. The proposed tives involves the use of earth-abundant, first-row transi- work will attempt to use redox active ligands capable of tion metals (e.g. iron) supported by redox active ligands. conferring multi-electron redox behavior on cheap and While first-row transition metals typically undergo abundant metals, thus potentially applicable to upgrad- one-electron redox changes, the coordination of redox ing glycerol based derivatives. active ligands to metals such as iron, facilitates the multi- electron redox changes that are required to promote the Benefit to National Security Missions desired coupling reactions. It is envisioned that the stor- age of extra reducing equivalents will more easily permit The proposed work will advance our understanding of the reductive coupling of ketonic substrates. the chemistry and chemical methods that directly bear upon energy security. Success in this area would seed To this end, we have already synthesized a series of the development of future approaches to more efficient both previously published as well as new iron complexes chemistries, potentially providing a bridge from basic containing redox-active ligands and have begun prelimi- science to global scale energy solutions. nary reactivity studies designed to establish the abil- ity of reduced iron complexes to affect the coupling of The LDRD funds requested for this project will provide ketones. The most promising complex thus far is a new an opportunity for our team to collect results in this [FeII(BIAN)Cl2] complex. [FeII(BIAN)Cl2] contains an iron high profile area of renewable energy. Results from this center supported by a bis(aryl)acenapthenequinonedii- exploratory work should prove invaluable in obtaining mine (BIAN) ligand. X-ray crystallographic characteriza- potential future funding from organizations such as DOE tion of this molecule reveals a tetrahedral geometry Office of Basic Energy Sciences (Catalysis Science) and around the iron center, and magnometry and Mössbauer DOE Office of Energy Efficiency and Renewnable Energy. data confirm the divalent oxidation state of iron within this molecule. Cyclic voltammetry reveals two succes- sive one-electron reductive processes; comparison with 193

Page 194

the cyclic voltammetry data of the BIAN ligand suggests that the first reduction occurs on the iron center while the second reduction is largely ligand-based. Exposure of [FeII(BIAN)Cl2] to chemical reductants in the presence of either cyclooctene (cod) or toluene results in the the formation of [Fe(BIAN)(cod)] and [Fe(BIAN)(tolu- ene)] species, respectively. Both species have been char- acterized via NMR spectroscopy, and the structure of the [Fe(BIAN)(toluene)] species has been confirmed via X-ray crystallography. Examination of both reduced iron species via spectroscopic and magnetic techniques is planned for the future. Computational studies aimed at elucidating their electronic structures are also underway. Finally, expo- sure of the [Fe(BIAN)(cod)] complex to a slight excess of ke- tones including benzophenone and acetophenone results in rapid formation of the anticipated coupled products as confirmed by 13C-NMR spectroscopy. Further efforts will attempt to extend this reactivity to both aldehydes and biomass-derived substrates. Future Work We will computationally screen existing ligands suitable for desired redox chemistries and attempt to make iron com- plexes supported by these. We will also design, prepare and synthesize other promising new ligand candidates. If successful in new ligand synthesis, we will also proceed to synthesize and characterize new iron complexes, and begin to test their redox chemistries with standard organic sub- strates that will serve as models for the biomass derived molecules. Conclusion Developing earth abundant metal-catalyzed processes to affect the low cost conversion of biomass derived feed- stocks remains a major challenge in the field of energy related chemistry. In part this is due to the inability of first row transition metals to participate in multi-electron redox processes. Success in this project would not only lead to a more fundamental mechanistic understanding of C-C coupling reactions with traditional organic substrates using these classes of metals, but also has the potential, if suc- cessful, to convert species derived from a large scale waste by-product of biodiesel (glycerol) into potentially useful fuel precursors and other chemicals. 194

Page 195

Chemistry and Material Sciences Exploratory Research Continuing Project Novel Chemical Architectures for Supercapacitor Electrolytes: Comparing In Situ Scattering Measurements to Theory and Simulation Cynthia F. Welch 20130681ER Introduction that quickly capture and release stored energy. If their We aim to increase the energy storage density in su- energy densities can be improved, given that superca- percapacitors, a family of devices well suited to the pacitors may deliver an order of magnitude more power intermittent power generation associated with many than chemical batteries, they may dramatically improve renewable sources. We propose that tailoring the size the performance of dynamic energy capture and reuse and shape of the electrolyte in supercapacitors to match systems. Our project aims to improve the energy densi- the electrode surface geometry will lead to enhanced ties of supercapacitors by tailoring the size and shape of energy densities, forming our central hypothesis. This the electrolyte to match the electrode surface geometry. hypothesis leads us to consider several nano-structured Specifically, we will explore certain dendritic chemis- organic electrolytes (NOEs) as candidates for use in tries, which possess all of the key properties needed in supercapacitors, as well as to study their behavior near a good electrolyte. In addition, several aspects of their and in the electrode pores. The novelty of our project physical chemistry recommend them for application in lies in optimizing the electrolyte rather than the elec- supercapacitors. These characteristics include precisely trode. Many key questions about the behavior of the controlled size and shape, broad range of available sizes, electrolytes in the harsh environments near electrodes high charge densities, low viscosities at relevant concen- must be answered for progress to be made. Our access trations, and the ability to capture impurities. to unique electrochemical sample cells that allow us to probe the in situ behavior of the electrolyte near elec- Progress trode surfaces and our recent theoretical advances for Our goals for Year 1 of the project are as follows: the proposed electrolyte class provide the key enabling • Build molecular dynamics (MD) models of charged developments. dendrimers in a salt-water solution. Study both single molecule and blend behaviors at various Thus, we have four major goals: relative concentrations and generations of growth.

1. Produce multi-scale descriptions of the NOE materi- Carry out the quantum chemical stability study to als in device settings introduce relevant fragmentation species into the
2. Discover the physical principles directing their spatial MD simulations. distribution in heterogeneous media • Experimentally validate both the MD and the chemi-
3. Explore their non-equilibrium route to this distribu- cal stability results via chromatography (gel perme- tion upon charging/discharging the device ation), spectroscopy (NMR, mass spec) and scatter- ing (light, small-angle neutron).
4. Encapsulate in quantitative expressions the physics dictating their performance • Use our new theory to extract an effective step length. This information will be used (in Year 2) to construct a field theoretical model to predict the lo- Benefit to National Security Missions calization of the individual dendrimers onto various Many processes involved in manufacturing and trans- pore surface geometries. porting goods suffer from inefficiencies due to wasted • Take the intra-molecular dimensions from the MD energy. Similarly, many renewable energy sources (such results to construct a coarse grained model. as wind and solar) yield only intermittent electricity generation. We therefore require advanced devices 195

Page 196

Over the past nine months, we have made significant prog- dielectric constant. Each of these parameters has spanned ress on each of these goals as well as on one of the goals values most relevant to our target application, supercapac- for Year 2. Here are more details regarding our progress itor devices. We are in the process of writing a paper that toward each of these goals: details the results of these studies, which suggest some interesting phenomena that may be unique to electrolytes We have constructed MD models of poly(propylene imine) with dendritic architectures. We will also build on the (PPI) dendrimers and carried out some single molecule results of these studies with additional BD simulations in studies. We have also begun the quantum chemical stabil- Year 2, focusing on different pore geometries. Our Institu- ity studies. MD models of poly(amido amine) (PAMAM) tional Computing allocation has been, and will continue to dendrimers will be constructed and studied over the next be, used for these studies. 3 months. These MD studies are computationally very expensive; therefore, we applied for LANL Institutional Future Work Computing time and have now been granted time on • Carry out Brownian dynamics (BD) simulations on the several Institutional Computing platforms over the next coarse grained models within the Debye-Hückel and 2 years. Part of our allocation will be utilized to complete implicit solvent approximations. Examine the behavior these atomistic MD studies, which will likely continue into of single and multiple molecules in the neighborhood Year 2 of the project. of surfaces with different pore geometries. The sur- faces will be modeled as static particles arranged in We have purchased several generations of poly(amido the desired topologies. Examine the behavior of blends amine) (PAMAM) dendrimers and begun the chemi- of nano-structured organic electrolytes (NOEs) in the cal stability studies. Characterization by gel permeation presence of electric fields and model (flat) electrode chromatography and NMR spectroscopy will be completed surfaces. Apply an external electric field and track the by the end of Year 1. Light scattering studies have been at- phase separation of the two species. tempted but were unsuccessful due to the small size of the • Using the information on the effective step length dendrimers, which is at the lower limit of detection for the obtained in Year 1, construct a field theoretical model instrument used. An alternate light scattering instrument using the Edwards path integral formulation to predict has been identified, and we will use this instrument over the localization of the individual dendrimers onto vari- the next 3 months to make another attempt at charac- ous pore surface geometries. Extend this approach to terization by light scattering. A beam time proposal for multiple molecules, solving for concentration within small-angle neutron scattering studies was submitted and, the pore. Compare the results with the CGM modeling if granted, will be executed at the end of Year 1 or begin- (1 above). ning of Year 2, depending on neutron beam availability and scheduling constraints. • Devise an extension of the Ginzburg-Landau theory for polymer blends to account for blends of the NOEs in One of us (P. Welch) has written a paper that details the solution and in the presence of an electric field. process of extracting an effective step length for den- • Experimentally challenge the simulations with X-ray drimers. This parameter will be used in Year 2 to construct and neutron reflectivity experiments of dendrimers at a field theoretical model to predict electrolyte / electrode flat electrode surfaces. interactions. The paper has recently received an LA-UR number and will be submitted within the next few weeks Conclusion to a peer-reviewed journal. This project will produce: 1) A new set of theories for predicting when specific NOEs will partition within pores Coarse-grained models of PAMAM and PPI have been con- of various geometries; 2) A new theory for predicting how structed. In addition, Brownian dynamic (BD) simulations the environmental parameters dictate the structure of the of these coarse-grained model dendrimers between two multilayer arrangement of the NOEs near the electrode; flat, oppositely charged electrodes have begun, which was 3) A new theory detailing the dependence of capacitance a goal originally slated for Year 2. These simulations have on the relationship between pore geometry and electro- the benefit of being less computationally expensive than lyte shape; 4) A multi-scale simulation model of a class of the MD studies of (1) above. Therefore, we have been able supercapacitor devices that may be used as the founda- to explore a large parameter space in the first nine months tion for a future program at LANL; 5) New insight into the of the project. Specifically, we have examined the effects of behavior of NOEs in complex environments. dendrimer concentration, dendrimer generation of growth (or size), electrode surface charge density, and solvent 196

Page 197

Publications Welch, P. M.. The escape of a charged dendrimer from an oppositely charged planar surface. To appear in Journal of Chemical Physics. Welch, P. M., C. F. Welch, and N. J. Henson. The flattening of dendrimers from solutions onto charged surfaces. ACS Macro Letters. 197

Page 198

Chemistry and Material Sciences Exploratory Research Final Report Superconducting Vortices in Magnetic Media Boris A. Maiorov 20110138ER Abstract netic systems in complex experimental conditions such As electricity demand continues to rise, and environ- as pulsed magnetic fields. mental constraints become more relevant, it is impera- Background and Research Objectives tive to enhance the performance of superconductors for applications oriented to energy efficiency and storage. As electricity demand continues to rise, and environ- Superconductors have great potential in transmission mental constraints become more relevant, it is impera- cables, high field magnets as well as in magnetic storage tive to enhance the performance of superconductors systems. For that they must be able to carry large electri- for applications oriented towards energy efficiency and cal currents without dissipation. The maximum dissipa- storage. In most applications, the superconductors are tionless current, called critical current density Jc, is set exposed to relatively high magnetic fields (H) that pen- by the pining force exerted by different types of defects etrate the material by creating superconducting vortices. on vortices, flux-lines that appear in type-II supercon- Any electric current that flows through the supercon- ductors. Magnetic defects can lead to the so-called mag- ductor exerts a Lorentz force (FL) on the vortex that is netic pinning. During this project we studied the effect proportional to the magnitude of the current. When the of a magnetic medium on the pinning and dynamics of current density, J, reaches a critical value, Jc, FL becomes the vortex state. strong enough to overcome the pinning force that anchors vortices. Vortex motion leads to energy dissipa- We developed new experimental and theoretical tools tion, i.e., the material acquires a finite resistivity for J>Jc. that allowed us to understand the interplay between The work in this project was aimed to find ways to re- superconducting vortices and magnetic media. To un- duce dissipation using magnetic pinning related routes. equivocally study the effect of the magnetic medium we studied system where the magnetic ordering transition Standard vortex core pinning comes from material takes place inside the superconducting phase, namely defects that create an attractive vortex potential. In par- ErNi2B2C and ThNi2B2C with Tc~11K and TN=6 and 2K ticular, magnetic defects can lead to so-called magnetic respectively with Tc and TN being the superconducting pinning. In principle, magnetic pinning can significantly and Néel transition temperature respectively. increase the value of Jc (in particular at high tempera- tures) as it can result in pinning energies larger than We calculated and measure the magnetic influence of those achieved by non-magnetic defects (core pinning). different magnetic phases on Jc. Among relevant results, However, the interplay between vortices and magnetic we found a Jc increase associated with TN in ErNi2B2C. media cannot be properly exploited without a compre- We analyzed the nature of the pinning enhancement as hensive study that includes both static and dynamical well as its origin in magnetic domains. Theoretically, a properties of the entire system. Thus, we centered our new model used to explain the already found increase in studies on the effect of a magnetic medium on the dy- Jc at a ferrimagnetic low-temperature phase in ErNi2B2C namics and pinning of the vortex state. was developed. A common problem in previous studies of magnetic The technical and theoretical advances made in this pinning is the inability of separating the core (local project are available to other areas of condensed matter suppression of the superconducting order parameter) physics in the Laboratory. The capabilities developed are and magnetic contributions. To unequivocally study the being used to explore other superconducting and mag- effect of the magnetic medium, it is necessary to be able 198

Page 199

to induce substantial changes in the magnetic properties. phase and TmNi2B2C that also has an Antiferromagnetic For this purpose, we studied different systems where the transition but at lower temperature and with a different magnetic ordering transition takes place inside the super- crystallographic orientation. A manuscript is in prepara- conducting phase. These studies have never been done tion. before. In parallel studies of the dynamic of vortices in parallel Scientific Approach and Accomplishments potential has allowed us to find a universal description including copper- and iron-based superconductors. Be- The idea is to explore the interaction between supercon- sides de minimum in the N values where E=C JN previously ducting vortices and magnetism. This project had two found we found a new regime that can be associated to components, a theoretical and experimental. Due to the the lock-in angle, this regime is characterized as maximum extremely challenging objectives of passing high amounts in N. Field dependence can be fitted reasonable well with of currents to observe the fast vortex movement, extreme- models too. Interestingly the same qualitative behavior is ly clean single crystals were needed and sample fabrica- found in iron based superconductors, measured in collabo- tion down to microns sized (without contamination or ration with IFW (Dresden). A manuscript is in preparation. crystalline degradation) was needed. ErB2Ni2C, TmB2Ni2C and LuB2Ni2C single crystals were obtained from Ames Lastly, our measurements in micro-bridges in clean sam- Lab, (Dr. Sergey Bud’ko, [email protected]). ErNiBC has ples have allow us to tailor superficial pinning by cutting several magnetic phases that coexist with superconductiv- some samples at arbitrary angles we were able to ‘gener- ity and LuB2Ni2C is non magnetic. We will compare and ate’ new maxima in Jc. contrast the results obtained in these single crystals. For achieving the required current densities we were able to We have studied the vortex dynamics in ferromagnetic su- fabricate micro bridges to inject high current densities in perconductors and shown that vortices attract each other these very clean crystals. This was done using Focused Ion at large distances due to polarization of the magnetic mo- Beam (FIB) at CINT facilities in Sandia (Figure 1). A separate ments. [2] Vortex clusters are then stabilized in the ground proposal was submitted to CINT for user facility access. state for a low vortex density. The motion of vortex clusters This proposal was approved and renewed. We also devel- driven by the Lorentz force excites magnons. These ma- oped photolithography masks that allow for short connect- gnon excitations become unstable at a threshold veloc- ing wires (these are made of Pt deposited in-situ by FIB-ID) ity above which domain walls are generated for a slow in order to reduce contact heating. The resulting micro- relaxation of the magnetic moments. The domain walls bridges show extremely sharp Tc (narrower than that of modulate the vortex configuration. This underlying dynam- bulk single crystals) as well as neither reduction of Tc nor ics of vortices and magnetic moments can be probed by increase of resistivity. This achievement solves tremen- transport measurements. We also discuss a novel way to dous efforts to reduce the cross section of single crystals measure the spectrum of magnetic excitation in magnetic that we made during this project. materials using motion of vortex lattice based on the pres- ent study [3]. We performed measurement of linear and non-linear elec- trical transport. In the superconducting phase the electric We also found a new type of vortex pinning by enhancing field (E=V/d V=voltage and d=distance) that indicates vor- the viscosity of vortex in magnetic superconductors with tex movement, as a function of current density amplitude long relaxation time of magnetization and large magnetic (J), which measures the force that acts on the vortices. susceptibility. [4] In the absence of current, vortices are dressed by nonuniform magnetic polarization and form We were able to find by performing transport and mag- vortex-polarons. Under a small current and consequently netization measurements of the critical current density Jc low Lorentz force, the magnetic polarization follows the in ErNi2B2C single crystals that show strongly enhanced vortex motion.(Figure 3) However, at long magnetic relax- vortex pinning at the Néel temperature TN and low ap- ation time of magnetization, there is additional dragging plied fields. The height of the observed Jc peak decreases force by the magnetization besides the Bardeen-Stephen with increasing magnetic field in clear contrast with that one, thus the effective viscosity of vortex is significantly of the peak effect found at the upper critical field. We also enhanced resulting in suppression of dissipation. For a performed angular transport measurements of Jc on this large current, the magnetic polarization cannot follow the compound. They reveal the correlated nature of this pin- vortex motion and the vortex-polaron dissociates, i.e. the ning enhancement, which we attribute to the formation of magnetization and vortex become decoupled. In the IV antiphase boundaries at TN. [1]. (Figure 2) characteristic, the decoupling transition shows as a volt- age jump and can be identified as a depinning transition. This also lead to explore the lower temperature magnetic 199

Page 200

The polaronic pinning mechanism successfully explains the observed enhancement of critical current in the ErNiBC su- perconductor at low temperatures. The polaronic pinning can be optimized in magnetic-superconducting multilayers. [5] We show also that vortex-polaron creep is suppressed at low temperatures. [ 6,7] Impact on National Missions The work achieved in this project is related to the Labora- tory’s Energy Security mission and Materials for the Future science pillar. The knowledge and technical advances produced during this three-year project will aid the LANL materials community in devising ways to harness the func- tionality provided by a superconducting state in order to address applied energy problems. Figure 2. (a) Transport measurements of Jc for H || c and θ = 30◦ . A new peak in Jc is observed at T = TN for H || c. The sketch gives the measurement geometry. (b) Contour plot of Jc as a function of temperature and magnetic field H || c showing the field range where the peak at T = T_N is present. Figure 1. Scanning Electron Microscope image of a micro-bridge of ErBNiC fabricate using Focused Ion Beam. The inset shows the superconducting transition that is extremely sharp and has not suffered any degradation during the fabrication process. (the colors were added for visualization purpose). Figure 3. Schematic view of the vortex lattice in the presence of free Ising magnetic moments along the a axis. The vortex lattice is tilted from the applied magnetic fields in the ac plane due to the polarization of the magnetic moments. The vertical columns show the vortex cores. The polarized magnetic moments are nonuniform in space due to the spatial modulation of the vortex lattice magnetic field. Due to the Lorentz force vortices move along the x axis. In moving lattice, there is a phase shift between the magnetic induction (dashed line) associated with the vortex lattice and the magnetization (solid line) caused by the retarda- tion in the response of magnetic moments to the vortex magnetic field. 200

Page 201

References Vortices and a Dissociation Depinning Transition in Magnetic Superconductors: The Example of ErNi2B2C.

1. Weigand, M., L. Civale, F. J. Baca, Jeehoon Kim, S. L.
2. Physical Review Letters. 102 (2): 027001. Bud’ko, P. C. Canfield, and B. Maiorov. Strong enhance- ment of the critical current at the antiferromagnetic Bulaevskii, L. N., and S. –Z. Lin. Prediction of Polaronlike transition in ErNi2B2C single crystals. 2013. Phys. Rev. Vortices and a Dissociation Depinning Transition in B. 87: 140506(R). Magnetic Superconductors: The Example of ErNi2B2C.
3. Phys. Rev. Lett.. 109: 027001.
4. Lin, S. –Z., L. N. Bulaevskii, and C. D. Batista. Vortex dynamics in ferromagnetic superconductors: Vortex Lin, S. Z., and L. N. Bulaesvskii. Enhancement of critical cur- clusters, domain walls, and enhanced viscosity. 2012. rent density in superconducting-magnetic multi-layers Phys. Rev. B. 86: 180506. with slow magnetic relaxation dynamics and large magnetic susceptibility. 2012. Physical Review B. 86
5. Lin, S. –Z., and L. N. Bulaevskii. Measuring spectrum of (6): 064523. spin wave using vortex dynamics. 2012. Phys. Rev. B. Lin, S. Z., and L. N. Bulaevskii. Measuring spectrum of spin 85: 134508. wave using vortex dynamics. 2012. Physical Review B. 85 (13): 134508 .
6. Bulaevskii, L. N., and S. –Z. Lin. Prediction of Polaron- like Vortices and a Dissociation Depinning Transition in Lin, S. –Z., C. Reichhardt, and A. Saxena. Manipulation Magnetic Superconductors: The Example of ErNi2B2C. of skyrmions in nanodisks with a current pulse and
7. Phys. Rev. Lett.. 109: 027001. skyrmion rectifier. 2013. Applied Physics Letters. 102:
9. Lin, S. –Z., and L. N. Bulaevskii. Enhancement of critical current density in superconducting/magnetic multilay- Lin, S. –Z., L. N. Bulaevskii, and C. D. Batista. Vortex dynam- ers with slow magnetic relaxation dynamics and large ics in ferromagnetic superconductors: Vortex clusters, magnetic susceptibility. 2012. Phys. Rev. B. 86: 064523. domain walls, and enhanced viscosity. 2012. Phys. Rev. B. 86: 180506.
10. N.Bulaevskii, L., and S. –Z. Lin. Self-induced pinning of Lin, S. –Z., and L. N. Bulaevskii. Quantum motion and level vortices in the presence of ac driving force in magnetic quantization of a skyrmion in a pinning potential in chi- superconductors. 2012. Phys. Rev. B. 86: 224513. ral magnets. 2013. Applied Physics Letters. 88: 060404.
11. Bulaevskii, L. N., and S. -Z. Lin. Polaron-like vortices, N.Bulaevskii, L., and S. –Z. Lin. Self-induced pinning of vor- dissociation transition and self induced pinning in mag- tices in the presence of ac driving force in magnetic netic superconductors. 2013. JETP. 144: 475. superconductors. 2012. Phys. Rev. B. 86: 224513.
12. Lin, S. –Z., C. Reichhardt, and A. Saxena. Manipulation Solovyov, V. F., Q. Li, W. Si, B. Maiorov, T. J. Haugan, J. L. of skyrmions in nanodisks with a current pulse and MacManus-Driscoll, H. Yao, Q. X. Jia, and E. D. Specht. skyrmion rectifier. 2013. Applied Physics Letters. 102: Influence of defect-induced biaxial strain on flux pin-
13. ning in thick YBa2Cu3O7 layers. 2012. Physical Review B. 86 (09): 094511.
14. Lin, S. –Z., and L. N. Bulaevskii. Quantum motion and Weigand, M., L. Civale, F. J. Baca, Jeehoon Kim, S. L. Bud’ko, level quantization of a skyrmion in a pinning potential P. C. Canfield, and B. Maiorov. Strong enhancement in chiral magnets. 2013. Applied Physics Letters. 88: of the critical current at the antiferromagnetic transi-

tion in ErNi2B2C single crystals. 2013. Phys. Rev. B. 87: 140506(R). 10. Solovyov, V. F., Q. Li, W. Si, B. Maiorov, T. J. Haugan, J. L. MacManus-Driscoll, H. Yao, Q. X. Jia, and E. D. Specht. Influence of defect-induced biaxial strain on flux pin- ning in thick YBa\\\{\\\}_\\\{2\\\}Cu\\\{\\\}_\\\{3\\\}O\\\{\\\}_\\\{7\\\} layers. 2012. Phys. Rev. B. 86: 094511. Publications Bulaevskii, L. N., and S. -Z. Lin. Polaron-like vortices, disso- ciation transition and self induced pinning in magnetic superconductors. 2013. JETP. 144: 475. Bulaevskii, L. N., and S. Z. Lin. Prediction of Polaronlike 201

Page 202

Chemistry and Material Sciences Exploratory Research Final Report Novel Anti-Perovskite Electrolytes for Superionic Lithium Transport Luc L. Daemen 20110139ER Abstract effect on the transportation battery market. We proposed a novel class of superionic solid electrolyte The project sought support to further develop our newly made of lithium rich anti-perovskites (LiRAP) to work invented class of solid Li+ electrolytes for high perfor- with metallic Li-anode and readily rechargeable cath- mance batteries. Lithium-rich Li3OCl, Li3O(Cl,Br) and odes. LiRAP solid electrolytes conduct Li+ ions well at related anti-perovskites Li3-x-Mx/2O(A1-zA’z)1 (where A high voltage and high current, providing much enhanced and A’ are halogens at the dodecahedral site and M is a energy density and power capacity for the solid-state divalent metal doped at the octahedral vertex) are novel battery. The new materials can form rather adaptable compounds that have been invented recently by the solid solutions with immense Li+ vacancies and lattice PIs (US & International patent pending; Zhao et al.) and imperfections in the crystal lattice for fast ionic trans- have demonstrated 3D superionic conductivity, a broad porting with low energy barriers. The Li+ mobility of the working window, economic viability, and environmental LIRAP increases to advanced superionic conduction as friendliness in preliminary studies. These new materials soft-phonon driven structural phase transitions occur at can form rather adaptable solid solutions with immense high temperatures. It overcomes current fluid electrolyte Li+ vacancies and lattice imperfections for fast ionic shortcomings of toxic leakage and short circuit burning transport. The novel anti-perovskites can accommodate and offers a much improved electrochemistry safety a large number of mobile Li+ ions in the crystal lattice solution. with a low energy barrier for ionic hopping. The ionic mobility increases to advanced superionic conduction as Background and Research Objectives phonon-softening structural phase transition occurs at Developing clean and renewable energy is one of the high temperature. most pressing tasks of humanity. Viability of sustainable energy depends on the effectiveness of electric energy Scientific Approach and Accomplishments storage (EES), manifested by the slow pace with which Summary of third-year and project accomplishments: the petroleum fuel is replaced with electricity in today’s vehicle transportation. High-performance batteries are Synthesis of a solid, Li-rich electrolyte with the anti- hence essential for a clean-energy economy. However, perovskite structure for battery and other electrochemi- the research and development of Li-based battery/ cal devices such as supercapacitors. Several synthetic capacitor materials for EES applications has progressed routes were developed: solid state reaction, high-pres- only incrementally for the last decade. Current technol- sure, “wet” synthesis in ionic liquids. ogy of lithium ion battery (LIB) may meet some require- ments for the emerging plug-in hybrid electric vehicle Clarification of the synthetic mechanism and identifica- market, but provides unsatisfactory performance in tion of defect phases forming during the synthesis. Post- terms of vehicle driving range and acceleration. The processing to eliminate non-perovskite phases. limitations are largely due to shortcomings of LIB’s fluid electrolytes, especially their inability of pairing with a Replacement of oxygen by sulfur to obtain novel Li3SX metallic lithium anode – an energy density booster with anti-perovskites (X=halogen) similar to the superionic which a solid electrolyte could team up. Materials that conductor Ag3SX. Aliovalent doping of the sulfur phases. possess superionic lithium conductivity and can serve The new class of sulfur-based materials is of importance as solid electrolytes hence have been long sought after. to sulfur-based baterries (anodes and cathodes). Chemi- Such materials will have transformational and disruptive cal synthesis, thermal analysis, crystallography, and 202

Page 203

electrochemical properties of these new compounds. Most originated from establishing a uniform contact between notably, Li3SF was synthesized when attempts at making the pellet and the electrodes. We found that the electrical Li3OF failed. Fluorine is less volatile than Cl or Br in these charge passed rather than time elapsed was responsible compounds. The new fluorine-based material will be much for the observed current decrease, since the current after more stable for battery applications. a several hour break was typically returning to its value measured before the break. In some cases, even a partial Development and characterization of layered antiper- current recovery was observed. ovskites, Li7O2X3 (X=halogen) with higher Li ionic conduc- tivity. The impedance spectroscopy experiments performed at various temperatures and bias voltages revealed that the Synthesis of the first set of hybrid, organic-inorganic anti- Li|Li3OCl|Li cell is best represented by an equivalent cir- perovskites: Li5O(TP)X where TP = terephthalate anion cuit comprising of two independent circuits in series corre- and X = halogen. These materials, inspired from the parent sponding to the electrolyte and its interface(s) with lithium compounds, are much less moisture-sensitive than the electrode(s) (Figure 1). The circuit representing the electro- inorganic anti-perovskites. We are partnering with MPA-11 lyte was further divided into two sub circuits correspond- to examine their use in Li air batteries. ing to the bulk material (resistance R2 and capacitance C1) and its grain boundaries. The latter are best described Deposition of the first multilayer structures by pulsed laser as a resistor (R4) in parallel with either constant phase deposition from a Li3OCl target and characterization of element (CPE1) or a pure capacitor. On the other hand, these structures. Important step toward device fabrication. the Li|Li3OCl interface is best represented by a capacitor describing the respective electrical double layer (C3) in Electrochemical study of the electrolyte/lithium metal parallel with a resistor (R3) characterizing the charge trans- interface to assess the chemical compatibility of lithium fer reaction. The transport phenomena associated with the anodes (high-energy and high-power density) with the new charge transfer, typically described as a Warburg element anti-perovskite electrolytes. (Wo), were found negligible under all conditions studied. Durability studies on the long term effects of current flow While the gradual current decrease following the initial in the antiperovskite electrolytes. period seemed concordant with the visible formation of Several difficulties arose, particularly the appearance of a grayish deposit on the surfaces of lithium electrodes non-perovskite phases during the synthesis. These phases facing the solid electrolyte and possible electrode pas- blocked Li transport and decreased ionic conductivity. Two sivation, no proof for such a hypothesis was obtained post-processing methods were developed to overcome this from electrochemical impedance spectroscopy. We found difficulty. This resulted in a gain of an order of magnitude that the main source of the gradual increase in the over- in Li conductivity. all cell impedance during cyclic voltammetry was not the deposit formed on the electrode surfaces but an increase Highlight: Durability studies (3rd year) in the material grain boundary resistance. For instance, in a ~100 hour long experiment at 120 °C with a 1.70 mm The effects of long term ionic current flow on Li3OCl dura- x 1 cm (thickness x diameter) Li3OCl pellet the following bility were studied in an electrochemical cell comprising of time independent parameters were obtained: (3.8 ± 0.9) a pressed pellet of the material between two flat elec- x 105 Ohm for the bulk material resistance, (7.8 ± 2.8) x trodes made of solid lithium. The current was generated by 10 10 for the grain boundary constant phase element and applying triangular wave voltage with an amplitude of ±5 V 0.74 ± 0.04 for its exponential, (4.8 ± 0.1) x 10 11 F for the and 100 mV s 1 sweep rate (cyclic voltammetry) or a con- cell electrical capacity, and (2.4 ± 0.2) x 10 10 F for the stant voltage of up to ±4 V (chronoamperometry) between Li|Li3OCl double layer capacity. During that time, the grain the electrodes. Full electrochemical characteristic of the boundary resistance increased more than two times from cell was measured at specific time intervals using electro- (7.3 ± 2.2) x 105 Ohm to (1.8 ± 0.3) x 106 Ohm (Figure 2). chemical impedance spectroscopy. Local “melting” leading to a smaller contribution of the Cyclic voltammetry was performed at several temperatures grain boundaries in the measured cell impedance or ranging from 70 °C to 160 °C. Irrespective of tempera- restructuring of the grain boundaries can be postulated ture, the measured current initially increased, reaching its at temperatures ≥ 140 °C. The effect is manifested by a maximum value after 20 to 40 cycles (1 - 2 hrs) and then decrease in the measured capacitance of grain boundaries decreased following an approximate logarithmic relation- with temperature and by a significant decrease in the grain ship versus time. The initial current increase most likely boundary resistance. The phenomenon has a stronger ef- 203

Page 204

fect on the cell impedance than the decrease of the bulk material resistance. For instance, after ~36 hours of cyclic voltammetry at 140 °C and 160 °C, the grain boundary resistance for a 3.56 mm x 1 cm (thickness x diameter) pel- let was found to be approximately 10 times lower at 160 °C than at 120 °C, whereas the respective ratio of the bulk material resistance was only ~3.2 (Figure3). Impact on National Missions Figure 2. Li3OCl grain boundary resistance as a function of time Impact on energy security, industrial partnership with of voltage cycling between +5 V and 5 V at 120 °C national laboratories, basic research in functional materials and across length scales. Research on anti-perovskites is now being actively pursued at the University of Texas, Austin in the laboratory of John Goodenough (the inventor of the lithium battery), at the University of Nevada - Las Vegas, and by a private com- pany, Pathion, Inc. based in the Bay Area. A UNLV student is currently pursuing his PhD research on anti-perovskite electrolytes at LANSCE. Early in 2012, an ARPA-E proposal to develop devices based on the materials developed in the LDRD-ER project was submitted to DOE with a view to short-term com- mercialization of the electrolyte materials developed at LANL. The proposal was approved by DOE-ARPA-E and research and development under this proposal started a few months ago. The industrial partners are K2Energy and Pathion. The LDRD-ER project played its intended role perfectly in providing seed funding to develop a novel product. The work done under LDRD-ER funding allowed very clearly Figure 3. Grain boundary and bulk material resistance of Li3OCl for the continuation, extension, and growth of this project versus temperature for a sample that underwent ~36 hours of with other universities and industrial partners. voltage cycling between +5 V and 5 V in a Li|Li3OCl|Li cell at 140 °C and 160 °C. Publications Xia, C., L. Li, Y. Tian, Q. H. Liu, Y. C. Zhao, L. J. Jia, and Y. D. Li. A high performance composite ionic conducting electrolyte for intermediate temperature fuel cell and evidence for ternary ionic conduction. 2009. JOURNAL OF POWER SOURCES. 188 (1): 156. Zhao, Y. S., and L. L. Daemen. Superionic Conductivity in Lithium-Rich Anti-Perovskites. 2012. JOURNAL OF THE AMERICAN CHEMICAL SOCIETY. 134 (36): 15042. Figure 1. Equivalent circuit of a Li|Li3OCl|Li cell 204

Page 205

Chemistry and Material Sciences Exploratory Research Final Report Quasiparticle Scattering for Multiscale Modeling of Electronic Materials Sergei Tretiak 20110212ER Abstract multiscale modeling of electronic properties on nanoma- This research program promises to develop new theo- terials with limited numerical effort. Our theory bridges retical methodologies allowing the quantum chemical traditional molecular description (spatially confined techniques that have proven themselves in providing wavefunctions, orbitals) with solid-state physics picture accurate and adequate results in intermediate size mol- (periodic delocalized wavefunctions, quasiparticles) and ecules to be extended to the world of macromolecules directly addresses nanostructures spanning intermedi- and superstructures. Consequently, the range of treat- ate sizes. Scattering calculations include two steps: i) able molecular systems are to be extended from ~1,000 conventional quantum-chemical calculations and/or atoms to at least 10,000-100,000 atoms sized nanostruc- analysis of experimental data of simple molecular frag- tures, without any substantial information loss or ac- ments (building blocks ~10-100 atoms in size) to retrieve curacy compromise. The theoretical modeling efficiently the quasiparticle properties; and ii) applying scattering uses previously LANL-developed quantum chemical theory to large nanostructures to obtain electronic prop- tools, which outperform similar techniques world-wide erties of interest. The resulting numerical effort scales (Tretiak and Piryatinski). Close synergistic interactions cubically with the number of molecular building blocks with experiment at LANL and word-wide allow dissemi- (not electrons!) in the superstructure. nating theoretical findings. Scientific Approach and Accomplishments Background and Research Objectives In the course of this project we have formulated the the- A quantitative understanding of matter at the nanoscale oretical foundation of the Exciton Scattering (ES) meth- is a grand-challenge for theory. Over the years, elec- odology as illustrated in Figure 1, and have applied this tronic structure calculations, such as density functional approach to many important electronic materials. First theory (DFT), transformed theoretical chemistry, surface of all, we have applied the ES approach to understand science and materials physics and have created a new excited-state structure of donor and acceptor substitut- ability to describe the electronic structure and inter- ed conjugated oligomers [1]. Intuitive relationships be- atomic forces in molecules with hundreds of atoms. tween the substituent’s electron withdrawing or donat- Still, these calculations require atomistic information ing ability and the ES parameters have been established. (electrons and orbitals). Consequently, numerical cost is A good agreement of the absorption spectra between high, effectively limiting the range of treatable systems the ES approach and the reference quantum-chemical by ~1000 atoms. On the other hand, design of functional computations demonstrates that the ES approach is nanostructures such as photovoltaic unit or nanoscale qualified for such conjugated push-pull systems [1]. Sec- electronic chip, requires quantum-mechanical model- ondly, we have extended the ES method to include sym- ing of extended systems with millions atoms in size. This metric triple and quadruple joints that connect linear challenge calls for development of multiscale approach- segments on the basis of the phenylacetylene backbone es. Notably, any reduction of computational complexity [2]. The obtained scattering matrices that characterize is highly non-trivial, since we need to take an advantage these vertices are used in application of our approach to of complex quantum-mechanical phenomena, such several test structures, where we find excellent agree- as delocalized wavefunctions, quantum confinement, ment with the transition energies computed by the and coherences, to achieve new functionalities. In this reference quantum chemistry. We introduce topologi- project we will develop theoretical framework based on cal charges, associated with the scattering matrices, the quasiparticle scattering concept, which would allow which help to formulate useful relations between the 205

Page 206

number of excitations in the exciton band and the number the nonlinear coherent response to a sequence of three of repeat units [2,3]. The obtained features of the scatter- ultrafast optical pulses and specifically focused on the ing phases are analyzed in terms of the observed excited so-called double-quantum coherence technique, which is state electronic structure [2]. To treat more complicated proved to be sensitive to the carriers correlations [10]. Fi- molecular systems using sophisticated ab initio approaches nally, we have investigated the exciton collective behavior we have introduce a Natural Atomic Orbital representation in ensemble of semiconductor quantum dots interacting to be used with the ES theory, which has broadly expanded through delocalized plasmon mode (Figure 2c). In this situ- applications of our multiscale techniques to a variety of ation exciton collective behavior can lead to the formation new materials [4]. of spontaneous coherence and under certain conditions the superradiant states as shown in Figure 2d [11]. We further formulated a universal multiscale scattering ap- proach. The scattering equations have been re-formulated Impact on National Missions in terms of an effective tight-binding model, which is a First and foremost our project upon completion will Hamiltonian describing the parent scattering problem [5]. provide novel computational capabilities critical for un- We further built into this Hamiltonian coupling to phonons derstanding light-induced dynamics in many technologi- and defects (following many previous solid state studies). cally relevant nanostructures. Consequently, we envision Electron-vibrational couplings have been extracted from extremely broad applications of developed tools, relevant the conventional quantum chemical calculations as we to the current and future LANL/DOE missions. Our proj- routinely did previously for many molecular systems such ect primarily addresses Energy and Earth Systems LANL as polymers and carbon nanotubes. Such reduced descrip- Grand challenge by providing computational means for tion (an effective lattice model), can be numerically solved, molecular materials suitable for clean energy (solar energy while it retains the accuracy sufficient to describe electron- capture and energy storage). Secondly, it strongly relates ic properties of the material in question. The lattice model to Materials: Discovery Science to Strategic Applications has been successfully tested on several molecular systems challenge by discovering emergent phenomena in complex containing disorder/defect effects and exciton-phonon systems. Consequently, proposed activities have potential interactions [6]. The outline of developed methodology for extending our program through LANL, and will place us as well as our numerous applications to organic electronic in the excellent position to respond to incoming National materials has been summarized in a large review article to Initiatives in energy and materials, particularly BES calls. appear in the Accounts of Chemical Research journal [7]. meta- A M Our scattering model is not limited to organic molecules n=1 i3n=2 T α aexp(ikxα) i1 since it provides a universal description of electronic pro- n=3 bexp(-ikxα) B ⇐⇒ cesses in many nanoscale materials. We have implemented ortho- i2 n=2 O our interband exciton scattering model into a numerical (a) n=1 (b) T code to simulate the processes of carrier multiplication � � (CM) (also referred in the literature as the multi-exciton (i) generation) in narrow gap, PbSe and PbS, nanocrystal (v) (vi) (ii) (iii) (iv) quantum dots. Using this approach we have performed a systematic study of the CM pathways (as illustrated in Figure 2a,b) in these materials [8,9]. Two fundamentally (vii) (viii) (c) (ix) (x) different processes of multiple-exciton photogenera- (e) P10 8 (d) tion (Figure 2a) and subsequent phonon assisted cooling 80S1,Ω(cid:30)3.37eV 80S2,Ω(cid:30)3.57eV 80S3,Ω(cid:30)3.82eV 80S4,Ω(cid:30)4.08eV 80S5,Ω(cid:30)4.34eV 0.8 60 60 60 60 60 0.6 (Figure 2b) are investigated and their contributions to 40 40 40 40 40 0.4 the CM quantum efficiency are determined. The analysis 20 20 20 20 20 0.2 shows that the exciton scattering induced impact ioniza- 1 1 20 40 60 80 1 1 20 40 60 80 1 1 20 40 60 80 1 1 20 40 60 80 1 1 20 40 60 80 0.0 tion dynamics is the main mechanism responsible for the FIG.1:IllustrationoftheESapproach.(a)structureofaconjugatedmolecule;(b)excitonscatteringpictureonthequasi-1D graphofthemolecule;(c)buildingblocksofphenylacetylene(PA)molecules;(d)linearPAmoleculeconsistsof10repeatunits; CM during both the photogeneration and the population F( t e h i) e ge m xuc o i l t e ro c ne u - le s c1 s a h tt. o e w rI n inllg in up ( as d t ) tt . errnasgtiivoennby othef ctonhtoeur pEloxtscoifttorannsit iSoncdaenttsityemraitnricges f(roEmSg)ro uandpsptatreotoaexcchite.d s(taate)i n structure of a conjugated molecule; (b) exciton scattering picture relaxation events. The results provide insight into experi- on the quasi-1D graph of the molecule; (c) building blocks of mental data reported in the literature. Another fundamen- phenylacetylene (PA) molecules; (d) linear PA molecule consists tal question that has been addressed in our study is the of 10 repeat units; (e) exciton-scattering patterns given by the spectroscopic signatures of the interband exciton scatter- contour plots of transition density matrices from ground state to ing and the Coulomb correlations that lead to this process. excited state in the molecule shown in (d). We have used our exciton scattering theory to calculate 206

Page 207

Orbital Representation for Optical Spectra Calculations in the Exciton Scattering Approach. 2012. Journal of Physical Chemistry Letters. 137: 9245 . 5. Li, H., C. Wu, S. V. Malinin, S. Tretiak, and V. Chernyak. Effective tight-binding models for excitons in branched conjugated molecules. 2013. Journal of Chemical Phys- ics. 139: 064109. 6. Li, H., M. Catanzaro, S. Tretiak, and V. Chernyak. Ana- lytical properties of exciton scattering in conjugated molecules and chemistry of substituents. Journal of Physical Chemistry Letters. 7. Li, H., C. Wu, S. V. Malinin, S. Tretiak, and V. Chernyak. Exciton scattering approach for optical spectra calcu- lations in branched conjugated macromolecules. To appear in Accounts on Chemical Research. 8. Velizhanin, K. A., and A. Piryatinski . Numerical Study of Carrier Multiplication Pathways in Nanocrystalline and Bulk Forms of PbSe. 2011. Physical Review Letters. 106: 207401. 9. Velizhanin, K. A., and A. Piryatinski . Probing Interband Coulomb Interactions in Semiconductor Nanostruc- tures with 2D Double-Quantum Coherence Spectrosco- py . 2011. Journal of Physical Chemistry B. 115: 5372. Figure 2. Investigated pathways of the exciton scattering leading 10. Velizhanin, K. A., and A. Piryatinski. Numerical analysis to the carrier multiplication processes in semiconductor quantum of carrier multiplication mechanisms in nanocrystalline dots. (a) Indirect and direct photogeneration processes. (b) Popu- and bulk forms of PbSe and PbS. 2012. Physical Review lation relaxation and Interband population transfer. (c) Semi- B. 86: 165319. conductor quantum dots (red) topping metal ellipsoids (blue) whose collective surface-plasmon modes lead to the interaction 11. Bittner, E. R., A. Piryatinski, and D. Dunlap. Thermody- between the excitons in the quantum dot. (c) Phase diagram namics of exciton collective behavior. Nano Letters. showing formation of the exciton superradiant states. Publications References Bittner, E. R., A. Piryatinski, and D. Dunlap. Thermodynam-

1. Li, H., C. Wu, S. Malinin, S. Tretiak, and V. Chernyak. ics of exciton collective behavior. Nano Letters. Excited States of Donor and Acceptor Substituted Con- jugated Oligomers: a Perspective from the Exciton Scat- Catanzaro, M., T. Shi, S. Tretiak, and V. Chernyak. Topologi- tering Approach. 2010. Journal of Physical Chemistry cal properties of scattering matrices on graphs: count- Letters. 1: 3396 . ing the number of standing waves in quasi-one-dimen- sional systems. Physical Review Letters.
2. Li, H., C. Wu, S. Malinin, S. Tretiak, and V. Chernyak. Exciton Scattering on Branching Centers in Conjugated Li, H., C. Wu, S. Malinin, S. Tretiak, and V. Chernyak. Excited Molecules. 2011. Journal of Physical Chemistry B. 115: States of Donor and Acceptor Substituted Conjugated Oligomers: a Perspective from the Exciton Scattering 5465 . Approach. 2010. Journal of Physical Chemistry Letters.
3. Catanzaro, M., T. Shi, S. Tretiak, and V. Chernyak. To- 1: 3396 . pological properties of scattering matrices on graphs: Li, H., C. Wu, S. Malinin, S. Tretiak, and V. Chernyak. Ex- counting the number of standing waves in quasi-one- citon Scattering on Branching Centers in Conjugated dimensional systems. Physical Review Letters. Molecules. 2011. Journal of Physical Chemistry B. 115: 5465 .
4. Li, H., V. Chernyak, and S. Tretiak. Natural Atomic 207

Page 208

Li, H., C. Wu, S. V. Malinin, S. Tretiak, and V. Chernyak. Ef- fective tight-binding models for excitons in branched conjugated molecules. 2013. Journal of Chemical Phys- ics. 139: 064109. Li, H., C. Wu, S. V. Malinin, S. Tretiak, and V. Chernyak. Exci- ton scattering approach for optical spectra calculations in branched conjugated macromolecules. To appear in Accounts on Chemical Research. Li, H., M. Catanzaro, S. Tretiak, and V. Chernyak. Analytical properties of exciton scattering in conjugated mole- cules and chemistry of substituents. Journal of Physical Chemistry Letters. Li, H., V. Chernyak, and S. Tretiak. Natural Atomic Orbital Representation for Optical Spectra Calculations in the Exciton Scattering Approach. 2012. Journal of Physical Chemistry Letters. 137: 9245 . Velizhanin, K. A., and A. Piryatinski. Numerical analysis of carrier multiplication mechanisms in nanocrystalline and bulk forms of PbSe and PbS. 2012. Physical Review B. 86: 165319. Velizhanin, K. A., and A. Piryatinski . Numerical Study of Carrier Multiplication Pathways in Nanocrystalline and Bulk Forms of PbSe. 2011. Physical Review Letters. 106: 207401. Velizhanin, K. A., and A. Piryatinski . Probing Interband Coulomb Interactions in Semiconductor Nanostruc- tures with 2D Double-Quantum Coherence Spectrosco- py . 2011. Journal of Physical Chemistry B. 115: 5372. 208

Page 209

Chemistry and Material Sciences Exploratory Research Final Report Solving the Active Site Conundrum in Oxygen Reduction Catalysis Piotr Zelenay 20110236ER Abstract precious metal and Pt-based ORR catalysts is unlikely to Given the extremely volatile and generally very high happen without “molecular-level” insight into the struc- price of Pt and other noble metals over recent years, the ture of the active site(s) and knowledge-based design of development of non-precious metal fuel cell electrocata- future catalysts. lysts that meet the Department of Energy targets of high These catalytic systems, derived from polyaniline (PANI) volumetric activity (330 A/cm3) and long-term stability salts of Fe and/or Co, and high surface-area carbon in a (5,000 hours with cycling) would significantly acceler- high temperature process at ca. 900°C have been shown ate energy independence and environmentally benign to possess the field-leading ORR activity (Figure 1). energy technologies. While non-precious metal catalysts However, for non-precious catalysts to become practical, for oxygen reduction reaction (ORR) have been stud- the activity needs to be improved by approximately 0.1 ied for decades, with only incremental improvements V, to the level exhibited by Pt-based oxygen-reduction made in the total activity, the active catalytic site in such catalysts (blue plot in Figure 1). In order to proceed at materials remains unknown. LANL-sponsored research, a rate faster than traditional Edisonian methods, this involving pyrolyzed graphitic-metal-nitrogen species, has will require learning what the active site in these ma- produced catalysts with the highest activity/durability/ terials is such that design principles can be devised to selectivity ever reported in non-precious metal electro- optimize active site exposure and concentration. Since catalysis. A focused theoretical and experimental effort characterization of the species involved in non-precious has been applied to develop science-based structure- metal catalysis is challenging, by virtue of the extreme activity relations for these poorly understood and conditions involved during synthesis, we applied molecu- difficult-to-characterize systems. Atomistic modeling was lar modeling techniques in concert with experimental employed in synergy with complementary experimental characterization. characterization techniques (electrochemistry, electron paramagnetic resonance, nuclear resonance vibrational Experimentally, our objective in this project was to trans- spectroscopy and Mößbauer) to elucidate the nature form bulk characterization techniques, e.g., EPR, NRVS, of the active site of non-precious ORR catalysts includ- and Mößbauer, into surface-specific tools by using gas ing structure-property relations, and to employ LANL surface probes (NO, alkyl peroxides). Molecular model- expertise in synthesis and characterization to develop ing provides a powerful complement to these experi- a non-precious metal catalyst that possesses optimized ments. ORR activity. Scientific Approach and Accomplishments Background and Research Objectives Our focus in this work was to determine features of the The cost of a Pt cathode oxygen reduction reaction ORR active site in the non-precious metal catalysts to (ORR) catalyst represents the greatest challenge for use as a basis for future rational design of new active large-scale introduction of polymer electrolyte fuel site centers that lead to minimal H2O2 production while cells (PEFCs) for automotive transportation. Except for facilitating high activities and reduction potentials for precious-metal “thrifting” (already close to its limit), the four-electron path. Our research approach was to replacement of Pt with non-precious metal ORR cata- use stochastic search techniques to provide a “short- lysts is the sole viable solution. Significant progress has list” of hypothetical active sites via molecular modeling, been achieved within the past decade with heat-treated and then to calculate, using quantum mechanics, their N-Metal-C catalysts but closing the gap between non- 209

Page 210

associated spectroscopic signatures and theoretical oxida- does not shift significantly for the different samples, indi- tion/reduction reaction activity. This latter activity was cating there is no significant change in the local environ- calculated from the energy barriers connecting the neces- ment. Subtracting the spectra of the NO-adsorbed sample sary intermediate states along the reaction path from O2 from the spectra of the reduced sample yields a weak sig- through to H2O and H2O2. The spectroscopic signatures nal with a hyperfine pattern that could be representative and active-site configurations were then compared with of Fe-NO. However the signal is not clearly distinguishable those obtained using the NRVS (nuclear resonance vibra- from background and so not definitive evidence of Fe-NO tional spectroscopy) technique, EPR (electron paramag- interaction (Figure 3a). netic resonance) and Mößbauer spectroscopy. Diffuse reflectance infrared Fourier transform spectros- Next, we present the synthesis of the electrocatalysts and copy (DRIFTS) results show changes in the carbon matrix the results of the experimental characterization work. as evidenced by the C≡C peak, and the FTIR results dis- Then we describe the modeling efforts and summarize the play changes in OH concentration, presumably related to findings. changes in surface chemistry (Figure 3b). There are no absorbance bands evident around 600 cm-1 as expected Synthesis of Optimized Electrocatalysts for Fe-NO, based upon NRVS (below). Commercial carbon blacks were used as supporting ma- terials in polyaniline (PANI-Fe) and cyanamide (CM-Fe) The Raman results are typical of graphite samples [8] but catalysts synthesis. In a typical approach [1-6], carbon do not directly inform about Fe-NO interactions. (Figure supports were treated in hydrochloric acid for 24 hours, to 3c,d) The data have been normalized to the G-peak at 1600 remove metal impurities, and oxidized in 70% nitric acid cm-1, and this highlights the increase of the D-peak in the at 80°C for 8 h for oxygen functionalization. 57FeCl3 and reduced and NO absorbed samples. The intensity increase 57Fe(CH3COO)2 were used as iron precursors in PANI-Fe of the D-peak is typical of increasing disorder in the carbon and CM-Fe catalyst synthesis, respectively. matrix [9]. These studies highlight the challenge in inter- rogating these non-precious metal catalysts using standard Addition of Chemical Probe Species (NO) spectroscopy. As-prepared Fe catalysts were electrochemically reduced in an O2-free wet box, as shown in Figure 2, to reduce Mößbauer Spectroscopy Fe(III) to Fe(II), facilitating binding of NO onto Fe(II) sites. The 80K 57Fe Mößbauer spectrum of as-prepared CM-Fe-C 0.5 M H2SO4 and DI-water were used to purify the NO is shown in Figure 3e-A. Three sextet species are present in before it entered the glove bag for NO treatment. Teflon- as-prepared CM-Fe-C representing ca. 40% of all the iron sealed vials containing solid-state Fe samples were put present with the major sextet species consistent with the into a N2-purged polyethylene foil glove-bag and NO flow formation of bulk iron nitride, FexN. A singlet iron spe- was connected into the vials by using needles. As the NO cies is also present, previously assigned in PANI-Fe-C to adsorption onto Fe sites is rapid, the duration of NO treat- superparamagnetic iron. The remaining iron species (ca. ment was around 2 min. Then the NO-treated Fe catalysts 54%) are well fit by a sum of three doublet iron species samples were stored in a liquid N2-filled container and consistent with the formation of mono- and/or dimeric shipped to the Advanced Photon Source (APS) at Argonne iron. Such species are predicted by molecular model- National Laboratory and the University of Rochester for ing as strong candidates for the active site. The overall NRVS and Mößbauer spectroscopy, respectively. Mösßbauer spectrum, as well as the individual sextet and doublet species present in CM-Fe-C are distinct from those Electron Paramagnetic Resonance (EPR) and Diffuse previously identified in PANI-Fe-C, indicating the effect of Reflectance Infrared Fourier Transform Spectroscopy changing the nitrogen precursor molecule. (DRIFTS) Six samples of oxidized, reduced and NO-exposed PANI-Fe Electrochemical reduction of as-prepared CM-Fe-C has a and CM-Fe catalysts were measured by electron paramag- dramatic effect on iron speciation as shown by the changes netic resonance (EPR) spectroscopy showing a signal at g of in the spectrum (Figure 3e-B). Detailed analysis of individ- ca. 4.3 from Fe3+ in a distorted octahedral symmetry [7]. ual contributions to the Mößbauer spectrum indicated the The EPR measurements did not show clear evidence of NO presence of the same three sextet species and the same adsorption. One difficulty was the presence of significant singlet species upon reduction. A significant change in the bulk Fe that resulted in low sensitivity, obscuring detec- doublet iron species was observed. The change required tion of surface Fe. Six new catalysts samples with a smaller four doublet species in order to obtain a reasonable fit. concentration of Fe were measured. As seen previously, all Changes in the iron speciation upon reduction occur only samples showed a signal at g = 4.3, as before. The position within the mono- and dimeric iron doublet species. The 210

Page 211

presence of a large isomer shift in the doublet due to the empirical PM6 calculations [11] as implemented in Gauss- high velocity shoulder on the sextet feature at ca. 3 mm/s ian ’09 [12]. PM6 allowed very rapid energy calculation. A is consistent with a high-spin iron(II) species. Upon reduc- handful of systems were tested using density functional tion, the amount of sextet (ca. 54%) and singlet (ca. 9%) theory [13-17] to help benchmark the accuracy of the iron increased while the amount of doublet iron (ca. 37%) relative energies emerging from the PM6 calculation. The decreased. Metropolis Monte Carlo algorithm [18] was implemented to screen the results. The flake was kept overall neutral in Treatment of electrochemically reduced CM-Fe-C with charge and the spin-state was set to a singlet state and ran- NO was performed in order to evaluate the presence of dom substitutions of C for vacancies, N or Fe atoms were surface or near-surface, gas-accessible iron species. The made. The energies of these random systems were evalu- Mößbauer spectrum of NO treated CM-Fe-C (Figure 3e-C) ated using PM6. According to Metropolis Monte Carlo, a exhibits changes in the doublet Mößbauer intensity rela- random perturbation (i.e. atom substitution or swap) from tive to reduced catalysts, consistent with the perturbation the currently selected structure was accepted if the follow- of iron species due to NO treatment. No changes are ob- ing criterion was met: -ΔE/kT > ln µ where µ is a random served in the sextet or singlet species. NO treatment leads number between 0 and 1. This method, combined with the to a change in the doublet iron speciation consistent with generation of random initial structures studied in parallel, the presence of gas accessible reduced iron species. As NO provided a robust method for exploring active-site configu- is an analog for O2 for iron ligation, such gas-accessible rational space. iron sites should also be capable of binding O2 during ORR. One prediction of the semi-empirical potential model was Nuclear Resonance Vibrational Spectroscopy (NRVS) the restructuring of C to form cyclopentane and larger C The 57Fe samples prepared from optimized PANI-Fe and rings at C vacancy sites (see Figure 5b). These structures CM-Fe fuel cell catalysts were analyzed at APS for NRVS ac- are known to be the stable vacancy structures in 2-D C sys- tivity. This technique provides the unique ability to probe tems [18-20], thus validating our computational method. vibrational spectra based upon the environment of specific Examination of the local atomic arrangements about the isotopes, in this case 57Fe, and thus avoids the challenges M (Fe) atoms in the model system shows that Fe cluster- of optical spectroscopy. The use of NO provides a way to ing at 2nd nearest neighbor sites is thermodynamically distinguish bulk from surface. Two sets of NRVS spectra preferred (Figure 5c,d). In this cluster, a local N atom (or, in were obtained according to two allocations of the com- other cases, a C atom) was perturbed from its intermediate petitive beam-time: June 2011 and April 2013 (Figure 4). lattice site, becoming a bridge atom linking two Fe atoms. In both cases there was observed a change in iron specia- N bridged Fe clusters were more stable than C bridged Fe tion upon reduction, whereas NO adsorption resulted in clusters. The consideration of N- or C-bridged binuclear weak changes. It was suggested that NO photolysis may site as the NPMC active site was unexpected and contrib- be the cause of the weak signal, as multiple scans lead to utes significantly to the literature on this topic, as most a reduced difference spectrum. Strategies to overcome hypothesized active sites are based on the “heme” Fe-N4 these challenges have been proposed, but beam-time center. The site we have elucidated has several catalytic did not permit further measurements. In the first analy- advantages: if breaking of the O-O bond is to occur (disso- sis (June 2011), a difference spectrum led to a significant ciative ORR pathway) then having two sites in close prox- peak around 600 cm-1, which was identified using density imity that can coordinate the O-O bond from each side is functional theory (below). However, in the repeat experi- advantageous. Furthermore, these sites will be templated ment in April 2013, this spectral feature was not observed, by the graphitic network itself. Another motif based on possibly due to NO photolysis. our search is the clustering of vacancies, N atoms, and the bridged two-Fe structures and the apparent preference of Modeling and Simulation these clusters to be positioned close to the flake edge. For Previous experimental results implied that a positive cor- this reason, we suggest that active sites will be much more relation between carbon graphitization, i.e. the C-sp2/sp3 prevalent and accessible at edge-sites of the graphitized ratio, and activity of synthesized NPMC [10]. Hence, the carbon matrix. model system adopted was a graphene “flake” (Figure 5a). This system consisted of 96 C-atoms (before vacancies) The motifs, identified and described, above have served and contained zig-zag and armchair H-coordinated edges, as starting points for more accurate, but computationally allowing the simulation of bulk and edge effects. Based on intensive, DFT studies and further optimization. The stabil- typical values for N and M concentrations used in NPMC ity of vacancies, N-coordinated vacancies, and Fe binding synthesis [3] we inserted 8 N and 2 M atoms into the on these defects was studied on 1-D graphene nanorib- model system. Calculations were performed using semi- 211

Page 212

bons as a function of distance from the nanoribbon edge ygen reduction reaction on precious metal catalysts in the [21]. It was found that N coordination stabilizes the local past. These results have been used to initiate an LDRD-DR Fe defect and that such defect complexes were most stable effort (FY13) to use graphene model systems to advance near the edge. These findings are important for catalysis catalytic activities through rational design. in two respects: (1) N is important in that it stabilizes the Fe defect leading to higher concentrations of these as- sumed active sites and (2) these sites are most abundant near the ribbon edge where they are most accessible, leading to less transport limited active sites. In addition, by moving a pair of N-coordinated Fe edge defects closer together, it was determined that Fe does prefer to cluster. Using a consistent thermodynamic model, we compared relative stability of nitrogen-coordinated transition-metal atom defects at graphene edges. We have found that the most stable structure is dependent on the nitrogen and transition-metal chemical potential and so, precursors and processing conditions are likely to change the structures that are formed. This theoretical model establishes a direct link between synthesis conditions (N-rich versus Fe-rich) and the molecular products obtained. By performing a phonon analysis, we determined the Figure 1. Metal- and nitrogen-modified catalysts have perfor- vibrational modes of active site candidate structures with mance characteristics approaching Pt-based systems (rotating- NO ligands and discovered that the Fe-NO bond stretch disk electrode data at 25°C). is dependent upon the local nitrogen coordination of the transition metal. The calculated frequencies for the FeN3 active sites agreed well with the experimentally deter- mined Fe-NO bond stretch frequency (using Nuclear Reso- nance Vibrational Spectroscopy, NRVS, in the June 2011 experiment) while the frequency of FeN4 sites did not. Impact on National Missions The development of alternative energy systems is an important component of the current Laboratory mission, as outlined in the LANL Grand Challenge in Energy and Earth Systems. In particular the Grand Challenge calls for “advances in prediction and controlling atom motion and electron distribution”, as well as “fundamental advances in […] devices for harvesting, storing and converting energy.” Furthermore, the connection between materials structure and ultimate activities relates to the Grand Challenge in Materials to exercise “intentional control of functionality through discovery and application of fundamental materi- als properties […] from the molecular level.” In the current work, we have significantly advanced the molecular level Figure 2. A newly developed experimental setup for an electro- understanding of the operation of these non-precious chemical reduction treatment of Fe catalysts in an oxygen-free metal fuel cell catalysts, highlighting a path forward for wet box (top) and schematic diagram of a schematic diagram of science-based optimization of the materials and synthe- the homemade experimental setup for the NO treatment of Fe sis procedures. Furthermore, we have, for the first-time, catalysts (bottom). provided an atomistic-level view of the structure of these active sites in these highly-efficient catalyst materials that provides the opportunity to apply the same level of rigor that has been applied to the theoretical analysis of the ox- 212

Page 213

Figure 3. (a) The results of subtracting the spectra of the NO ad- sorbed on PANI-Fe from the spectra of reduced PANI-Fe. The hy- perfine pattern is consistent with a Fe-NO species but the signal is barely distinguishable from the baseline. (b) DRIFTS spectra of CM-Fe showing changes in the C≡C and O-H absorbance values. The vibrations are typical of the catalyst and surface chemistry, respectively. (c,d). Raman spectra of catalyst samples normalized to the G-peak at 1600 cm-1. Data attest to an increase in the D- peak relative to the G-peak indicating and increase in disorder in Figure 5. (a) Example of a perfect graphene flake with zig-zag the carbon matrix. (e). 80 K 57Fe Mößbauer spectra (dots) and (top and bottom) and armchair (left and right) edges. (b) Exam- best fit (black line) for (A) as-prepared CM-Fe-C, (B) reduced CM- ple of potential random initial flake structures with 2 vacancies, 8 Fe-C, and (C) reduced CM-Fe-C treated with NO. N, and 2 Fe. Larger spheres represent Fe, blue spheres represent N, and white spheres represent H, C is in gray. (c,d) Two examples of highly stable structures found using the stochastic search al- gorithm displaying 2nd NN Fe/N/vacancy clusters near the flake edge. References

1. Metropolis, N., A. W. Rosenbluth, M. N. Rosenbluth, and A. H. Teller. Equations of State Calculations by Fast Computing Machines. 1953. Journal of Chemical Phys- ics. 21: 1087.
2. Kresse, G., and J. Hafner. Ab inito molecular dynam- ics for liquid metals. 1993. Physical Review BPhysical Review B. 47.
3. Kresse, G., and J. Hafner. Ab initio molecular-dynamics simulation of the liquid-metal–amorphous-semicon- ductor transition in germanium. 1994. Physical Review BPhysical Review B. 49 (20): 14251.
4. Rao, A. Srinivasa, R. Ramakrishna Reddy, T. V. Ramak- rishna Rao, and J. Lakshmana Rao. Electron paramag- netic resonance and optical absorption spectra of Fe3+ ions in alkali cadmium borosulphate glasses. 1995. Solid State Communications. 96 (9): 701.
5. Kresse, G., and . . Efficient iterative schemes for ab ini- tio total-energy calculations using a plane-wave basis set. 1996. Physical Review BPhysical Review B. 54. Figure 4. NRVS spectra obtained with two different samples of 57Fe non-precious metal catalysts exposed to NO.
6. Perdew, J. P., K. Burke, and M. Ernzerhof. Generalized 213

Page 214

gradient approximation made simple. 1996. Physical H. Garzon, and P. Zelenay. Titanium dioxide-supported Review LettersPhysical Review Letters. 77. non-precious metal oxygen reduction electrocatalyst. 2010. Chemical CommunicationsChemical Communica- 7. Perdew, J. P., K. Burke, and M. Ernzerhof. Erratum: Gen- tions. 46 (40): 7489. eralized gradient approximation made simple. 1997. Physical Review LettersPhysical Review Letters. 78. 18. Jaouen, F., E. Proietti, M. Lefevre, R. Chenitz, J. Dode- let, G. Wu, H. T. Chung, C. M. Johnston, and P. Zelenay. 8. El-Barbary, A. A., R. H. Telling, C. P. Ewels, M. I. Heg- Recent advances in non-precious metal catalysis for gie, and P. R. Briddon. Structure and energetics of the oxygen-reduction reaction in polymer electrolyte fuel vacancy in graphite. 2003. Physical Review BPhysical cells. 2011. Energy & Environmental Science. 4 (1): Review B. 68 (14): 144107. 114. 9. Sammalkorpi, M., A. Krasheninnikov, A. Kuronen, K. 19. Wu, G., H. T. Chung, M. Nelson, K. Artyushkova, K. Nordlund, and K. Kaski. Mechanical properties of L. More, C. M. Johnston, and P. Zelenay. Graphene- carbon nanotubes with vacancies and related defects. Enriched Co9S8-N-C Non-Precious Metal Catalyst for 2004. Physical Review BPhysical Review B. 70 (24): Oxygen Reduction in Alkaline Media. 2011. ECS Trans- 245416. actionsECS Transactions. 41 (1): 1709. 10. Ferrari, A. C., J. C. Meyer, V. Scardaci, C. Casiraghi, M. 20. Wu, G., C. M. Johnston, N. H. Mack, K. Artyushkova, Lazzeri, F. Mauri, S. Piscanec, D. Jiang, K. S. Novoselov, M. Ferrandon, M. Nelson, J. S. Lezama-Pacheco, S. S. Roth, and A. K. Geim. Raman Spectrum of Graphene D. Conradson, K. L. More, D. J. Myers, and P. Zele- and Graphene Layers. 2006. Physical Review Letters- nay. Synthesis-structure-performance correlation for Physical Review Letters. 97 (18): 187401. polyaniline-Me-C non-precious metal cathode catalysts for oxygen reduction in fuel cells. 2011. Journal of Ma- 11. Ferrari, A. C.. Raman spectroscopy of graphene and terials Chemistry. 21 (30): 11392. graphite: Disorder, electron–phonon coupling, doping and nonadiabatic effects. 2007. Solid State Communi- 21. Holby, E. F., and C. D. Taylor. Control of graphene na- cations. 143 (1–2): 47. noribbon vacancies by Fe and N dopants: Implications for catalysis. 2012. Applied Physics Letters. 101 (6): 12. Stewart, J. P.. Optimization of parameters for semiem- 064102. pirical methods V: Modification of NDDO approxima- tions and application to 70 elements. 2007. Journal of Publications Molecular Modeling. 13 (12): 1173. Choi, J. H., C. M. Johnston, D. Cao, and P. Zelenay. Se-mod- ified Ru nanoparticles as ORR catalysts. Part 2: evalu- 13. Chung, H. T., C. M. Johnston, F. H. Garzon, and P. Zele- ation for use as DMFC cathodes. 2011. J. Electroanal. nay. A Non-Precious Electrocatalyst for Oxygen Reduc- Chem. 662: 267–273 . tion Based on Simple Heat-Treated Precursors. 2008. ECS TransactionsECS Transactions. 16 (2): 385. Chung, H. T., G. Wu , and P. Zelenay. Nanostructure vs. Re- activity of Non-Precious Metal Oxygen Reduction Cata- 14. Wu, G., Z. Chen, K. Artyushkova, F. H. Garzon, and P. lysts. To appear in Hydrogen and Fuel Cells Zing Confer- Zelenay. Polyaniline-derived Non-Precious Catalyst for ence 2013. (Napa, California, July 12-15, 2013). the Polymer Electrolyte Fuel Cell Cathode. 2008. ECS Chung, H. T., J. H. Won, and P. Zelenay. Active and Stable TransactionsECS Transactions. 16 (2): 159. Carbon Nanotube/Nanoparticle Composite Electro- 15. Frisch, M. J., H. Schlegel, G. Scuseria, M. Robb, J. catalyst for Oxygen Reduction. 2013. Nat. Commun. 4: 1922. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. Petersson, and H. Nakatsuji. Gaussian-09, Revision A. Chung, H. T., and P. Zelenay. Nitrogen-doped carbon tubes 01. Gaussian. 2009. Inc, Wallingford. for polymer electrolyte fuel cells and other electro- chemical applications (invited lecture). Invited presen- 16. Chung, H. T., C. M. Johnston, K. Artyushkova, M. Fer- tation at 13th International Symposium on Polymer randon, D. J. Myers, and P. Zelenay. Cyanamide-derived Electrolytes (ISPE-13). (Selfoss, Iceland, 26-31 August, non-precious metal catalyst for oxygen reduction. 2012). 2010. Electrochemistry CommunicationsElectrochem- istry Communications. 12 (12): 1792. Elbaz, L., G. Wu, and P. Zelenay. Heat-treated non-precious- metal-based catalysts for oxygen reduction. 2013. In 17. Wu, G., M. A. Nelson, N. H. Mack, S. Ma, P. Sekhar, F. Electrocatalysis in fuel cells: A non- and low-platinum 214

Page 215

approach (Springer-Verlag London Limited). Edited by 2013). Shao, M.. , p. 213. London: Springer. Johnston, C. M., D. Cao, J. H. Choi, P. Babu, F. Garzon, and Ferrandon, M., A. J. Kropf, D. J. Myers, K. Artyushkova, U. P. Zelenay. Se-modified Ru nanoparticles as ORR cata- Kramm, P. Bogdanoff, G. Wu, C. M. Johnston, and P. lysts. Part 1: synthesis and analysis by RRDE and in Zelenay. Multi-Technique Characterization of a Polyan- PEFCs. 2011. J. Electroanal. Chem. 662: 257–266 . iline-Iron-Carbon Oxygen Reduction Catalyst. 2012. J. Phys. Chem. C. : 16001. Shi, Z., H. Liu, K. Lee, E. Dy, J. Chlistunoff, M. Blair, P. Zele- nay, J. Zhang, and Z. S. Liu. A Theoretical Study of Ferrandon, M., X. Wang, A. J. Kropf, D. J. Myers, G. Wu, C. Possible Active Site Structures in Cobalt-Polypyrrole M. Johnston, and P. Zelenay. Stability of Iron Species in Catalysts for Oxygen Reduction Reaction. 2011. J. Phys. Heat -Treated Polyaniline-Iron-Carbon Polymer Electro- Chem. C. 115: 16672. lyte Fuel Cell Cathode Catalysts. To appear in Electro- chim. Acta. : doi:10.1016/j.electacta. Wu, G., C. M. Johnston, N. H. Mack, K. Artyushkova, M. Nelson, M. Ferrandon, J. S. Lezama-Pacheco, S. D. Holby, E. F., C. D. Taylor, G. Wu, and P. Zelenay. Active site Conradson, K. L. More, D. J. Myers, and P. Zelenay. modeling: Non-precious metal based catalysts for ORR. Synthesis-structure-performance correlation for poly- Presented at 222nd Meeting of the Electrochemical aniline-Me-C non-precious metal cathode catalysts for Society. (Honolulu, Hawaii, 7-12 October, 2012). oxygen reduction in fuel cells. 2011. J. Mater. Chem.. 21: 11392. Holby, E. F., C. D. Taylor, G. Wu, and P. Zelenay. Modeling Carbon Based Non-Precious Metal Catalyst Active Site Wu, G., K. L. More, C. M. Johnston, and P. Zelenay. High- Structures for the Oxygen Reduction Reaction in PEFC Performance Electrocatalysts for Oxygen Reduction De- Cathodes. To appear in 245th National Meeting of the rived from Polyaniline, Iron, and Cobalt. 2011. Science. American Chemical Society. (New Orleans, April 9, 332: 443. 2013). Wu, G., K. L. More, P. Xu, H. L. Wang, M. Ferrandon, A. Holby, E. F., G. Wu, P. Zelenay, and C. D. Taylor. Modeling J. Kropf, D. J. Myers, S. Ma, C. M. Johnston, and P. of non-precious metal catalyst active sites (invited Zelenay. Carbon-nanotube-supported graphene-rich lecture). Invited presentation at Los Alamos National non-precious metal oxygen reduction catalyst with Laboratory Materials Theory Seminar. (Los Alamos, enhanced performance durability. 2013. Chem. Com- New Mexico, 25 July, 2012). mun.. 49 (32): 3291. Holby, E. F., G. Wu, P. Zelenay, and C. D. Taylor. Modeling Wu, G., M. Nelson, H. Chung, and P. Zelenay. Co9S8-N-C of non-precious metal catalyst active sites (invited lec- Non-Precious Metal Catalyst for Oxygen Reduction ture). Invited presentation at University of New Mexico in Alkaline Media. To appear in 220th Meeting of the Energy Technology Seminar. (Albuquerque, New Mexi- Electrochemical Society. (Boston, October 9-14, 2011). co, 10 September, 2012). Wu, G., and P. Zelenay. Nanostructured Nitrogen-Carbon- Holby, E. F., G. Wu, P. Zelenay, and C. D. Taylor. Metropolis Transition-Metal Catalysts for Oxygen Reduction Reac- Monte Carlo search for non-precious metal catalyst tion. 2013. Acc. Chem. Res. 46 (8): 1878–1889. active sites candidates. 2013. ECS Transactions. 50 (2): 1839. Zelenay, P.. Non-precious metal catalysts for oxygen reduc- tion (invited lecture). Invited presentation at Tongji Holby, E. F., J. L. Kneebone, S. L. Daifuku, G. Wu, H. T. University, State Research Center for Fuel Cell Vehicle Chung, M. L. Neidig, P. Zelenay, and C. D. Taylor. Struc- and Powertrain System Engineering Technologies. ture of Fe-Nx-C defects in Non-Precious Oxygen Reduc- (Shanghai, China, 6 September, 2011). tion Reaction Catalysts Elucidated via First-Principles Modeling and Mössbauer Spectroscopy. Nature Mate- Zelenay, P.. Non-precious Metal catalysts for oxygen reduc- rials. tion (invited lecture). Invited presentation at Nanjing University, Department of Materials Science. (Nanjing, Holby, E. F., and C. D. Taylor. Control of graphene nanorib- China, 8 September, 2011). bon vacancies by Fe and N dopants: Implications for catalysis. 2012. Applied Physics Letters. 101: 06412. Zelenay, P.. Non-precious metal catalysts for oxygen reduc- tion (invited lecture). Invited presentation at Dalian Holby, E., G. Wu, P. Zelenay, and C. Taylor . Modeling Non- Institute of Chemical Physics, Chinese Academy of Sci- Precious Metal Catalyst Structures and Their Relation- ence. (Dalian, China, 15 September, 2011). ship to ORR Activity. To appear in 224th Meeting of the Electrochemical Society. (San Francisco, 27 October Zelenay, P.. Non-PGM electrocatalysis of oxygen reduction (invited lecture). Invited presentation at Tsinghua Uni- 215

Page 216

versity, Institute of Nuclear and New Energy Technol- 2011). ogy. (Beijing, China, 19 September, 2011). Zelenay, P., G. Wu, Q. Li, and H. T. Chung . The Effect of Zelenay, P.. Non-precious metal catalysts for oxygen reduc- Carbon Derived from Different Nitrogen Precursors on tion in fuel cells: In pursuit of platinum (invited lec- Oxygen Reduction Activity of Non-Precious Metal ORR ture). Invited presentation at Peking University, College Catalysts. To appear in First International Conference of Engineering. (Beijing, China, 20 September, 2011). on Electrochemical Materials and Technologies for Clean Sustainable Energy. (Guangzhou, China, July 5-9, Zelenay, P.. Non-precious metal oxygen reduction catalysts 2013). for fuel cells (invited lecture). Invited presentation at University of Rochester, Department of Chemical Engi- Zelenay, P., H. T. Chung, Q. Li, G. M. Purdy, and G. Wu . Ad- neering. (Rochester, NY, 26 October, 2011). vantages and Limitations of Non-Precious Metal ORR Catalysts in Various Environments. To appear in 64th Zelenay, P.. Precious metal-free oxygen reduction catalysts Annual Meeting of the International Society of Electro- for polymer electrolyte fuel cells (invited lecture). chemistry. (Santiago de Querétaro, 8-13 September, Invited presentation at University of California, UCSB 2013). Chemical Engineering Seminar Series. (Santa Barbara, CA, 3 November, 2011). Zelenay, P., H. T. Chung, Q. Li, and G. Wu. Oxygen reduc- tion at non-precious metal electrocatalysts: Beyond Zelenay, P.. State of the art in non-precious metal catalysts the standard PEFC cathode (invited lecture). Invited for fuel cells (invited lecture). Invited presentation at presentation at 3rd CARISMA International Conference 221st Meeting of the Electrochemical Society. (Seattle, on Medium and High Temperature Proton Exchange Washington, 6-10 May, 2012). Membrane Fuel Cells. (Copenhagen, Denmark, 3-5 September, 2012). Zelenay, P.. Oxygen reduction on non-precious metal cata- lysts: How close are we to a viable system? (invited lecture). Invited presentation at Gordon Research Con- ference on Fuel Cells. (Smithfield, Rhode Island, 5-10 August, 2012). Zelenay, P.. High-Performance Electrocatalysts for Oxygen Reduction Derived from Polyaniline, Iron, and Cobalt. To appear in American Physical Society March Meeting in 2013. (Baltimore, Maryland, March 18-22, 2013). Zelenay, P.. Non-Precious Metal Oxygen Reduction Cata- lysts: Are We Closing on the Performance Gap to Plati- num?. To appear in Hydrogen + Fuel Cells 2011. (Van- couver, Canada, May 15-18, 2011). Zelenay, P., G. Wu, H. T. Chung, M. Blair, E. F. Holby, C. D. Taylor, and M. L. Neidig. Insight into the possible na- ture of the active catalytic site in non-precious metal fuel cell ORR catalysis. Presented at 222nd Meeting of the Electrochemical Society. (Honolulu, Hawaii, 7-12 October, 2012). Zelenay, P., G. Wu, H. T. Chung, M. W. Blair, E. F. Holby, C. D. Taylor, and M. L. Neidig. Oxygen reduction on non- precious metal fuel cell catalysts (invited keynote lec- ture). Invited presentation at 63rd Annual Meeting of the International Society of Electrochemistry. (Prague, Czech Republic, 19-24 August, 2012). Zelenay, P., G. Wu, H. T. Chung, and C. M. Johnston . Elec- trocatalysis of oxygen reduction on non-precious metal fuel cell catalysts (invited keynote lecture). Invited presentation at Electrocatalysis: Present and Future (ELCAT Meeting). (Alicante, Spain, 14-17 November, 216

Page 217

Chemistry and Material Sciences Exploratory Research Final Report Innovative Process for Making Ultra-thin Dielectrics Quinn R. Marksteiner 20110246ER Abstract and low loss. This project set out to fabricate nonlinear dielectrics Scientific Approach and Accomplishments that exhibited a high degree of nonlinearity (dielectric constant vs. electric field) and low dielectric loss. In The first step to developing high quality materials was to addition, we attempted to fabricate these nonlinear develop useful diagnostics. At the beginning of the proj- materials into very thin slabs (~100 um), which could be ect, we were already able to measure the low frequency used in very high frequency applications such as nonlin- (< 10 MHz) dielectric constant and loss of the material as ear transmission lines. We developed the capability to a function of temperature. In the first half of the proj- diagnose and fabricate high quality, low loss nonlinear ect, we developed a resonant cavity method to measure materials using Barium Strontium Titanate. the high frequency (300 MHz – 2.5 GHz) dielectric con- stant and loss as a function of temperature [3]. We also Background and Research Objectives developed a test stand to measure the nonlinear depen- BSTO is a ferroelectric material that has a variety of dence of dielectric constant as a function of temperature interesting electromagnetic properties [1]. When the and electric field strength. material is below the Curie temperature, it is in the fer- These diagnostics were then used to scan a set of pa- roelectric state. In this state the material is very lossy (a rameters related to the chemical mixture and the fabri- large fraction of electromagnetic energy is converted to cation process of the materials. After a large amount of heat when it interacts with the material), and the mate- data-taking, it was found that the most important factor rial exhibits hysteresis. Above the Curie temperature, that determined the quality of BSTO was the chemical the material enters the paraelectric state. In this state, purity of the starting materials [3]. The combination of hysteresis disappears and the material becomes less and diagnostic development and optimization of LANL led to less lossy as the temperature increases above the Curie LANL having the capability to fabricate large quantities point. In this paraelectric state, the material is useful for of high quality BSTO. a wide variety of electromagnetic devices, because if its high dielectric constant, high nonlinearity, and relatively After developing the capability to diagnose and fabri- low loss at high frequency. Figure 1 shows our mea- cate high quality bulk ceramics, we begin to work on the surement of dielectric constant and loss vs. temperature, fabrication of thin laminates of materials, which could and Figure 2 shows our measurement of nonlinearity. be used to increase the frequency output of a NLTL. The fabrication of thin materials proved to be more difficult In particular, this project was focused on developing a than we had anticipated. The supporting material (MgO) technique for fabricating high quality ceramic blocks of tends to have a slightly different coefficient of thermal Barium Strontium Titanate (BSTO) that could be used expansion, which leads to small cracks forming in the in a Nonlinear Transmission Line (NLTL). A nonlinear BSTO during the baking process. Because BSTO has a transmission line is a periodic array of nonlinear materi- very high dielectric constant, even a microscopic crack als that can convert a capacitive discharge into a narrow- can significantly degrade the performance of the mate- band, high frequency radiation [2]. The frequency and rial. performance of a NLTL can be improved if the capaci- tance in a individual block of BSTO can be brought down. Although we did not succeed in fabricating thin lami- Because of this, a key goal of this project was fabricating nates of BSTO, we did succeed in fabricating laminates very thin slabs of BSTO that still were highly nonlinear 217

Page 218

and multilayer laminates of thick materials. These mul- tilayer laminates can be directly metalized, which greatly simplifies the process of constructing a device which con- sists of a large number of electrically connected slabs of BSTO. Figure 3 shows a picture of one of these laminates. We have also expanded our capability to diagnose materi- als, by developing a resonant cavity that can measure the high frequency properties of materials under a DC bias. This work was completed in the last few days of the proj- ect, and will be published. Impact on National Missions The capabilities developed in this project can support many national missions, particularly with regards to the Figure 2. Measurement of dielectric constant vs. bias development of NLTL technology for civilian and military applications. The output of an NLTL is very high power [2], and has the potential to be used as a nonlethal electro- magnetic weapon. BSTO is a material that has a wide variety of applications, many of which are of interest to the lab and the nation. For example, BSTO have been used in dielectric wakefield accelerators, which could lead to significant cost saving for future accelerators such as the MaRIE XFEL. Figure 3. Picture of multilayer laminate. References

1. Tagantsev, A. K., V. O. Sherman, K. F. Astafiev, J. Ven- katesh, and N. Setter. Ferroelectric materials for micro- wave tunable applications (vol 11, pg 5, 2003). 2005. JOURNAL OF ELECTROCERAMICS. 14 (3): 199.
2. IKEZI, H., S. S. WOJTOWICZ, R. E. WALTZ, J. S. DEGRASS- IE, and D. R. BAKER. HIGH-POWER SOLITON GENERA- TION AT MICROWAVE-FREQUENCIES. 1988. JOURNAL OF APPLIED PHYSICS. 64 (6): 3277.
3. Chen, C. F., Q. R. Marksteiner, G. King, T. A. Wynn, M. B. Treiman, D. A. Dalmas, A. Llobet, W. B. Haynes, D. R. Guidry, and P. A. Papin. Slip casting of sol-gel-synthe- Figure 1. Measurement of dielectric constant and loss vs. tem- sized barium strontium zirconium titanate ceramics. perature.
4. JOURNAL OF MATERIALS SCIENCE. 48 (17): 5788.
5. Publications
6. Chen, C. F., Q. R. Marksteiner, G. King, T. A. Wynn, M. B. Treiman, D. A. Dalmas, A. Llobet, W. B. Haynes, D. R. Guidry, and P. A. Papin. Slip casting of sol-gel-synthe- sized barium strontium zirconium titanate ceramics.
7. JOURNAL OF MATERIALS SCIENCE. 48 (17): 5788. 218

Page 219

Chemistry and Material Sciences Exploratory Research Final Report Embedding Plasmonic Nanostructures at Semiconductor Interfaces for Enhancing Photovoltaic Efficiency Todd L. Williamson 20110290ER Abstract modes at the particle’s surface, which is synthetically de- The remarkable optical properties and tunability over fined by their shape, size, and composition. Using these visible wavelengths of plasmonic nanostructures make properties, nanostructured metal particles will permit them attractive candidates for enhancing the perfor- the development of next generation materials that may mance of photovoltaic (PV) devices. Measureable im- be employed in numerous sophisticated technologies. provements in performance and efficiency of inorganic Conventional optoelectronic devices, such as photovol- semiconductor-based PV devices have been observed by taics (PV) are based on successful charge separation at placing plasmonic nanostructures on either their front or p-n semiconductor interfaces. Recent approaches to back surface. Ideally, plasmonic nanostructures should improving efficiencies involve multi layer stacks, use of be embedded into the active p-n junction in PV devices new and exotic materials, and careful selection of mate- to achieve new, higher levels of efficiency. The goal of rials with appropriate bandgap energies. However, there this project was to investigate new mechanisms for en- are currently no meaningful approaches that address hancing PV efficiency by embedding plasmonic nanopar- routes for enhancement at the p-n interface. In this ticles at the p-n junction in InGaN-based semiconductor work, we have sought to merge conventional optoelec- PV devices whose tunable bandgaps cover nearly the tronic semiconductor device geometries with advanced entire solar spectrum making them ideal materials for nanostructured metal particle synthetic techniques to high efficiency hybrid PVs. Our overall target was to produce novel optoelectronic materials that exhibit su- create the first PV devices with plasmonic NPs buried perior performance characteristics. The primary limiting at active semiconductor p-n junction interfaces with factor of these nanostructure architectures has been the tailored properties for higher conversion efficiencies. successful fusion of high-quality, low temperature thin These PV devices take advantage of near field interac- film growth techniques with chemically synthesized, uni- tions to manipulate their local electronic fields resulting form, optically active nanomaterials. This work involved in enhanced performance. The fundamental advances a combination of fabrication and synthetic capabilities in materials science and photophysics made through available at LANL to produce novel hybrid plasmonic- this work will drive the development of novel PV device photovoltaic architectures. This unique approach of architectures that far out perform current implementa- embedding plasmonic nanotructures at an active p-n tions. semiconductor interface enables direct manipulation of plasmon-exciton couplings that we desired to utilize for Background and Research Objectives enhanced device performance. Figure 1 shows a sche- The emerging field of plasmonics has been facilitated by matic of a concept device from this work. dramatic advances in control over the design and imple- Scientific Approach and Accomplishments mentation of nanostructured materials. One distinctive aspect of these advanced materials is their exceptional The fabrication of buried plasmonic NPs at semiconduc- optical properties relative to bulk and/or molecular tor interfaces involves the combination of vacuum-based counterparts. More specifically, metal nanoparticles are thin film growth and solution-based chemical synthesis. increasingly being utilized for diverse applications that Leveraging these two techniques, we have attempted to range from biological sensing to implementations of sub- build novel device architectures for exploiting enhanced wavelength optical lenses that circumvent the traditional photon-plasmon-exciton couplings using four research diffraction limit. The key to these materials success has thrusts: been the efficient control and excitation of plasmonic 219

Page 220

Nanoparticle synthesis: The targeted plasmonically active investigated using e-beam lithography to directly pattern PV devices inherently involve a wide variety of NP architec- structures onto the surface. This process is slower than tures available through several synthetic approaches. The the colloidal approach, but offered the flexibility of a wider most rudimentary particle designs involve colloidal Au and array of geometries and a cleaner process to improve elec- Ag that are prepared using standard wet chemical syn- trical metrics. thetic techniques allowing plasmon excitation across much Buried Interface characterization: The general approach of the visible spectral region (400nn to 600nm). Increasing we have followed to form buried nanostructured PV in- the particle complexity using nanoshell particles allows terfaces relies heavily on a thorough understanding of the access to a wider range of plasmon energies. Control of photophysics associated with both plasmons and semicon- overall particle size and dispersion are important for ensur- ductor excitons. To this end, thorough charac-terization of ing that well defined plasmon resonances are available at all thin film growth stages (and their functionalization with the semiconductor interfaces to observe their interaction plasmonic NPs) is vital for comprehending how the NPs with the local electronic band structure. maintain their properties during semiconductor deposition Semiconductor Film Deposition: The targeted semiconduc- and on the quality of the overgrown semiconductor films. tor system for this work is the InxGa1-xN ternary alloy sys- Initially, plasmonic nanostructures on the semiconductor tem. This important semiconductor system has a tunable layers were extensively characterized SEM prior to over- direct bandgap ranging from 0.7 eV (1800 nm) to 3.4 eV growth of the capping semiconductor layer to determine (365 nm), a range that encompasses nearly the entire solar their surface coverage. Enhanced PV performance should spectrum. In order to take full advantage of plasmon- require minimal particle coverage (e.g. on the order of semiconductor interactions, the resonance frequency and a few percent of a ML) this minimizes lattice defects in the semiconductor bandgap need to be carefully chosen overgrown semiconductor layers. XRD was used to exam- and optimized to yield the maximum enhancement. InGaN ine film crystallinity. Following structural characterization, is the only known direct bandgap semiconductor alloy the devices’ electrical properties both with and without system that has the bandgap tunability over this critical illumination were studied. energy range that favorably overlaps with readily achiev- Accomplishments: This project has had a number of no- able plasmon resonance energies. Our film deposition table accomplishments: approach utilizes energetic neutral atom beam lithography & epitaxy. ENABLE is a MBE-type film growth technique Growth and n and p type doping of InGaN over a wide wherein energetic nitrogen atoms are used to initiate the range of compositions relevant to the goals of this project. chemistry required for growth of epitaxial nitride films. Of See Figures 2 & 3 for electrical data of relevant films. importance is the ENABLE system’s capability of growing semiconductor films at low temperature, critical for suc- Growth and evaluation of control junctions (i.e. without cessfully incorporating active plasmonic NPs at the semi- NPs) of InGaN conductor p-n interfaces. Typical substrate temperatures during growth are ~500 - 600°C, well below the melting Deposition of uniform colloidal NPs on the surface of temperature of the NPs, but high enough to desorb any InGaN insulating ligands from the colloidal synthesis. ENABLE is used for growing both the initial InGaN film (p- or n-type) Deposition of e-beam lithography patterned plasmonic and the capping film (opposite doping) on top of the plas- structures on InGaN monic structures. Plasmonic structures will be deposited Overgrowth of InGaN on top of previously grown InGaN in between growth steps. that has been patterned with NPs. The overgrown films Nanoparticle Deposition: Some of the plasmonic sensitiz- did not show any structural degradation by XRD ers are prepared via solution phase colloidal synthesis, and Impact on National Missions require efficient means of transfer onto the semiconductor surface using different deposition approaches depending This work will directly impact LANL’s emerging core mis- on the particle type and the surface’s properties. After sions in energy and materials science. The fundamental evaluating a number of methods, we determined that understandings into plasmon-exciton interactions and placing a solution containing the particles on the desired their direct manipulation are of broad interest to both the InGaN wafer and then dragging a slide across the surface nanomaterials and optoelectronics communities. Apply- led to the best distribution and uniformity of coated NPs. ing principles from these the emerging fields of plasmonics In addition to deposition of colloidal NPs, we have also and photovoltaics based on InGaN alloys, this work can have direct implications on the next generation of PV tech- 220

Page 221

nologies that are of significant interest to energy indepen- dence, national security, and environmental missions that are ongoing at LANL and the DOE complex. Direct impact in LANL threat reduction and energy security missions via applications in sensing and improvements in semiconduc- tor-based energy harvesting materials is also possible. This project positions LANL for numerous external follow-on funding calls from various agencies. Figure 1. Schematic diagram of a proposed hybrid plasmon- Figure 3. Electrical data from p-type InGaN with ~25% In content photovoltaic device. and a electronic bandgap of ~2.2 eV. Publications Hoffbauer, M. A., T. L. Williamson, J. J. Williams, J. L. Fordham, K. M. Yu, W. Walukiewicz, and L. A. Reichertz. In-rich InGaN thin films: Progress on growth, compositional uniformity, and doping for device applications. 2013. JOUR- NAL OF VACUUM SCIENCE & TECHNOLOGY B. 31 (3): -. Williams, J. J., T. L. Williamson, M. A. Hoffbauer, A. M. Fischer, S. M. Goodnick, N. N. Faleev, K. Ghosh, and C. B. Honsberg. Inducing a junction in n-type InxGa(1-x)N. 2013. JOURNAL OF VACUUM SCIENCE & TECHNOLOGY B. 31 (3): -. Williams, J. J., T. L. Williamson, M. A. Hoffbauer, Y. Wei, N. N. Faleev, and C. Honsberg. Growth of high crystal quality InN by ENABLE-MBE. Physica Status Solidi C. Williamson, T. L., A. L. Salazar, J. J. Williams, and M. A. Hoffbauer. Improvements in the compositional uniformity of In-rich InxGa1-xN films grown at low temperatures by ENABLE. 2011. PHYSICA STATUS SOLIDI C: CURRENT TOPICS IN SOLID STATE PHYSICS, VOL 8, NO 7-8. 8 (7-8): -. Figure 2. Electrical data from n-type InGaN with ~25% In content Williamson, T. L., J. J. Williams, J. C. D. Hubbard, and M. A. and a electronic bandgap of ~2.2 eV. Hoffbauer. High In content InxGa1-xN grown by energetic neutral atom beam lithography and epitaxy under slightly N-rich conditions. 2011. JOURNAL OF VACUUM SCIENCE & TECHNOLOGY B. 29 (3): -. 221

Page 222

Chemistry and Material Sciences Exploratory Research Final Report Bio-analogue Catalysts: Evolved Aptazymes for the Hydrolysis of Organophosphorous Compounds Robert F. Williams 20110338ER Abstract structures. Isolated from high diversity combinatorial Aptamers are nucleic acid binding species capable of rec- libraries of synthetic nucleic acids by a rapid iterative se- ognizing a wide range of target molecules and biologi- lection process, aptamers can be selected and obtained cal structures with varying conformational flexibilities. for effective binding recognition as well as for a designed Isolated from high diversity combinatorial libraries of catalytic function (aptazyme). Systematic evolution of synthetic nucleic acids by a rapid iterative selection pro- ligands by exponential enrichment (SELEX) is an in vitro cess, aptamers have been selected for effective binding selection technique that utilizes combinatorial chemistry recognition as well as for a designed catalytic function to produce nucleic acid ligands with high binding affin- (enzyme mimic or aptazymes). This catalytic function ity and specificity toward a desired target. The attrac- is universal for multiple classes of phosphorus bonds tion of nucleic acid-based combinatorial chemistry lies found in chemical warfare nerve agents (CWAs). Our first in its potential for amplification and directed evolution. selection protocols sought to find aptamers to hydro- Mimicking the process of natural selection in the labo- lyze a P-O bond as an initial target. Additional aptamer ratory setting, repeated cycles of selection for specific libraries were constructed to target other phosphorus high-affinity binding with increasing stringency, followed bonds such as P-S and P-N bonds and optimum apta- by polymerase chain reaction (PCR) amplification can zymes were selected for these specific bond types with lead to the isolation of oligomers ‘evolved’ to bind tightly varying degrees of success. Initially, our three aptazyme and selectively to a target of choice. These selected classes were designed individually to function on P-O, aptamers have demonstrated a remarkable affinity for P-S, and P-N bonds; however, isolated aptazymes were their targets, with dissociation constants typically in only successful for P-O and moderately successful for P-S the micromolar to sub-nanomolar range. For all of their bond types. These three bond types represent all known diversity, however, the ability of nucleic acids to perform classes of G and V organophosphorus CWAs (e.g. sarin biological functions such as binding and catalysis is lim- (GB) and VX) as well as the majority of pesticides in com- ited when compared to the rich chemical functionality mercial use. We believe future inclusion of non-natural of proteins. To enhance aptamer capabilities, synthetic nucleosides in the aptamer sequence will allow for en- chemistries for modifying nucleotides have been devel- hanced hydrolysis of P-S bond types as well as hydrolysis oped to increase the potential of aptamers for binding, of the P-N bond type. Our selected apatazymes for P-O catalysis, and stability. bonds are efficient catalysts that hydrolyze a compound While this project was designed to develop novel cata- with that bond type to eliminate the compound’s toxic lysts as enzyme mimics, these catalysts are not designed effects. While initial aptazymes have been shown to to hydrolyze the phosphorous-fluorine bond of organo- effectively catalyze the hydrolysis of organophosphorus phosphorous (OP) compounds (e.g. nerve agents) as has compounds with P-O bond types, additional develop- been traditionally done for destruction or decontamina- ment of aptazymes for P-S and P-N bond types is needed tion of such OPs. Rather, these enzyme mimics, called to enhance and increase their binding, catalytic activity, aptazymes, were designed to target other specific phos- and overall stability. phorous bonds that are ubiquitous in OP compounds such as P-O, P-S, and P-N bonds. Breaking or hydrolyzing Background and Research Objectives these bonds, instead of the reactive P-F bond, produces Aptamers are nucleic acid (DNA; deoxyribonucleic acid a highly unstable intermediate that rapidly and sponta- or RNA; ribonucleic acid) binding species capable of rec- neously decomposes to an inert, non-toxic product. This ognizing a wide range of target molecules and biological 222

Page 223

approach allows a universal capture and breakage of bonds lenge our aptazyme libraries. Figure 1 shows the structures to phosphorous found in all chemical weapon agents of a substituted cyclic phosphoroamidate (TS model) or (CWAs) as well as all related OP toxic industrial chemicals a phosphate diester (GS model) used for this approach. and materials (TICs/TIMs) such as pesticides. Develop- After coupling to an immobilized support, using standard ment of a robust toolbox, which specifically targets bonds coupling chemistries, a nucleic acid combinatorial library for catalytic breakage, has important ramifications for the containing 50-mers or higher, as required, was developed chemical and pharmaceutical industries, for therapeutic and exposed to the immobilized analogs under controlled applications, and for our fundamental understanding of conditions of pH, temperature, and ionic strength. A typical biological catalysis. uncatalyzed hydrolysis half-life at pH 7 ranges from 10-30 hours for the most reactive class of compounds, depending Construction of DNA aptazymes was predicated, partially, upon the phosphorous ligands allowing sufficient time to on the recognition of the robust role these molecules have select for binding and enhanced hydrolytic activity. Re- played in the origins of enzymatic function in the evolu- moval of unbound nucleic acid species, release of bound tion of life and their specific abilities to make and break aptamers, amplification of selected aptamers, and re- phosphate bonds. With the high structural and functional exposure to bound targets was sequentially repeated with diversity space available, we proposed that unique apta- increasing stringency to produce aptamer structures with zymes would identify, recognize, and hydrolyze specifically optimal binding and, ultimately, hydrolytic rate. Kinetic targeted side chains of OP compounds. Consequently, this competition between the selected aptamers and BChE was project has produced rationally designed, highly effective used to evaluate aptamer efficacy. The sequence of the biological catalysts that mimic naturally occurring enzymes; selected aptamers allows us to synthesize optimal apta- yet, are more easily manufactured, effectively employed in zymes, de novo, with high fidelity and low cost. the field for remediation, can be developed for biomimetic sensor applications, and may be useful in prophylactic As shown pictorially in Figure 2, the synthesized combi- roles for both the environment and personnel. Addition- natorial library, with a diversity of approximately 1015, ally, their continued study will allow development of more was bound to a streptavidin gel through a biotin moiety complex enzyme mimics by providing a more fundamental and remained immobilized on the gel unless the desired understanding of molecular recognition and the role of bond of the phosphonate, phosphate, or transition state conformational flexibility in enzyme catalysis. analog target molecule was hydrolyzed (Figure 3). The initial library was constructed by template-directed ex- Also, this project has addressed a critical LANL mission by tension of 5’-biotin-d(GGAAAAA)r(linker-target-linker) developing a new technical means to reduce the global d-(GGAAGAGATGGCGAC)-3’ on 5’-GTGCCAAGCTTAC- threat of existing and emerging CWAs and thwart terrorist CG-N50-100-GTCGCCATCTCTTCC-3’ (where N = G, A, T, or attacks that might employ them by providing a new tool to C). Extension products were purified by non-denaturing remediate an exposure event. polyacrylamide gel electrophoresis (PAGE). Optimization of the library and increased catalytic efficacy utilized concen- Scientific Approach and Accomplishments tration modifications, deletions, and/or changes in ionic Our goals for this project were the following: strength. Ineffective catalysis obtained in the initial N50 libraries; consequently, the random oligonucleotide region To construct, using a combinatorial approach, highly ef- was increased from N50 to N100. Additionally, modifica- ficient and selective bio-analogue catalysts comprised of tions to extend the linker portion on the aptamer side of DNA oligomers (aptamers) that will break a phosphorous the linker-target-linker region were done until binding/ (P-O, P-S, or P-N) type bond. These designed enzyme mim- hydrolysis was observed. Both linkers were comprised of ics or aptazymes will exhibit molecular recognition and ~2 kDa polyethene glycol (PEG) units with a short oligo- have the ability to degrade, specifically and catalytically, nucleotide sequences on either the 3’ side for Linker 1 or OP compounds with the above types of bonds. 5’ side for Linker 2. Additionally, Linker 2 terminated on the 5’ side with a biotin moiety for attachment to a streptavi- To optimize the catalytic turnovers of these aptazymes by din solid support for immobilization. The 5’ side of Linker elucidating and understanding the conformational con- 1 and the 3’ side of Linker 2 were functionalized to allow straints that controls the observed catalysis. easy covalent incorporation of the target molecules shown In order to accomplish these goals we attempted to syn- in Figure 3. The ligation reaction to couple the target with thesize two general types of analogs that mimic either the the linkers utilized the high yielding and extremely effi- transition state (TS) or the ground state (GS) structure of cient Cu(I)-catalyzed Huisgen [2+3] dipolar cycloaddition a P-X bond targeted for breakage and use them to chal- reaction between an organic azide and a functionalized 223

Page 224

acetylene to form a 1,2,3-triazole (Click Chemistry) under phosphorous through another polyethylene glycol linker mild physiologic conditions. Typically, the azide functional- (PG-Linker1) through an alkyl bond to the phosphorous ity was attached to PEG Linker 1 (PEG-N3) and the acety- (PG-Linker1-C-P). The PG-Linker1-P bond is not hydrolyz- lene moiety was incorporated into the ligand side chain able. The second key accomplishment was the synthesis of attached to the phosphorous coupled to Linker 1 (Figure the complete phosphonate containing polyethylene linkers 3). PEG Linker 2 was prepared with a terminal acetylene on both sides of the molecule. One terminated with biotin moiety (PEG-C≡CH) and attached to a ligand side chain that and the other terminated in an azide functional group (N3- contains an N3 (Figure 3). Although this procedure can PG-Linker1-C-P-O-Linker2-B. When the complete complex result in dimers, performing the initial coupling reaction, construct was attached to the DNA aptamer library using regardless of choice, with a large excess of linker generally copper mediated click chemistry, a ground state transition prevented this from occuring. model for hydrolysis was produced (DNA-Library-CCN-PG- Linker1-C-P-O-Linker2-B). A third key accomplishment was Only when the correct hydrolysis event occurs can the 3’ design and synthesis of the DNA aptamer library terminat- end of the molecule be uncoupled from the biotin; thus ing in an alkyne moiety for the click coupling to the azide allowing removal of the hydrolyzed 3’ molecule from the terminated moiety containing the phosphorus described gel. It was ultimately found that using a column of pre- above. Additionally, the primer molecules for PCR amplifi- attached libraries resulted in poor recognition although cation of the selected DNA aptamers were also designed; hydrolysis was observed. This was due to our inability to however, additional design work on these primers to fur- control the spacing of each library unit resulting in cross- ther optimize the amplification of aptamers that show hy- reaction by adjacent library units on the column. The drolytic activity is needed. The fourth key accomplishment solution to this problem was to use a dilute solution of the was a redesign of the selection strategy to take advantage libraries and then separate all the non-hydrolyzed library of temperature as a selection stringency parameter and to units by chromatography on a streptavidin column (vide eliminate the cross-linking interactions caused by high den- infra). After elution of the 3’-molecule-linker-target moiety, sity placement of the complete complex construct (DNA- these oligonucleotide sequences, which showed hydrolytic Library-CCN-PG-Linker1-C-P-O-Linker2-B) on a streptavidin- activity, were amplified by standard PCR, purified, and linked solid support. As mentioned above, if the complex again subjected to primer template-directed extension to constructs are too close together they can interact with an allow for a subsequent round of selection. As more rounds adjacent DNA aptamer to hydrolyze its neighbor phospho- of selection were undertaken, more stringent criteria, such nate rather that the phosphonate to which it is actually at- as variations in concentration, metal ion co-factor, salt, tached. This produces aptamers that are not truly compe- temperature, and exposure time, were evoked to enhance tent with correct conformations to the pool of successful the desired catalytic activity and specificity. To guaran- aptamers. However, dilute solutions of the DNA aptamer tee that the aptamer distinguishes between a target and library complexes, first heated to ~85 deg C to denature a known closely related molecule (protein, peptide) we the aptamers followed by metal addition and cooling to al- employed alternate rounds of positive selection taking only low the DNA aptamers to refold prior to any reaction, was those molecules that bind with negative selection for the a successful strategy. Our fifth key accomplishment was related, but undesired molecular targets (taking only those the selection and isolation of four competent aptamers molecules that do not bind). that hydrolyzed the P-O bond of the ground state model. These aptazymes, characterized by sequencing and NMR, The first key accomplishment of this project was the design showed rates of hydrolysis that were approximately 104 and implementation of a synthetic route to a phosphonate to 105 higher than the base hydrolysis rate of diisopropyl- compound that is coupled to a polyethylene glycol linker fluorophosphate in a standard pH7 buffer. One additional that ends in a biotin moiety. This was required in order to aptamer that showed a 102 rate enhancement for meth- select a specific phosphorous-ligand (oxygen, sulfur, or ni- amidophos (P-S bond) was also isolated. trogen) bond that will be selectively hydrolyzed by the DNA aptamers. The initial target molecule was a phosphonate Review board recommendations at the end of the second that contains the P-O-Linker2-B bond linkage in the ground year suggested our focus in the third year be redirected state. The oxygen is connected to both the phosphorous from our original design protocol to take advantage of add- and the polyethylene glycol linker that ends in the biotin ing functionalized dUTP’s in the DNA cocktail to produce moiety (Linker2-B). Aptamers that successfully hydrolyzed aptazymes with enhanced binding and catalysis resulting the target P-O bond were recovered for subsequent ampli- from the enhanced functionalities on the nucleic acids fication. On the other side of the phosphorus containing (e.g more protein like). Selection against a transition state molecule, the DNA aptamer library was attached to the analogue was also suggested to be a more important 224

Page 225

goal since we had already demonstrated hydrolysis with a ground state model. Unfortunately, problems with person- Linker 1 Target Linker 2 5’ B nel turnover and delays in the synthesis of the transition P-X state model compounds did not allow us to complete an optimized selection before the project ended. Neverthe- less, hydrolysis was demonstrated but at rates lower than N – N observed with the aptazymes generated with the ground 50 100 3’ state model. While additional work in this area is war- DNA Library ranted to further enhance the catalytic efficacy of these aptazymes and to test them against a wider variety of OP Figure 2. S cheme Describing the Design of the Combinatorial Li- compounds, our results are extremely promising. brary of DNA that will Hydrolyze the Target P-X Bond. The 5’ End is Connected to a Biotin (B) to Bind to Streptavidin and Impact on National Missions Linker 1 and 2 are Described in Figure 3. The 50 to 100 Random The development of technologies that allow for the rapid Sequence Deoxyribonucleic Acid Library. and specific identification/elimination of chemical threat agents in the environment and mitigation of adverse effects on exposed individuals are essential for threat reduction and mitigation. A primary mission goal of the Laboratory is to develop the technical means to reduce the global threat of existing and emerging chemical and biological weapons of mass destruction and to thwart terrorist attacks that might employ them by providing the necessary science and technology. Consequently, develop- ment of effective countermeasures to organophosphorus CWAs, TICs, TIMs, and related compounds are critical to LANL’s nonproliferation activities and mission to prevent or reduce the capability for rogue nations or terrorists to do our nation harm. Our systematic experimental approach Figure 3. Design of Linker1-Target-Linker2 Used in Selection has provided new tools for the discovery of enzyme-like Scheme for the Ground State Model (Right) and the Transition function, a core objective for our understanding Complex State Model (Left). Systems of biological function. X, Y, or Z = O, C, N, S O O Y P P X Linker Linker Z Y Z X Linker Linker Cyclic Phosphoamidates Functionalized CW Agent or a or Phosphate Diester as a Related Chemical Threat Agent as Transition State Model for a Ground State Model for P-X P-X Bond Hydrolysis Bond Hydrolysis Figure 1. Representations of a Transition State Model and a Ground State Model for DNA Aptamer Selection where the P-X Bond is the Target Bond to Cleave. 225

Page 226

Chemistry and Material Sciences Exploratory Research Final Report Developing a Mild Catalytic Route for the Reduction of N to NH 2 3 John C. Gordon 20110358ER Abstract can then be optimized to improve catalyst efficiency and The research in this project concerns the development efficacy through a feedback loop of experimental and of catalysts capable of supporting the conversion of di- theoretical investigations. nitrogen (N2) into ammonia under mild conditions using Scientific Approach and Accomplishments the earth-abundant metal iron (Fe). Largely due to the demands of food production, industrial generation of Towards our aims above, we have been using calcula- ammonia (NH3) consumes more than 1 % of the world’s tions to guide both our design strategies and synthetic energy output as a result of the energetically in-tense efforts related to target catalytic species. We have been conditions employed. Enzymatic ammonia production able to isolate a very interesting (and rare) example of a occurs under ambient temperature and pressure, and di-iron complex containing a bridging N2 ligand, in which translation of this efficiency to a synthetic system would the bound N2 fragment appears to be highly activated significantly decrease energy production demands. We (based on metric data from the X-ray structure of this aim to pursue this goal through a heretofore unexplored molecule, as well as from Raman spectroscopy). Recent reaction manifold designed to maximize nitrogen activa- calculations are pointing towards the significant reduc- tion and minimize the number of intermediates neces- tion of the bond order within the bound N2 fragment, sary for ammonia production. The interdependent use thus consistent with our original hypothesis that a ze- of experiment and theoretical calculations will be used rovalent di-iron complex should significantly weaken the to realize secondary objectives of advancing scientific N2 ligand and promote its susceptibility towards further understanding of dinitrogen functionalization in addition functionalization. We have also learned more about to probing the boundaries thereof, understanding the ef- electron density within this molecule and how this might fect of varied electronic structure on nitrogen reactivity, impact the activation of the N2 moiety and its propen- and exploring new Fe coordination chemistry. sity to undergo reactivity to form NH3 or other reduced nitrogen products. Background and Research Objectives In parallel with computational work, we have been The research outlined in this project seeks to lower the investigating the chemistry of the Fe2N2 fragment with energetic and economic footprints of NH3 production reagents that should be capable of cleaving the bound via the development of base metal catalysts that are N2 ligand, thus forming another of the intermediates capable of converting N2 into NH3 under mild condi- necessary to complete our targeted catalytic cycle (a tions. We are approaching this goal by incorporating trivalent iron nitride fragment). lessons learned from previous studies on the bonding and reactivity of N2 and using them to design catalysts We have also invoked the use of Mossbauer spectros- that maximize N2 activation and minimize the number of copy as a probe of electronic structure within these iron potential intermediates involved in catalytic turnover in complexes, as this is a tool that will help rationalizeboth order to confer robustness to the catalytic cycle. the reactivity of these molecules as well as their elec- tronic structure and bonding properties pointed towards In pursuit of this objective, we also aim to advance the by calculation. current understanding of how electronic configuration and ligand steric bulk affect the ability of iron (Fe) to As a result of our work, we are beginning to gain a more reductively cleave N2, functionalize the resulting nitride, detailed picture of the electronic structure of such and release NH3. These catalyst design characteristics 226

Page 227

classes of metal-bound N2 molecules. In particular we are gaining more insight into which oxidation states and spin states of iron are required for N2 activation, and what kinds of ligands are need to support such molecules. Impact on National Missions Developing a process to affect the low cost reduction of N2 to NH3 under mild conditions remains a major chal- lenge in the field of modern chemistry. Although the fundamental chemistries of this transformation have been demonstrated, an economically viable catalyst system that also exhibits reasonable turnover numbers has thus far been elusive. Research in this area of catalysis for energy will both advance our fundamental knowledge about the stepwise reduction of N2 at a cheap and abundant transi- tion metal center and potentially lay the groundwork for a less energy intensive NH3 production scheme than the current Haber-Bosch process. This work directly supports our mission in energy security. Publications Mukhopadhyay, T. K., R. K. Feller, F. N. Rein, N. J. Henson, N. C. Smythe, R. J. Trovitch, and J. C. Gordon. Investi- gation of Formally Zerovalent Triphos Iron Complexes. 2012. ChemCommun (Chemical Communications). 48: 8670. 227

Page 228

Chemistry and Material Sciences Exploratory Research Final Report Determination of Fluid Properties at Carbonate Interfaces - An Integrated Experimental and Theoretical Approach Donald D. Hickmott 20110363ER Abstract geologic reservoirs. Surprisingly, the fundamental phys- Phenomena at the interface between carbonate miner- ics of mixed-volatile fluid interactions with carbonate als and CH4-CO2-H2O-rich fluids at elevated pressure and other minerals at the Ps and Ts of energy extrac- (up to 200 MPa) and temperature (up to 400o C) control tion remain poorly understood; this is largely due to a number of energy-relevant processes, particularly the extreme difficulties of both doing experiments and extraction of hydrocarbons from unconventional oil developing accurate theoretical methods at such condi- and gas reservoirs and sequestration of CO2 in deep tions. Thus, the overarching objective of this research is geologic formations. We developed and tested a novel to use integrated experimental and theoretical methods high pressure-temperature experimental cell for neutron to better understand high P-T carbonate/fluid interface reflectometry of fluid-solid interfaces to 200 MPa and properties, particularly density and related transport 200o C. The cell has wide application in Earth and energy properties, as a function of P-T. Density is a particularly science and in corrosion studies. Neutron and x-ray important property to understand because it can be studies of calcite thin films in contact with calcite-sat- related to wettability, which controls fluid flow in Earth’s urated fluids have revealed positive density anomalies crust. Understanding fluid/mineral interfaces at a fun- at ambient pressure and temperature consistent with damental level opens up the possibility of manipulating hydrophilic (highly wetting) interactions. Synchrotron surface structures and free energies and thereby control- x-ray studies also reveal structural rearrangement of the ling reaction driving forces and kinetics during fluid-rock films due to contact with fluids; both amorphization and interactions. Surface diffusion is crucial in high-surface crystallographic re-orientation are observed. Theoretical materials such as shales, due to the potential control on calculations using molecular dynamics and kinetic Monte fluid flow (CO2 infiltration or hydrocarbon extraction) by Carlo methods simulate the density anomalies and show dynamic diffusive processes. that the diffusion of CaCO3 monomers on calcite surfac- Scientific Approach and Accomplishments es with increasing temperature (300-800o K) at elevated pressure is complex. Such diffusion shows a reversal in We pursued a combined experimental and theoretical rate at 600o K related to minimization of the desorp- approach to address this key problem. On the experi- tion free energy between 600-700o K. Simulations also mental side we developed novel high P-T capabilities for suggest that two dimensional surface clusters below a neutron investigations of interfaces and began to apply critical size of a few nanometers preferentially exist as these to calcite (CaCO3) thin films, whereas on the theo- disaggregated calcite chains; this has ramifications for retical side we used state-of-the-art molecular dynamics surface growth and dissolution mechanisms of calcite. methods to model fluid and interface processes and dy- namic transport phenomena at calcite/water interfaces Background and Research Objectives at high P-T. Understanding and controlling interface processes for We developed a novel elevated P-T cell for neutron carbonate minerals is crucial for a range of energy and reflectometry (Figure 1) [1]. Neutron reflectometry is environmental applications. Carbonate minerals interact a sensitive method for studying processes at interfaces with mixed-volatile H2O-CO2-CH4-rich fluids at elevated with angstrom-scale resolution, but has never been ap- pressures (P) and temperatures (T) during extraction of plied at the elevated Ts and Ps of hydrocarbon extraction hydrocarbons from carbonate reservoirs, during produc- or carbon sequestration. The new cell is constructed of tion of geothermal energy, at some subsurface nuclear aluminum and can support both aqueous and mixed- waste sites, and during sequestration of CO2 into deep 228

Page 229

volatile fluids; measurements can be made in-situ while water interface over a range of Ts (300-800o K) at elevated conditions are changing, allowing the study of dynamic P (~ 100 MPa) [4]. Water density profiles perpendicular to processes. The Q range of the cell is 0.31 angstrom-1. Stud- the calcite surface were determined and showed an elevat- ies of corrosion of an aluminum film demonstrate the abil- ed density region. The density curve was more affected by ity to probe surface degradation mechanisms and interface T than P with sharp density peaks associated with surface reactions at elevated P-T conditions. This cell is capable of structured water (< 6 angstroms) at lower Ts (< 400o K), studying density profiles near interfaces at Ts up to 200o C suggesting less water ordering at higher Ts. These calcula- and Ps up to 200 MPa; this P-T range represents conditions tions are consistent with the neutron reflectometry results. in the Earth at depths up to ~ 6.5 kilometers under a range At ~ 100 MPa water density decreases monotonically from of geothermal gradients. The cell represents a new capabil- 1.03 to 0.48 g/cm3 from 300o to 800o K and this change ity for the Los Alamos Neutron Scattering Center (LANSCE) in density affects the solvation of calcite molecules in and for the nation; it should be applicable to a range of solution. Calculations of free energy barriers for an extra Earth science problems and to problems in materials sci- CaCO3 molecule to desorb from the calcite surface over a ence such as high P-T corrosion, catalysis, and materials range of Ts (300-800o K at ~ 100 MPa) shows that the ratio synthesis. of the desorption barrier to the thermal energy initially decreases with T and reaches a minimum at 600o K. Above We applied neutron reflectometry to calcite thin films that T the trend reverses due to the relatively faster in- (Figure 2). A scattered-length-density contrast-matched crease in the barrier which is a result of poor water solva- deuterated water/hydrogenous water mixture was used tion. This unusual phenomenon also affects the dynamical to highlight density contrasts within a few angstroms of behavior of the extra CaCO3 molecule at the interface. We the calcite surface. Initial data shows that the near-surface calculated the residence times in different surface states region in the fluid has elevated density (at room P-T) (i.e. surface and bulk) to study the diffusion of the CaCO3 compared to bulk fluid, suggesting hydrophilic interactions monomer at the interface. From the distribution of dura- and rapid interface transport at these conditions. The thin tion times and the corresponding diffusion coefficients in films were produced using an atomic-layer chemical vapor different states, we generated a family of dynamic trajecto- deposition (ALCVD) method [2]. We characterized these ries and computed the effective diffusivity as a function of thin films with a range of analytical techniques, including T. The effective diffusivity increases with T to 600o K then scanning electron microscopy, surface profilometry, and begins to decrease (Figure 4) , indicating that the adsorbed synchrotron-based surface x-ray scattering methods at the species diffuses relatively less through the bulk liquid than Advanced Photon Source (APS)(Figure 3). The APS results on the interface. The calculations also suggest surpris- showed: 1) that the calcite thin films were polycrystalline ing T-dependent crystallographic orientation dependence and that their crystallographic orientation was dependent of diffusion rate on the (1014) surface. The extent of the of the conditions during ALCVD; and 2) that the calcite bulk-mediated diffusion strongly depends on the ambient thin films re-oriented during water exposure. Films depos- condition for the calcite/water system. ited at 250o C on an alumina buffer layer amorphized in contact with calcite-saturated fluids whereas those depos- It has been recently shown that amorphous CaCO3 plays ited directly on single-crystal quartz crystallographically a critical role in the formation of other polymorphs of reoriented [3]. Films deposited at higher temperatures CaCO3 [5][6]. In order to identify the equilibrium con- are more strongly subject to fluid-induced amorphization. figuration of aqueous calcite (1014) interfaces at different These results suggest careful diagnostics prior to or during P-T, we applied the parallel tempering method. In parallel high P-T neutron reflectometry are vital. Future elevated tempering, multiple MD trajectories evolve at different P-T neutron reflectometry studies of these calcite films are Ts. Periodically, any two trajectories can swap configura- planned. tions using a Metropolis-like Monte Carlo algorithm. Since high-T trajectories explore larger regions of phase space, The use of inelastic neutron (INS) scattering to probe the exchanging configurations can significantly improve sam- disposition of methane on calcite surfaces was explored. pling at low Ts. As an added bonus, one can show that all INS provided insights into the vibrational spectra of meth- the trajectories correctly sample their respective canonical ane monolayers on calcite compared to bulk methane. ensemble. We used parallel tempering as implemented in Such observations may be useful in optimizing methane the LAMMPS package to simulate aqueous calcite (1014) extraction from carbonate-rich unconventional gas reser- interfaces. Various surface coverages (100, 70 and 50%) voirs. were studied. We used a chain of 64 Ts ranging from 300o to 1000o K. The results show that incomplete layers are Molecular dynamics (MD) and kinetic Monte Carlo simu- unstable. Instead, the calcite surface tends to form disag- lation methods were used to model the calcite (1014) / 229

Page 230

gregated calcite chains as shown in Figure 5 (a) [top and side views of the original surface are shown in Figure 5 (b) and (d) respectively.] For the same coverage (70%), this result was confirmed by instead starting the simulation from an energetically stable structure [c.f. Figure 5 (c)]. These results have interesting ramifications. They suggest a transition in the morphology of surface clusters can be expected as a function of their size: small clusters would be found in a disaggregated state, but they would eventu- ally adopt a conventional layer structure with increasing size. This could have interesting consequences for calcite growth and dissolution, as the unraveling of such disaggre- gated calcite structures at steps or kink sites could affect the surface morphology and the growth mode. For geo- logical systems, this surface destabilization process could Figure 1. Cross section of aluminum high pressure-temperature cell for neutron reflectometry. be common, especially when the P-T vary dramatically. Also, sudden changes in super-saturation could result in rapid precipitation, where clusters (instead of monomers) would deposit on the surface. This suggests that a complex dynamical process, where the interplay between cluster formation and precipitation and crystallization from an amorphous state, would determine the final morphology. This phenomenon is not only fundamental from the geo- logical point of view, but it could also play a crucial role in the mineralization processes that take place in living organ- isms. While surprising, these observations are consistent with literature results. It has been shown that small CaCO3 particles (<3.8nm in diameter) are amorphous, but that larger ones are crystalline [5]. In our simulation, the cell Figure 2. (left) The reflectivity curve for scattered-length-density size is about 2 nm, which is consistent with the amorphous (SLD) matched water in contact with calcite thin film deposited (disaggregated calcite) structure of the surface. on Al2O3 substrate. Heavy line represents best-fit calculated SLD profile. (right) Real-space best fit profile. Impact on National Missions Chi Scan of Calcite (104) Carbonates, particularly calcite, are extremely important Plane in Air Dry (In Air) 60nI In solution Run 1 in a range of energy and environmental applications. n et In solution Run 2 s I w n i t t e h r a c c a ti lc o it n e s i o s f a h k y e d y r o p c r a o r c b e o ss n , d c u a r r in b g o n e i x c t , r a a n cti d o a n q , u p e a o r u ti s c u fl l u a i r d ly s 105 FWHM = 6.9° 40 u o c( yti unconventional extraction, of oil and gas from carbonate nt s) 20 ) st n reservoirs and in CO2 sequestration. The new capabilities c o u Calcite (104) (high P-T cell and theoretical advances) and understanding y ( 40 80 120 sit Chi (degrees) of interface behavior for carbonates derived from the com- n Quartz (001) bination of experiment and theory in this project will allow I nt e 104 Quartz (002) more rigorous models of chemical interactions during CO2 sequestration and natural gas extraction. Such models may enable optimization of these processes, benefiting the energy security of the nation 1.0 1.5 2.0 2.5 Q (Å-1) Figure 3. SG-XRD pattern of 250 C atomic layer chemical vapor deposition thin calcite film on an Al2O3 buffer layer on single crystal quartz in air (dry) and in solution (Runs 1 and 2). A chi scan of the calcite (104) plane showing a FWHM of 6.9 is in the upper left corner. 230

Page 231

3. Wang, P., M. R. Hudak, A. Lerner, R. K. Grubbs, S. Wang, Z. Zhang, E. Karapetrova, D. D. Hickmott, and J. Majewski. X-ray scattering of calcite thin films depos- ) s p ited by atomic layer deposition: studies in air and in o / A 26.5 Ao calcite saturated water solution. Thin Solid Films. ( y t 19 iv 4. Lu, C. Y., D. Perez, D. D. Hickmott, and A. F. Voter. is u 12 Insights into microscopic diffusion processes at a solid/ f f iD 9 fluid interface under supercritical conditions: a study of the aqueous calcite (10¯14) surface. 2012. Journal of Physical Chemistry C. 116: 25934. 300 800
4. Raiteri, P., and J. D. Gale. Water is the key to nonclassi- Temperature (K) cal nucleation of amorphous calcium carbonate. 2010. Journal of the American Chemical Society. 132 (49): Figure 4. Calculated calcite/fluid interface diffusivities at a range 17623. of temperatures for varying water thicknesses.
5. Demichelis, R., P. Raiteri, J. D. Gale, D. Quigley, and D. Gebauer. Stable prenucleation mineral clusters are (a) (b) liquid-like ionic polymers. 2011. Nature Communica- tions. 2 (590): 1. Publications Hickmott, D. D., A. Lerner, P. Wang, J. Majewski, and M. Taylor. Neutron reflectometry at elevated pressures and temperatures - novel P-T cell and preliminary ex- (c) periments. Presented at American Geophysical Union Annual Meeting. (San Francisco, 4-9 Dec. 2011). Lu, C. Y., D. Perez, D. D. Hickmott, and A. F. Voter. Insights into microscopic diffusion processes at a solid/fluid interface under supercritical conditions: a study of the aqueous calcite (10¯14) surface. 2012. Journal of Physi- cal Chemistry C. 116: 25934. (d) Lu, C. Y., D. Perez, D. D. Hickmott, and A. F. Voter. Insights into microscopic diffusion processes at a solid/fluid interface under supercritical conditions: a study of the aqueous calcite (1014) surface. Presented at American Chemical Society. (New Orleans, 7-11 April 2013). Figure 5. Parallel tempering simulation of a hydrated calcite interface: a) Equilibrium structure of surface with 70% cover- Wang, P., A. H. Lerner, M. Taylor, J. K. Baldwin, R. K. Grubbs, age; b), c), d) structures used to initialize the simulation. Brown J. Majewski, and D. D. Hickmott. High-pressure and spheres: Ca; red spheres: O; blue spheres: C; Yellow dots: water high-temperature neutron reflectometry cell for solid- molecules. fluid interface studies. 2012. European Physical Journal Plus. 127 (7): 76. References
6. Wang, P., A. H. Lerner, M. Taylor, J. K. Baldwin, R. Wang, P., J. Majewski, A. Lerner, M. Taylor, and D. Hick- Grubbs, J. Majweski, and D. D. Hickmott. High pressure mott. High Pressure and High Temperature Neutron Reflectometry Cell for Solid-fluid Interface Studies . and high temperature neutron reflectometry cell for
7. In American Conference on Neutron Scattering. solid-fluid interface studies. 2012. European Physical (Washington, D.C., 24-28 June 2012). , p. 5. Warren- Journal - Plus. 127 (7): 12076. dale, PA: Materials Research Society.
8. Nilsen, O., H. Fjellvag, and A. Kjekshus. Growth of Wang, P., M. R. Hudak, A. Lerner, R. K. Grubbs, S. Wang, Z. calcium carbonate by the atomic layer chemical vapour Zhang, E. Karapetrova, D. D. Hickmott, and J. Majew- deposition technique. 2004. Thin Solid Films. 450 (2): ski. X-ray scattering of calcite thin films deposited by

231

Page 232

atomic layer deposition: studies in air and in calcite saturated water solution. Thin Solid Films. 232

Page 233

Chemistry and Material Sciences Exploratory Research Final Report Photo-triggerable Immolative Polymers: A New Modality for Radiation Dosimetry Robert D. Gilbertson 20110421ER Abstract radiation through the fragmentation of a chemical bond, We have been developing a chemical method for ra- generating products that can be readily detected, as diation detection. Our approach involves the use of a an odor or as a highly fluorescent product. By relying chemical reaction that is initiated by the absorption of on a chemical reaction rather than on the emission of ionizing radiation. The material utilizes nanoparticle light (scintillation) or generation of an electrical signal scintillators to absorb the incident radiation. Upon (semiconductor detector), this method does not require absorption, a radical chain process is initiated by the the use of materials that are optically transparent or excited nanoparticles that results in the cleavage of a electrically conductive. The initial work has focused on chemical bond releasing a small molecule that can be understanding the fragmentation chemistry well enough detected in various ways. In this fashion, we have dem- to enable the design of self-immolative polymers whose onstrated the generation of carboxylic acids or amines depolymerization reactions can be initiated by ionizing from ester or carbamate precursors, by exposing these radiation induced bond fragmentation. By choosing a constructs to gamma or X-ray sources. polymer that is non-fluorescent when polymerized but is composed of highly fluorescent monomers, depolymer- This class of polymers is non-fluorescent, and can be ization of this material will result in the turn on of the readily depolymerized by endgroup cleavage. Once the fluorescence of the monomers that are generated when polymer unzips, the aniline monomers that are gener- the polymer is exposed to ionizing radiation. The inten- ated are highly fluorescent. Thus, this reaction sequence sity of the fluorescence should be proportional to the can be used as “turn on” radiation sensor. By using quantity of radiation absorbed by the material (Figure 1). neutron-absorbing materials as the radiation antennae, this method should be tunable to neutron detection as Scientific Approach and Accomplishments well. The initial phases of the project have involved a large synthetic effort for the preparation of model com- Background and Research Objectives pounds, immolative polymers, and scintillating nanopar- The long-term goal of this project is to develop a new ticles. We first prepared a monomeric model compound radiation dosimetry method with a high sensitivity for in order to test the proof of concept of the project. gamma and neutron radiation. Specifically, the target That is, could we initiate cleavage of a polymer system applications of this methodology are personal dosimetry (a monomer in the model case) using the response of and long dwell time measurements. The work takes ad- scintillating nanoparticles to gamma-ray irradiation. To vantage of a recently developed class of functional poly- this end a monomeric immolative model compound with mers that have been called “self-immolative” polymers. a picolinium trigger was synthesized. In addition to the The defining feature of self-immolative polymers is that model compound we also prepared lanthanum fluoride the polymer backbone, and in some cases the poly- scintillating nanopaticles with a cerium dopant. In the mer side chains, can be completely depolymerized in a proof of concept experiment a solution of the model controlled manner when subjected to a single triggering compound and scintillating nanoparticles was exposed event such that a large number of copies of a single “re- to gamma-ray irradiation for 30 minutes resulting in 7% porter” molecule are released after a triggering event. fragmentation as measured by integration of proton NMR signals. An additional proof of concept experi- We have developed a chemical system based on these ment was carried out using a monomeric model com- self-immolative polymers that responds to ionizing pound with a masked fluorescent reporter. When this 233

Page 234

compound was exposed to gamma-ray irradiation in the have shown that a radical chain mechanism is likely occur- presence of scintillating nanoparticles, 18% fragmentation ring in triggering process (radiation→light→photo trigger). was observed by NMR and a fluorescent compound was We have shown that small molecule cleavage of pico- produced (Figure 2). linium esters and carbamates is viable with the systems of nanoparticle, electron shuttle and ascorbic acid developed While this experiment demonstrated the proof of con- in this work. We have determined that the dose rates cept of our proposal we observed fragmentation levels needed for observable chemical reaction are much lower much higher than expected based on the radiation flux than in other materials that detect radiation via chemical and assuming 100% efficiency in our triggering system. bond cleavage. From this work it is further apparent that This led us to investigate the triggering and degradation further development of the polymer chemistry will be nec- mechanism more closely. Upon further investigation we essary to observe immolative effects in a dosimetry device. made several more interesting observations: 1) the reac- Studies on oligomers will be the first step to understanding tion occurred non-scintillating lanthanide nanoparticles how this system can be used to induce depolymerization. were used. 2) Citrate radical species are observed by EPR It is apparent that quantification of the extent of bond upon irradiation of the nanoparticles with an X-ray source. fragmentation will be related to the total dose the material 3) The cleavage reaction continues after the sample is receives, a fundamental requirement of dosimetry. removed from the gamma-ray source (and kept in the dark. 4) The reaction is very sensitive to the nature of the We have demonstrated the proof of concept of the funda- stabilizing ligand on the nanoparticle: citrate works best, mental premise of this proposal and the concept offers the diethylene glycol shows very small (~1%) of cleavage, potential for many different applications. We have begun unstabilized particles give no observable cleavage. Taken to extend this research to other types of radiation. The together all of these observations led us to believe that a use of neutron activated triggers (10B and 7Li) is being free radical chain reaction mechanism was responsible for pursued to develop new materials for neutron detection. triggering the polymer degradation. We anticipate our results will lead to new funding op- portunities and we will continue to strive to understand With the results of successful proof of concept experi- the energy transfer efficiency of the triggering process ments in hand we began the synthesis of gamma-ray (radiation→light→photo trigger), the triggering mecha- triggered immolative polymers which should show signal nisms and how they can be tailored to multiple stimuli, amplification proportional to their molecular weight when the depolymerization process including ambient environ- cleavage of the trigger is initiated. Polymers up to thirty ment effects and time constraints, and the optimal readout repeat units in length were produced using slight modi- process. fications of the literature procedures (Figure 3). When a solution of this polymer was subjected to gamma-ray Impact on National Missions irradiation in the presence of scintillating nanoparticles no This work directly supports existing laboratory programs fragmentation was observed in the NMR spectra. We sur- in the development of new radiation detector materials, mised that impurities present in the polymer solution (due as well as providing foundation resources for the develop- to the difficulties of polymer purification) led to inhibition ment of new radiation detector systems. Based on the of the cleavage and/or polymer fragmentation reactions. results of this effort and the potential for signal amplifica- tion we have submitted several additional research propos- In order to determine the root cause of the lack of frag- als outside the laboratory to NSF, NA-22, NA-42, and GS mentation and have better analytical accuracy in doing sponsors including proposals using immolative polymers so, a plan was devised to synthesize a model polymer of 2 for latent fingerprint visualization, explosives detection, repeat units. To date, this model polymer has been syn- and as an aid to enhance photo-dynamic therapy in cancer thesized, however it has not yet been exposed to gamma- treatment. We anticipate that positive results from this ER ray irradiation. We anticipate that this experiment will be work will enhance our efforts to further develop this work carried out and that the results will available in short order. and obtain support from additional funding sources out- side the laboratory. These proposals have not been funded Our overarching research goal was to develop a radiation as yet. However, recent experience has demonstrated that dosimetry method based on self-immolative polymers these sponsors as well as DHS and DTRA are interested in that has a high sensitivity (low minimum detectable dose) supporting new concepts at relatively high funding levels, for X-ray and gamma radiation, long-term stability, and a but that initial results, such as those obtained through simple readout method. We have demonstrated the prop LDRD projects, are crucial in demonstrating the wisdom of concept of this idea by the gamma-ray induced cleavage of a particular approach. Under this strategy, a single of a picolinium ester (Figure 2). Within this framework, we 234

Page 235

LDRD-ER might spawn two or three new external projects, depending on what additional development is required for individual applications. We are optimistic that we can develop new funding in the future using the results from this LDRD-ER project to generate research proposals for the aforementioned sponsors as well as in the GS arena. Figure 1. Schematic of a radiation triggered self-immolative polymer degradation. Figure 2. Gamma-ray induced cleavage of a model compound. Figure 3. NMR spectra are consistent with a 20 repeat unit poly- mer. Publications Boncella, J. M., L. P. Spencer, and J. B. Edson. Photorelease of primary aliphatic and aromatic amines by visible- light-induced electron transfer. 2011. Organic Letters. 13 (23): 6156. Cape, J. L., J. B. Edson, L. P. Spencer, M. S. DeClue, H. J. Ziock, S. Maurer, S. Rasmussen, P. A. Monnard, and J. M. Boncella. Phototriggered DNA Phosphoramidate Ligation in a Tandem 5 ‘-Amine Deprotection/3 ‘-Im- idazole Activated Phosphate Coupling Reaction. 2012. BIOCONJUGATE CHEMISTRY. 23 (10): 2014. Edson, J. B., D. S. Kuiper, R. D. Gilbertson, and J. M. Boncel- la. Low level gamma-ray radiation detection: carbox- ylic acid release from picolinium esters. . Angewandte Chemie International Edition.. 235

Page 236

Chemistry and Material Sciences Exploratory Research Final Report Stable Uranium Benzyne Complex Jaqueline L. Kiplinger 20130793ER Abstract Besides this initial report, the chemistry of uranium ben- The primary focus of this project was to isolate and zyne compounds was largely unexplored until recently structurally characterize a stable uranium benzyne com- when Evans and co-workers reported a similar insertion plex in both solution and the solid-state. The solid-state reaction to yield the metallocycle complex 4 by reacting structure would provide an important benchmark for (C5Me5)2U(C6H5)2 (1) with iPrN=C=NPri [3]. During the future theoretical studies on the bonding and electronic course of this work, Hayton and co-workers reported a structure in this compound. The electronic structure of structurally characterized uranium complex containing a compound, or “what the electrons are doing”, governs tethered benzyne ligands but did not explore its reactiv- all its nonnuclear properties, including the chemistry it ity [4]. displays with other substrates. Exploring the chemistry As shown in Figure 1, we recently discovered that of uranium compounds provides an improved under- pyridine undergoes C-H activation with the uranium standing of chemical reactivity, covalency, and the rela- diphenyl complex (1) to give the novel uranium phenyl tive roles of 5f- and 6d-orbitals in chemical bonding for η2-pyridyl complex, (C5Me5)2U(C6H5)[η2-(N,C)-NC5H4] the actinides. This project supports LANL’s core capa- (5). The dialkyl complexes (C5Me5)2U(CH3)2 and bility in actinide science, the Lab’s Plutonium Strategy, (C5Me5)2U(CH2C6H5)2 also react with pyridine to give and LANL’s Energy Security and Global Security mission the structurally related uranium methyl and benzyl η2- areas. pyridyl complexes, (C5Me5)2U(CH3)[η2-(N,C)-NC5H4] and (C5Me5)2U(CH2C6H5)[η2-(N,C)-NC5H4], respec- Background and Research Objectives tively. By extension, we initially thought this chemistry Benzyne (C6H4) is a reactive species, which features was taking place with the diphenyl complex (1). How- rich and diverse chemistry with numerous organic ever, it was possible that this chemistry could also be compounds. Unfortunately, these reactions tend to be taking place with the uranium benzyne intermediate (2) difficult to control. However, metals are able to stabilize as illustrated in Figure 1. In support of this proposal, highly reactive organic fragments and also to activate we discovered something completely unexpected when stable molecules toward selective attack. Although sev- the uranium diphenyl complex (1) was reacted with eral benzyne complexes of transition metals have been furan. Instead of reacting by the anticipated C-H cleav- reported [1], few are known for the f-element metals. age pathway, we obtained an unusual 1H NMR spectrum Early work by Marks’ group at Northwestern showed and a preliminary (highly disordered) X-ray structure that that the uranium diphenyl complex, (C5Me5)2U(C6H5)2, is consistent with the furan-stabilized uranium complex (1) eliminated benzene at room temperature to gener- 6 as illustrated in Figure 1. Our preliminary data sug- ate the benzyne intermediate (C5Me5)2U(benzyne) gests that we have serendipitously synthesized the first (2) (Figure 1) [2]. The benzyne complex was very reac- example of a stable uranium benzyne complex. tive and could not be isolated. However, its existence Since our structural data were not concrete, we per- was determined by the chemistry it performed. Once formed a reaction with our proposed furan-stabilized generated, the benzyne complex reacted with diphenyl- uranium benzyne complex (C5Me5)2U(C6H4)(C4H4O) acetylene to afford the uranindene complex (3). Similar (6). It is known that the “reactive” uranium benzyne chemistry was reported for the thorium diphenyl com- complex 2 reacts with diphenyl acetylene at room tem- plex, (C5Me5)2Th(C6H5)2, but the chemistry required perature to give the complex 3 as illustrated in Figure temperatures above 100 °C. 236

Page 237

1 [2]. When complex 6 is reacted with diphenylacety- ylation of 6 should initially form the dihapto-acyl insertion lene at 60-75 °C, complex 3 is produced. This is compel- complex 8. Actinide acyl complexes have been reported ling evidence that we have essentially placed a uranium and resemble “anchored” Fisher carbene complexes in benzyne complex “in a bottle.” This is very exciting as this spectroscopic and chemical properties. However, complex observation suggests that our benzyne complex is rather 8 is interesting as it features a second reactive site. Migra- stable compared to complex 2. This enhanced stability will tory insertion of a second equivalent of CO should give the provide us with an unparalleled opportunity to probe the novel bis-acyl complex 9. It will be interesting to examine chemistry and ultimately understand the electronic struc- the benzyne complex’s stability toward oxidation chemis- ture and bonding of the actinide benzyne functionality. try. By increasing the oxidation state of the uranium metal, the electronic structure, and therefore the chemistry, of Scientific Approach and Accomplishments the benzyne functional group is expected to change. Our Initially, we prepared the uranium diphenyl complex, LANL team showed that reaction of uranium(IV) com- (C5Me5)2U(C6H5)2, (1) using Marks’ published procedure plexes with copper salts is a mild way to synthesize the which involves the reaction of 3 equivalents of phenyl corresponding uranium(V) systems. Accordingly, reac- lithium, Li(C6H5) with (C5Me5)2UCl2 in diethyl ether at tion of 6 with CuI should give the uranium(V) benzyne -78 °C (Figure 2). However, the synthesis was not repro- iodide complex (10). Similarly, pyridine N-oxide has been ducible and generated a mixture of the desired complex shown to promote two electron oxidation of uranium with 1 and a lithium “ate” complex, (1·LiCl(Et2O)x) as shown concomitant oxygen-atom transfer. Thus, reaction of 6 in Figure 2. This hindered our progress and prevented us with pyridine N-oxide could yield the uranium(VI) benzyne from reproducibly accessing the target uranium benzyne oxo complex 11. However, an alternative pathway involv- complex (C5Me5)2U(C6H4)(C4H4O) (6). To address this ing C-H activation of the pyridine N-oxide may yield the problem, we turned to the diphenyl magnesium reagent, unusual metallacycle 12. If this latter pathway occurs, then Ph2Mg(1,4-dioxane) and reaction of this compound with other oxidants such as silver nitrate or nitrous oxide will be (C5Me5)2UCl2 afforded pure (C5Me5)2U(C6H5)2, (1). This employed to yield complex 11. It would also be interest- result represents a new route to (C5Me5)2U(C6H5)2, (1) ing in future studies to perform simple substitution of the and is a significant improvement over Marks’s original furan donor ligand in complex 6 as a means to tune the 1981 procedure. From this new synthesis we were able to reaction chemistry displayed by the benzyne fragment. obtain the first crystal structure of (C5Me5)2U(C6H5)2, (1) (Figure 2). Impact on National Missions Exploring the chemistry of uranium improves our under- We anticipated that the uranium benzyne complex would standing of the electronic structure of the light actinides. chemically behave differently than transition metal ben- Uranium is a member of the actinide series, which marks zyne complexes. The closest analogues are the early tran- the emergence of 5f electrons in the valence shell. The sition metal benzyne complexes, which have demonstrated roles of these 5f electrons in the various pure phases, chemical behavior consistent with an ortho-metallated alloys, compounds, and molecules can vary from being phenyl complex description [1]. Using this bonding de- delocalized (bonding) to being localized (nonbonding). The scription, the benzyne ligand in complex 6 is dianionic. electronic structure of a material, or “what the electrons Chemistry is the best way to understand the electronic are doing”, governs all its nonnuclear properties. It dictates nature of the benzyne functional group and we planned the arrangement of the atoms in a molecule and is respon- to survey the chemistry of complex 6 using the reactions sible for its response to light, the magnetic properties, and outlined in Figure 3. the bonds formed with other materials. This project sup- ports LANL’s Energy Security and Global Security mission Unfortunately, despite numerous efforts, to date we have areas as well as the Lab’s Plutonium Strategy. not been able to obtain a crystal structure of the furan- stabilized uranium benzyne complex (C5Me5)2U(C6H4) (C4H4O) (6). This is still a work in progress. The reactions we intended to perform were designed to examine the ability of complex 6 to undergo insertion reactions with a variety of unsaturated substrates to form metallacycles. For example, 6 will likely insert benzonitrile to give the ketimide metallacycle complex 7. Multiple insertions could also take place with complex 6. An ideal substrate to ex- amine this would be carbon dioxide as it is small. Carbon- 237

Page 238

References

1. Buchwald, S. L., and R. B. Nielsen. Group 4 metal com- plexes of benzynes, cycloalkynes, acyclic alkynes, and alkenes. 1988. Chemical Reviews. 88: 1047.
2. Fagan, P. J., J. M. Manriquez, E. A. Maatta, A. M. Seyam, and T. J. Marks. Synthesis and properties of bi s(pentamethylcyclopentadienyl) actinide hydrocarbyls and hydrides. A new class of highly reactive f-element organometallic compounds. 1981. Journal of the Amer- ican Chemical Society. 103: 6650.
3. Evans, W. J., J. R. Walensky, and J. W. Ziller. Insertion of carbodiimides and organic azides into actinide-carbon bonds. 2009. Organometallics . 28: 3350.
4. Seaman, L. A., E. A. Pedrick, T. Tsuchiya, G. Wu, E. Jaku- Figure 1. Synthetic procedures for the generation and chemistry bikova, and T. W. Hayton. Comparison of the reactivity of a reactive uranium benzyne complex and the synthesis of a stable uranium benzyne complex. of 2-Li-C6H4CH2NMe2 with MCl4 (M=Th, U): Isolation of a thorium aryl complex or a uranium benzyne com- plex. 2013. Angewandte Chemie, International Edition. 52: 10589. Figure 2. Improved synthetic procedure for the preparation of pure uranium diphenyl complex, (C5Me5)2U(C6H5)2, (1). Also shown is the molecular structure of complex 1. Figure 3. Proposed Synthetic outline to probe the chemistry of a stable uranium benzyne complex. 238

Page 239

Chemistry and Material Sciences Exploratory Research Final Report Characterization of Products from Hydrolysis of UF6 Marianne P. Wilkerson 20130803ER Abstract the chemical behavior of these systems under the LDRD- The goal of this project is to develop a fundamental DR, Molecular Forensic Science of Nuclear Materials. A understanding of signatures from products of uranium reaction chamber was developed in which to conduct hexafluoride (UF6) hydrolysis reactions. Uranium hexa- UF6 hydrolysis (Figure 1). This reaction chamber pro- fluoride is a ubiquitous material in the nuclear enter- vides the ability to introduce a stream of UF6 into a prise. Due to its volatility, it is essential for isotope sepa- humidified space or conversely to allow humidity to be ration, and as such, is often a ‘fingerprint’ for breakout added to a chamber filled with UF6. The initial studies from reactor- or fuel-grade to weapons-grade uranium. revealed that the morphology and chemical speciation A number of studies in the open literature reveal com- of deposited materials are dependent upon atmospheric plex chemical reactivity between UF6 and water vapor, conditions of temperature and humidity. Images reveal and suggest that relative humidity and temperature play significantly different morphologies between products a role in the chemical and physical composition of the deposited on carbon tape versus aluminum (Figure 2). hydrolysis products and the evolution of these products Furthermore, these materials readily changed morpho- over time. Additional work suggests that the deposited logically within days of release, as illustrated in Figure 3. products may exhibit morphology dependent upon initial Careful study of the chemical and morphologic char- UF6 concentration and chemically interact with deposi- acteristics of hydrolyzed UF6 will enable more effec- tion surfaces. Our limited understanding of the com- tive means of developing tools to detect and quantify plexity of these systems, however, precludes our ability release of UF6. Specifically, the studies proposed here to develop viable and quantitative signatures from these will achieve this fundamental understanding through the products. Characterization of these materials produced following: under a variety of parameters, including relative humid- ity, temperature, concentration of UF6, and composition Generation of hydrolysis products through reactivity of surface depository, will lead to more effective analy- controlled as a function of: ses of detectable signatures from UF6 hydrolysis. • UF6 concentration Background and Research Objectives • Water vapor pressure Many technical references report that UF6 reacts with water as follows: • Temperature • Rate of water vapor and/or UF6 release UF6 + 2H2O → UO2F2 + 4HF • Reactivity of hydrolysis products with surface deposi- Additional work reveals that the reactivity of UF6 with tory water is complex, forming a variety of uncharacterized reaction intermediates and products, and is likely de- Chemical changes in hydrolysis products as a function of pendent upon environmental and release conditions.1,2 • Temperature However, these studies do not address the chemical compositions of these materials, information that is • Relative humidity important for detection of released UF6. • Time Paffett, Wagner and Wilkerson initiated studies to probe 239

Page 240

A series of hydrolysis products will be prepared and ana- suggested the formation of different oxidation states of lyzed through morphologic imaging via Scanning Electron uranium following one week of aging, such that U(IV) was Microscopy, and characterization of chemical speciation produced under conditions of low temperature (20°F), through X-ray diffraction analysis and infrared absorption while U(VI) was the predominant product following aged spectroscopy. under conditions of high temperature (90°F). Under this LDRD-Reserve, powder X-ray diffraction measurements Scientific Approach and Accomplishments were then conducted, and analysis reveals the presence The technical approach includes the following: of both UO2F2·xH2O and UOF4 (Figure 4). The stronger intensity of lines from UOF4 is of notable interest. • Hydrolysis of UF6 on the bench scale as a function of temperature, relative humidity, and composition of The majority of the experimental work was not com- surface depository pleted due to the September 2013 timeline of a separately funded project utilizing the capability developed for this • Store products under controlled temperature and rela- project (vide supra). Therefore, a portion of the funds tive humidity as a function of time (2/3) were returned to the LDRD Program Office, and it is proposed that the experimental portion of this work will • Image morphology of material for characterization of be completed in FY2014. size and surface homogeneity Impact on National Missions • Characterize chemical speciation through infrared The impact of this project is the ability to correlate en- absorption and X-ray spectroscopies vironmental conditions (temperature, relative humidity) Bench scale hydrolysis of UF6 will provide a means for and depository surface composition with UF6 hydrolysis safely conducting these reactions safely and in a controlled products. As stated by Andy Erickson (PADGS), “the contin- environment, permitting the ability to conduct a large ued pursuit of Weapons of Mass Destruction (WMD) by number of experiments using varied reaction conditions numerous proliferant nations is not showing any signs of that one might expect in different natural environments. slowing down. As such, the need to accurately determine Vessels developed under a separate WFO project will en- whether a nation is pursuing WMD or simply a nuclear able the ability to control exposure of the samples to time, fuel cycle is important, as is the ability to locate previously temperature and relative humidity. Measurements of mor- unknown/undeclared uranium facilities… By providing phology will reveal particle size and homogeneity of the funding for the proposed work, it demonstrates the labora- initial and aged hydrolysis products. Chemical speciation tory’s investment in an area of key interest to sponsors, a (powder X-ray diffraction analysis for phase type and long welcome occurrence in the current funding environment. range order; diffuse reflectance for vibrational signatures; In addition to answering key scientific questions, the work synchrotron-based studies for characterization of amor- is timely, as significant government effort is ongoing to de- phous components of the material) will yield complete velop new signatures for monitoring countries of concern.” characterization. Each of these factors will enable more effective detection capabilities for identifying UF6 hydroly- sis products and understanding the temporal behavior of these materials. Through the initial portion of this project, an extensive lit- erature search of publications relevant to hydrolysis of UF6 was conducted. An outline for a publication for peer re- view was prepared. In addition, these results were shared with potential sponsors of this work. (Please contact M. Wilkerson for additional information.) The feasibility of characterizing chemical speciation from these particulates was tested. Through initial study under the LDRD-DR Molecular Forensic Science of Nuclear Mate- rials, an experiment was conducted at Beam Line 2-3 at the Figure 1. An experimental chamber was designed for laboratory Stanford Synchrotron Radiation Lightsource to measure scale experiments to hydrolyze UF6 under controlled temperature X-ray Absorption Spectra from aged material deposited and relative humidity. onto aluminum, shown previously in Figure 3. Initial data 240

Page 241

References

1. Bostick, W. D., and W. H. McCulla. Sampling, character- ization, and remote sensing of aerosols formed in the atmospheric hydrolysis of uranium hexafluoride. 1985. Journal of Environmental Science and Health. A20 (3):
3. Kips, R.. Development of uranium reference particles for nuclear safeguards and nonproliferation control. Figure 2. Material released from hydrolyzed UF6 (1 g UF6, 19% 2007. relative humidity, room temperature) was directly deposited onto carbon tape (a) and onto aluminum (b). Publications Wilkerson, M., M. T. Paffett, and G. L. Wagner. Character- ization of Products from Hydrolysis of UF6. To appear in Manuscript in preparation. Figure 3. Morphologic images were collected from material released from hydrolyzed UF6 (1 g UF6, 19% relative humidity, room temperature) and then aged at the indicated temperatures and relative humidities after one week. Figure 4. Powder X-ray diffraction analyses of hydrolyzed UF6 deposited onto aluminum reveals the presence of UOF4 and UO2F2-xH2O. 241

Page 242

Chemistry and Material Sciences Exploratory Research Final Report Re-invigorating LANL Quantum Computing Research Antoinette J. Taylor 20130809ER Abstract Because the basic building block of a QC, the qubit, has Quantum computing holds great promise of exceptional not been established, there are still many physics experi- speed up of computations for applications ranging from ments to be done to improve on the existing qubits, as national security to pharmaceutical design. Because well as models and simulations needed to understand the basic building block of a Quantum Computer (QC), the many complex real-world issues surrounding devel- the qubit, has not been established, there are still many opment of a QC. physics experiments to be done to improve on the exist- In this project we will pursue proof-of-principal experi- ing qubits, as well as models and simulations needed mental research on topological QC and superconducting to understand the many complex real-world issues qubits, as well as fundamental QC theoretical studies. surrounding development of a QC. In this project, we We believe we have assembled an exceptional team to propose to reinvigorate QC research at LANL, exploit- execute the proposed research. The proposed research ing its world-class capabilities condensed matter phys- builds upon areas where LANL has existing world class ics (theory and experiment), photonics and QC theory, capabilities and will materially advance QC research to enable program development opportunities in this since LANL retains a strong QC theory team, and many arena. Our proposed research includes proof-of-principal materials capabilities not available elsewhere. The pro- experimental research on spins in quantum dots, topo- posed work will allow us to re-invigorate our activities in logical quantum computing (QC), photonics, and super- QC. conducting qubits, as well as fundamental QC theoreti- cal studies. The proposed research builds upon areas Scientific Approach and Accomplishments where LANL has existing world-class capabilities that will Fabrication of basic constituents of superconducting materially advance QC research. We have assembled an qubits and study of their physical properties: The basic exceptional team to execute the proposed research. The component of a QC is a qubit. All types of superconduct- proposed work will allow us to re-invigorate our activi- ing (SC) qubits rely in two phenomena, flux quantization ties in QC and seek external funding that is available in and coherent tunneling across Josephson junctions (JJ). the near term (FY14) and not later, hence the urgency of Leonardo Civale, Boris Maiorov and Quanxi Jia worked FY13 LDRD funding. together to perform magnetization and transport studies to explore flux quantization in SC circuits and the proper- Background and Research Objectives ties of Josephson junctions. Quantum computing holds great promise of exceptional speed up of computations for applications ranging from The basic idea behind the first experimental study was national security to pharmaceutical design. The field to use our Magnetic Properties Measurement System became critical for national security in 1993 when Peter (MPMS) SQUID magnetometer to measure the magnetic Shor developed an algorithm that provided exponential flux Φ produced by the persistent supercurrents circulat- speed up for factoring large numbers, an algorithm with ing in SC rings as a function of the magnetic field H. The direct applicability to cryptography. From the mid-1990s values of Φ are quantized in units of the flux quantum to the mid-2000s Los Alamos was arguably the strongest Φ0=2.07x10-7 Gcm2, and the response is periodic in H single institution for quantum computing (QC) research. with periodicity H=Φ0/A, where A is the ring area. The Since that time the Lab has lost some key personnel, but MPMS is not sensitive enough to detect the signal of a a strong core of experimentalists, theorists, and model- single ring, thus we proposed to pattern arrays of 104 to ers remains and can be tapped to advance the field. 242

Page 243

106 identical rings. The first goal was to implement up- neered structures utilizing topological insulators. A materi- grades in the MPMS to (1) reduce the noise level, and (2) al that is a topological insulator (TI) possesses an insulating increase the field resolution. L.C. visited Quantum Design, ground state that is topologically distinct from the vacuum; the manufacturer of our SQUID magnetometer to discuss this necessarily ensures a metallic state at the surface. these plans. Regarding the noise, they explained that the Unfortunately, all known TI materials possess impurities in main source is the current circulating through the heater the bulk which destroy the topological protection of the of the sample temperature control, and showed us how surface state. Our route to circumventing this problem is to to reduce it by modifying some software parameters. We investigate systems that are topological insulators due to tested this method and we have been able to reduce the strong correlations, rather than conventional band insula- noise in standard measurements from the typical ~ 2 x tors. Strong correlations can result in much larger insulat- 10-7 emu to ~ 5 x 10-8 emu. It is possible to increase the ing gaps in the bulk that would add protection to a metallic sensitivity by one more order of magnitude by maintain- surface state, as well as provide a simple mechanism for ing the sample in a fixed position (rather than scanning it generating magnetism and/or superconductivity. Our ex- through the detection coils) and processing the raw SQUID perimental efforts led by Filip Ronning examined whether signal. We learned to do this without the need to incorpo- specific Samarium (Sm) based compounds may act as good rate additional electronics, and we are currently testing the host states for Majorana Fermions. Majorana Fermions method. Regarding the field resolution, they suggested a should have topologically protected non-abelian statistics simple modification to the magnet circuit (change of the that some day may be used for quantum computation. We control resistors) that will allow us to increase the field res- performed three complimentary studies. olution from the standard 0.15 Oe to around 0.01 Oe. We First we measured specific heat of 6 powdered SmB6 will implement this hardware change in the near future. In samples with varying particle size from 40 um to a 1 mm3 parallel, we have discussed the details of the fabrication of single crystal. This compound was selected because it is a the arrays of rings with the colleagues at CINT (SNL) that well known example of a “Kondo insulator”, a narrow gap will pattern them by photolithography. We purchased the material with an energy-gap brought about due to many- Nb ingot needed and are presently designing the masks. body correlations stemming from rare-earth magnetism In the second study we fabricated and studied individual interacting with charge carriers. The varying surface to high Tc superconductive JJs of different characteristics. Un- volume ratio for different samples should enable us to de- til now, studies of JJs for QC have been almost exclusively termine whether a predicted surface state exists. Instead, focused on low Tc superconductors. The key idea is to take we observed a decrease in the heat capacity at the lowest advantage of the fact that high-temperature superconduc- temperatures in the more finely powdered samples with tor (HTS) films develop a natural JJ when deposited on higher surface area. This effect is likely due to a thermal bi-crystal substrates, eliminating the need for multi-step JJ decoupling of the sample at the lowest temperatures. fabrication processes. Although bi-crystal JJs can only be Future measurements on LaB6 would help confirm this used to build some types of qubits (e.g. phase and charge), interpretation. the work conducted under this project allows us to estab- Second, we used a single crystal piece of SmB6 and at the lish the basic fabrication needs and setup the measure- CINT core in Sandia deposited a 80 nm thick film of Al on ments capabilities. We have already deposited YBa2Cu3O7 its surface. Subsequently, a 100nm wide trench in the Al (YBCO) films on two SrTiO3 (100) bi-crystal substrates by was cut using a FIB, thereby creating a Josephson junction Pulsed Laser Deposition. We have chosen bi-crystals with device. If a topologically protected surface state exists in angles of 30°, typically used for YBCO SQUIDs, and 36.8°, SmB6, then this Josephson junction should possess Majo- which will produce JJs with different parameters (lower rana fermions hence modifying its current-phase relation. coupling, and consequently smaller critical currents). Our transport measurements on this preliminary sample We are currently in the process of patterning the bridges showed only a very weak superconducting signature (1% of needed for the nonlinear electrical transport studies (I-V the transport). I-V measurements revealed unstable resis- characteristics as a function of H and T) on these samples. tance values. The origin of this behavior is likely in part due Topological quantum computation: The recent discovery to the surface roughness observed in our SEM measure- of a new class of materials known as topological insulators ments of order 1 um. Future studies will require enhanced provides renewed enthusiasm for the prospect of topo- fabrication efforts to reduce sample roughness. logically protected qubits for quantum computation. The Lastly, with Cristian Batista we calculated the electronic source of this excitement lies in the potential realization of band structure for SmS. This material, like SmB6, is a corre- Majorana fermions, which should exist in properly engi- lated semi-conductor with an incomplete gap at low tem- 243

Page 244

peratures, which may be due to a topologically protected quantum (unitary) circuits can be related to condensed surface state. By inspection the electronic band structure matter problems via mappings to “Feynman” Hamiltoni- has a band inversion that is a minimal requirement for ans. That is, the ground state of Feynman Hamiltonians possessing a topologically protected surface state. Using have large amplitude with the state output by a quantum this band structure as input Cristian Batista is creating a circuit. Rolando Somma’s efforts in support of this project tight binding Hamiltonian and calculating the topologically focused on generalizing this mapping to non-unitary cir- invariant quantities to determine if indeed a protected cuits. This opens the door to using exact methods, devised surface state must exist in this material. for computing ground state properties, for the simulation of non-unitary circuits. Fast quantum computation and simulation methods: While quantum computers can efficiently simulate Schrödinger’s Diego Dalvit has worked in collaboration with Adolfo del evolution equation, efficient quantum methods to simu- Campo (T-4 JRO Fellow) on issues associated with find- late simple classical processes, such as those determined ing shortcuts to adiabaticity. They also examined ideas by partial differential equations, are unknown. Examples to apply similar techniques to problems where the target are the simulation of the Master or heat equations. The state is not the ground state with a given excited state. He first goal of this project’s computational component was has also used some of the resources to invite Dr. Federico to design quantum-computation inspired methods for the Spedalieri from the Information Sciences Institute (USC efficient simulation of such equations. Our approach is School of Engineering), who is the lead person running the to construct quantum Hamiltonians whose ground states supercomputing D-Wave machine at USC. Dr. Spedalieri encode the solution to the corresponding time-dependent visited LANL during September 10-13, and presented a equation. We also sought efficient quantum methods to quantum lunch on September 12. prepare the ground states. One such approach is by means of adiabatic quantum computation, in which the ground Impact on National Missions state is prepared by slow changes in the Hamiltonian inter- Quantum computing holds great promise of exceptional actions. speed up of computations for applications ranging from national security to pharmaceutical design. The field be- Together with Director’s funded postdoc, Jie Ren (T-4), and came critical for national security in 1993 when Peter Shor a visiting summer student, Fuxiang Li, (from Texas A&M developed an algorithm that provided exponential speed University) Nikolai Sinitsyn performed calculations for the up for factoring large numbers, an algorithm with direct effect that they call the Braid Phase Transition. The latter applicability to cryptography. The proposed work will in- is a topological phase transition, induced by the measure- vigorate quantum computer science and technology at Los ment apparatus. They discovered that this effect occurs Alamos National Laboratory, enabling researchers at the when coupling to the apparatus becomes relatively strong, Laboratory advance forefront technology in this arena. The i.e. this transition indicates the appearance of quantum proposed work is tightly tied to the Laboratory’s national Zeno effect. This effect sheds new light on the physics security and global security missions. of the measurement process by proving that quantum measurement can be considered as a topological critical phenomenon. The results are currently being written as a manuscript. Importantly, it should be possible to observe this unusual phase transition at LANL using the recently developed capability in the spin noise spectroscopy. Cold atoms provide promising candidates for the quantum simulation of complex many-body Hamiltonians. Two chal- lenges in this scheme are cooling to extremely low tem- peratures, and overcoming the effects of noise and fluctua- tions. Postdoc Armin Rahmanisisan’s research in quantum simulations has focused on these challenges. He has proposed a scheme based on the optimal control of uni- tary evolution, which may help extend the limits of atomic cooling beyond the current state of the art. Additionally, He has been investigating the heating effect due to the inherently noisy fluctuations of quantum Hamiltonians and their effects on measurable observables. It is known that 244

Page 245

Chemistry and Material Sciences Postdoctoral Research and Development Final Report Attosecond Probing of Dynamical States in Solids - Graphene George Rodriguez 20100633PRD4 Abstract optical excitation in materials has proven difficult due to By combining the time resolved angle resolved photo- the extremely short timescales involved. The powerful emission spectroscopy setup (trARPES) [1] at LANL with technique of trARPES represents the first such experi- ultrashort laser pulse generated attosecond pulse trains, ments on this unique material and has placed LANL in a the physics of electron dynamics and transport in gra- leading position in this rapidly emerging field. phene was studied in the time domain. Graphene was Combining trAPRES with attosecond pulse trains allows chosen due to the technological interest in its potential the opportunity to measure the electron transport in applications and unique electronic properties [2-4]. A graphene with resolution limited only by the number question of central importance in graphene is how the of pulses in the attosecond pulse train. ARPES mea- carriers behave after optical excitation. The trARPES sures the simultaneous emitted angle and energy of setup gives the ability to answer this through a measure- electrons from a sample. Static ARPES measurements ment of the energy and momentum resolved dynam- have already measured the band structure within the ics of the unique linear band structure in graphene, linear regime of graphene [8, 9]. In our experiments, an although in a transient way. By promoting carriers attosecond pulse train comprised of XUV photons was above the ground state of the system through optical created through the high harmonic generation process. photoexcitation followed by extreme ultraviolet (XUV) To the attosecond train, an infrared (IR) pulse was added photoemission of the now excited state in a pump-probe as a pump for the graphene samples. Figure 1(a) shows configuration, the time evolution of the carrier dynamics a model of the band structure behavior in graphene could be directly recorded. Additionally, since the pulses following optical excitation. At t=0 when the pump and in the probe pulse train were cutting-edge sub-femtosec- probe beams are overlapped, the optical pulse creates a ond (fs) pulses, the time domain response was measured band of promoted electrons above the Dirac point and a with very high temporal resolution – approaching state- corresponding band of holes below. After a fast ther- of-the-art technology. malization process, the electrons and holes redistribute themselves covering a wider energy range than the Background and Research Objectives original band of promoted carriers. Most of this redistri- The technique of trARPES has already demonstrated bution has already been theoretically explained [10-12]. its utility in measuring complex correlated electron As the delay between the pump and probe is increased, materials [5]. The electronic structure of materials is snapshots of the population decaying back to the ground measured as the time evolution of both occupied and state via electron-electron and electron-phonon interac- unoccupied states within the complex band structure. tions are made. A direct measurement of the dynamics Luckily, graphene has a relatively simple structure in the involved in this process was the goal of this work. While vicinity of the Fermi edge with a linearly dispersive set of there has been some theoretical work discussing the bands forming the so-called “Dirac point” [6]. While this redistribution of carriers in graphene samples [10, 12], band structure makes theoretical calculations easier, the an accurate experimental result was realized for the continuous nature of the band structure opens the door first time by this work. The technique of trARPES with for many electron-electron and electron-phonon inter- attosecond pulse trains has greatly increased our under- actions. Scattering of these “quasiparticles” results in standing of the dynamical processes in both graphene a very fast excited state lifetime in comparison to other and other similar Dirac materials. materials with large bandgaps [7]. Accurately measuring the time domain lifetimes of quasiparticles formed from 245

Page 246

Scientific Approach and Accomplishments static photoemission spectra for graphene mounted on a In this work we studied the strongly correlated electron substrate. Silicon dioxide was chosen as the substrate to system in graphene on the fastest timescale currently minimize modifications to the electronic structure through possible, namely an attosecond pulse train. Such short substrate interactions [19, 20]. By using a pump pulse with pulses allowed us to accurately measure the relevant photon energy below the work function of the material quasiparticle dynamics and transient spectral functions in (1.55 eV) to first excite the sample, the time-evolution both the energy and time domains simultaneously. Due to of the carrier and hole dynamics could be measured by the low photoelectron count rates available we chose to simply varying the delay between the 1.55 eV pump pulse look at the graphene samples in an angle integrated mode and the XUV probe pulse. The inset in Figure 1(b) shows although the possibility for angle resolved experiments the signal within the linear portion of graphene’s band also exists. We looked at two important questions in the structure. studies of graphene (described below), both of which are By measuring the population of emitted photoelectrons of significant scientific interest. in the vicinity of the Dirac point, a map of the material’s 1.) Tracing ultrafast separation and coalescence of car- time-dependent electron transport was produced as seen riers in graphene with time-resolved photoemission: In in Figure 2(a). By extracting the data for graphene from graphene, an important question is how the photoexcited Figure 2(a), we could assign a time varying temperature carriers behave on ultrashort timescales. Although more and chemical potential to the material. The results of this recent time-resolved studies have provided a wealth of are shown in Figure 2(b). For early times, an ultrafast and information [13-15], fundamental questions concerning large temperature change was measured due to thermal the quantum descriptions of the transient electron-hole effects of the laser pulse on the graphene sample. This plasma remain. More precisely, two alternative views of temperature then decayed slowly over a picosecond times- the dynamical relaxation are possible. For the first pos- cale due to a bottleneck effect of relaxing electrons. More sibility, a very rapid thermalization amongst both the elec- interestingly, the chemical potential exhibited a non-zero trons and holes leaves the system at an elevated tempera- behavior at early times followed by decay to zero. This ture, the resultant transient state being well-described by represents a separated carrier distribution more indicative a Fermi distribution with unaltered chemical potential, μ, of the μ* model. At later times, the distributions coalesce but at raised temperature, T. We will refer to this system into a single distribution centered at zero more closely as the T* model. A second possibility is that the nonequi- matching the T* model. This demonstrates that both mod- librium carriers form transient populations mimicked by els occur in the samples but are strongly dependent on the two separate Fermi distributions with distinctly different timescales involved. The μ* model is more prevalent in ul- chemical potentials and temperatures. We will call this sys- trafast systems and at early times while the T* model is ob- tem the μ* model. On one hand, the μ* description of the vious in relatively slower experiments. Since the trARPES photoexcited electrons and holes allows for phenomena technique allows a direct sampling of the time evolution such as terahertz lasing or using graphene in tunable lasers of the carrier distributions, only this technique combined [2, 16]. On the other hand, the observation of phenomena with an attosecond pulse train was capable of measuring such as ultrafast photoluminescence and studies of carrier these results. Additional details of these experiments can multiplication in graphene [15,16] imply fundamentally dif- be seen in References [21, 22]. ferent quantum descriptions of the photo excited electron 2.) Direct measurement of quasiparticle lifetimes in gra- hole plasma, more in line with the T* model. phene in the vicinity of the Dirac point: In a similar experi- In the experiments, an attosecond pulse train of XUV pho- ment to the previous one, graphene samples were again tons with energy ~30 eV was used. The photons were then excited with optical photons and probed with ultrashort incident on a graphene sample grown via the chemical attosecond pulse trains. In this experiment however, the vapor deposition method with very few defects [17, 18]. quasiparticle lifetimes were measured and compared with Since the photon energy was well above the work function theory. With this method, we observed a very obvious of the material, a spectrum of the emitted photoelectrons excitation energy dependence on the relaxation time of could be measured by observing the electron’s resultant graphene. At high energies above the Dirac point, the kinetic energy. This spectrum was a reconstruction of the lifetimes were found to decay much faster than those close band structure of the graphene and since our method for to the Dirac point as seen in Figure 3. This was attributed time-resolved photoemission is a true single photon pro- to the reduced phase space volume available for electron- cess, an accurate reconstruction of the electrons and holes phonon interactions close to the Dirac point in graphene’s could be measured directly. Figure 1(b) shows a plot of the unique zero-gap band structure [6, 7]. This phenomenon 246

Page 247

was further studied by directly measuring the relaxation times of carriers and holes generated in an energy range of +/- 1 eV around the Dirac point. We limited our photon energy to this range since the energy band structure is linear and we can eliminate intraband scattering processes. Our experimental results showed that after excitation, the carriers and holes redistribute themselves on a few fs tim- escale through various methods [10, 11]. After thermaliza- tion, the excited carriers cascade back to the ground state with those closest to the Dirac point relaxing last. Our experimental results were also compared to a full tight-binding model to give further insight. The model includes many-body interactions which represent the coupling of electronic states to the lattice [23, 24]. These results along with the most recent experimental measure- ments can be seen in Figure 3 and in References [25, 26]. The resultant theoretical approach gave qualitatively iden- tical results to the experiments. Additional measurements that include the angular dependence of the photoelec- Figure 1. (a) A model of the carrier distributions in the linear trons may yield more insight on which portions of the band region of graphene. The red color represents filled states and the structure are responsible for the quasiparticle dynamics white represents empty states. (b) A plot of the static photo- while even shorter pulse durations can probe the initial emission signal from the graphene sample mounted on a silicon thermalization processes after optical excitation. dioxide substrate. The inset shows the signal on a log plot in the vicinity of the linear portion of the band structure. Impact on National Missions The research in this project was aimed at fundamental understanding of electronic structure and behavior of the complex correlated electronic material of graphene and with attosecond prulse trains setting the ultimate tempo- ral resolution. It supported fundamental experiments as a foundation for future energy technologies by improving our understanding of new material properties. Hence, this project embraced the Laboratory’s Materials Strategy, as well as the LDRD Grand Challenge in Materials. Broadening our knowledge about electrical transport and the physics of low dimensional electron systems is relevant to the DOE energy security mission, and ultimately National Security, by providing the scientific base for next generation energy efficient technologies. The scientific capabilities demon- strated in this project enhanced the Laboratory’s portfolio in these studies for DOE Basic Energy Sciences program in Material Sciences and Engineering. Figure 2. (a) The photoemission signal above the Dirac point in graphene as a function of delay. (b) A plot of the time varying chemical potential (black squares) and temperature (red squares) as extracted from the data in (a). 247

Page 248

8. Zhou, S. Y., G. -H. Gweon, A. V. Fedorov, P. N. First, W. A. De Heer, D. -H. Lee, F. Guinea, A. H. Neto, and A. Lanzara. Substrate-induced bandgap opening in epitax- ial graphene. 2007. NATURE MATERIALS. 6 (11): 916.
9. Ohta, T., A. Bostwick, T. Seyller, K. Horn, and E. Roten- berg. Controlling the electronic structure of bilayer graphene. 2006. SCIENCE. 313 (5789): 951.
10. Rana, F.. Electron-hole generation and recombination rates for Coulomb scattering in graphene. 2007. PHYSI- CAL REVIEW B. 76 (15): 155431.
11. Winzer, T., and E. Malic. Microscopic study of the ef- ficiency of Coulomb and phonon induced relaxation channels in graphene. 2011. Phys. Status Solidi B. 248: Figure 3. The measured quasiparticle lifetime is shown as the 2615. black squares. The blue dashed line represents a Lorentzian fit to the experimental data. The solid red line is the comparison with 12. Kim, R., V. Perebeinos, and P. Avouris. Relaxation of the theoretical model. optically excited carriers in graphene. 2011. PHYSICAL REVIEW B. 84 (7): 075449. References
12. Dakovski, G. L., Y. Li, T. Durakiewicz, and G. Rodriguez. 13. Sun, D., Z. Wu, C. Divin, X. Li, C. Berger, W. de Heer, Tunable ultrafast extreme ultraviolet source for time- P. First, and T. Norris. Ultrafast Relaxation of Excited and angle-resolved photoemission spectroscopy. 2010. Dirac Fermions in Epitaxial Graphene Using Optical Dif- Review of Scientific Instruments. 81 (7): 073108. ferential Transmission Spectroscopy. 2008. PHYSICAL REVIEW LETTERS. 101 (15): 157402.
13. Sun, Z., D. Popa, T. Hasan, F. Torrisi, F. Wang, E. J. R. Kelleher, J. Travers, V. Nicolosi, and A. C. Ferrari. A 14. Breusing, M., S. Kuehn, T. Winzer, E. Malic, F. Milde, N. stable, wideband tunable, near transform-limited, Severin, J. P. Rabe, C. Ropers, A. Knorr, and T. Elsaesser. graphene-mode-locked, ultrafast laser. 2010. Nano Ultrafast nonequilibrium carrier dynamics in a single Research. 3: 653. graphene layer. 2011. PHYSICAL REVIEW B. 83 (15):
15. Gusynin, V. P., and S. G. Sharapov. Unconventional inte- ger quantum Hall effect in graphene. 2005. PHYSICAL 15. Lui, C. H., K. F. Mak, J. Shan, and T. Heinz. Ultrafast REVIEW LETTERS. 95 (14): 146801. Photoluminescence from Graphene. 2010. PHYSICAL REVIEW LETTERS. 105 (12): 127404.
16. Zhang, Y., Y. Tan, H. Stormer, and P. Kim. Experimental observation of the quantum Hall effect and Berry’s 16. Ryzhii, V., M. Ryzhii, A. Satou, T. Otsuji, A. A. Dubinov, phase in graphene. 2005. Nature. 438 (7065): 201. and V. Ya. Aleshkin. Feasibility of terahertz lasing in optically pumped epitaxial multiple graphene layer
17. Rohwer, T., S. Hellmann, M. Wiesenmayer, C. Sohrt, structures. 2009. JOURNAL OF APPLIED PHYSICS. 106 A. Stange, B. Slomski, A. Carr, Y. Liu, L. M. Avila, (8): 084507. Kalla&uml, S. Mathias, L. Kipp, K. Rossnagel, and M. Bauer. Collapse of long-range charge order tracked by 17. Dorn, M., P. Lange, A. Chekushin, N. Severin, and J. P. time-resolved photoemission at high momenta. 2011. Rabe. High contrast optical detection of single gra- Nature. 471 (7339): 490. phenes on optically transparent substrates. 2010. JOURNAL OF APPLIED PHYSICS. 108 (10): 106101.
18. Park, C. H., F. Giustino, C. Spataru, M. Cohen, and S. G. Loiue. Angle-resolved photoemission spectra of 18. Ferrari, A. C., J. C. Meyer, V. Scardaci, C. Casiraghi, M. graphene from first principles. 2009. Nano Letters. 9: Lazzeri, F. Mauri, S. Piscanec, D. Jiang, K. S. Novoselov,
19. S. Roth, and A. K. Geim. Raman spectrum of graphene and graphene layers. 2006. PHYSICAL REVIEW LETTERS.
20. Bostwick, A., T. Ohta, T. Seyller, K. Horn, and E. Ro- 97 (18): 187401. tenberg. Quasiparticle dynamics in graphene. 2007. NATURE PHYSICS. 3 (1): 36. 19. Knox, K., S. Wang, A. Morgante, D. Cvetko, A. Locatelli, 248

Page 249

T. O. Mentes, M. A. Nino, P. Kim, and R. M. Osgood. tion of carrier distributions in graphene. Presented at Spectromicroscopy of single and multilayer graphene Annual Meeting of the Conference on Lasers and Elec- supported by a weakly interacting substrate. 2008. tro-Optics (CLEO) 2012. (San Jose, CA, 6-11 May 2012). PHYSICAL REVIEW B. 78 (20): 201408. Gilbertson, S., G. Dakovski, T. Durakiewicz, J. X. Zhu, A. D. Mohite, A. Dattelbaum, and G. Rodriguez. Time re- 20. Wang, Y. Y., Z. H. Ni, T. Yu, Z. X. Shen, H. M. Wang, Y. H. solved carrier distributions in graphene. Presented at Wu, W. Chen, and null. Wee. Raman studies of mono- American Physical Society March Meeting 2012. (Bos- layer graphene: The substrate effect. 2008. JOURNAL ton, MA, 27 Feb. - 2 Mar. 2012). OF PHYSICAL CHEMISTRY C. 112 (29): 10637. Gilbertson, S., G. Dakovski, T. Durakiewicz, J. Zhu, K. Dani, 21. Gilbertson, S., G. Dakovski, T. Durakiewicz, J. Zhu, K. A. Mohite, A. Dattelbaum, and G. Rodriguez. Tracing Dani, A. Mohite, A. Dattelbaum, and G. Rodriguez. ultrafast separation and coalescence of carrier distribu- Tracing ultrafast separation and coalescence of carrier tions in graphene with time-resolved photoemission. distributions in graphene with time-resolved photo- 2012. Journal of Physical Chemistry Letters. 3 (1): 64. emission. 2012. Journal of Physical Chemistry Letters. Gilbertson, S., T. Durakiewicz, J. Zhu, A. Mohite, A. Dat- 3 (1): 64. telbaum, and G. Rodriguez. Direct measurement of quasiparticle lifetimes in graphene using time-resolved 22. Gilbertson, S., G. Dakovski, J. X. Zhu, K. Dani, A. Dat- photoemission. 2012. JOURNAL OF VACUUM SCIENCE telbaum, and G. Rodriguez. Separation and recombi- & TECHNOLOGY B. 30 (3): 03D116. nation of carrier distributions in graphene. Presented at Annual Meeting of the Conference on Lasers and Electro-Optics (CLEO) 2012. (San Jose, CA, 6-11 May 2012). 23. Neto, A. H. Castro, F. Guinea, N. M. Peres, K. S. No- voselov, and A. K. Geim. The electronic properties of graphene. 2009. REVIEWS OF MODERN PHYSICS. 81 (1): 109. 24. Piscanec, S., M. Lazzeri, F. Mauri, A. C. Ferrari, and J. Robertson. Kohn anomalies and electron-phonon inter- actions in graphite. 2004. PHYSICAL REVIEW LETTERS. 93 (18): 185503. 25. Gilbertson, S., T. Durakiewicz, J. Zhu, A. Mohite, A. Dattelbaum, and G. Rodriguez. Direct measurement of quasiparticle lifetimes in graphene using time-resolved photoemission. 2012. JOURNAL OF VACUUM SCIENCE & TECHNOLOGY B. 30 (3): 03D116. 26. Gilbertson, S., G. Dakovski, T. Durakiewicz, J. X. Zhu, K. Dani, A. Mohite, A. Dattelbaum, and G. Rodriguez. Time resolved carrier distributions in graphene. Pre- sented at American Physical Society March Meeting 2012. (Boston, MA, 27 Feb. - 2 Mar. 2012). Publications Gilbertson, S.. Attosecond observation and control of au- toionization in helium. Invited presentation at Interna- tional Symposium on (e,2e), Double Photoionization and Related Topics & 16th International Symposium on Polarization and Correlation in Electronic and Atomic Collisions. (Dublin, Ireland, 4-6 Aug. 2011). Gilbertson, S., G. Dakovski, J. X. Zhu, K. Dani, A. Dattel- baum, and G. Rodriguez. Separation and recombina- 249

Page 250

Chemistry and Material Sciences Postdoctoral Research and Development Final Report Mechanistic Investigation of C-C Bond Forming Reactions for the Production of Long Chain Hydrocarbons Enrique R. Batista 20110542PRD1 Abstract Scientific Approach and Accomplishments Due to instabilities in the global supply of crude oil, the Detailed mechanistic results derived from theory focus- development of bio-fuels is a top priority. A critical step ing on the C-C bond-forming step in this process were toward realization of high energy density bio-fuels is the produced for the piperidine catalyzed aldol condensation development of organic fuels that consist of a narrow between HMF and acetone. The catalytic aldol con- distribution of longer hydrocarbon chains, which have densation process was determined to proceed through higher energy density than current bio-fuels like ethanol. a six-step mechanism involving a two-step substrate The key challenge in the synthesis of long hydrocarbon activation, a C-C bond forming step, a two-step release chains is the formation of new carbon-carbon bonds. of products/catalyst regeneration followed by an addi- One fundamental organic transformation ubiquitous in tional dehydration of the products. In order to arrive at modern organic chemistry, the aldol condensation, is a this conclusion we had to analyze all possible reactions carbon-carbon bond-forming reaction involving ketones involving the catalyst and the two substrates as well as and aldehydes (species with a C-O double bond, or car- the involvement of the solvent molecules during each bonyl group). Our experimental collaborators have en- step of the reaction. This last consideration is usually joyed some success with a new type of catalyst that can treated in a very approximated fashion or ignored all affect this reaction. A mechanistic understanding can together. We found that many of the steps along the lead to new options for improvement and optimization. reaction pathway require proton transfers, which are cat- We will use computational tools to analyze the atomistic alyzed by clusters of water molecules requiring inclusion aspects of the reaction mechanism and use this under- of explicit solvent. In addition, this mechanism proceeds standing to test new possibilities that will then be visited through a very reactive zwitterionic intermediate previ- by the experimentalists for testing in the laboratory. ously believed to react with itself. Our more accurate approach utilizing explicit solvent molecules was used to Background and Research Objectives clarify this confusion in the literature about intermediate In an effort to upgrade bio-renewables to high-energy- steps of this nature in similar reactions. Consideration density fuels and chemical feedstocks we are developing of those confused intermediate structures would make a strategy to convert readily available cellulose-derived one believe that a key step in the catalytic process is not polyols (such as hydroxymethyl furfural HMF) to high- possible. density C8-C15 hydrocarbons. This first involves chain This work has been further expanded to examine the lengthening through C-C bond-forming aldol conden- C-C bond-forming step in a related aldol condensation sations followed by systematic defunctionalization process utilizing bio-derived ethyl levulinate in place involving such challenges as catalytic ring-opening of of acetone. This procedure also requires an acetic acid furans, removal of oxygen atoms through dehydration co-catalyst and is performed in the absence of water. and reduction and removal of the resulting unsatura- While the general mechanism has not changed these tion through hydrogenation. The project consists in a new reaction conditions present several new problems: tightly coupled effort with experimental groups, that are

1. Unlike acetone, ethyl levulinate has multiple active synthesizing catalyst molecules, and our computational cites requiring understanding of the observed regiose- work aimed at understanding the underlying mecha- lectivity; 2) While the general mechanism is believed nisms of catalysis. Our effort will inform experimental to be the same, water, which was found to be critical in groups and help them better tailor their efforts. the previous studies, is no longer present; 3) The role 250

Page 251

of the co-catalyst is unknown. Through an examination of References the possible reaction pathways we have determined that 1. Keith, J. M., J. K. Kim, L. Alexander, R. Wu, R. L. Mar- while reaction at multiple sites in ethyl levulinate leads to tin, E. R. Batista, R. Michalczyk, B. Scott, S. K. Hanson, thermodynamically stable products, the observed product A. D. Sutton, J. C. Gordon, and L. A. P. silks. Solvent- is the result of substantially lower kinetic barrier. The other free Organocatalysis Applicable to the Carbon Chain two issues were found to solve each other as the acetic Extension of Carbohydrate Derivatives: Application to acid was found to act as a proton transfer catalyst in the the Production of Transportation Fuels. J. of Organic place of water. Chemistry. Finally, we have moved into studies of O-O bond activation 2. Keith, J. M.. Insertion of Molecular Oxygen into Rh-H: mediated by transition metal complexes. These systems Sterically Induced Mechanistic Crossover Between are also of interest in energy fuel cells. A fuel cell recovers Radical-Chain and Hydrogen-Atom-Abstraction Path- stored energy in the form of hydrogen via a four-proton, ways. Angew. Chem. Int. Ed.. four-electron reduction of oxygen at the cathode. In an effort to understand this mechanistically complex chemis- 3. Keith, J. M., T. S. Teets, and D. G. Nocera. O2 Insertion try, much work has been performed over the last quarter into Group 9 Metal-Hydride Bonds: Evidence for Oxy- century to develop molecular analogues for the study of gen Activation through the Hydrogen-Atom-Abstrac- oxygen reduction. In biology, oxidase enzymes such as tion Mechanism. 2012. Inorg. Chem.. 51 (17): 9499. cytochrome c oxidase perform this same chemistry using oxygen as the terminal oxidant. By employing oxygen in Publications this same way, a rich array of aerobic oxidation chemistry Daly, S., J. Keith, E. Batista, K. Boland, S. Kozimor, R. Martin, has been developed employing palladium catalysts. In all and B. Scott. Probing NiS2PR2 electronic structure to generate insight relevant to minor actinide extrac- of these chemistries, understanding the precise nature of tion chemistry. 2012. Inorganic Chemistry. 51 (14): the reaction of O2 with the given metal complex is crucial 7551. to further understanding and advancing these systems. Keith, J. M.. Insertion of Molecular Oxygen into Rh-H: Steri- The initial work has been presented at the Gordon Con- cally Induced Mechanistic Crossover Between Radical- ference on reaction mechanisms, and at the American Chain and Hydrogen-Atom-Abstraction Pathways. . Chemical Society National meeting in Denver, CO. A joint Angew. Chem. Int. Ed.. experimental/theory paper has been accepted in the peer-reviewed journal Current Organic Chemistry with an Keith, J. M., J. K. Kim, L. Alexander, R. Wu, R. L. Martin, exhaustive description of the whole catalytic process. In E. R. Batista, R. Michalczyk, B. Scott, S. K. Hanson, A. D. Sutton, J. C. Gordon, and L. A. P. Silks. Solvent-free addition a new joint experimental/theory manuscript is in Organocatalysis Applicable to the Carbon Chain Exten- preparation detailing the mechanistic differences involved sion of Carbohydrate Derivatives: Application to the with substrate modification, changes in solvent and the Production of Transportation Fuels. J. of Organic Chem- addition of co-catalysts. The oxygen activation work was istry. published in two papers in the journal Inorganic Chemistry. Keith, J. M., J. K. Kim, L. Alexander, R. Wu, R. L. Martin, E. Impact on National Missions R. Batista, R. Michalzyk, B. L. Scott, S. K. Hanson, A. D. This project includes organic aldol catalysis for carbon- Sutton, L. A. P. Silks, and J. C. Gordon. Aqueous organo- carbon chain extension. Optimization of this process catalysis applicable to the carbon chain extension of carbohydrate derivatives: Application to the produc- will result in significant impact to the fuels and specialty tion of transportation fuels. To appear in Current Or- chemical sectors. Because the approach to producing ganic Chemistry. liquid fuels is based on bio-mass, this project contributed to the carbon-neutral fuel cycle impacting environment Keith, J. M., T. S. Teets, and D. G. Nocera. O2 Insertion into and climate missions. The development of efficient cata- Group 9 Metal-Hydride Bonds: Evidence for Oxygen lytic mechanisms for production of biofuels has also been Activation through the Hydrogen-Atom-Abstraction identified as a research need by the Office of Science BES Mechanism. 2012. Inorg. Chem.. 51 (17): 9499. report on “Clean and Efficient Combustion of 21st Century Keith, J., E. Batista, R. Martin, R. Wu, L. Pete. Silks, and J. Transportation Fuels”. Having established credibility in the Gordon. Catalyzed conversion of non-food biomass areas of biomass conversion is important for EERE and BES to fuels: Probing the mechanism of the initial C-C missions in the development of new cycles for biofuels. bond forming step. 2011. In 242nd National Meeting of the American-Chemical-Society (ACS) ; 20110828 - 251

Page 252

20110901 ; Denver, CO. (Denver, CO, Aug. 28 - Sep. 1st, 2011). Vol. 242, p. 1. Denver, CO: ACS. Keith, J., and E. Batista. Theoretical Examination of the Thermodynamic Factors in the Selective Extraction of Am3+ from Eu3+ by Dithiophosphinic Acids. 2012. In- organic Chemistry. 51 (1): 13. 252

Page 253

Chemistry and Material Sciences Postdoctoral Research and Development Final Report Exploiting the Fermi Surface to Understand Quantum Criticality in Complex Electronic Materials Eric D. Bauer 20110549PRD1 Abstract from the nearly degenerate kinetic and potential energy Standard models for simple metals and insulators often of the f-electrons. The two extreme limits of f-electron fail for strongly interacting electronic systems in which behavior, i.e., magnetic order or itinerant paramanget- correlations between d- or f-electrons and conduction ism, due to competition between Coulomb repulsion electrons promote new and unexpected states of matter. (potential energy) and orbital overlap with neighbor- An increasingly common example of emergent phenom- ing ligands (kinetic energy), respectively, are well- ena in these complex electronic materials (e.g., heavy understood phenomena with in standard models for fermions and high-Tc cuprates) is found when a magnetic metals. An increasingly common example of emergent phase transition is tuned to absolute zero temperature phenomena in these complex electronic materials (e.g., by an external parameter, such as doping, pressure, or heavy fermions and high-Tc cuprates) is found when magnetic field. At the resulting quantum critical point, a magnetic phase transition is tuned to absolute zero new and distinct states are found. In particular, jiggling temperature by an external parameter, such as doping, of spins of the f-electrons (“spin fluctuations”) that are pressure, or magnetic field. At the resulting quantum abundant and violent at the quantum critical point lead critical point, new and distinct states are found, which to unusual behavior that do not follow the expectations cannot be understood by knowing the properties of the for a simple metal. This type of behavior has continued two parent magnetic or nonmagnetic states. This type of to defy a complete description after nearly 20 years behavior has continued to defy a complete description and poses one of the most significant challenges to the after nearly 20 years and poses one of the most signifi- condensed matter community. A crucial yet unresolved cant challenges to the condensed matter community. issue of quantum criticality is what happens to the Fermi While many theories have been proposed, these mod- surface (the surface of constant energy that dictates the els do not describe all the properties of even a single motion of the valence electrons) at the quantum criti- quantum critical material. A solution to this quantum cal point. In general, the Fermi surface is expected to be criticality conundrum will come from carefully designed small in the magnetic state, where the f-electrons are experiments that distinguish between existing theories located close to the atomic core (localized) and evolves and guide in the development of new ones. A crucial yet into a large Fermi surface when the f-electrons on the unresolved issue of quantum criticality is what happens other side of the quantum critical point. The two leading to the Fermi surface (the surface of constant energy models of quantum criticality have distinct predictions that dictates the motion of the valence electrons) at the for this small-to-large Fermi surface crossover—either quantum critical point. The synthesis and character- it is continuous or occurs abruptly. Director’s Funded ization of ultra-high purity f-electron compounds that postdoctoral Fellow Ryan Baumbach will synthesize and exhibit quantum critical behavior, mainly involving Ce, characterize f-electron compounds that exhibit quantum and U, provides the means to quantitatively determine critical behavior to quantitatively determine the evolu- the evolution of the Fermi surface volume through direct tion of the Fermi surface volume. This research will lay measurements including de Haas van Alphen and Hall the scientific foundation for understanding quantum effect. This research will lay the scientific foundation for criticality. understanding quantum criticality and will play a central role in understanding and exploiting functional complex Background and Research Objectives electronic materials of the 21st century. In f-electron materials, novel states of matter emerge 253

Page 254

Scientific Approach and Accomplishments from the static susceptibility. The anisotropy of the spin- Director’s Funded Ryan Baumbach’s research for this lattice relaxation rates indicates that the anisotropy of spin project is focused around two main goals: 1) to search fluctuations are also Ising-like along the c-axis. The abso- for promising quantum critical f-electron materials and lute values and temperature dependence of spin-lattice to characterize their properties and 2) explore the Fermi relaxation rate suggest a well-localized 4f picture for this surface of these materials near the quantum critical point. system: The local moments of Ce ions fluctuate as a result Below is a summary of his accomplishments towards these of ferromagnetic intersite coupling above the ferromag- goals. netic transition and individually fluctuate due to intrasite coupling well above Tc [4]. These measurements provide Search for New Quantum Critical Systems important clues for the microscopic origin of the magne- Baumbach discovered a new class of materials with the tism in CeRu2Ga2B and point to future directions to tune chemical formula LnT2M2X (Ln = La-Nd, T = Ru, Os; M = this new class of magnets to a quantum critical point. Al, Ga; X = B, C). These compounds crystallize in a filled variant of the CeMg2Si2–type structure and are closely Explore the Fermi surface of these materials near the related to the ubiquitous ThCr2Si2-type structure, which quantum critical point is associated with a large number of systems that exhibit In solids containing elements with f-orbitals, the interac- non-Fermi-liquid behavior and quantum criticality. As such, tion between f-electron spins and those of itinerant elec- the LnT2M2X materials provide an exciting new avenue trons leads to the development of low-energy excitations for the study of correlated electron physics. Attention has with a heavy effective mass. These excitations are funda- been focused on the CeT2M2X materials, as these are mental to the appearance of unconventional superconduc- the most likely to exhibit quantum critical behavior. Three tivity and non-Fermi liquid behavior observed in actinide- compounds: CeRu2Al2B, CeRu2Ga2B, and CeRu2Ga2C and lanthanide-based compounds. In particular, massive have been synthesized. Contrary to what is often observed mobile electrons are expected to emerge from complex for Ce-based compounds, these three compounds show interactions in a periodic array of f-electron atoms, the pronounced local magnetic moment behavior, resulting Kondo lattice, only at low temperatures— which has never in ferromagnetic ordering of the Ce-ions at strikingly high observed directly before. Scanning tunneling microscopy temperatures: i.e., Curie temperatures Tc = 12.8 K, 16.1 K, has been used to detect the emergence of heavy excita- and 16.7 K for CeRu2Al2B, CeRu2Ga2B, and CeRu2Ga2C, tions with lowering of temperature in a prototypical family respectively. In addition, CeRu2Al2B undergoes antifer- of CeMIn5 (M=Co, Rh) heavy fermion compounds, which romagnetic ordering at a Neel temperature TN = 14.1 K. provides first direct visualization of heavy fermion band The application of a magnetic field H to CeRu2Al2B results structure and developing ground state (Figure 1). Further- in a rich phase diagram that includes three magnetically more, these measurements reveal signature of quantum ordered phases: (1) antiferromagnetic, (2) canted antifer- criticality, opening up an entirely new window for un- romagnetic, and (3) ferromagnetic, indicating that there derstanding this decades-old conundrum found in many are several finely tuned exchange interactions [1-2]. classes of energy-saving materials [5]. The unit cell volumes for these three compounds decrease Baumbach’s efforts produced ultra-high purity samples of in the order CeRu2Al2B, CeRu2Ga2B, CeRu2Ga2C, com- the Ce2T3X5 (T = transition metal and X = Si, Ge) family of mensurate with the increase in Tc. This trend suggests that correlated electron compounds, which includes the pres- the Ce 4f electrons in these materials localized and mag- sure-induced superconductor Ce2Ni3Ge5. Of these materi- netic, supporting the point of view that in order to access als, the antiferromagnets Ce2Rh3Ge5 and Ce2Pt3Si5 are the quantum critical point, it is necessary to further reduce particularly interesting (Neel temperature TN = 5.5 K and the unit cell volume. To this end, the temperature (T)– 6.3 K, respectively). Recent studies at the National High pressure (P) phase diagrams were studied up to pressures Magnetic Field Laboratory in Los Alamos show that in ad- of 25 kbar. The evolution of properties with chemical and dition to having similar ordering temperatures and associ- applied pressure suggests that CeRu2Al2B is furthest from ated magnetic 4f-electron entropy their antiferromagnetic the possible quantum phase transition and CeRu2Ga2C is states are suppressed to zero temperature at similar mag- the nearest to one [3]. netic fields (critical fields Hc = 23 T and 36 T, respectively as shown in Figure 2), suggesting comparable magnetic To further characterize the microscopic properties of energy scales in these compounds. In contrast, while the this new class of materials, nuclear magnetic resonance pressure needed to access a quantum critical point (QCP) (NMR) studies were performed on the tetragonal Ce- in Ce2Rh3Ge5 is extremely low (Pc = 5 kbar), the Néel tem- based ferromagnet CeRu2Ga2B. The Knight shifts show perature for Ce2Pt3Si5 is insensitive to pressures up to 15 an Ising-type anisotropy along the c axis, similar to results 254

Page 255

kbar. These results imply that although these compounds are markedly similar, the mechanism that drives the QCP in Ce2Rh3Ge5 is not present in Ce2Pt3Si5. In order address this difference, it is essential to investigate their behavior in the vicinity of their field and pressure driven quantum phase transitions. Preliminary measurements to quantify the shape of the Fermi surface for H > Hc for both of these compounds, have been carried out at low temperatures and high magnetic fields for both of these compounds, as illustrated in (Figure 3). Future measurements at lower temperatures will be performed in April, 2013 to observe Figure 2. Temperature – magnetic field (T-H) phase diagrams for the Fermi surface change across the quantum critical point Ce2Rh3Ge5 and Ce2Pt3Si5. For H < 9 T, data were gathered using in these Ce2M3X5 materials. electrical resistivity, specific heat, and magnetic susceptibility techniques in DC magnetic fields. Data points at higher magnetic Impact on National Missions fields were collected in pulsed magnetic fields up to 60 T. The study of quantum criticality is one of the true frontiers in modern science. Understanding the essential interac- tion among strongly correlated electron materials is a necessary step for the Laboratory and for DOE to fulfill its missions in energy security. This research will provide the scientific foundation for understanding and exploiting func- tional complex electronic materials of the 21st century. CeCoIn 5 Figure 3. Main panels: Fourier transform of the oscillations in Ce2Rh3Ge5 (left panel) and Ce2Pt3Si5 (right panel) for H > Hc revealing different frequencies corresponding to various Fermi surface sections at several representative temperatures. Insets: Raw tunnel diode oscillator data at 0.69 K (after polynomial background subtraction) showing quantum oscillations. References

1. Baumbach, R. E., T. Shang, M. Torrez, F. Ronning, J. D. Thompson, and E. D. Bauer. Local moment ferromag- netism in CeRu2Ga2B. 2012. Journal of Physics-Con- densed MatterJournal of Physics-Condensed Matter. 24 (18).
2. Baumbach, R. E., H. Chudo, H. Yasuoka, F. Ronning, E. D. Bauer, and J. D. Thompson. CeRu2Al2B: A local- mo- ment 4 f magnet with a complex T-H phase diagram.
3. Physical Review BPhysical Review B. 85 (9).
4. Baumbach, R. E., X. Lu, F. Ronning, J. D. Thompson, and E. D. Bauer. Pressure tuned ferromagnetism in Ce- Ru2M2X (M = Al, Ga; X = B, C). 2012. Journal of Phys- Figure 1. Transition from a small to large Fermi surface in Ce- ics-Condensed MatterJournal of Physics-Condensed CoIn5 determined via scanning tunneling spectroscopy showing Matter. 24 (32). the development of the heavy fermion state with decreasing temperature. 4. Sakai, H., Y. Tokunaga, S. Kambe, R. E. Baumbach, F. 255

Page 256

Ronning, E. D. Bauer, and J. D. Thompson. NMR study for 4f-localized ferromagnet CeRu2Ga2B. 2012. Physi- cal Review BPhysical Review B. 86 (9). 5. Aynajian, P., E. H. da Silva Neto, A. Gyenis, R. E. Ba- umbach, J. D. Thompson, Z. Fisk, E. D. Bauer, and A. Yazdani. Visualizing heavy fermions emerging in a quantum critical Kondo lattice. 2012. NatureNature. 486 (7402): 201.. Publications Aynajian, P., E. H. da Silva Neto, A. Gyenis, R. E. Baumbach, J. D. Thompson, Z. Fisk, E. D. Bauer, and A. Yazdani. Vi- sualizing heavy fermions emerging in a quantum criti- cal Kondo lattice. 2012. NatureNature. 486 (7402): 201. Baumbach, R. E., H. Chudo, H. Yasuoka, F. Ronning, E. D. Bauer, and J. D. Thompson. CeRu2Al2B: A local- mo- ment 4 f magnet with a complex T-H phase diagram. 2012. Physical Review BPhysical Review B. 85 (9). Baumbach, R. E., T. Shang, M. Torrez, F. Ronning, J. D. Thompson, and E. D. Bauer. Local moment ferromag- netism in CeRu2Ga2B. 2012. Journal of Physics-Con- densed MatterJournal of Physics-Condensed Matter. 24 (18). Baumbach, R. E., X. Lu, F. Ronning, J. D. Thompson, and E. D. Bauer. Pressure tuned ferromagnetism in CeRu2M2X (M = Al, Ga; X = B, C). 2012. Journal of Physics-Con- densed MatterJournal of Physics-Condensed Matter. 24 (32). Sakai, H., Y. Tokunaga, S. Kambe, R. E. Baumbach, F. Ron- ning, E. D. Bauer, and J. D. Thompson. NMR study for 4f-localized ferromagnet CeRu2Ga2B. 2012. Physical Review BPhysical Review B. 86 (9). 256

Page 257

Chemistry and Material Sciences Postdoctoral Research and Development Final Report Non-aqueous Organic Materials for Bio-defense James M. Boncella 20110580PRD2 Abstract chemical and biological tools. Hydrophobic, protic, and Room-temperature ionic liquids (RTILs) are uniquely aprotic ionic liquids have been designed, synthesized, composed of cation-anion pairs that form non-volatile, and tuned depending upon requirements of the biologi- low-density, thermally stable liquids at room tempera- cal system, and applied to diverse problems related to ture, and can be rationally designed such that they antibiotic-resistant pathogens, biodefense, and bioen- interact with biomolecules in highly advantageous ways. ergy. We have investigated and reduced to practice the use of We have developed a new general strategy for produc- RTILs to stabilize, solubilize, and eliminate polymorphs ing ionic liquid formulations of cationic pharmaceutical of small molecule drugs, penetrate skin for transdermal agents that is more robust than the current state of the drug delivery, eliminate biofilms of pathogenic bacte- art and features a simplified synthetic method. Many rial, which are antibiotic-resistant due to their nearly drugs form crystalline polymorphs that differ in solubil- impenetrable extracellular matrix, and stabilize carbonic ity in the acidic stomach environment. Some drugs are anhydrase, an enzyme useful in some alternative energy poorly absorbed due to low solubility of the crystalline strategies, These results are relevant to the LANL mission form. Increased bioavailability through drug formulation because we have shown defeat of antibiotic-resistant with ionic liquids would mean smaller effective dosages biofilms, and eight-fold increased thermal stability of and would eliminate batch variations due to polymor- drugs to temperature extremes and sunlight exposure. phism. In addition to circumventing crystal polymor- The work has received a follow-on grant to address phism issues, ionic liquid formulations of medicines are defense-related needs for skin-based delivery of phar- also useful because they can be designed to stabilize maceutical agents. Three patent applications and four drugs to high temperature or UV irradiation, and im- peer-reviewed publications have also arisen from this prove or control solubility and bioavailability of the work. drug. We have produced thermally stabilized forms of small molecule drugs as well as trans-dermally bioavail- Background and Research Objectives able forms of pharmaceuticals that must otherwise be Room-temperature ionic liquids (RTILs), are uniquely administrated i.v. in a hospital setting (Figure 1). Materi- composed of cation-anion pairs that form non-volatile, als synthesized have been evaluated on stability vs. time low-density, thermally stable liquids at room tempera- and pH, purity, melting/glass transitions, water-octanol ture, and interact with biomolecules in highly advan- partitioning, and cost effectiveness. tageous ways. For example, RTILs, which are neither hydrophobic nor hydrophilic in the traditional definition, Ionic liquids are not currently used for transdermal ap- can stabilize proteins, solubilize substrates, and improve plications because they have not been sufficiently stud- enzyme or biocatalyst selectivity and activity. These non- ied for drug delivery and cytotoxicity. Our basic research aqueous fluid materials can be used to address many gave rise to a funded project to address these bottle- areas related to the LANL mission by rational design for necks. The funded project is advancing the design of pre- application in biological processes. RTILs are unique ma- liminary topical formulations that were developed based terials as they are primarily organic in nature, enabling on FDA-approved and GRAS (generally recognized as safe rapid modification to address specific applications. This ingredients). The properties of these formulations are ability to tune the properties of RTILs renders them ex- being optimized and the investigation is leading to a bet- tremely powerful new materials and allows investigation ter understanding of the science and design criteria of into the basic science that drives their functionality as 257

Page 258

controlling transdermal efficacy and improving toxicity to starting from four different metallochlorides (ZnCl2, FeCl3, biofilms. This research has also given rise to investigations CoCl2, and SnCl4). The use of Lewis-acidic ionic liquids of the efficacy of ionic liquids in treating infections due to succeeds without production of side products or impuri- pathogenic biofilms. ties because the chloride of the pharmaceutical agent combines with ZnCl2 to produce zinc chloride anions and Scientific Approach and Accomplishments oligomers (Figure 2) that contribute to the disorder and The scientific approach has involved the synthesis of over reduced melting temperatures of the system. We also real- 100 ionic liquids and deep eutectic solvents. We began by ize benefits of eliminating the use of ion exchange columns using existing principles to design materials with desir- due to the extra time it takes to pass reagents through the able room temperature properties such as a low melting column, the loss of a certain percentage of reagent to the temperature, low viscosity, high or low conductivity, high stickiness of the column and the cost of the column resin. or low water solubility, low toxicity, or high lipophilicity. Completed analyses on these materials (Figure 3), includ- We transitioned to relying on our own observations about ing glass transition temperatures, decomposition tem- structure-activity relationships, outlined below, for design- peratures, and viscosity for some ionic liquids synthesized ing new materials. as described, include thermal gravimetric analysis for The accomplishments of this project have included four determination of thermal stability, differential scanning peer-reviewed papers. At least two manuscripts destined calorimetry for identification of glass transitions and melt- for peer-review are in preparation, which will bring the ing temperatures, Raman spectroscopy for detection of total publication number to six. Three patent applications zinc chloride speciation, infrared spectroscopy for observa- have been filed for materials developed during this proj- tion of cation-anion interaction in solution, water-octanol ect. Specifically, one important accomplishment involved partition coefficients for determination of lipophilicity, the identification of a novel, unpublished, and eminently mass spectrometry for evaluation of purity and speciation, patentable synthetic strategy for reliably synthesizing ionic viscosity measurements, cyclic voltammetry, and determi- liquids from pharmaceutical salts. Our strategy, one that nation of electrical conductivity. These results have been was previously not identified by the ionic liquid and phar- published in peer-reviewed journals. maceutical formulation research communities, predictably The strategy has produced an agent with an 8-fold increase produces ionic liquids from small molecule drugs. Our own in shelf life at room temperature and a 10-fold increase in initial experiments with the state-of-the-art synthesis of shelf life at 40°C (Figure 4), which also has increased water pharmaceutical ionic liquids revealed issues not readily solubility as compared with the commercially available, revealed by published literature. The standard syntheses crystalline HCl salt. With another agent, we can reduce the involved metathesis reactions or activation of components number of cells in biofilms derived from clinically isolated, by ion exchange columns. The metathesis reactions inevi- pathogenic P. aeruginosa by three orders of magnitude. tably produced side products, such as NaCl, which were difficult to remove from the final ionic liquid. The use of Other accomplishments of this project have included the ion exchange columns also sometimes introduced impuri- identification of ionic liquids that allow the activity of the ties, either through material leaching from the columns enzyme carbonic anhydrase. This enzyme is of interest for themselves or because the organic molecules degraded in CO2 conversion and in biofuels applications. Ionic liquids the presence of the strongly acidic or basic ion exchange must have low viscosity to allow diffusion into and out of resins. The formation of ionic liquids containing active the enzyme active site and may not denature or inhibit the pharmaceutical agents using our newly introduced method enzyme. Scientific design criteria that were developed to has been successful for every pharmaceutical agent that promote carbonic anhydrase activity in ionic liquids includ- we have tried, as long as the melting point of the standard ed prioritization of ionic liquids with no “active” protons pharmaceutical salt is less than 300°C. Our strategy for and ionic liquids containing imidazolium derivatives as the new formulations is defined by four features: 1) inorganic- cationic component. Moreover, materials with bistrifluo- organic anion-cation pairs instead of organic-organic pairs, romethylsulfonimide were ultimately eliminated from con- 2) IL formation involving chlorometallate speciation in sideration. Imidazolium derivatives promote the activity Lewis acids, rather than proton transfer reactions, 3) GRAS- of carbonic anhydrase, as has been shown by x-ray crystal- based reagents to speed regulatory approval, and 4) use lography, and bistrifluoromethylsulfonimide is a suspected as starting materials of virtually any active pharmaceutical inhibitor of the enzyme. Preliminary, unpublished results agent with melting point <300°C. The synthetic method suggest that four ionic liquids with potential for use in a has produced 30 new pharmaceutically relevant ionic carbonic anhydrase system are 1-hexyl-1-methylimidazo- liquids, including an agent with potent anti-biofilm activity, lium chloride, ethylammonium nitrate, propylammonium 258

Page 259

nitrate, and 1,3-dimethylimidazolium dimethylphosphate. national security concerns related to administration of prophylactic agents in the case of a CBRNE attack on the Another accomplishment of the project is the develop- general population. ment of ionic liquid and deep eutectic solvent materials for promoting transdermal delivery of small molecule drugs. The creation of new materials from active pharmaceuti- We are the first to examine the interaction between neat cal compounds is a new area of study with relevance to ionic liquids and mammalian skin. Work is ongoing, but military and government applications. We have devel- preliminary investigations have revealed two excellent oped these new, biologically-active materials within the delivery agents, both of which include a terpenoid anion. two-year project, building a capability that is relevant to Tetradecyltrihexylphosphonium cations and choline cat- several missions. The work has developed our ability to ions have both been employed in this work to design and understand and design these new agent tool kits, and has produce ionic liquids and deep eutectic solvents with low addressed areas where significant intellectual property has viscosity, high lipophilicity, low conductivity, low toxicity to been generated. human primary lung cells, and high ionic strength. OH m A e n n o t t h o e f r i o a n cc ic o l m iq p u l i i d sh a m nd e n d t e o e f p t h eu is t e p c ro ti j c e s c o t l i v s e t n h t e m de at v e e r l i o a p ls - HN HO O H O 2N HO +H3N O NH O 2 NH 2 for elimination of clinically derived, pathogenic biofilms of O Pseudomonas aeruginosa. We are the first to investigate O O O S O O O Ionic liquid the toxicity of neat ionic liquids in pathogenic biofilms. The antibiotic, cation- use of neat ionic liquids instead of solvated ionic liquids anion pairs, GRAS/ FDA approved greatly improves dissolution of the biofilm. A neat ionic liquid contains only cations and anions and has strong wound treatment (no permeation) hydrogen bond-disrupting capabilities, in contrast with a solvated ionic liquid, which is essentially a highly con- dermis centrated aqueous solution and has reduced potential to break down the biofilm extracellular matrix. transdermal delivery (high permeation) Three collaborations have been initiated due to this proj- ect. A partnership with biofilm experts at Northern Arizona Figure 1. Transdermal drug delivery by IL. University (Prof. Andrew Koppisch) has allowed investi- gation of the antimicrobial effects of our ionic liquids in clinically isolated, biofilm-forming pathogenic bacteria. An effort involving students from Dixie State College of Utah (Prof. Rico Del Sesto) will allow investigation into the pre- vention of biofilm growth on ionic liquid coated catheters, army uniforms, metal surfaces, and fiberglass plates. Fi- nally, a team effort with transdermal drug delivery experts at the University of California-Santa Barbara (Prof. Samir Mitragotri) is beginning to identify mechanisms for ionic Figure 2. Ethambutol, a pharmaceutical used in the treatment of liquid penetration of skin. For the collaboration with UCSB, tuberculosis, is shown along with the zinc chloride species identi- we have won a two-year grant from this year’s highly com- fied by Raman spectroscopy that contribute to the fluidity of the petitive UC Lab Fee program, a result of the rapid progress ionic liquid formulation. of this project to garner further interest in more applied aspects of these materials. Impact on National Missions Drugs formulated as ionic liquids will be more stable at extremes of temperature and under exposure to sunlight and should therefore prove more useful for use by the warfighter. Other applications of this work will include addressing defense-related needs for rapid and facile skin- based or oral delivery of antimicrobial, anti-tubercular, and other pharmaceutical agents, as well as addressing 259

Page 260

Figure 3. Summary of properties, including glass transition temperatures, decomposition temperatures, and viscosity for a selection of ionic liquids produced during this project. Lovejoy, K. S., G. M. Purdy, S. Iyer, T. C. Sanchez, A. Rob- ertson, A. T. Koppisch, and R. E. Del Sesto. Tetraalkyl- phosphonium-Based Ionic Liquids for a Single-Step Dye Extraction/MALDI MS Analysis Platform. 2011. ANA- LYTICAL CHEMISTRY. 83 (8): 2921. Lovejoy, K. S., L. E. Davis, L. M. McClellan, A. M. Lillo, J. D. Welsh, E. N. Schmidt, C. K. Sanders, A. J. Lou, D. T. Fox, A. T. Koppisch, and R. E. Del Sesto. Evaluation of ionic Figure 4. Demonstration of increased thermal stability of am or- liquids on phototrophic microbes and their use in bio- phous chlorometallate formulations of pharmaceuticals. (A) fuel extraction and isolation. 2013. JOURNAL OF AP- Thermal stability as measured by mass loss (loss of HCl) over 4 PLIED PHYCOLOGY. 25 (4): 973. h at 120°C. The solid blue line represents the amorphous formulation and the dotted red line represents the unstabilized small molecule drug. (B) Demonstration of a predicted 8-fold increase in shelf life due to formulation of standard chloride salt pharmaceuticals as metal chloride ionic liquids. These ionic liquid species form due to metal chloride speciation of ZnCl<sub>2</ sub> and other pharmaceutically safe salts. Publications Lovejoy, K. S., A. J. Lou, L. E. Davis, T. C. Sanchez, S. Iyer, C. A. Corley, J. S. Wilkes, R. K. Feller, D. T. Fox, A. T. Koppisch, and R. E. Del Sesto. Single-Pot Extraction- Analysis of Dyed Wool Fibers with Ionic Liquids. 2012. ANALYTICAL CHEMISTRY. 84 (21): 9169. Lovejoy, K. S., C. A. Corley, E. K. Cope, M. C. Valentine, J. G. Leid, G. M. Purdy, J. S. Wilkes, A. T. Koppisch, and R. E. Del Sesto. Utilization of Metal Halide Species Ambigu- ity to Develop Amorphous, Stabilized Pharmaceutical Agents As Ionic Liquids. 2012. CRYSTAL GROWTH & DE- SIGN. 12 (11): 5357. 260

Page 261

Chemistry and Material Sciences Postdoctoral Research and Development Final Report Metal Catalysts for Oxidation of Lignocellulose and Reduction of Carbon Dioxide Susan K. Hanson 20110581PRD2 Abstract desired reaction) is required to produce fuels or other With demand increasing for fossil fuels and concern industrial chemicals from lignocellulose or carbon diox- growing over carbon dioxide emissions, the develop- ide, but the current catalytic technologies are lacking. ment of alternative energy sources is a national security More effective and selective catalysts are needed to fully priority. Lignocellulose or carbon dioxide might function utilize these renewable feedstocks. This project investi- as a source of renewable feedstocks for a range of com- gated the synthesis of new catalysts for the conversion modities; however, technical challenges remain in the of lignocellulose and carbon dioxide to useful chemicals catalytic conversion of lignocellulose or carbon dioxide and fuels. for use in biofuel and chemical applications. This proj- Disassembly of lignin ect has focused on replacing traditional catalysts based Non food-derived biomass (lignocellulose) is attracting on precious metals with alternative catalysts based attention as a potential renewable feedstock for produc- on inexpensive, earth-abundant metals. Vanadium- tion of chemicals and fuels. Efficient transformation of catalyzed aerobic oxidation has been demonstrated as lignocellulose into more valuable products remains a an approach to break down lignin model compounds, major challenge, particularly due to the difficulties asso- suggesting the potential of vanadium catalysts for con- ciated with breakdown of lignin. Although some promis- verting lignin into more valuable chemicals. A new class ing selective conversions of cellulose to useful materials of cobalt catalysts has been developed for hydrogena- have recently been reported, less progress has been tion and dehydrogenation reactions. The cobalt cata- made with lignin, which is typically incinerated in the lysts were shown to reduce several different functional paper and forest industry. New methods to depolymer- groups, including C=C, C=O, and C=N bonds, under mild ize lignin would represent a major breakthrough towards conditions, providing an inexpensive alternative to the the production of bio-derived chemicals and fuels. precious metal catalysts typically used for these reac- tions. The development of inexpensive, earth-abundant The strategy explored in this project was to use an earth- catalysts for both of these processes is an important abundant transition metal (vanadium) catalyst to break technical advance in our ability to more efficiently use apart lignin. This approach could lead to selective meth- renewable resources. ods to transform lignin into more useful chemicals and fuels. We selected vanadium for investigation because in Background and Research Objectives oxidation reactions, vanadium complexes are known to From 2007 to 2035, global energy consumption is break carbon-carbon and carbon-hydrogen bonds. projected to increase by 49%. Demand for fossil fuels, fuel costs, carbon dioxide emissions, and concerns over New vanadium(V) catalysts were synthesized, charac- global warming are on the rise. The development of terized, and evaluated for the oxidation of lignin model alternative energy sources is thus becoming an increas- compounds. Specifically, vanadium(V) complexes of the ingly urgent priority for national security. Non-food de- tridentate bis(phenolate)pyridine ligand H2BPP (H2BPP = rived biomass (lignocellulose) and carbon dioxide are the 2,6-(HOC6H2-2,4-tBu2)2NC5H3) and the bis(phenolate) only renewable carbon feedstocks that could be used to amine ligand H2BPA (H2BPA = N,N-bis(2-hydroxy-4,5-di- replace petroleum as a source of carbon-based fuels and methyl¬benzyl)¬propyl-amine) were prepared (Figure 1). chemicals. However, converting either lignocellulose or Complexes (BPP)VV(O)(OiPr) and (BPA)VV(O)(OiPr) were carbon dioxide to useful chemicals and fuels is a ma- evaluated as catalysts for the aerobic oxidation of 4-me- jor challenge. A catalyst (a chemical which enables the 261

Page 262

thoxybenzyl alcohol and arylglycerol β−aryl ether lignin tional group tolerance of more commonly used precious model compounds. The catalytic activities and selectivities metal catalysts. In this project, iron and cobalt complexes of the bis(phenolate) complexes were compared to other were investigated as catalysts for the hydrogenation of related catalysts. Decomposition of the vanadium catalyst organic substrates. Specifically, a cobalt-based catalytic sys- resulting from oxidation of the bis(phenol)amine ligand tem for the hydrogenation of alkenes, aldehydes, ketones, was observed during the attempted aerobic oxidation of and imines was developed (Figure 2). The cobalt system is lignin model compounds. However, the vanadium complex remarkably active, operating under mild conditions (room with the bis(phenolate)pyridine ligand did catalyze the temperature or 60 oC). The cobalt catalyst is effective for aerobic oxidation of lignin model compounds. Overall, in a wide range of substrates and tolerates alcohol, amine, considering a complete catalytic cycle, the evidence sug- and carboxylic acid functionalities and water. Mechanistic gests that while the more electron rich bis(phenolate) li- experiments indicated that the active catalyst is a cobalt(II) gand set may accelerate the reaction of vanadium(IV) with hydride complex. These findings have major ramifications air, the opposite trend is expected for the alcohol oxidation for future catalyst development, suggesting the potential step, which would be promoted by more electron deficient of cobalt(II) hydride complexes, which have been largely ligands. Designing more active vanadium catalysts for aero- neglected, to serve as highly active catalysts. A paper bic oxidation reactions will require balancing the electronic describing these findings was featured on the cover of the requirements of both steps. A paper describing these find- chemistry journal Angewandte Chemie International Edi- ings has been published. Another goal of the project was tion (Figure 3). to develop a water-soluble vanadium complex and test its The new cobalt complexes were also evaluated as catalysts ability to catalyze aerobic oxidation in aqueous solution. for alcohol dehydrogenation. Alcohol dehydrogenation has Several different approaches were attempted in order to applications in energy technologies, like hydrogen storage prepare a water-soluble vanadium analogue, and one such and production, as a selective and low-temperature route catalyst bearing 8-hydroxyquinoline-5-sulfonate ligands to generate hydrogen from biomass-derived alcohols and was prepared. This complex was tested for the oxidation carbohydrates. However, previous examples of homoge- of lignin model compounds, and a number of other sub- neous catalysts capable of the acceptorless dehydrogena- strates. Our results indicate that certain benzylic alcohols tion of alcohols have been limited exclusively to precious can be oxidized by air in water, using the vanadium cata- metals (Ru, Rh, Ir, Os). In this project, the first example of lyst. Overall, these results are a promising indication that a non-precious metal (cobalt) catalyst for the acceptorless it may be possible to develop an environmentally friendly dehydrogenation of alcohols was developed. The cobalt method to break down lignin, using water as a solvent, air catalyst displays comparable activity to previously reported as the oxidant, and a vanadium catalyst. ruthenium catalysts, is active for a wide range of aliphatic Developing hydrogenation catalysts based on earth- and aromatic alcohols, and generates the carbonyl prod- abundant metals ucts in high yields, affording hydrogen as the only reaction Non-food biomass and carbon dioxide are potential renew- by-product. Deuterium labeling studies were used to con- able feedstocks, but both contain more oxygen-based firm that the alcohol dehydrogenation proceeds through a functional groups than traditional petroleum-derived cobalt-hydride intermediate. feedstocks. To use biomass or carbon dioxide as precursors The new cobalt catalysts for hydrogenation and dehydroge- for chemicals or fuels, removal of the oxygen functionality nation were an important step forward in the development is needed. Although considerable progress has been made of cost-effective, viable routes to use alternative carbon in the catalytic reduction of carbon dioxide and biomass sources like biomass or carbon dioxide. However, the feedstocks, many of the best current processes employ mechanisms of these reactions were not known, and it was precious metal hydrogenation catalysts (e.g. platinum, clear that a better fundamental understanding of the el- ruthenium, rhodium, or iridium). Replacing these precious ementary steps involved would be instrumental for rational metal catalysts with earth abundant metals (cobalt or iron) catalyst design and optimization. Additional experimental would be a key breakthrough in terms of developing more studies were carried out that explored the mechanisms of cost-effective and sustainable processes. the cobalt-catalyzed hydrogenation and dehydrogenation This project focused on the design of earth-abundant reactions. Investigations into the role of metal-ligand co- metal catalysts for hydrogenation based on iron and co- operativity in the reactions, as well as the identification of balt. Prior to this work, few examples of homogeneous Fe, potential catalytic intermediates and resting states (includ- Ni, or Co hydrogenation catalysts were known, and these ing an unusual stable cobalt(III) aryl hydride product of C-H lacked the high activity, broad substrate scope, and func- bond activation) provided new insights into the catalytic 262

Page 263

cycles. Notably, our studies suggested that the hydrogena- tion of alkenes and the dehydrogenation of alcohols pro- ceeded by different mechanisms, and that both catalytic cycles based on cobalt(II) and cobalt(I)/(III) oxidation states are viable. This part of work has been published as a full article in the Journal of American Chemical Society. Over- all, the design of effective earth-abundant metal catalysts for the hydrogenation of C=C and C=O bonds is a key step Figure 2. Cobalt catalyst developed for the hydrogenation of C=C, forward in our ability to more efficiently and cost-effective- C=O, and C=N bonds. ly utilize renewable resources like lignocellulose or carbon dioxide. Impact on National Missions In a recent 2013 address to the nation, President Obama stated that “the only way for America’s energy supply to be truly secure is by permanently reducing our dependence on oil…We have to discover and produce cleaner, renew- able sources of energy with less of the carbon pollution that threatens our climate. And we have to do it quickly.” In order to effectively use renewable sources of carbon (non-food biomass or carbon dioxide) as fuels, new cata- lytic technologies are needed. In demonstrating that it is possible to use inexpensive, earth-abundant metals like vanadium and cobalt as catalysts for these applications, this work has provided an important conceptual and tech- nological advance. The work is relevant to Office of Basic Energy Sciences mission priorities. DoD relevance is driven by the fact the DoD is aggressively seeking new energy sources for its deployed operations. (a) (b) Thesubstitutionofcobaltforpreciousmetalsincatalysisisavariationonthe ancientalchemicalthemeoftransmutingbasemetalsintopreciousones.Intheir Communication(DOI:10.1002/anie.201206051),S.K.Hansonetal.describethe cobalt-catalyzedhydrogenationofalkenes,aldehydes,ketones,andiminesunder mildreactionconditions.Theprecatalystisacationiccobalt(II)alkylcomplex. (TheartworkisbyJoshSmith,LosAlamosNationalLaboratory.) Figure 3. Cobalt catalysis featured on the cover of Angew. Chem. Int. Ed. Publications Zhang, G., B. L. Scott, R. Wu, L. A. Silks, and S. K. Hanson. Figure 1. Structures of bis(phenolate)pyridine vanadium complex Aerobic oxidation reactions catalyzed by vanadium (a, left) and bis(phenolate)amine vanadium complex (b, right). complexes of bis(phenolate) ligands. 2012. Inorganic Chemistry. 51 (13): 7354. Zhang, G., B. L. Scott, and S. K. Hanson. Mild and homoge- neous cobalt-catalyzed hydrogenation of C=C, C=O, and C=N bonds. 2012. Angewandte Chemie International Edition. 51: 12102. Zhang, G., K. V. Vasudevan, B. L. Scott, and S. K. Hanson. 263

Page 264

Understanding the mechanisms of cobalt-catalyzed hydrogenation and dehydrogenation reactions. 2013. Journal of the American Chemical Society. 135: 8668. Zhang, G., and S. K. Hanson. Cobalt-catalyzed acceptorless alcohol dehydrogenation: synthesis of imines from al- cohols and amines. 2013. Organic Letters. 15: 650. 264

Page 265

Chemistry and Material Sciences Postdoctoral Research and Development Final Report Controlled Synthesis and SERS Responses of Metal Nanostructures with Complex Morphologies Hsing-Lin Wang 20110759PRD4 Abstract to enhance the Raman signal of absorbed molecules. In this project, we aimed at synthesis of metal nanopar- This technique has been widely explored as a powerful ticles (1D) and hybrid nanocomposites (2D) with strong spectroscopic probe that can provide non–destructive plasmonic properties through an extremely facile syn- and ultra–sensitive detection of chemical and biologi- thetic platform that is cost effective and time efficient. cal molecules up to single molecule levels. Most recent This synthetic platform allowed us to fabricate metal activities have focused on the SERS properties of metal nanoparticles with a wide range of sizes, structures, and nanoparticle dimers since the space between the dimers morphologies. These metal nanoparticles have novel can provide large electromagnetic field enhancements. optical, sensing, and catalytic properties, resulting from However, the focus of nanoscience and nanotechnology the complex particle morphology, which led to surface, has gradually shifted from the synthesis of individual enhanced Raman scattering. Results obtained from this components to their assembly into larger systems with study allowed us to establish correlations between the well–defined structures. Assemblies of nanoscale build- metal structures and the optical/catalytic properties. ing blocks with complex structures usually presents The surface enhanced Raman scattering (SERS) tech- collective optical, electronic, catalytic, and magnetic nique utilizes surface plasmons generated by the greatly properties that are distinctly different from the corre- enhanced electromagnetic field at the rough metal sponding individual nanoparticles. Self–assemblies of surfaces. SERS spectroscopy has been recognized as a metal nanoparticles based on rational control of non-co- powerful tool for chemical analysis and sensing appli- valent interactions provide powerful tools for the design cations due to its high sensitivity and resolution. [1, 2] of hierarchical structures at length scales from nano– to The development of the SERS technique has enabled a micrometer size. Therefore, the large–scale controllable series of applications that was previously not accessible. synthesis of metal structures with specific morphologies Another objective of this project was to develop highly is expected to lead to a greater understanding/validation efficient SERS substrates based on conducting polymer of underpinning SERS mechanisms involved in complex mediated electrodeless deposition of metal nanoparti- assemblies with enhanced optical and electronic proper- cles [3] (Figure 1). These substrates were extremely sen- ties. sitive and showed promise toward detection of several The goal of this project was to demonstrate controlled chemicals related to food safety and explosive detection. synthesis of self-assembled metal structures with spe- Moreover, the confinement of SERS on a single metal cific complex morphologies through a facile and novel particle has shown promise in better understanding technique trough which metal nanoparticles with a the catalytic reactivity and demonstrated their use as a wide range of structures, sizes, and morphologies can heterogeneous catalyst. We also show examples of plas- be synthesized. We proposed to study the SERS re- mon assisted catalytic reaction on our SERS substrate sponse of these metal nanoparticles for sensitive de- toward better understanding the reaction mechanisms tection of chemical and biological analytes. We aimed induced by laser excitation. to understand the underpinning mechanism of how metal nanoparticles self-assemble into larger, complex Background and Research Objectives systems with well–defined structures (Figure 2). We Surface–enhanced Raman scattering (SERS) utilizes the synthesized these metal structures using reducing and greatly enhanced electromagnetic fields generated by directing agents without polymer surfactant or capping the excitation of surface plasmons at metal surfaces agents. Structure–property relationships and tunable 265

Page 266

photonic responses resulting from the directed assembly wavelength-dependent kinetics has been demonstrated of nanoparticles were interpreted. by using three laser excitation wavelengths. The reactiv- ity under 633 nm proceeded much slower than that under Another important goal of this project was to demonstrate 532 nm or 514.5 nm excitation. Further decrease of the catalytic properties of these SERS substrates for various laser energy turned off the reaction completely. Our re- chemical reactions. Using surface plasmon to assist chemi- sults revealed consistency with previous observation as we cal reactions has tremendous implications for energy ap- use a relatively low power (0.05 mW) of 633 nm excitation plications. SERS particles irradiated under laser light result laser, conversion of p-NTP to DMAB proceeded extremely in the formation of “hot electrons” which can then reduce slow as compared to that using the high laser power (0.2 the absorbates immobilized on the metal particle surface. mW). We used p-nitrothiophenyl (p-NTP) as model molecules to show dimerization of p-NTP to form azobenzene molecules The SEM micrograph of a single micron sized particles on SERS substrate. We have shown for the first time, the comprised of numerous nanosheets and the SERS spectra detailed reaction mechanisms of this dimerization reac- of p-NTP as a function of time are shown in Figure 4. The tion, revealing the catalytic nature of our SERS substrates, surface Plasmon intensity decrease of the ν(NO2) band at which can be applied to reactions with significant energy 1335 cm-1, and increase of the ν(N=N) bands at 1380 and implications. 1440 cm-1 can be distinguished over a time scale of several hours. The diminished NO2 peak and concomitant increase Scientific Approach and Accomplishments of the N=N band absorption provided unequivocal evi- Synthesis and characterization dence of the chemical reaction catalyzed by surface plas- These as–synthesized metal structures have been analyzed mons. The control experiment carried out under the same with electron microscopy, transmission electron micros- conditions without the usage of metal particles showed no copy, and X–ray diffraction. Mapping and imaging using formation of N=N band frequency. Such drastic change in SERS spectra of chemical and biological molecules on the reaction clearly associates the reaction itself with the the metal structures have been measured with a Raman surface plasmon as a catalytic substrate. The modification microscope apparatus. In order to conduct a systematic of the plasmonic properties of the metal particles through study of how small molecules impact the self–assembly variation of the particle morphology was expected to have of metal nanoparticles into larger systems, small inorganic a huge impact on their catalytic properties. This work has and organic acids, amino acids, and inorganic salts have been published in Chemical Communications as featured been employed in the synthesis of the metal structures to front cover article.[5] develop control over the size and morphology of the metal particles. SERS comparisons, polarization–dependent SERS Impact on National Missions responses and plasmonic properties of the metal struc- Our project aimed at developing a facile and novel small tures with different morphologies were also performed. molecule–directed synthesis of metal nanoparticles and their self-assemblies with well–defined structures. These Chemical sensing studies could open up a new avenue for fabricating com- Through the use of a wide range of metal nanoparticles plex assembled architectures with emergent properties we synthesized, we have shown applications of these that were previously inaccessible. The spontaneous orga- nanoparticles as well as nanocomposites toward detec- nization of nanoparticles into larger systems is of scientific tion of chemicals. A clear example is the detection of importance as understanding the assembly process on the melamine down to 2 ppm sensitivity (Figure 3). As of nanometer scale is of great interest for bottom–up fabri- today, our SERS substrate is the only substrate capable of cation of functional devices. The advantages of fabricat- detecting melamine with sensitivity reaching 2 ppm. Such ing self–assembled metal complex structures is clear: (1) results have potentially huge impacts on food safety as Superior SERS responses and morphological control over our technology represents a new technique of inspecting conventional approach via manipulating the organization toxins in dairy products and the method is fast and cheap. of small nanoparticles. The emergent, superior optical and This work was published as a featured front cover article in catalytic properties can lead to significant reduction of the Journal of Materials Chemistry.[4] energy use for chemical reactions. Moreover, the cata- lytic properties also show promise in biofuel and biomass Catalytic applications production with much lower cost and faster reaction time. We have demonstrated catalytic reactions driven by a (2) A greater understanding of the self–assembly mecha- single particle on which the kinetics and products can be nism that will allow optimization of the SERS response with monitored by SERS spectra. The conversion of p-nitrothio- sensitivities that rival fluorescence at the single molecule phenyl (p-NTP) into dimerized azobenzene (DMAB) and the 266

Page 267

detection level, which will be tremendously important for detection of explosive and toxins and will therefore con- tribute significantly to the area of threat reduction. Figure 1. Schematic representation of a SERS substrate comprised of numerous nanosheet assembly on top of a conducting polymer layer. Figure 4. (top) Laser induced dimerization reaction, converting p-nitrothiophenyl (p-NTP) into dimerized azobenzene (DMAB), and (bottom) Time-dependent SERS spectra of p-nitrothiophenol (pNTP) under continuous exposure to the 633 nm laser. The spec- tra were collected for a single Ag particle, with an integration time of 2 s, and a laser power of 2 mW. Figure 2. Schematic illustration of the synthesis of metal particles with well–defined morphologies aided by reducing and direct- References ing agents. Variations of the molar ratio between reducing and

1. Lantman, E. M. van Schrojenstein, T. Deckert-Gaudig, directing agents, and the timing of adding directing agent to the A. J. G. Mank, V. Deckert, and B. M. Weckhuysen. solution can result in metal particles with four distinct morpholo- Catalytic processes monitored at the nanoscale with gies. tip-enhanced Raman spectroscopy. 2012. NATURE NANOTECHNOLOGY. 7 (9): 583.
2. Casadio, F., M. Leona, J. R. Lombardi, and R. Van Duyne. Identification of Organic Colorants in Fibers, Paints, and Glazes by Surface Enhanced Raman Spec- troscopy. 2010. ACCOUNTS OF CHEMICAL RESEARCH. 43 (6): 782.
3. Wang, H. L., W. G. Li, Q. X. Jia, and E. Akhadov. Tailoring conducting polymer chemistry for the chemical deposi- tion of metal particles and clusters. 2007. CHEMISTRY OF MATERIALS. 19 (3): 520.
4. Zhang, B., P. Xu, X. M. Xie, H. Wei, Z. P. Li, N. H. Mack, Figure 3. SERS spectra of melamine in different concentrations X. J. Han, H. X. Xu, and H. L. Wang. Acid-directed syn- taken on core–shell Ag wire structures: (a) 200 ppm, (b) 100 thesis of SERS-active hierarchical assemblies of silver ppm, (c) 50 ppm, (d) 20 ppm, (e) 10 ppm, and (f) 5 ppm. Mea- nanostructures. 2011. JOURNAL OF MATERIALS CHEM- surements were conducted with a 10 s exposure and 1 mW laser ISTRY. 21 (8): 2495. power. Inset shows the relationship between Raman intensity of the 679 cm-1 band and concentration of melamine solution. 5. Kang, L. L., P. Xu, B. Zhang, H. H. Tsai, X. J. Han, and H. L. Wang. Laser wavelength- and power-dependent plasmon-driven chemical reactions monitored using single particle surface enhanced Raman spectroscopy.
5. CHEMICAL COMMUNICATIONS. 49 (33): 3389 267

Page 268

Publications face-Enhanced Raman Spectroscopy. 2013. JOURNAL He, J. J., X. J. Han, J. Yan, L. L. Kang, B. Zhang, Y. C. Du, C. K. OF PHYSICAL CHEMISTRY C. 117 (19): 10007. Dong, H. L. Wang, and P. Xu. Fast fabrication of homo- Li, Q., P. Xu, B. Zhang, G. Wu, H. T. Zhao, E. G. Fu, and H. L. geneous silver nanostructures on hydrazine treated Wang. Self-supported Pt nanoclusters via galvanic re- polyaniline films for SERS applications (vol 14, pg 4952, placement from Cu2O nanocubes as efficient electro- 2012). 2012. CRYSTENGCOMM. 14 (24): 8737. catalysts. 2013. NANOSCALE. 5 (16): 7397. He, J. J., X. J. Han, J. Yan, L. L. Kang, B. Zhang, Y. C. Du, C. K. Li, Q., P. Xu, B. Zhang, H. Tsai, S. J. Zheng, G. Wu, and H. L. Dong, H. L. Wang, and P. Xu. Fast fabrication of homo- Wang. Structure-Dependent Electrocatalytic Properties geneous silver nanostructures on hydrazine treated of Cu2O Nanocrystals for Oxygen Reduction Reaction. polyaniline films for SERS applications. 2012. CRYS- 2013. JOURNAL OF PHYSICAL CHEMISTRY C. 117 (27): TENGCOMM. 14 (15): 4952. 13872. He, J., X. Han, J. Yan, L. Kang, B. Zhang, Y. Du, C. Dong, H. Li, S. W., P. Xu, Z. Q. Ren, B. Zhang, Y. C. Du, X. J. Han, N. Wang, and P. Xu. Fast fabrication of homogeneous H. Mack, and H. L. Wang. Fabrication of Thorny Au silver nanostructures on hydrazine treated polyaniline Nanostructures on Polyaniline Surfaces for Sensitive films for SERS applications. 2012. CRYSTENGCOMM- Surface-Enhanced Raman Spectroscopy. 2013. ACS AP- CRYSTENGCOMM. 14 (15): 4952. PLIED MATERIALS & INTERFACES. 5 (1): 49. Huang, Y. F., H. S. Shih, C. W. Lin, P. Xu, D. J. Williams, K. J. Mack, N. H., J. A. Bailey, S. K. Doorn, C. A. Chen, H. M. Gau, Ramos, D. E. Hooks, and H. L. Wang. Morphology Con- P. Xu, D. J. Williams, E. A. Akhadov, and H. L. Wang. trol of Cu Crystals on Modified Conjugated Polymer Mechanistic Study of Silver Nanoparticle Formation on Surfaces. 2012. CRYSTAL GROWTH & DESIGN. 12 (4): Conducting Polymer Surfaces. 2011. LANGMUIR. 27 1778. (8): 4979. Huang, Y. F., Y. I. Park, C. Kuo, P. Xu, D. J. Williams, J. Wang, Park, Y. S., Y. F. Chen, Y. Ghosh, A. Pirytinski, P. Xu, N. H. C. W. Lin, and H. L. Wang. Low-Temperature Synthesis Mack, H. L. Wang, V. I. Klimov, J. A. Hollingsworth, and of Au/Polyaniline Nanocomposites: Toward Controlled H. Htoon. Strong Photon Bunching in Individual Nano- Size, Morphology, and Size Dispersity. 2012. JOURNAL crystal Quantum Dots Coupled to Rough Silver Film. OF PHYSICAL CHEMISTRY C. 116 (20): 11272. 2012. In 2012 CONFERENCE ON LASERS AND ELECTRO- Huang, Y., H. Shih, C. Lin, P. Xu, D. J. Williams, K. J. Ramos, OPTICS (CLEO). , p. -. D. E. Hooks, and H. Wang. Morphology Control of Cu Park, Y. S., Y. Ghosh, P. Xu, N. H. Mack, H. L. Wang, J. A. Hol- Crystals on Modified Conjugated Polymer Surfaces. lingsworth, and H. Htoon. Single-Nanocrystal Photolu- 2012. CRYSTAL GROWTH & DESIGNCRYSTAL GROWTH minescence Spectroscopy Studies of Plasmon-Multiex- & DESIGN. 12 (4): 1778. citon Interactions at Low Temperature. 2013. JOURNAL Huang, Y., Y. I. Park, C. Kuo, P. Xu, D. J. Williams, J. Wang, C. OF PHYSICAL CHEMISTRY LETTERS. 4 (9): 1465. Lin, and H. Wang. Low-Temperature Synthesis of Au/ Park, Y. S., Y. Ghosh, Y. Chen, A. Piryatinski, P. Xu, N. H. Polyaniline Nanocomposites: Toward Controlled Size, Mack, H. L. Wang, V. I. Klimov, J. A. Hollingsworth, and Morphology, and Size Dispersity. 2012. JOURNAL OF H. Htoon. Super-Poissonian Statistics of Photon Emis- PHYSICAL CHEMISTRY CJOURNAL OF PHYSICAL CHEM- sion from Single CdSe-CdS Core-Shell Nanocrystals ISTRY C. 116 (20): 11272. Coupled to Metal Nanostructures. 2013. PHYSICAL RE- Jeon, S. H., P. Xu, B. Zhang, N. H. Mack, H. Tsai, L. Y. Chiang, VIEW LETTERS. 110 (11): -. and H. L. Wang. Polymer-assisted preparation of metal Wu, G., K. L. More, P. Xu, H. L. Wang, M. Ferrandon, A. nanoparticles with controlled size and morphology. J. Kropf, D. J. Myers, S. G. Ma, C. M. Johnston, and P. 2011. JOURNAL OF MATERIALS CHEMISTRY. 21 (8): Zelenay. A carbon-nanotube-supported graphene-rich 2550. non-precious metal oxygen reduction catalyst with en- Kang, L. L., P. Xu, B. Zhang, H. H. Tsai, X. J. Han, and H. L. hanced performance durability. 2013. CHEMICAL COM- Wang. Laser wavelength- and power-dependent plas- MUNICATIONS. 49 (32): 3291. mon-driven chemical reactions monitored using single Xu, P., E. Akhadov, L. Wang, and H. Wang. Sequential chem- particle surface enhanced Raman spectroscopy. 2013. ical deposition of metal alloy jellyfish using polyaniline: CHEMICAL COMMUNICATIONS. 49 (33): 3389. redox chemistry at the metal-polymer interface. 2011. Kang, L. L., P. Xu, D. T. Chen, B. Zhang, Y. C. Du, X. J. Han, Q. CHEMICAL COMMUNICATIONSCHEMICAL COMMUNI- Li, and H. L. Wang. Amino Acid-Assisted Synthesis of CATIONS. 47 (38): 10764. Hierarchical Silver Microspheres for Single Particle Sur- 268

Page 269

Xu, P., E. Akhadov, L. Y. Wang, and H. L. Wang. Sequential chemical deposition of metal alloy jellyfish using poly- aniline: redox chemistry at the metal-polymer inter- face. 2011. CHEMICAL COMMUNICATIONS. 47 (38): 10764. Xu, P., K. Chang, Y. I. Park, B. Zhang, L. L. Kang, Y. C. Du, R. S. Iyer, and H. L. Wang. Conjugated polymer medi- ated synthesis of nanoparticle clusters and core/shell nanoparticles. 2013. POLYMER. 54 (2): 485. Yan, J., X. Han, J. He, L. Kang, B. Zhang, Y. Du, H. Zhao, C. Dong, H. Wang, and P. Xu. Highly sensitive surface- enhanced Raman spectroscopy (SERS) platforms based on silver nanostructures fabricated on polyaniline membrane surfaces. 2012. ACS Applied Materials and InterfacesACS Applied Materials and Interfaces. 4 (5): 2752. Yan, J., X. Han, J. He, L. Kang, B. Zhang, Y. Du, H. Zhao, C. Dong, H. Wang, and P. Xu. Highly Sensitive Surface- Enhanced Raman Spectroscopy (SERS) Platforms Based on Silver Nanostructures Fabricated on Polyaniline Membrane Surfaces. 2012. ACS APPLIED MATERIALS & INTERFACESACS APPLIED MATERIALS & INTERFACES. 4 (5): 2752. Yan, J., X. J. Han, J. J. He, L. L. Kang, B. Zhang, Y. C. Du, H. T. Zhao, C. K. Dong, H. L. Wang, and P. Xu. Highly Sensi- tive Surface-Enhanced Raman Spectroscopy (SERS) Platforms Based on Silver Nanostructures Fabricated on Polyaniline Membrane Surfaces. 2012. ACS APPLIED MATERIALS & INTERFACES. 4 (5): 2752. Zhang, B., B. T. Zhao, S. H. Huang, R. Y. Zhang, P. Xu, and H. L. Wang. One-pot interfacial synthesis of Au nanopar- ticles and Au-polyaniline nanocomposites for catalytic applications (vol 14, pg 1542, 2012). 2012. CRYSTENG- COMM. 14 (24): 8735. Zhang, B., B. T. Zhao, S. H. Huang, R. Y. Zhang, P. Xu, and H. L. Wang. One-pot interfacial synthesis of Au nanopar- ticles and Au-polyaniline nanocomposites for catalytic applications. 2012. CRYSTENGCOMM. 14 (5): 1542. Zhang, B., B. Zhao, S. Huang, R. Zhang, P. Xu, and H. Wang. One-pot interfacial synthesis of Au nanoparticles and Au-polyaniline nanocomposites for catalytic applica- tions. 2012. CRYSTENGCOMMCRYSTENGCOMM. 14 (5): 1542. Zhang, B., P. Xu, X. M. Xie, H. Wei, Z. P. Li, N. H. Mack, X. J. Han, H. X. Xu, and H. L. Wang. Acid-directed synthesis of SERS-active hierarchical assemblies of silver nano- structures. 2011. JOURNAL OF MATERIALS CHEMISTRY. 21 (8): 2495. 269

Page 270

Chemistry and Material Sciences Postdoctoral Research and Development Final Report Developing a Materials Strategy for d-electron Heavy Fermion Systems Filip Ronning 20120749PRD2 Abstract Scientific Approach and Accomplishments The physics of strongly correlated d-electron systems is Ni Ni discovered a new family of superconductors based generally believed to be fundamentally the same as their on Fe (the so-called 10-3-8 family). She grew single f-electron counterparts with simply larger energy scales crystals, which allowed for an accurate determination due to the more extended d-orbital wave functions and of the temperature-composition phase diagram [2,3]. thus larger hybridization. Indeed, for either a single d or In addition, nuclear magnetic resonance measurements f element localized in a metallic host the Kondo effect were performed, which demonstrated the presence of (the screening of magnetic moments by conduction magnetism in these materials – a likely ingredient for electrons) will occur. When a lattice of f-elements are the mechanism of superconductivity [1]. In addition, Ni embedded in a local host, the Kondo effect occurs in a Ni doped iron and cobalt into non-magnetic analogs of collective fashion giving rise to electronic excitations that another member of the Fe-based superconductors with have masses up to 1,000 times that of a free electron. the chemical formula SrNi2As2. Transport and thermo- However, while a lattice of f-electrons embedded in a dynamic measurements revealed a Kondo effect oc- metallic host gives rise to “heavy” electrons, analogous curring with iron doping, as was our goal, but not with behavior has not been realized in d-electron intermetal- cobalt. This suggests an additional connection between lics although very strong Coulomb interactions due to the Fe-based superconductors, and the f-electron based proximity to insulating states leads to similar phenom- heavy electron superconductors, which had not been ena in transition metal (3d-electron) oxides. By doping previously considered, and was one of our objectives. iron and cobalt into a non-magnetic host material (Sr- Finally, Ni Ni reinvestigated the uranium based heavy Ni2As2) we demonstrated the single ion Kondo effect, a electron superconductor U2PtC2. We found evidence for necessary precondition for heavy fermion behavior. Fur- a nearby magnetic state when the system is alloyed with thermore, by examining the Fe rich analog in the 10-3-8 rhodium. This may play an important role for elucidating compound we found unconventional superconductivity. the superconducting mechanism in this material. This work sheds new light on the formation of strong electronic correlations and the emergent phenomena Impact on National Missions that result from it. One of the three pillars of the LANL materials strategy is focused on developing the understanding required to Background and Research Objectives achieve controlled functionality of strongly correlated We are internationally recognized for developing an systems that will produce emergent phenomena. We understanding of the emergent phenomena such as achieved partial understanding by recreating the chemi- superconductivity that occurs in strongly correlated (f-) cal conditions to create heavy electron physics (well electron systems. We are now in a position to move be- known in f-electron materials) in d-electron materials. yond observation to control of these properties. In order This work has potential benefit to nuclear deterrence, to demonstrate control we must be able to recreate the global threat reduction, and most immediately, energy chemical conditions for d-electron systems to mimic the security through the potential development of rare behavior of f-electron systems. This was the challenge earth replacement magnets and novel superconducting and goal of this project. We aimed to find new forms of materials. matter (including superconductivity) in d-electron based systems, and to gain a better understanding of how the Kondo effect works in a lattice. 270

Page 271

References

1. Zhou, T., G. Koutroulakis, J. Lodico, N. Ni, J. D. Thomp- son, R. J. Cava, and S. E. Brown. Antiferromagnetic order in Ca10(Pt3As8)(Fe2As2)5 observed by 75As NMR. 2013. Journal of Physics: Condesned Matter. 25:
3. Ni, N., W. E. Straszheim, D. J. Williams, M. A. Tanatar, R. Prozorov, E. D. Bauer, F. Ronning, J. D. Thompson, and R. J. Cava. Temperature-concentration phase diagram of (Ca1-xLax)10(Pt3As8)(Fe2As2)5 superconductors.
4. Physical Review B. 87: 060507.
5. Gao, P., L. Sun, N. Ni, J. Guo, Q. Wu, C. Zhang, D. Gu, K. Yang, S. Jiang, R. J. Cava, and Z. Zhao. Scaling of pressure-induced and doping-induced superconductiv- ity in the Ca10(PtnAs8)(Fe2As2)5 arsenides. Nature Communications. Publications Gao, P., L. Sun, N. Ni, J. Guo, Q. Wu, C. Zhang, D. Gu, K. Yang, S. Jiang, R. J. Cava, and Z. Zhao. Scaling of pres- sure-induced and doping-induced superconductivity in the Ca10(PtnAs8)(Fe2As2)5 arsenides. Nature Com- munications. Ni, N., W. E. Straszheim, D. J. Williams, M. A. Tanatar, R. Prozorov, E. D. Bauer, F. Ronning, J. D. Thompson, and R. J. Cava. Temperature-concentration phase diagram of (Ca1-xLax)10(Pt3As8)(Fe2As2)5 superconductors.
6. Physical Review B. 87: 060507. Zhou, T., G. Koutroulakis, J. Lodico, N. Ni, J. D. Thompson, R. J. Cava, and S. E. Brown. Antiferromagnetic order in Ca10(Pt3As8)(Fe2As2)5 observed by 75As NMR. 2013. Journal of Physics: Condensed Matter. 25: 122201. 271

Page 272

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project Multiferroic Magnetoelectric Hybrid Inorganic-Organic Frameworks Michael R. Fitzsimmons 20110716PRD2 Introduction cation achieved by changing frequencies used to broad- Metal-organic-frameworks are crystalline compounds cast information in real-time, and the detection of steel consisting of metal atoms or clusters coordinated to moving under the ocean’s surface. organic ligands with extended structure in at least one direction. A unique potential of metal-organic-frame- Benefit to National Security Missions works is to exhibit combinations of properties that might Our work will yield compounds called multiferroics ex- not be possible in purely inorganic and organic systems. hibiting a strong magnetoelectric effect by tailoring the This potential is realized by tailoring the polymeric ligand metal-organic-framework structures. In a field domi- structures comprising the framework and the metal nated by materials that contain the toxic element lead, ion captured inside the framework that will allow us metal-organic-frameworks will open new opportunities to circumnavigate the mutually exclusive problem of for the production of lead-free (i.e., green) multifer- achieving ferromagnetism and ferroelectricity in a single roic compounds customized for specific technological inorganic compound. applications, including but not limited to data storage, photo-voltaic, sensors, and transducers. Devices using We will study structurally disordered metal-organic- multiferroic behavior will be faster, smaller and more frameworks which upon application of an external sensitive to their environment than present technol- stimulus, i.e. temperature or pressure, result in a polar- ogy. New applications include: secure communication ordered structure and ferroelectric behavior. Magnetic achieved by changing frequencies used to broadcast in- transition metals will be used to achieve simultaneous formation in real-time, and the detection of steel moving ferromagnetic ordering. Many known examples will also under the ocean’s surface. be studied and optimized. For example, we will study potential multiferroic behavior in [Co2(C3H3N2)4]n (zni In addition, Prashant’s work in material science will phase). We will investigate how the lattice distortions, enhance the reputation of the Lujan Center in the area dopant metals, magnetic and electric field, and local of the application of neutron scattering techniques in bonding environments affect multiferroic behavior and research programs. The Lujan Center is funded as a na- coupling between two ferroic orders in pressure induced tional user facility for neutron scattering. amorphous and crystalline phases of this material. Further, chiral ligands will be used to create novel polar Progress materials. The key accomplishment of this project during the last 12 months would be the realization of giant coupling The proposed work will rely heavily on our expertise between magnetism and electric polarization, a much in metal-organic-framework synthesis, x-ray based sought after property necessary to satisfy ever growing characterization techniques, and electric and magnetic need for better data storage media and other functional transport measurements, and will make extensive use properties otherwise not possible, in metal-organic- of neutron scattering. Our work will yield compounds, frameworks (MOFs) or hybrid inorganic-organic materi- multiferroics, exhibiting a strong magnetoelectric effect als. by tailoring the metal-organic-framework structures. Devices using multiferroic behavior will be faster, smaller MOFs are crystalline materials consisting of metal atoms and more sensitive to their environment than present or clusters coordinated to organic ligands with extended technology. New applications include: secure communi- 272

Page 273

structure in at least one direction. A unique potential of derstand the spin states in technologically relevant hybrid MOFs is to exhibit combinations of properties that might multiferroic materials. not be possible in purely inorganic and organic systems. This potential is realized by tailoring the polymeric ligand In addition, work will also be carried out at the National structures comprising the framework and the metal ion High Magnetic Field Lab. Materials studied with PNR will captured inside the framework that allows us to circum- be studied with high magnetic field at the pulse field facil- navigate the mutually exclusive problem of achieving ferro- ity. We will do pulse-field measurements of electric polar- magnetism and ferroelectricity in a single inorganic com- ization to quantify the coupling between electric polariza- pound. We accomplished this in (CH3)2CH2Mn(HCOO)3 tion and magnetism. a manganese formate framework in which Mn atoms are bridged together by the most basic carboxylate ligand Conclusion formic acid. This framework contains disordered organic Our work will yield compounds, multiferroics, exhibiting a cations in the pores. It is the ordering of these cations that strong magnetoelectric effect by tailoring the metal-organ- leads to net electric polarization in this material below ic-framework structures. In a field dominated by materials 180K. Due to the paramagnetic nature of the materials, the that contain the toxic element lead, metal-organic-frame- magnitude of electric polarization can be enhanced by an works will open new opportunities for the production of applied magnetic field. This is an important breakthrough lead-free (i.e., green) multiferroic compounds custom- given that ferroelectricity and ferromagnetism are usually ized for specific technological applications, including but two mutually exclusive properties in a single phase mate- not limited to data storage, photo-voltaic, sensors, and rial. transducers. Devices using multiferroic behavior will be faster, smaller and more sensitive to their environment In addition to the scientists at the magnet lab (TA-35), we than present technology. New applications include: secure have also worked with theoretical division. Together we communication achieved by changing frequencies used to are developing novel MOF materials which would display broadcast information in real-time, and the detection of a strong magnetoelectric coupling via new mechanisms steel moving under the ocean’s surface. (PRL 108, 097202, 2012). This work is currently in progress where we are applying bottom-up assembly for the realiza- Publications tion of the structure required as described in the paper. Li, W., Z. Zhang, E. Bithell, A. Batsanov, P. Barton, P. Saines, P. Jain, C. Howard, M. Carpenter, and A. Cheetham. Future Work Ferroelasticity in a metal-organic framework For the next year, we will continue developing materials perovskite; towards a new class of multiferroics. 2013. with magnetoelectric coupling at close to room tempera- Acta Materialia. 61: 4928. ture. Part of this effort will be to understand the origin of Stroppa, A., P. Barone, P. Jain, J. M. Perez-Mato, and this coupling in mixed valence iron metal organic frame- S. Picozzi. Hybrid Improper Ferroelectricity in a work (MOF) that according to our preliminary results is Multiferroic and Magnetoelectric Metal-Organic magnetoelectric. Furthermore, there are no well-estab- Framework. 2013. Advanced Materials. 25: 2284. lished methods to quantify this coupling between mag- netism and electric polarization, primarily due to the lack Tan, J., P. Jain, and A. Cheetham. Influence of ligand of materials with this extraordinary behavior. As part of field stabilization energy on the elastic properties of this project, as we develop new materials with this effect, multiferroic MOFs with the perovskite architecture. we hope to quantify this effect and establish benchmarks 2012. Dalton Transactions. 41: 3949. similar to the existing ones in the field of ferroelectrics Thomson, R. I., P. Jain, A. K. Cheetham, and M. A. (e.g. electric polarization values of potassium dihydrogen Carpenter. Elastic relaxation behavior, magnetoelastic phosphate and urea). We will also attempt to make thin coupling, and order-disorder processes in multiferroic films of MOFs, which is an architecture required before metal-organic frameworks. 2012. Physical Review B. these materials could be used for data storage or sensor 86. applications. We will use polarized neutron reflectometry to characterize magnetism of the thin films. Wang, Z., P. Jain, K. Choi, J. van Tol, A. Cheetham, H. Kroto, H. Koo, H. Zhou, J. Hwang, E. S. Choi, M. Whangbo, and N. Dalal. Dimethylammonium copper formate (CH3) Prashant will also work with Michael Fitzsimmons on (2)NH2 Cu(HCOO)(3): A metal-organic framework polarized neutron reflectometry (PNR). This will be a new with quasi-one-dimensional antiferromagnetism and technique for Prashant, hence a growth opportunity, which magnetostriction. 2013. Physical Review B. 87. he will apply to his future research as an academic to un- 273

Page 274

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project Catalysts Containing Earth Abundant Metals for Energy Applications John C. Gordon 20110752PRD3 Introduction security. In particular, agencies such as OBES (ctalysis Catalytic transformations impact an enormous range program) and EERE (BETO) will likely be interested in of DOE mission areas, for example in the conversion of such technological approaches. crude petroleum and biomass into clean burning fuels and materials. Electrocatalysts control the conversion of Progress fuels into energy in fuel cells and batteries and play im- In the past year, the project “Catalysts Containing Earth portant roles in the photocatalytic conversion of energy Abundant Metals for Energy Applications” has been ad- into chemicals and materials. Catalysts are crucial to vanced from an initial, exploratory concept to the point creating new, energy-efficient routes for the production where we have now developed new, functional cobalt of basic chemical feedstocks and value-added chemicals. catalytic systems. In the process, several ligand systems Environmental applications of catalysis science include were synthesized and examined for complex formation. minimizing unwanted products and transforming toxic 6,6′′-Di(Mes)-2,2′:6′,2′′-terpyridine (Mes = 2,4,6-tri- chemicals into benign ones, such as the transformation methylphenyl) was prepared and the iron(II) and iron(0) of chlorofluorocarbons into environmentally acceptable derivatives obtained, however the low yield and high refrigerants. expense of the ligand precursor 6,6′′-dibromo-2,2′:6′,2′′- terpyridine forced us to abandon this approach. The pro- It is well known that Nature utilizes iron-based catalysts ligand 2,6-bis(Mes)-pyridine was synthesized but failed contained within the active sites of enzymes to allow a to ligate with Co, Fe or Mn. Attempts were made to wide array of difficult catalytic bond transformations to prepare 1,10-bis(Mes)-phenanthroline but these failed. occur. Unfortunately our mechanistic understanding of The 1,4-diimine family of ligands RN=C(R’)-C(R’)=NR these processes is generally poor. Therefore, expanding were prepared for R = Mes, Dipp (2,6-di-iso-propyl- our knowledge of rudimentary synthetic and mechanis- phenyl), Ad (1-adamantyl); R’ = Me, H. Subsequently tic molecular iron chemistry is therefore a worthy goal the new transition metal species M(RN=C(R’)-C(R’)=NR) towards the design and scale-up of efficient catalysts Cl2 and M(RN=C(R’)-C(R’)=NR)2 (M = Co, R’ = H , R = capable of promoting energy related processes impor- Ad, Dipp; M = Mn, R’=H, R = Dipp) were prepared and tant to humanity. Thus focus of this particular research fully characterized. Structural studies using X-ray dif- will be on the chemistry and development of cheap and fraction methods revealed divalent metal centers and earth abundant metal (e.g. iron , cobalt, nickel) cata- anionic ligand non-innocence in the case of the re- lysts pertinent to a variety of reactions with implications duced bis(ligand) complexes. The reactivity of these in global scale energy storage and production on the compounds is under investigation. The new species terawatt scale. Fe(AdN=C(H)-C(H)=NAd)Cl2 was also prepared, and was used as a platform to investigate the feasibility of the Benefit to National Security Missions synthesis of a multiply-bound terminal imido species. These findings will inform the future design of new Unfortunately, sustained efforts towards this goal did not catalysts relevant to energy, such as in the conversion of produce the desired compound, although the new spe- methane to methanol and other global-scale reactions cies Fe(AdN=C(H)-C(H)=NAd)(N(SiMe3)2)Cl was obtained with direct energy applications, improvements to which during this process and fully characterized. will ultimately benefit society. This is directly aligned with DOE and LANL missions in National and Energy A new bulky ligand, Mes2NH, has been synthesized 274

Page 275

by Buchwald coupling; however, no tractable transition Future Work metal derivatives have been currently identified from It is anticipated that by judicious choice of of molecular synthetic efforts. Reaction of the bulky amide LiN(SiMe3) targets, we will be able to prepare and characterize mo- Ad with FeCl2 yielded an unusual tetrameric species lecular compounds that exhibit new earth abundant metal \{Fe(NSiMe3Ad)\}4(O). (e.g. Fe, Co) based reactivity, including inert bond-activa- tion reactions (e.g. C-H, Si-H, N-N, and H-H bond cleavage) The pyridine-diamide ligand [2,6-(DippN(H)CH2)2NC5H3] and other novel small molecule transformations pertinent was prepared and attempts were made to synthesize new to energy related processes. Fe and Co catalysts based on this system. Unfortunately, We will continue our work into next FY on the design, syn- ligation with Fe results in multiple C-C and C-N bond cleav- thesis and characterization of catalyst targets. The most age, yielding an unusual dimeric species in which one arm promising candidates will be screened for reactivity to- of the pincer ligand has been broken and two ligands have wards small molecule synthons, such as N2, CO, CO2, etc. been fused together via a cis carbon-carbon double bond. These findings will inform the future design of new cata- At this point the mechanism for the formation of this spe- lysts relevant to energy processes, such as in the conver- cies is unclear. However this result effectively terminates sion of methane to methanol, the conversion of nitrogen investigations into this species as a potential catalyst. into ammonia, and other global-scale chemical processes. Similar work with Co has yielded crystals suitable for X-ray Conclusion diffraction studies; we are currently waiting for repairs on It is anticipated that by judicious choice of of molecu- this instrument in order to carry out structural studies. lar targets, we will be able to prepare and characterize Given the failure rate of several of these routes to poten- molecular compounds that exhibit new iron-based reac- tial new catalysts, attention shifted to synthesizing novel tivity, including inert bond-activation reactions (e.g. C-H, variants on a functioning Co system that had been recently Si-H, N-N, and H-H bond cleavage) and other novel small discovered in our group by Drs Susan Hanson and Guoqi molecule transformations pertinent to energy related Zhang. The new variation on the “PNP” donor ligand in- processes. These findings will inform the future design of volved a tolyl rather than ethylene linker in the backbone, new catalysts relevant to energy, such as in the conver- which was expected to decrease the flexibility of the ligand sion of methane to methanol, the conversion of nitrogen skeleton and change the basicity of the central amine into ammonia, and other global-scale chemical processes, donor. The new complex Co[PNP]CH2SiMe3 was obtained improvements to which will ultimately benefit society. and characterized, and showed a longer Co-N bond, and a lesser degree of sp2-hybdrization at the amide nitro- gen, compared with the ethylene linked analogue. These structural and electronic variations led to varied reactivity, and whilst addition of H(OEt2)2BArF to the ethylene linked complexes leads to a protonated amine center in the form \{Co[PN(H)P]CH2SiMe3\}BArF with the aryl linked ligand, Co[PNP]BArF was observed to form instead. Co[PNP]BArF has been shown to be an effective catalyst for hydrogena- tion and isomerization of olefins; hydrosilation reactivity is currently under investigation. The analogous iron deriva- tives, Fe[PNP]CH2SiMe3 and Fe[PNP]Cl have been pre- pared and characterized, however have not been shown to be active as hydrogenation catalysts. Modification of these catalysts is underway. In conclusion, a variety of ligands, both known and hith- erto unknown, have been prepared and examined for their ability to support molecular catalysts based on first row transition metals. Although many of these systems failed to prove stable, we have discovered that the PNP pincer sys- tem supports a diverse range of catalytic systems. Future work will focus on exploring these systems. 275

Page 276

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project Direct Tracking of Charge Carriers in Heterostructured Nanowires Rohit P. Prasankumar 20110753PRD4 Introduction nanoscale and ultrashort time scales, which will contrib- Silicon nanowires (SiNWs) have attracted much attention ute to institutional efforts in nanoscience at LANL. because of their great potential for nanophotonic and nanoelectronic applications. These quasi-one-dimen- Benefit to National Security Missions sional (1D) nanosystems have unique properties that This research will develop the new capability of ultrafast depend on their size, shape, and alignment. While time- optical microscopy, which is a general technique that integrated and time-resolved photoluminescence (PL) can be applied nearly any biological, chemical, or physi- experiments have given some insight into the influence cal system. The generality of this technique therefore of these parameters on NW properties, these measure- makes it quite interesting to several programs within the ments have not had sufficient time resolution to resolve DOE Office of Science, and could also appeal to other several important carrier relaxation processes in these programs within DOE, DOD, NIH, and NIST. In addition, systems, including electron-phonon coupling and inter/ the understanding of carrier transport across interfaces intravalley scattering. Therefore, there remains a lack of in nanowires gained here will have a significant impact basic understanding of how light interacts with individ- on solar cell and solid-state lighting applications, as well ual NWs on an ultrashort time scale, which is critical to as the use of these nanowires as high speed nanoscale fully realize their promise for various applications. transistors. Finally, the goals of this project align well with the “Basic Understanding of Materials” mission, To date, inhomogeneities in NW ensembles have made and to a lesser extent, the “Renewable Energy” mission it difficult to unambiguously extract their underlying at LANL. physics. In this project, by isolating individual nanow- ires, we avoid complications resulting from the broad Progress size and alignment distribution in nanowire ensembles, In the past year, we developed a two-dimensional (2D) allowing us to use ultrafast optical microscopy to di- smart pixel detector to create an ultrafast optical wide rectly examine carrier dynamics and carrier velocities in field microscope (UOWFM), capable of rapidly acquiring individual NWs with high temporal and spatial resolution wide field microscopic images with high temporal- and in a non-contact manner. Moreover, we can use a hybrid spatial-resolution. As compared to most optical micros- NW-nanoantenna structure in a terahertz (THz)-pump, copy measurements, which are nearly always per- optical-probe experiment to generate THz field strengths formed in a time-integrated mode, our ultrashort pulsed on the order of tens to hundreds of kV/cm at the an- laser-based technique provides femtosecond temporal tenna focus. This will enable us to directly photoexcite resolution over a broad energy range. This is realized by transient electron dynamics, allowing us to monitor the combining wide field optical microscopy and ultrafast resulting carrier populations with ultrafast time resolu- optical spectroscopy. tion and observe novel phenomena such as the dynamic Franz-Keldysh effect. As a first step, we acquired time-resolved images of a gold-patterned amorphous silicon film and a single Our ultrafast optical and THz experiments should thus silicon nanowire (NW) to demonstrate the validity of this open pathways to directly study carrier dynamics and novel concept. Two different femtosecond laser systems charge transport in quasi-1D nanosystems. This research were used to study these different samples, demonstrat- has potential application to NW-based devices, opto- ing the versatility of our approach. In both cases, we electronics, and sensitive photodetection on both the 276

Page 277

used a unique 2D smart pixel detector array (provided by Future Work Heliotis AG) with a 50X (0.50 NA) objective and a zoom Building on our previous work, we expect to investigate ul- lens with variable magnification to image these materials. trafast carrier dynamics in various single nanowire sam- This 2D array detector performs with sensitivity compara- ples, such as semiconductor NW heterostructures, NWs ble or better to that of a conventional single pixel detector integrated with plasmonic structures, and NWs integrated in these experiments, which allows our UOWFM system with other materials, such as graphene. We will do this to behave like a ‘lock-in’ camera; essentially, this tech- by extending our system to incorporate a microstructured nique allows one to “look” through an optical microscope fiber to generate a white light continuum, which will allow with extremely high time resolution. The time-resolved us to tune the probe wavelength throughout the visible optical images of a single Si NW acquired using UOWFM and near-infrared relative to a given material’s band gap. thus provided valuable insight into carrier dynamics with This will also allow us to directly measure carrier diffu- sub-micron optical resolution, while time-resolved images sion velocities and coefficients in a wide range of samples, of a patterned Si film demonstrated our ability to capture focusing on nanowire-based structures but also extend- information over a large area within a short acquisition ing our capabilities to other materials, such as complex time. Importantly, our nanowire imaging results show oxide heterostructures and metamaterials. This aspect of that our technique works for very small targets with sub- our research thus has potential applications in NW-based micron spatial resolution, which should be very helpful for devices, optoelectronics, and sensitive photodetection by nanomaterials research, particularly given the rapid image combining measurements at both micrometer distance acquisition time. and femtosecond time scales to reveal the intrinsic proper- ties of these quasi-one-dimensional nanosystems. Overall, ultrafast optical wide field microscopy combines the non-contact nature and spatial resolution of conven- In parallel, we will use our newly built THz system to tional optical microscopy with the temporal resolution of explore the possibility of directly pumping semiconduc- ultrafast spectroscopy to rapidly and sensitively acquire tor NWs with ultrashort, intense THz pulses. We will spatially and temporally resolved images of nearly any facilitate this by incorporating a single NW into a hybrid sample. Furthermore, ultrafast optical wide field micros- NW-nanoantenna structure, which will enhance the THz copy can, in principle, be used to acquire time-varying field strength enough to produce a measurable signal level images of almost any material that can be imaged with a (this would normally be difficult since the NW volume is conventional optical microscope, and can thus be expected extremely small compared to the THz wavelength). By per- to have many future applications in a variety of physical, forming THz-pump, optical-probe experiments on the hy- chemical, and biological systems. Our future work will thus brid NW-nanoantenna structure, we can selectively excite include further optimizing the 2D array detector for this and track electron dynamics with THz peak field strengths unique application and applying this technique to track on the order of tens of kV/cm and monitor the resulting space and time varying processes in a variety of systems. evolution of the electron population with ultrafast time This will contribute to LANL’s leadership in nanoscience resolution. This will be a significant advance over typical and nanotechnology while supporting the collaboration optical-pump/optical-probe experiments, which investigate between LANL and Sandia National Laboratories (through both hole and electron dynamics created simultaneously the Center for Integrated Nanotechnologies). Finally, this using band-to-band photoexcitation, and may allow us to forefront science and capability development will attract observe unique phenomena such as the dynamic Franz- high quality students, postdocs and collaborators to LANL. Keldysh effect. In parallel, we built a time-domain terahertz (THz)-probe Conclusion spectroscopy system, based on an amplified 1 kHz laser, Silicon nanowires (SiNWs) have attracted much atten- and designed a special hybrid plasmonic NW-nanoantenna tion because of their great potential for nanophotonic device structure that will enable us to study the response and nanoelectronic applications. However, to fully realize of individual semiconductor NWs after both optical and this promise, an understanding of non-equilibrium carrier THz photoexcitation. This THz nonlinear spectroscopy dynamics in these quasi-one-dimensional systems, espe- capability can be applied to a wide range of problems in cially on an ultrashort time scale, is critical. Here, we will photonics and electronics, advancing both fundamental use ultrafast optical and terahertz (THz) spectroscopy to science while potentially impacting engineering applica- explore carrier dynamics in individual semiconductor NWs. tions. 277

Page 278

This will provide deep insight into the intrinsic properties 2013. Optics Express. 21: 8763. of semiconductor NWs, enabling their optimization for a variety of applications. Publications Prasankumar, R. P.. Tracking charge carriers through space and time in single semiconductor nanowires. Invited presentation at CINT User Workshop. (Albuquerque, 19-20 September 2012). Prasankumar, R. P.. Tracking charge carriers through space and time in single semiconductor nanowires. Invited presentation at SSL EFRC coffee hour. (Albuquerque, 14 June 2012). Seo, M. A., J. Yoo, D. E. Perea, S. A. Dayeh, S. T. Picraux, A. J. Taylor, and R. P. Prasankumar. Tracking ultrafast carrier dynamics in single semiconductor nanowire heterostructures. 2013. In 18th International Conference on Ultrafast Phenomena. (Lausanne, Switzerland, 8-13 July 2012). , p. 4030. London: Oxford University Press. Seo, M. A., J. Yoo, S. A. Dayeh, S. T. Picraux, A. J. Taylor, and R. P. Prasankumar. Mapping carrier diffusion in single silicon core-shell nanowires with ultrafast optical microscopy. 2012. Nano Letters. 12: 6334. Seo, M. A., J. Yoo, S. A. Dayeh, S. T. Picraux, A. J. Taylor, and R. P. Prasankumar. Tracking charge carriers through space and time in single semiconductor nanowires. Presented at Conference on Lasers and Electro-Optics. (San Jose, CA, 6-11 May, 2012). Seo, M. A., J. Yoo, S. A. Dayeh, S. T. Picraux, A. J. Taylor, and R. P. Prasankumar. Mapping carrier diffusion in single silicon core-shell nanowires with ultrafast optical microscopy. 2013. In 7th International Conference on Materials for Advanced Technologies. (Singapore, 30th June-5th July, 2013). Vol. 8, p. 128. Singapore: World Scientific. Seo, M. A., S. A. Dayeh, P. C. Upadhya, J. A. Martinez, B. S. Swartzentruber, S. T. Picraux, A. J. Taylor, and R. P. Prasankumar. Understanding ultrafast carrier dynamics in single quasi-one-dimensional Si nanowires. 2012. APPLIED PHYSICS LETTERS. 100 (7): 071104. Seo, M. A., S. A. Dayeh, P. C. Upadhya, S. T. Picraux, J. Martinez, B. S. Swartzentruber, A. J. Taylor, R. P. Prasankumar, and IEEE. Polarization anisotropy of transient carrier dynamics in single Si nanowires. 2011. In Conference on Lasers and Electro-Optics (CLEO) ; 20110501 - 20110506 ; Baltimore, MD. Seo, M. A., S. Boubanga-Tombet, J. Yoo, Z. Ku, A. V. Gin, S. T. Picraux, S. R. J. Brueck, A. J. Taylor, and R. P. Prasankumar. Ultrafast optical wide field microscopy. 278

Page 279

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project Fabrication of an All-carbon Solar Cell Andrew M. Dattelbaum 20120732PRD1 Introduction a hydrogen fuel cell. We observed an enhancement in Unlike other nanomaterials that degrade in air and loss the ability of graphene oxide after the ozonation process their functionality, graphene oxide is stable in air. Thus, was completed. A manuscript based on this result was graphene oxide is a unique nanoscale material that may written and is submitted. be functionalized at the atomic level to control the prop- erties of bulk films prepared from the nanoscale building Use of graphene oxide as a nanoscale supercapacitor blocks. We will demonstrate that chemical modification We developed a technique that uses a laser to pattern of graphene oxide can be used to control the resulting films of graphene oxide to make them act as superca- properties of bulk films prepared from the functionalized pacitors. The technique uses a mask to simplify the material. Such functionalization can be used to make fabrication process, as well as facilitate assembly of the graphene oxide that is optimal for new photovoltaic ma- supercapacitor devices, which is the biggest challenge terials, supercapacitors or fuel cell membranes. in micro-scale-device fabrication. Devices with three different sizes were fabricated (0.25 mm2, 2.5 mm2, and 25 mm2). An unusual dependence of the capacitance as Benefit to National Security Missions the area changed was observed (exponential instead of This work supports the energy security mission of Los linear, as we initially expected, i.e. the capacitance was Alamos National Laboratory. Specifically, we will synthe- better than we thought it would be). We are currently size new carbon nanomaterials that may be useful for trying to understand this behavior through a series of a variety of applied energy applications in fuel cells, or- systematic studies. ganic photovoltaic architectures, and supercapacitors. If successful, the work may be useful to a variety of spon- Crosslinking of graphene oxide sors in the DOD or Intelligence Community. The work also supports the mission of the Center for Integrated We are chemically crosslinking graphene oxide sheets Nanotechnologies, which is a DOE Basic Energy Sciences to make the bulk films prepared from GO more stable funded user facility located in part at Los Alamos. The in fuel cell and supercapacitor devices. We are trying to Center’s goal is to promote nanoscience accessibility crosslink GO with various chemistries that involve poly- across the country by providing access to the expertise mers like poly(ethylene) imine and poly(ethylene) glycol. and equipment at the facility. This work will eventually If successful, we will have a much stronger GO-based lay the foundation for new capabilities that will be avail- film that can sustain extreme conditions such as me- able to future users keeping us at the cutting edge of chanical stretching and water/organic solvent flooding. nanoscience work. Future Work Progress • Set up a synthesis program at LANL to prepare re- Graphene oxide (GO) ozonation for polymer electrolyte duced graphene oxide that can be incorporated into membrane application in Fuel cells a solar cell device. In collaboration with the fuel cell group in MPA-11, we • Optimize the overall cell performance by chemical investigated the ozonation process of graphene oxide. doping or functionalization of the graphene oxide to We studied the structure of ozonated graphene oxide us- tailor the interface for controlled characterization of ing a variety of techniques including solid state nuclear the charge transfer mechanisms in the solar cell. magnetic resonance, and it ability to conduct protons in 279

Page 280

• Prepare functionalized graphene oxide thin films and work with colleagues in MPA-11 who will determine if they are viable materials for fuel cells. • Synthesize graphene oxide films and pattern them to test their use as supercapacitors. Conclusion If successful, this work would establish a new class of organic photovoltaic devices based on all-carbon nanoma- terials. Due to the ability to make large quantities of gra- phene oxide and its doping flexibility, devices made from this material may lead to more stable and efficient solar cells, supercapacitors or membranes for fuel cells than what is achievable from currently available technology. 280

Page 281

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project Chemically Modifying the Uranyl Ion Jaqueline L. Kiplinger 20120750PRD2 Introduction compounds could find relevance in the nonproliferation The uranyl ion, [UO2]2+, is the most ubiquitous form of community. uranium. The metal-oxygen bonds in this linear mole- cule are incredibly strong, making uranyl complexes very Progress stable. Because uranyl complexes are so stable, they are We attempted to prepare stable uranyl complexes of the resistant to modifications–modifications that could lead type, [L]UO2Cl (where L = bulky ligand). Unfortunately, to easier separations, or improved nuclear fuel repro- the ligand we chose ended up reacting with the normally cessing, for example. In this context, chemically modify- inert UO2 fragment. We will be looking at bulky ligands ing a uranyl complex by connecting a carbon ligand to to try and circumvent this new reaction pathway. it is a very challenging but important goal. A molecular We prepared stable thorium complexes of the type uranyl complex containing a metal-carbon bond, also [L]2ThCl2 (where L = bulky ligand) and uranium com- called an organometallic uranyl complex, would be of in- plexes of the type [L]2UCl2 (where L = bulky ligand). We terest because solid-state uranium carbide materials can discovered new synthetic methods for the high-yielding be used in the new generation of safer nuclear reactors. preparation of thorium metallocene dichloride com- We propose to prepare stable organometallic uranyl plexes. We started to examine the reaction chemistry of complexes using a straightforward method. These well- both systems and developed routes to the first thorium defined molecular compounds will be valuable because azide complexes. they enable a range of experimental and theoretical studies that are not available using solid-state materials. We successfully applied our new method for preparing actinide fluoride bonds by reacting the [L]2UCl2 and Benefit to National Security Missions [L]2ThCl2 complexes with trimethyltin fluoride (Me3Sn- The goals of this project are (1) to synthesize and charac- F). This chemistry yielded the corresponding fluoride terize a rare uranyl fluoride complex and (2) to use this complexes [L]2UF2 and [L]2ThF2. uranyl fluoride complex to synthesize a uranyl complex containing a uranium–carbon bond. Within the field of Finally, we discovered a route to the first terminal actinide chemistry, this work will have a considerable im- uranium(III) fluoride complexes. These compounds pact, not only because examples of these complexes are are remarkable stable, which is in contrast to previous so uncommon, but also because the methods we pro- predictions. We began exploring the chemistry of these pose will be adaptable for widespread use by the chem- molecules. istry community, beyond uranyl complexes and across the actinides. More broadly, this chemical achievement All complexes will be fully characterized using miscella- will be a major step forward in understanding how to neous spectroscopic techniques and X-ray diffraction. chemically modify the uranyl ion, and the proposed work could lead to essential improvements in process- Future Work ing and separation. Finally, this effort will generate a We will continue to explore avenues to prepare stable handful of well-defined complexes of uranyl difluoride. uranyl complexes of the type, [L]UO2Cl (where L = bulky UF6 reacts with water to form uranyl fluoride, which is a ligand). We will be looking at other bulky ligands to try hallmark for the presence of UF6 activities. Character- and stabilize a monomeric uranyl fluoride complex. izing and studying the chemistry of these uranyl fluoride We will continue to explore the chemistry of the 281

Page 282

[L]2ThF2 and [L]2UF2 (where L = bulky ligand) complexes. We will finish and publish a manuscript on the new syn- thetic method for the high-yielding preparation of thorium metallocene dichloride complexes. We will finish and publish a manuscript on the synthesis of the first thorium azide complexes. We will continue to explore the generality of the our new method for preparing actinide fluoride bonds by react- ing uranium chloride, bromide and iodide complexes with complexes with trimethyltin fluoride (Me3Sn-F). After this scope is accomplished we will file a patent application. Finally, we will continue to explore the chemistry and prop- erties the first terminal uranium(III) fluoride complexes, which we recently prepared. Conclusion The main project goals are (1) to synthesize and character- ize a rare uranyl fluoride complex and (2) to use this uranyl fluoride complex to synthesize a uranyl complex containing a uranium–carbon bond. Within the field of actinide chem- istry, this work will have a considerable impact, not only because examples of these complexes are so uncommon, but also because the methods we propose will be adapt- able for widespread use by the chemistry community, beyond uranyl complexes and across the actinides. Publications Monreal, M. J., R. J. Wright, J. T. Golden, D. E. Morris, B. L. Scott, P. P. Power, and J. L. Kiplinger . Thorium(IV) and uranium(IV) halide complexes supported by bulky β-diketiminato ligands. 2013. Organometallics (Invited Article – Special Issue: Recent Advances in Organo-f- Element Chemistry). 32: 1423. Travia, N. E., M. J. Monreal, B. L. Scott , and J. L. Kiplinger. Thorium-mediated ring-opening of tetrahydrofuran and the development of a new thorium starting material: Preparation and chemistry of ThI4(DME)2. 2012. Dalton Transactions. 41: 14441. Travia, N. E., M. J. Monreal, B. L. Scott, D. E. Morris , and J. L. Kiplinger. Developing the next generation thorium iodide starting materials. Invited presentation at 243rd American Chemical Society National Meeting. (San Diego, California, 25-29 March 2012). Travia, N. E., M. J. Monreal, B. L. Scott, and J. L. Kiplinger. A new starting material for thorium chemistry: ThI4(DME)2. Presented at Gordon Research Conference on Inorganic Chemistry. (University of New England, Biddeford, Maine, 17-22 June 2012). 282

Page 283

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project Designing and Probing Novel Materials by Pressure Tuning of Nanocrystals Hongwu Xu 20120753PRD2 Introduction ticles and their building blocks, with great potential for Nanocrystals, which are sized between 1 and 100 nm technological applications. and are structurally defect-free, possess unique prop- erties compared to their bulk counterparts. Pressure Benefit to National Security Missions tuning of nanocrystals is a novel approach to design, This project represents a research topic with great develop and probe new nanostructured materials. At scientific value and tremendous potential for techno- high hydrostatic pressure and/or deviatoric stress, nano- logical applications. Through the state-of-the-art in-situ crystals may not only undergo structural transformations high-pressure small- and wide-angle synchrotron X-ray to other phases but also coalesce into one-, two- and scattering measurements of series of lead chalcogen- three-dimensional superlattices and mesostructures, ide nanocrystals, this work will shed important lights potentially leading to new physical properties. on structural mechanisms underlying the formation of novel chalcogenide nanocrystals and their assembled su- Using lead chalcogenide (e.g., PbSe) as a model sys- perlattices at high hydrostatic pressure and/or deviatoric tem, this project focuses on investigations of pressure- stress. More broadly, the results will have applications/ induced behavior of nanocrystals mainly using small- implications for controlled nanoparticle synthesis and angle and wide-angle synchrotron X-ray scattering for design and fabrication of novel structured nano- (SAXS and WAXS) coupled with diamond-anvil cell (DAC) materials with potential, enhanced or newly emerging techniques. Major tasks include: 1) synthesize series of properties. The use of pressure will enable systematic lead chalcogenide nanocrystals of various particle sizes, manipulation of the nanophases and their assembled morphology and capping ligands using wet-chemical and architectures (e.g., interparticle spacing), thereby en- other methods; 2) characterize the obtained nanopar- abling tuning of their properties for specific applications. ticle samples using powder X-ray diffraction, scanning and transmission electron microscopy and other meth- This research is expected to open up a new direction in ods; and 3) employ several high-pressure techniques high-pressure nanomaterials science by exploring new including SAXS, WAXS, Raman, absorption and emission phenomena and tuning mechanisms for a wide spec- spectroscopies to determine pressure-tuned structural trum of nanoparticles and their building blocks. Ex- stability, phase transformations, mechanical and opti- amples of potential applications include catalysts, solar cal properties of random “amorphous” and assembled cells, optoelectronic and thermoelectric devices. Using ordering nanocrystals. In particular, the combined in-situ pressure as a tool/variable, this research seeks funda- high-pressure SAXS and WAXS techniques will allow mental understanding of the structures and properties examination of structural variations in both nanocrystals of novel nanomaterials, and is tied to missions in DOE (atomic structures, WAXS) and their assembled architec- office of Science and NSF, and to all the three themes of tures (mesoscale superlattices, SAXS). LANL’s Materials Grand Challenge (defects and interfac- es, extreme environments, and emergent phenomena), The results will shed light on structural mechanisms as well as MaRIE. underlying the formation of novel high-pressure nano- phases and their assembled superlattices. This research Progress is expected to open up a new direction in high-pressure During the past year, we have made tremendous prog- nanomaterials science by exploring new phenomena ress in this project. Specifically, we investigated the roles and tuning mechanisms for a wide spectrum of nanopar- 283

Page 284

of organic capping ligands in controlling the formation of anvil techniques. The purpose was to study the effects colloidal lead selenide (PbSe) nanoparticle assemblies (or of nanoparticle size on the cubic-to-orthorhombic phase supercrystals), determined the temperature effects, and transition at high pressures. With decreasing the particle examined size-dependent phase transition behavior at high size from 9.7 nm to 5.3 nm, the phase transition pressure pressure. increases from 7 GPa to 11 GPa, which is referred to as the Hall-Petch effect. However, as the particle size is further Roles of capping ligands decreased to 3.8 nm, instead of the conventional cubic-to- Although the presence of organic ligands outside inor- orthorhombic transition, the rock-salt PbSe structure be- ganic, colloidal nanoparticles is essential to nanoparticles came amorphous at about 13 GPa. This behavior is similar formation, their roles in controlling self-assembly of the to that of PbTe nanoparticles, reported by us recently: Z. nanoparticles are not well understood. In this work, we Quan, Z. Luo, Y. Wang, H. Xu, C. Wang, Z. Wang, and J. Fang specifically designed our experiments to examine how cap- (2013), Pressure-induced switching between amorphiza- ping ligands affect the symmetry/structure of assembled tion and crystallization in PbTe nanoparticles. Nano Letters superlattice. We used monodisperse PbSe nanoparticles (revised). with sizes of 9.7 nm, 9.4 nm, 7.3 nm, 5.3 nm and 3.8 nm, which are coated with the same organic ligand - oleic acid. Future Work Small-angle X-ray scattering (SAXS) measurements show During FY14, we will perform the following tasks: that these nanoparticles self-assembled into supercrystals via evaporation-induced processes. We observed different

1. Conduct additional in-situ high-pressure small-angle packing behaviors for PbSe nanoparticles with different siz- and wide-angle X-ray scattering experiments of PbSe es: The smaller nanoparticles (3.8 and 5.3 nm), which have nanoparticles using diamond anvil cell (DAC) tech- higher ratios of ligand/particle (and are thus softer) tend niques at CHESS, Cornell University. The combined to form the body-centered-cubic (bcc) superstructures, small- and wide-angle synchrotron measurements will whereas the larger nanoparticles (7.3, 9.4 and 9.7 nm) allow examination of structural variations, phase trans- with lower ligand/particle ratios adopt the face-centered- formations and mechanical properties (compressibility cubic (fcc) superstructures. Entropic considerations suggest or expandability) of both PbSe nanocrystals (atomic- that these nanoparticles would form the closest packed level) and their assembled architectures (meso-scale) fcc structure. The formation of the bcc supercrystals from as a function of pressure, pressure medium, and smaller nanoparticles (3.8 and 5.3 nm) was unexpected, nanoparticle size. One goal is to determine the fun- indicating the important role that organic ligands may play damental mechanisms underlying the formation of in the self-assembly process. In addition, the fact that the high-pressure nanocrystal phases and the coalesced nanoparticles can orient to produce single supercrystals mesostructures or superlattices. Another goal is to use during the aging period confirms the importance of ligand- pressure as a synthetic tool to obtain novel nanopar- ligand interactions in directing the assembly process. ticle assembly morphologies such as nanowires, nanosheets, and nanoporous framework structures. Temperature effects
2. Extend our studies of single-component nanoparticle We further exploited the roles of capping ligands on the as- self-assemblies to binary-component systems. Po- sembled superlattices by varying temperature. A series of tential systems include lead chalcogenide and metal SAXS measurements were conducted on PbSe supercrystal or magnetic nanoparticles, in which novel couplings/ samples (3.8 and 5.3 nm) after they had been heated at interactions are expected to occur and, as a result, increased temperatures from 25 to 150 °C. The results new physical phenomena may emerge. Various fac- show that the supercrystal phases underwent a gradual tors controlling the co-assembly behavior, including transition from the bcc to fcc structure, implying that the assembly method, ratio of different components, type bcc superstructure was metastable and was stabilized due of metal/magnetic nanoparticles, and external forces to kinetic reasons. Moreover, as it was unlikely that the (especially pressure), will be investigated. The relations PbSe core particles would undergo much change at these of superlattices to the resulted physical properties will relatively low temperatures, the bcc-to-fcc transition may also be explored. solely arise from structural changes in ligands as well as their interactions between nanoparticles. 3. Further analyze the SXAS/WAXS data of PbSe nanopar- ticles collected in FY13, conduct additional experi- Pressure effects ments (such as electron microscopy) if needed, sum- We conducted high-pressure wide-angle X-ray scattering marize the results and write papers for publication. (WAXS) experiments of PbSe nanoparticles using diamond- 284

Page 285

Conclusion Through the state-of-the-art in-situ small- and wide-angle synchrotron X-ray scattering measurements of series of lead chalcogenide nanocrystals, this research will provide important insights into rational design of related struc- tured nanomaterials with enhanced or newly emerging properties. Specific results include: 1) development of novel routes/procedures for controlled synthesis of lead chalcogenide nanocrystals; 2) characterization of their compositions, sizes, shapes and surface structures; and 3) determination of the formation mechanisms and governing factors responsible for structural stability of high-pressure nanophases and their assembled architectures. Publications Quan, Z. W., Z. P. Luo, Y. X. Wang, H. W. Xu, C. Y. Wang, Z. W. Wang, and J. Y. Fang. Pressure-Induced Switching be- tween Amorphization and Crystallization in PbTe Nanopar- ticles. 2013. NANO LETTERS. 13 (8): 3729. 285

Page 286

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project Functional Soft Materials by Assembling Metallo-biopolymers Reginaldo C. Rocha 20120765PRD3 Introduction and fundamental understanding of novel materials for The extensive use of polygalacturonic acids and alginates energy and environment sciences and technologies. in textiles, food, and biomedical industries is based on the high electrostatic charges of the polysaccharide With the development of biopolymer-based functional structures and their ability to form multilayer gel-like (nano)materials and their target applications in the areas structures in the presence of metal cations. Evidently, of renewable energy, biosensing, regenerative medicine, ideal information-rich materials should be able to optical and electronic devices, and self-healing systems, translate positional and orientational orders beyond this research is highly relevant to the DOE (SC/BES), DOD the single-molecule level to the next hierarchical level. (DARPA), and DHHS (NIH), in addition to possible deploy- This project introduces a designed motif of higher-level ment and utility to NASA and Intelligence missions. DHS biopolymer structures where helical polypeptide chains and DTRA also have a vital interest in biodegradable strongly associate by specific sequestration and bind- polymers and wound-healing biocompatible materi- ing of metal ions. Not only will metal centers bring their als, which represent an underlying theme in this work. rich optical, electronic, and redox capabilities to such Therefore, through the creation of capabilities and in- biopolymers but also they can play the role of dynamic volvement in this new field, our project will lead to fund- cross-linkers by assembling multiple polymer strands to- ing opportunities from several government agencies. gether. Similarly, secondary structures can be conceived This research also ties directly into the following LANL’s of in the context of the so-called “egg-box model” by LDRD Grand Challenges: “Materials: Discovery Science replacing intra-stand hydrogen bonds with site-specific to Strategic Applications,” “Complex Biological Systems,” placement of metal-based cross-linkers. and “Energy & Earth Systems,” in addition to its align- ment with our DOE’s Center for Integrated Nanotech- Metal-ligand interactions can capture all the salient fea- nologies (CINT) missions in nanoscience and nanotech- tures of protein-protein interactions (specificity, direc- nology. tionality, symmetry, and reversibility) on a much smaller surface than needed by non-covalent interactions. The Progress reversibility of metal-ligand coordination also allows for A central goal of our proposed work was to construct a the self-correction necessary to produce assemblies with library of metal-biopolymer complexes by combining the long-range order. Importantly, chemical control and tun- rich information content of biopolymers with the tun- ability are inherent in this approach, as metal-directed able redox, magnetic, and optical properties of metal- self-assembly is dependent on external stimuli and can ligand units. Through this approach, we have designed be dictated by the choice of metal ions and binding related classes of constructs in which metals can be motif. The potentiality of resulting novel materials is integrated into biopolymer structures in three different determined by the tunability of specific functionalities ways: by tethering to the polymer backbone as a pre- and responsiveness, as well as the dynamic nature of the formed complex (class I), by coordination with a pre- supramolecular metallo-biopolymers. conjugated ligand in a “plug and play” fashion (class II), or by direct coordination of metal ions with polypeptide Benefit to National Security Missions residues (class III). We have investigated these possibili- This research supports our National Laboratory missions ties by synthesis and characterization of representa- by providing advancements toward the development tive members for each of these classes, as summarized 286

Page 287

below in our progress to date. bioconjugation reactions using mass spectrometries (both electrospray ionization and matrix-assisted laser desorp- For classes I and II, we have implemented different bio- tion/ionization). To follow on such promising results, we conjugation methods to take advantage of our new ligand are set out to evaluate these assemblies as to stimuli- units functionalized with -NH2/-COOH groups complemen- responsive characteristics as a function of their emergent tary to the side-chain functionalities of biopolymers that redox and optical properties. are suitable for carbodiimide coupling chemistry. In class I, we have used alterations in substituent groups on transi- In class III, an exciting accomplishment was the develop- tion metal-terpyridine complexes to modulate their strong ment of a method for constructing hybrid organic-inorganic metal-to-ligand charge-transfer (MLCT) absorptions across nanocomposites from metal ions such as copper(II)/silver(I) the visible spectral region and associated redox potentials. and natively unstructured polypeptides. Characteriza- tion by scanning electron microscopy (SEM) clearly shows We have also developed a suite of transition-metal com- that some of the tested metal ions induce the formation plexes in which one fixed bidentate ligand having conjuga- of well-defined nanostructures (e.g. “nanoflowers”) in tion sites is partnered with other ligands that are varied phosphate buffer. Our initial studies seem to indicate that to achieve improved photophysical and redox properties; the complexes formed between polymers and metal ions these properties can also be tuned by medium conditions, become nucleation sites for primary crystals. This interac- such as pH or solvent. Incorporation of redox and optical tion then leads to the growth of micrometer-sized flower features from metal-ligand moieties into the integrated particles featuring the nano-sized petal shapes. Energy- assemblies was clearly shown by electrochemical and pho- dispersive X-ray spectroscopy (EDX) confirmed that Cu(II) is tophysical studies. the main constituent of the structure, and X-ray photoelec- tron spectroscopy (XPS) showed that the oxidation state of The most striking results were 1) a large perturbation of copper (+2) does not change upon composite formation. the metal-based oxidation potential by hundreds of mV Wide-angle X-ray diffraction (XRD) has also been used and upon bioconjugation, and 2) a three-fold increase in inten- revealed that the crystal pattern is similar to Cu3(PO4)2 sity for the MLCT-based photoluminescence upon coacer- generated from the mixture of CuSO4 in phosphate buffer. vation of the polypeptides above their transition tempera- This is an important finding because metal-biopolymer ture. Based on systematic structure-property studies, we composites encompass materials where blending is a tech- have surmised that coacervation induces structural restric- nological approach to generating a complete landscape of tions with increased rigidity imposed on the metal-ligand biomaterials with specific sets of properties. moiety. Combined with the insulation by the surrounding polymer, such structural factors lead to the unique optical Future Work and electronic effects observed. In this first year of the project, libraries of metal-ligand complexes were created and used as electronically and op- In class II, we have adopted the “plug and play” approach tically active components into the construction of metallo- where the side-chain length of the functional groups of a biopolymers that were categorized in three classes. By terpyridyl unit has been varied systematically to procure a making these novel materials available and structurally spectrum of reactive ligands with variable exposure. This characterized, we can next focus on exploring the func- functionalized side chain was shown to readily bind to a tionalities emerging from the synergy between such metal series of lanthanide ions, resulting in highly photolumines- sites (transition metals and lanthanides) and biopolymers. cent metallo-biopolymers both in liquid solutions and in Parallel to the detailed understanding to be gained from thin films. structure-property relationships, the design and assembly of new derivatives optimized for stimuli-responsive func- The emission spectra of these integrated assemblies cover tions will be possible through systematic photophysical, the entire visible range, raising exciting possibilities for electrochemical, and microscopic studies as a function of applications into light-emitting materials with individual se- key factors such as chemical nature of metal ions, ligand lection of specific light hues or combined into the genera- environments, and composition of polypeptides. In ad- tion of white light. In both classes I and II, the successful dition to our ongoing experimental characterizations, conjugation of metal-ligand components with polypeptides computational modeling will be initiated to support the as well as the introduction of redox/optical properties elucidation of emergent properties arising from the inter- into specific polymer sites has been evidenced by several play between inorganic moieties and polypeptide struc- techniques such as nuclear resonance spectroscopy, gel tures in the integrated metallo-biopolymer assemblies (as electrophoresis, and time-dependent monitoring of the described in the Progress section). 287

Page 288

Conclusion In our metal-directed self-assembly approach, higher-order topologies will be created from association of tailored polypeptide modules. These topologies will arise from incorporation of metal-ligand sites into such modules and dependence of target metallo-biopolymer ensembles on the chemical nature of metal ions as well as ligands and coordination environments. We also aim at developing a detailed understanding of key factors determining func- tion of resulting biomaterials via systematic studies of structure-property relationships. In addition to high-impact fundamental contributions to new areas in soft nanomate- rials, outcomes of this project are expected to have great potential for applications in optoelectronics, sensing, and self-healing (bio)materials. Publications Balog, E. R., K. Ghosh, Y. Park, H. Wang, R. C. Rocha, and J. S. Martinez. Elastin-like Polymers for Stimuli- Responsive Opto-Electronic Materials. Presented at Materials Research Society (MRS) Spring Meeting. (San Francisco, CA, April 1-5, 2013). Ghosh, K., E. R. Balog, J. S. Martinez, and R. C. Rocha. Coacervation-Induced Photoluminescence Enhancement in Bio-Polymer Assembies of Elastin- like Polypeptides with Photoactive Metal Complexes. Chem. Comm.. Ghosh, K., E. R. Balog, J. S. Martinez, and R. C. Rocha. Optical and Redox Properties of Novel Conjugates from Elastin-Like Polypeptides and Photoactive Transition- Metal Complexes. Presented at Materials Research Society (MRS) Spring Meeting. (San Francisco, CA, April 1-5, 2013). Ghosh, K., E. R. Balog, J. S. Martinez, and R. C. Rocha. Metal-Biopolymer Assemblies and Composites From Elastin-like Polypeptides. Presented at Gordon Research Conferences - Polymers Conference. (South Hadley, MA, June 9-14, 2013). Ghosh, K., E. R. Balog, P. Sista, J. S. Martinez, and R. C. Rocha. Flower-Shaped Nanopatterns from Metal- Biopolymer Assemblies. Biomater. Sci.. Ghosh, K., P. Sista, and R. C. Rocha. Metallo-Biopolymers: Bioconjugation Strategies and Applications. Polym. Chem.. Sista, P., K. Ghosh, J. S. Martinez, and R. C. Rocha. Polythiophenes in Biological Applications. To appear in J. Nanosci. Nanotech.. 288

Page 289

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project nTuned Optical Properties of Low-dimensional Carbon Nanomaterials for Energy Harvesting Stephen K. Doorn 20120766PRD3 Introduction cal functionalization would open up a pathway for incor- Low-dimensional carbon nanomaterials such as gra- poration of these materials into large area, inexpensive, phene and carbon nanotubes (CNTs) offer intriguing flexible energy harvesting electronics. Understanding optical properties which can potentially be incorporated their optically excited states and relaxation dynamics into energy harvesting applications. However, their will clarify the competition between charge and energy implementation requires in-depth understandings of transfer and recombination in these materials, which the materials optical properties such as fundamental is crucial in determining their potential for photovolta- nature of the optically excited states and their relaxation ics. The results will provide insight for design of novel dynamics. Furthermore, tuning of these states by chemi- energy harvesting devices utilizing the unique optical cal doping will provide flexibility to device design. The properties of low-dimensional carbon nanomaterials, primary aim of this research is to evaluate the potential and makes an important contribution to LANL’s mission of oxygen doping of these materials for charge/energy on sustainable energy. The goals of the project also transfer processes at the core of photovoltaic needs by directly address development of new routes to control- first understanding the nature of their optical states, ling materials functionality. The same characteristics and to demonstrate their ideal chemical structures for and underlying behaviors that we will study as related to photovoltaics. We aim to answer the question: “Can photovoltaic applications also form the basis for applica- effective functionalization strategies lead to unique opti- tions of these materials in sensing, imaging, optoelec- cal properties of graphene oxide and CNTs for energy tronics, and photonics. These then also support LANL harvesting applications?” national security missions and are relevant to programs related to DOE, DOD, DHS, and NIH. With similar origins for optical properties based in oxygen doping, optical properties of graphene oxide Progress (GO) and oxygen doped nanotubes may display parallel Our two research thrusts of i) graphene and ii) CNTs (car- behaviors that lead to better understanding of optical bon nanotubes) were conducted in parallel to facilitate response in both. For GO we will focus on chemically the understanding of the underlying origin of the oxygen exfoliated flakes as material. Steady-state and time-re- doping induced optical states in low-dimensional carbon solved optical studies will be performed on this material. nanomaterials. Details of progress made for each thrust Our focus will be on defining how chemical modification are as follows: alters optical response. For oxygen-doped carbon nano- tubes, we will also apply steady state and time-resolved Graphene spectroscopies. Direct imaging and spectroscopy at the We fabricated graphene oxide (GO) single nanosheet single nanotube and even dopant-site level will also be optoelectronic devices as platforms for understanding pursued. Of special interest will be determining the the core photophysics of the low-dimensional carbon surface mobility of oxygen dopants, how these states in- nanomaterials during energy harvesting device opera- teract withe mobile excitons, and what are their surface tions. State-of-the-art electron-beam lithography and dynamics. thin film metal deposition capabilities at the Center for Integrated Nanotechnologies (CINT) were used for the fabrication. Single nanomaterial device fabrication Benefit to National Security Missions requires time and effort compared to the assembled Fine tuning of graphene/CNTs with well designed chemi- 289

Page 290

film counterpart; however, the obtained information is Future Work free from averaging effects, which is essential for the Development and optical studies of Graphene Oxide (GO) fundamental study of the device. Device characterization We will establish approaches to variable functionaliza- using Scanning Photocurrent Microscopy (SPCM) provided tion chemistry of GO, with the goal of tunable control over photocurrent distribution of the material with high spatial its optical properties. Additional control will be pursued resolution. The SPCM characterization was then correlated by defining size and shape distributions using density with complementary optical properties such as photolumi- gradient centrifugation approaches. Optical spectra and nescence (PL) and Raman spectral mapping measured at relaxation behaviors for different functionalization will be the same positions. The results showed clear correlations established. Development of simple thin film devices will which gives us insight as to how to design efficient gra- also be established as a platform for probing phto-induced phene-, and possibly CNT-based photovoltaics. Specifically, charge generation. Capacitive photocurrent spectroscopy regions of strong photocurrent intensity matched well will be used for determining the nature and evolution of with areas producing high PL, indicating that the origin of photogenerated charges. photocurrent is most likely the oxygen functional groups of GO. (It is known that oxidized regions of GO have strong Development and optical studies of oxygen-doped carbon PL.) Consistency between low PL intensity region with high nanotubes Raman intensity region also confirms that high PL intensity The ability to create oxygen doped carbon nanotubes will region has higher density of oxygen. Further investigations be established. Incorporation into new gel matrices for on the origin of high photocurrent in GO will be performed enabling certain types of optical measurements will also be using well-controlled device geometries which contains n- pursued. Of particular interest will be probing of exciton and p-doped regions. optical properties within these doped materials. Studies will include time-resolved, imaging, and temperature de- Carbon Nanotubes (CNTs) pendent measurements at both the ensemble and single- We have obtained unique optical properties of CNTs via tube levels. Initial focus will be on establishing migration controlled oxygen doping. By using ozonated water as an behaviors of dopant-induced states and how these interact oxygenating agent, new optical states with a clear red shift with optically generated excitons. Ultimately, incorpora- of photoluminescence (PL) were observed for CNT suspen- tion into electro-optic devices will also be pursued. sions compared to undoped samples. The shift was ob- served for both mixed chirality and chirality separated (6,5) Conclusion CNTs. The type of surfactant used to wrap and disperse Fine tuning of graphene/CNTs with well designed chemical CNTs in aqueous suspension was crucial in enabling the functionalization would open up a pathway for incorpora- doping chemistry to proceed. It is speculated that surfac- tion of these materials into large area, inexpensive, flexible tant dependent morphology of the wrapped CNT surface, energy harvesting electronics.Understanding their optically determined by the affinity of surfactants to CNTs, plays excited states and relaxation dynamics will clarify the com- an important role in achieving the appropriate density of petition between charge and energy transfer and recom- oxygen doping. Investigations on the nature of the shifted bination in these materials, which is crucial in determining optical state are underway by performing temperature their potential for photovoltaics. The results will provide dependent PL of chirality enriched (6,5) CNT suspensions, insight for design of novel energy harvesting devices as well as using single molecule spectroscopy to probe utilizing the unique optical properties of low-dimensional the single optical site induced by the oxygen doping. Our carbon nanomaterials, and makes an important contribu- temperature dependent measurements are indicating the tion to LANL’s mission on sustainable energy as well as to dopant sites act as trap sites for excitons, with a trapping the world’s energy needs. energy of about 250 meV. Single-tube and single-site im- aging and spectroscopy studies are laying the groundwork for further probing the trapping nature of these dopants and how it may lead to novel PL saturation behavior. Mag- neto-PL and electroluminescence (EL) on oxygen doped CNTs are also planned to gain insight into the nature of the sites and their role in improving CNT optical properties for energy harvesting applications. 290

Page 291

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project Enhanced Structural Robustness in Metal/Nonmetal Nanocomposites using Bio- inspired Structure Design Amit Misra 20120769PRD3 Introduction Development of scientific principles of designing materi- Nanocomposites of a metal (copper) and a non-metal als from the atomic scale for predictable performance (carbon or carbide or nitride) will be synthesized using at extreme conditions such as high stress, high radiation physical vapor deposition following bio-inspired design dose, etc also supports the mission of LANL’s MARIE and of sandwich structures. The crystallinity and morphology DOE-NE. of the non-metallic phase and the relative thicknesses of the two phases will be varied. The strength and crack Progress growth resistance will be measured using a variety of The primary focus has been on the synthesis of metal- nanomechanical testing such as hardness, stiffness, com- nonmetal nanocomposite systems following biologi- pression and notched bend tests. Scanning and transmis- cally-inspired design that would exhibit a considerably sion electron microscopy will be used to characterize the improved mechanical performance, beyond the limits structure and morphology of the constituent phases and imposed by traditional disciplines. To this end we have the interaction of defects such as dislocations and cracks studied the synthesis parameters for Cu-TiN (copper- with the metal / non-metal interfaces. titanium nitride), with the goal of a strong yet ductile material system. The metal (Cu) in this system is soft and The scientific impact of this research will be that, for the ductile, while the non-metal (ceramic TiN) is hard and first time, metal-carbon composites possessing a combi- brittle. Moreover unlike other metal-ceramic systems nation of high strength, high resilience and high fracture Cu has a low affinity towards nitrogen, and hence the toughness will be developed. While the current state- Cu-TiN interface is expected to be chemically sharp and of-the-art relies on the “small is stronger” paradigm to relatively weak in shear. increase the hardness of soft metals via refinement of grain size, this will push the frontiers in designing materi- Nanolayered composites of Cu-TiN have been synthe- als with unprecedented levels of strength and fracture sized using physical vapor deposition (PVD). Most of toughness. the effort in this year till date has been concentrated on developing high quality Cu-TiN multilayer systems by ad- Benefit to National Security Missions justing the deposition parameters such as temperature, The proposed work is directly aligned with the grand substrate bias, etc. Current results suggest that a combi- challenges and scientific issues outlined in the BESAC nation of high temperature (300 C) and high bias on the report of DOE, Office of Science titled “Directing matter TiN can generate low-porosity dense multilayer films. and energy: five grand challenges for science and the Testing of such miniaturized samples (thicknesses in imagination” and the Basic Research Needs (BRN) work- the micrometer range) poses significant challenges. shop series, specifically: Advanced Nuclear Energy Sys- Among the experimental techniques available at these tems (ANES); and Materials Under Extreme Environment length scales, nanoindentation, with its high resolution (MUEE). Some specific scientific questions from these load and depth sensing capabilities, shows the greatest reports that are pertinent to this LDRD project are: How promise due to its ease of experimentation and versatil- do we make hard matter that heals damage or defects? ity. Towards this end we have utilized a novel spherical and How mechanically strong can we make materials yet nanoindentation analysis technique which is able to keep them light-weight? transform the raw load-displacement data into meaning- ful indentation stress-strain curves (this technique was originally developed by the postdoc fellow as part of his 291

Page 292

graduate work). More specifically, the use of these inden- nanoindentation. The crack growth studies, including in tation stress-strain curves makes it possible to analyze the situ experiments in a scanning electron microscope, will be initial loading segments of spherical indentation (from the focus of the work in the second year. shallow indentation depths of 50-100 nm). Thus the char- acterized property (e.g. indentation yield) corresponds to Conclusion the intact material at the indentation site, and can be used The key expected results are (i) correlation between the to differentiate between inherent differences in the local synthesis parameters and the crystallinity, scale and mor- material structure at the indentation site. This has enabled phology of the non-metallic phase in the metal matrix, and the measurement of local mechanical properties including (ii) mechanical properties (strength, stiffness, toughness) the loading and unloading elastic moduli, the indentation for the different morphologies of the nanocomposites. yield points, as well as some of the post-yield character- Collectively, these results will provide the scientific prin- istics. Work in the first year has been focused on validat- ciples to design metal / non-metal nanocomposites that ing this system on metal-metal (Cu-Nb) interfaces before possess unique combination of strength and toughness not proceeding to the more complex metal-nonmetal systems. achieved in any existing engineering material. Significant progress has also been made in designing another novel micro-mechanical testing system – that of compression testing of micro-pillars containing these multilayered nano-composites. Such micrometer sized pil- lars are typically produced by removing material around a selected region of interest using a focused-ion beam (FIB). However this technique typically requires tremendous resources in terms of sample preparation and operator time, due to the large amount of material that needs to be removed by FIB. Work in the first year has focused on reducing the sample preparation time needed for such an effort. This was achieved with the use of lithographic poly- mer photomasks during the PVD process. Etching off the polymer after the deposition process leaves only the mul- tilayers deposited in a predetermined pattern. Since the amount of material needed to be removed by FIB is now significantly decreased, this reduces the sample prepara- tion time by more than a third, and makes bulk production of micro-pillars possible. Future Work Using physical vapor deposition capabilities in the Center for Integrated Nanotechnologies (CINT) at LANL, nanocom- posites foils of a metal such as copper and a non-metal such as carbon or carbide or metal nitride will be made. Deposition parameters such as deposition rate, tempera- ture and relative thickness of the metal and non-metal layers will be controlled to produce either amorphous or crystalline form of carbon in the form of continuous or discontinuous layers in the copper matrix. The as-depos- ited foils will be vacuum annealed to further change the structure, e.g., crystallize the amorphous carbon or change morphology from continuous to island-like or grow the island-size or improve crystallinity for nitride layers. The as-deposited and annealed foils will be characterized using scanning and transmission electron microscopy to eluci- date the morphology and crystallinity of the nano-phases. The initial nanomechanical characterization will include hardness, stiffness and compressive flow strengths using 292

Page 293

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project Nanoscopic Recrystalization Dynamics and Phase Behavior in Liquid Crystals and Polymers Juan Duque 20120770PRD3 Introduction Benefit to National Security Missions In photothermal processes, photons of light are ab- These projects will introduce a new capability to the sorbed and transferred into thermal energy. This is an laboratory called photothermal heterodyne imaging excellent technique for metallic nanoparticle detec- technique. This technique is a highly sensitive method tion, since it exploits the weaker size dependence on to probe temperature effects at the nanoscale to investi- nanoparticle absorption over Rayleigh’s scattering. A gate the recrystallization dynamics and the glass transi- laser in resonance with the nanoparticle plasmons is tion temperature of different materials important in the used to obtain an optimal photothermal effect. An imag- developing of new meso-materials for energy conver- ing technique that takes advantage of the photothermal sion, photovoltaics, and sensing in relevance to the DOE, effect is the photothermal heterodyne imaging (PHI) DOD, DHS, and NIH. technique. A probe beam with minimal nanoparticle absorption is used to measure the induced change in Progress the index of refraction by a modulated heating beam. We are studying phase transitions at the nanoscale, This technique has been used to image extremely small focusing on transitions that can be probed through tem- nanoabsorbers, even single molecules, much smaller perature changes, such as thermotropic liquid crystalline than the wavelength of light. phase transitions, polymer glass transitions, and melting transitions. Originally, we planned to study these transi- This unique effect offers the ability to heat the nanopar- tions only through a photothermal imaging technique ticle matrix at the nanoscale, which allows us the op- that we brought to Los Alamos National Laboratory. In portunity to answer some very interesting questions. brief, photothermal imaging is a technique that allows The PHI technique is able to detect changes in refrac- nanoparticles to be visualized when they absorb photons tive index and thus it was able to determine that at the and convert the energy to thermal energy. The thermal nanoscale, the nematic-to-isotropic phase transition energy creates a scattering field that scatters a non- occurs at a lower temperature. The exact reason for this absorbing laser. This thermal energy is highly localized depression is due to a combination of interfacial tension surrounding the nanoparticle, thus allowing a thermally and confinement. However, since this is a liquid crystal- induced phase transition to occur nanoscopically in line system, the nanoparticles must be fixed to the sur- the materials just surrounding the material. With this face, thus the transition is dependent on the interactions technique, we can monitor any changes in the critical of the liquid crystal with the surface. A glass transition temperature, or energy, needed to induce the phase temperature can remove the surface effects. Nanopar- transition and can imply as to the nature of the sur- ticles can be suspended and fixed within a polymer at rounding material, for example the local alignment and temperatures below the glass transition temperature interactions with the nanoparticle. In the past twelve and then heated to initiate the transition. Thin film months, we have made headway in improving our ability work suggests there should be a large depression in this to understand the nanoparticle interactions with the transition temperature; however, thin films have both surrounding material by introducing two new charac- a substrate and air interface. This information may lead terization techniques that are superimposed on the to better theories and more accurate parameters for mo- same diffraction limited spot. Fluorescence is extremely lecular interactions during a phase change. sensitive to environmental changes, thus it will be used to probe fluorescent molecule near the surface of the 293

Page 294

nanoparticles as the phase transition occurs. Lastly, Raman spectroscopy will be used to probe the local alignment as a function of temperature during the phase transition, which will give us better understanding of the recrystallization dynamics. Future Work • Development and optimization of a new experimental technique called photothermal heterodyne imaging (PHI) capable of imaging samples down to 1nm. fre- quency dependent experiment will be done at tem- peratures near the phase transition. Lower frequencies will allow more time for the liquid crystal to recrystal- lize and will result in greater signal. The multiple laser sources available at LANL (from 275 to 1000 nm) will increase the sensitivity of the measurement since the noise as a function of frequency is wavelength depen- dent. • We will establish for the first time an approach to control/monitor the change of index of refraction with temperature, at the liquid crystalline phase transi- tion, in order to enhance the PHI sensitivity. The high sensitivity of the PHI technique offers a low-powered method to probe the phase transition dynamics of nanomaterial with minimal external interactions. Pre- liminary results will provide information on the ther- motropic liquid crystalline phase transition dynamics. Conclusion The photothermal heterodyne imaging technique is a very useful tool to visualize individual nanoabsorbers for a wide range of applications, such as protein tracking, nanoparti- cal characterization, and nanoparticle localization in three- dimentional environments. We aim on determining the recrystallization characteristic time in liquid crystals as the temperature approaches the phase transition temperature and the role surface interactions play on the phase transi- tion temperatures in liquid crystals and polymers. 294

Page 295

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project The Fate of the Three Dimensional Electron Far Beyond the Quantum Limit Ross D. Mcdonald 20120772PRD4 Introduction surface states. To date these surface states have been We plan to carry out detailed study of magnetoresis- exceedingly hard to observe largely because of chemical tance and/or magnetic susceptibility measurements stability and environmental exposure of the surface dur- on semimetal graphite and bismuth in a pulsed field ing measurement. Zengwei’s high field measurements magnetic field up to 100 T as the angle between applied on sulphur doped bismuth tellurium reveal this to be a field and the crystalline axis is varied. The high quality significantly more stable surface state with significantly samples are commercially available. At low temperature, higher electronic mobility than observed in other TI magnetoresistance and/or magnetic susceptibility easily materials. Furthermore field orientation dependance track quantum oscillations when the Landau levels pass reveals the quantization of the Hall voltage to originate the Fermi level and abnormal transition. Single-particle from the 2D surface state. A high profile publication theory without coulomb interaction will be utilized to manuscript is in preparation. Further pulsed field mea- compare with our results for two materials. surements are underway aimed at employing a combi- nation of electric and magnetic fields - i.e. using a gate voltage to tune the surface carrier density whilst probing Benefit to National Security Missions the carrier density with the orbital quantizing effect of the magnetic field. Exploration of low carrier density metals in extremely high (pulsed) magnetic fields and at low temperatures Zengwei has also undertaken pulsed magnetic field is directly relevant to MaRIE because of the need to de- torque magnetometry measurements to investigate the velop and refine high-frequency conductivity probes at physical properties of related strong spin-orbit materi- high data acquisition rates compatible with the sub-mi- als that are too insulating for restively measurements crosecond duration of transient events in these extreme in high magnetic fields. This is an on going activity in environments. collaboration with the University of Cambridge that is yielding very interesting results. Furthermore, investigating the novel states of three dimensional metals when extreme magnetic fields drive them beyond the quantum limit will increase our un- Future Work derstanding of the fundamental interactions between Apply existing measurement techniques electrons in metals, which may impact other lab mis- (magnetization,magnetic torque,conductivity DC-rf sions who seek a more fundamental understanding of frequencies) to the measurement of low carrier density strategic metals (c.f. unknown band structure and Fermi metals in 65T (short) pulsed magnetic fields. surface of plutonium). Refine these techniques for the particular challenges of measuring low carrier density samples e.g. high restivity Progress and large magnetoresistance. During the last year Zengwei has extended his proposed Implement these improvements in measuring the mate- study of low carrier density three dimensional metals rials to the highest available non-destructive magnetic in pulsed magnetic fields to the topologically non-trivial fields using the 100 T multi-shot magnet system. metallic material bismuth tellurium. This material system Publish results in high impact journals. is an important example of a topological insulator system Be invited to give high profile talks. that has been proposed to support two dimensional 295

Page 296

Conclusion Within two years, we will finish detailed study of tempera- ture- and angle-dependent phase diagram in bismuth and graphite up to 100T. We will uncover the nature of intriguing transitions from strongly correlated electrons far beyond QL. Our experimental results will contribute to better theoretical calculations. Collaboration between experiment and theory would significantly improve our understanding of the fate of the 3D electron gas beyond the QL. This research could shed light on how electrons interact. Based on this, we can extend comprehension to other more complicated strongly correlated systems such as heavy fermion systems, high Tc superconductors etc. Publications Lin, X., Z. W. Zhu, B. Fauque, and K. Behnia. Fermi Surface of the Most Dilute Superconductor. 2013. PHYSICAL REVIEW X. 3 (2): -. Zhu, Z. W., A. Collaudin, B. Fauque, W. Kang, and K. Behnia. Field-induced polarization of Dirac valleys in bismuth. 2012. NATURE PHYSICS. 8 (1): 89. Zhu, Z. W., B. Fauque, L. Malone, A. B. Antunes, Y. Fuseya, and K. Behnia. Landau spectrum and twin boundaries of bismuth in the extreme quantum limit. 2012. In PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA. Vol. 109, 37 Edition, p. 14813. 296

Page 297

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project Access to Industrially Important Chemical Compounds by Direct Hydrogenation of Biomass Derived Molecules Containing Carbonyl Fragments John C. Gordon 20120774PRD4 Introduction industries. This area however remains a significant chal- The reduction of carboxylic acids has relied mostly on lenge. In particular, the hydrogenation of ureas, organic the (stoichiometric) use of reagents such as LiAlH4, carbonates, carbamates and formates is of significant NaBH4 and their derivatives (i.e. forcing conditions). global interest since these classes of compounds can Recent progress in the area of homogeneous, metal- be produced from CO2 and CO. Milder hydrogenations catalyzed hydrogenations of carboxylic acids and their applicable to CO2 and CO themselves could also pro- derivatives has been made however. Appropriately vide alternative routes to methanol, an important fuel designed bifunctional catalysts have demonstrated util- and synthetic building block in its own right. Success in ity in the area of polar functional group hydrogenations, this chemistry could therefore have significant positive as their metal-ligand cooperation effects promote mild impact for LANL, vis-à-vis the application of sustainable activation of small molecules (in this case H2). Bifunc- approaches to new polymers and materials synthesis, in tional catalyst promoted hydrogenations using H2 are energy applications, and in health related (pharmaceuti- now realized to be a practical tool in synthetic organic cal) R&D. chemistry in both academia and industry. The industrial outlook for this approach to hydrogenation is bright Progress because of its operational simplicity, scope, economic vi- In the last few decades, bifunctional molecular catalysis ability, and the growing awareness of the need for green based on metal–ligand cooperativity has been developed chemistry approaches. and has become a general strategy for effecting highly efficient molecular transformations in organic chemistry. This research project will thus specifically target the Originally developed for the asymmetric transfer hydro- rational design of new bifunctional metal complexes (M genation of ketones, metal–ligand bifunctional catalysts = Ru, Ir) capable of mild, selective and direct hydrogena- (in which non-innocent ligands directly participate in tion reactions of less electrophilic carbonyl groups. Par- substrate activation) are also now applicable to the ticular emphasis will be placed on these molecules with hydrogenation of less electrophilic polar functionalities. respect to their ability to hydrogenate cheap, biomass- We are focusing on the synthesis and characterization derived platform chemicals containing polar carbonyl of new earth abundant metal complexes supported by moieties into more useful (and higher value) molecular mixed donor (multidentate) ligands that are capable of building blocks. The hydrogenation of esters bearing promoting the hydrogenation of polar substrates. alpha-fluorine atoms will also be pursued, as their hydro- Although the project has just recently begun, we have genation products are of significant importance within already synthesized and characterized a series of ligands the pharmaceutical industry. containing different phosphorus, nitrogen and sulfur donor arms. These ligands will now be used to construct Benefit to National Security Missions catalysts containing metals such as iron, ruthenium, The development of new methods for mild and selective cobalt and copper. Catalysts will then be tested for their hydrogenations of polar functionalities contained within propensity to hydrogenate a variety of of carbonyl con- cheap and widely available biomass derived compounds taining fragments, notably esters, carboxamides, ureas would provide routes to highly desirable building blocks as well as potentially to less-electrophilic CO2 as a final for the pharmaceutical, polymer and optoelectronic target. 297

Page 298

Future Work We will utilize calculations to guide the design, synthesis and reactivity of various new bifunctional metal complex- es (M = Ru, Ir) capable of mild, selective and direct hydro- genation reactions of polar carbonyl containing moieties such as esters, carboxamides, and carbonic acid deriva- tives. These are important classes of biomass derived compounds, often proving difficult to hydrogenate. Since biomass-based routes are expected to make a significant impact on the production of bulk chemicals within 10–30 years, these studies will provide us with a deeper under- standing of the factors involved in hydrogenating these classes of compounds into other useful chemical building blocks. Conclusion The development of new methods for mild and selective hydrogenations of polar functionalities contained within cheap and widely available biomass derived compounds would provide routes to highly desirable building blocks for the pharmaceutical, polymer and optoelectronic indus- tries. This area however remains a significant challenge. In particular, the hydrogenation of ureas, organic carbonates, carbamates and formates is of significant global interest since these classes of compounds can be produced from CO2 and CO. Milder hydrogenations applicable to CO2 and CO themselves could also provide alternative routes to methanol, an important fuel and synthetic building block in its own right. 298

Page 299

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project Stimuli Responsive, Functional Biopolymers: Quinic Acid-based Polymers and Their Assemblies Hsing-Lin Wang 20130778PRD1 Introduction novel strategies for the design and synthesis of nanoas- This project aims to develop functional biomolecules semblies with adaptive control to enable the materials with emergent (mechanical, optical, and electronic) with immense technological relevance in areas of optics, properties that are not accessible through conventional electronics, photovoltaic, and biomedical devices that synthetic methods. The approach uses molecules that impact our missions in energy security and threat reduc- are known to form polymers compatible with biologi- tion. cal systems with side chains that are active to which functional groups can be attached. Depending on the Progress distribution and density of these functional groups on The Goal of this project is to investigate the develop- the side chains, a wide spectrum of mechanical and ment of functional biopolymers in which the existing optical properties, and electronic structures can be ac- functionalities can be coupled with oligomer, dye or cessed. Our goal is to exploit the emergent properties, other complex biomacromolecule for sensing and elec- and through the analysis of structure-property relation- tronic applications. We chose to use quinic acid (QA) as ships, obtain valuable insight regarding the design of a monomer to build the functional biopolymers due to biopolymers with enhanced functionality that can be its natural occurrence, chirality, poly functionality and used toward sensing, bioimaging, drug delivery, optical, well-established chemistry. In this project, we proposed electronic and energy applications. Such an approach the development of QA-based biopolymers to explore has not yet been studied in a systematic manner mainly their properties/responses against various stimuli such because multidisciplinary knowledge ranging from or- as temperature, pH etc. ganic synthesis to materials chemistry to biochemistry is needed. We synthesized the intermediates to obtain the polym- erization precursors. It is a ten-step synthesis to prepare The challenge is to develop synthetic chemistry routes the polymerization precursor. We have completed 3 that are facile and effective. Further, we must control steps out of total 10steps. Each molecule is synthesized, the distribution of functional groups on the side chains purified using silica gel column chromatography and in order to render desired properties. Dr. Pradeep carefully characterized using NMR techniques to ensure Cheruku is a researcher who has the unique combination that we have obtained the correct product with the of skills and knowledge to execute the project and offers desired purity. the best opportunity for success. Our plan is to conjugate these polymers with various Benefit to National Security Missions fluorophores such as conjugated oligomers and other We expect that these new classes of biopolymers may fluorescent dyes and follow their optical properties as exhibit high temperature thermal transitions and strong we expect these optical properties are the function of mechanical properties due to the high degree of hydro- the stimuli applied on these polymers. Toward this aim, gen bonding and rigid carbocyclic backbone. We hope to we wanted to develop new fluorophores derived from show the correlation between nanoassembly structures biomolecules, amino acids, such as tyrosine and argi- and their associated properties to obtain valuable in- nine. In this context, I have synthesized five analogs of sights on the interplay between structure, dynamics and tyrosine derived fluorescent conjugated oligomers and functions of the materials. We also expect to develop studied their optical properties. These molecules were 299

Page 300

synthesized starting from the iodination of the natural explore their properties against various stimuli such as amino acid tyrosine. Corresponding diiodo tyrosine was temperature, pH etc. employed in palladium catalyzed Heck-coupling reactions with various electron – withdrawing and – donating groups The copolymer syntheses are carried out by copolymeriza- at the para-position of the aryl ring. This allows us to fine tion of methacrylate conjugates of QA (QA monomer and tune the electronics and the dipoles of the molecules chromophore in the presence of a radical initiator like AIBN which play a crucial role in the formation of the aggregates to synthesize QA-biopolymer on a poly(methacrylic acid) which in turn directly impacts their optical properties. All backbone). The FITC dye on the polymer acts as a spectro- products were purified using silica gel column chromatog- scopic reporter on the state of the polymer conformation. raphy. Characterizations and purity assessments were done Conformational changes within the polymer trigger the by using NMR and HPLC (for water soluble molecules). arrangement/proximity of the fluorophores to put them These newly developed fluorophores bear amine (-NH2) in different environments and thus lead to a modulation and carboxylic acid (-COOH) functionalities as they are of fluorescence. The hydroxyl groups on the QA of the derived from natural amino acids. These functionalities polymer can be further functionalized to their correspond- will be used to conjugate to the polymers. Currently, my ing carboxylic acid(anionic)/amine(cationic) groups. This efforts are focused on the optimizing the bioconjugation may lead to pH responsive biopolymers which can shrink using the model molecules before we actually couple them or swell at certain pHs. to the desired polymer. Interestingly, these newly devel- oped fluorophores exhibit properties such as PH-sensitivity Conclusion (shows different emission colors under acidic and basic We expect synthesis of a new class of biopolymers with environments), redox-sensitivity (fluoresense is quenched thermal transitions and strong mechanical properties due upon oxidation and regenerated upon reduction) and also to the high degree of hydrogen bonding and rigid carbocy- solvatochromisam in different solvents. I am now compil- clic backbone. We hope to show the correlation between ing all these interesting properties and writing up a decent nanoassembly structures and their associated proper- manuscript for the publication. We are to expecting submit ties to obtain valuable insights on the interplay between our first manuscript in this project soon. structure, dynamics and functions of the materials. We also expect to develop novel strategies for the design and We have also synthesized one analog derived from amino synthesis of nanoassemblies with adaptive control to en- acid arginine. This analog is structurally similar to the tyro- able materials with immense technological relevance in sine derived molecules except the hetero atoms (nitrogen) areas of optics, electronics, photovoltaic, and biomedical with in the conjugated backbone. Due to the presence of devices that impact our missions in energy security and the nitrogen atoms in the conjugated backbone, we are threat reduction. expecting these molecules to show different optical prop- erties. These molecules are synthesized by following an entirely different synthetic route. Arginine is treated with the acetyl acetone to construct the core nitrogen-contain- ing aromatic ring bearing activated methyl groups. Upon treating with a strong base, these methyl groups undergo aldol-type condensation with the benzaldehyde to give the corresponding conjugated fluorophore. Again, the alde- hyde can be varied to obtain structurally different fluores- cent molecules. We are currently exploring the synthesis of various analogs. Future Work We propose to investigate the development of functional biopolymers in which the existing functionalities can be coupled with oligomer, dye or other complex biomacro- molecules for sensing and electronic applications. We also propose the use of quinic acid (QA) as a monomer to build the functional biopolymers due to its natural oc- currence, chirality, polyfunctionality and well-established chemistry. We plan to develop QA-based biopolymers and 300

Page 301

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project NMR Study of Quantum States of Matter Joe D. Thompson 20130780PRD1 Introduction Benefit to National Security Missions There are thousands of examples of a transition at finite Unconventional superconductivity holds promise for temperature from a magnetically ordered state to a mag- exciting new energy and defense technologies, but how netically disordered state. At such a classical phase tran- and where that superconductivity might appear cannot sition, thermal fluctuations due to the finite temperature be predicted. Empirically, experiments have shown that environment drive the transition. But, it also is possible this superconductivity frequently appears as a magnetic theoretically to have a transition from magnetically or- transition is tuned to zero temperature, but the nature dered to disordered states at absolute zero temperature of the quantum excitations and how they might induce where there are no thermal fluctuations. In this case the superconductivity are major unanswered questions. transition is driven solely by rules of quantum mechan- Fundamentally new understanding of the mechanism of ics. Though theoretically possible, there are a growing unconventional superconductivity and its possible rela- number of experiments suggesting that such a quantum- tionship to quantum fluctuations is needed to be able to driven transition exists, but none of these examples can predict how and when superconductivity appears. Suc- be explained completely with existing theory. Further, cess in this project will provide basic new understanding experiments during the last decade have found that an of materials and how quantum fluctuations induce new unconventional form of superconductivity tends to ap- quantum states of electrons. With unconventional su- pear in the vicinity of a quantum magnetic transition, but perconductivity also found recently in plutonium-based the possible relationship between quantum fluctuations materials, superconductivity itself can be used to inform at this zero-temperature transition and unconventional a much clearer perspective of states of matter than can superconductivity remains an outstanding question. arise from the electronic complexity of Pu. Nuclear magnetic resonance (NMR) techniques are a Progress particularly powerful probe of the nature of quantum Postdoc hire-on date was delayed unexpectedly until July fluctuations and their possible role in producing uncon- 1, 2013. ventional superconductivity and will be used to study a prototypical example CeRhIn5 whose magnetic transition Future Work can be tuned by applied pressure toward zero tempera- Nuclear magnetic resonance (NMR) techniques will be ture where superconductivity emerges. The role of inten- used to explore and understand the nature of quantum tionally introduced disorder on the nature and effect of fluctuations that evolve from a zero temperature mag- quantum fluctuations also is an important open question netic transition. The compound CeRhIn5 is an excellent that will be addressed by systematically replacing Ce by example of a system whose magnetic transition can be La and In by Sn in CeRhIn5. tuned by pressure to a zero temperature and will be used as a prototype for testing theoretical models of the These experiments will provide critical tests of theoreti- signatures for and consequences of quantum fluctua- cal models of quantum phase transitions, which ulti- tions for the formation of electronic states from them. mately should enable a predictive methodology for the discovery of superconductors for energy conservation Conclusion applications. This project addresses the fundamental, unanswered question of what happens when a magnetic transition 301

Page 302

is tuned to absolute zero temperature. One possibility is that quantum fluctuations of the magnetic order at zero temperature produce an attractive interaction that creates an unconventional form of superconductivity or that these fluctuations induce some new, unanticipated quantum state. Success in this project will make significant strides in advancing our understanding of the consequences of quantum fluctuations that arise out of a zero-temperature magnetic transition.. 302

Page 303

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project Efficient Carbon Nanotube Growth on Graphene-metal Surfaces Stephen K. Doorn 20130785PRD2 Introduction multi-ferroic materials of tecnological interest. Interac- The project goal is to optimize the synthesis conditions tions with nanoelectronic materials such as nanotubes that promote the catalytic conversion of graphene into can introduce emergent electronic behaviors that may high yield few-walled (1-2 layers) carbon nanotubes enhance the properties of interest. (CNTs). The approach will overcome typical drawbacks of nanotube synthesis that include use of harsh chemical Benefit to National Security Missions treatments of growth substrates and include the need The work is directly relevant to DOE Office of Science in- for multi-step purification processes. These challenges terests in the fundamental science of and development will be overcome by using graphene sheets as an in- of multifunctional nanoscale materials. The work will novative new substrate for growth that allows one-step also probe the fundamentals of carbon nanomaterial in- synthesis of composite materials. Growth can also be at- teractions with other materials, also of interest to DOE/ tained at low-temperatures (~200 degrees C vs the typi- SC. The nanotube/graphene materials are of significant cal 800 degrees or higher) which is essential for enabling interest for energy storage and energy harvesting, giving development of composites based on multiple materials the work mission relevance for both energy security with very different ranges of temperature compatibility. and the environment (climate and energy impacts). The effort is also directly related to development and basic Another innovative aspect is to use dip-pen lithography understanding of nanomaterials with ultimate impact on capability for fine-control over placement and morphol- electronics, sensing, and imaging needs as well. ogies of metallic nanoparticle growth catalysts. Optimi- zation of our growth processes will require variation and Progress study of the influence of multiple growth parameters. To date limited progress is available as the postdoc has These will include modulation of catalyst particle size, only been at Los Alamos for one month. However, sig- composition, and placement. nificant progress has been made in preparing the labora- tory space for both graphene and nanotube synthesis. A We will also study the effects of variable temperatures new IWD covering the expanded scope of the nanotube and reactant gas composition and flow rates. The resul- synthesis work is nearly in place. The growth furnaces tant synthesis will directly provide nanotube/graphene have been put in place and Enkeleda has already gener- composites that will be studied for their energy harvest- ated some large area graphene films that will be used for ing and storage potential. Low temperature growth generating multi-component hybrid structures. processes are anticipated to also allow the use of gold nanoparticles for growth as well, which will be the key Future Work to enabling development of nanotube-nanowire hybrids The goal is to identify the synthesis conditions required (not yet attained) of interest for optical and electronics for high yield few-walled carbnon nanotube (CNT) needs. growth on graphene-metal surfaces using a chemical vapor deposition approach. Few-layer and large area The knowledge gained in these first synthesis directions graphene sheets will be decorated with various types will also allow us to taylor synthesis conditions to pursue of metal nano-particles (nps) and their deposition will growth of other novel composite materials including car- be accomplished using two different methods; dip pen bon nanotube hybrids with complex metal oxides. Such nano-lithography (DPN) and a simple route involving a oxide thin films form the basis for superconducting and 303

Page 304

mixture of the metal nanoparticle solution with graphene. Using DPN to precisely control the arrangement and the number of nps on the graphene sheet is very innovative and essential since the concentration of nps is known to af- fect the hydrogen/carbon ratio in the reaction mixture and nanotube formation. The influence of both techniques on the size and dispersion of nps, which are important factors in controlling the nanotube diameter, will also be investi- gated. Given the complex nature of the growth kinetics of nano- tubes with controlled wall numbers, the catalytic conver- sion of graphene into single-wall or double-wall CNTs will be a breakthrough in nature. Optimization of the synthesis conditions for few-walled CNT growth with controlled properties will be achieved by varying the following: catalyst size and composition (combination of two differ- ent metals at various loadings), reaction temperature, and type/flow rates of gases. To increase the single-wall nano- tube yield, while preventing Ostwalt ripening, water vapor will be added in the CVD set up. Performing in-situ TEM ex- periments will be a crucial step towards understanding the complex role of graphene during the catalytic nanotube growth. Microscopy (TEM, SEM, AFM) and Raman spec- troscopy analysis will be performed to study the morpho- logical properties of nps and realize their interaction with the graphene surface during the nanotube formation. Conclusion The primary technical goal is to develop novel synthesis approaches to improve the optical and electronic proper- ties of few-walled (targeting 1-2 layers) carbon nanotubes. The methods we anticipate using will also result in genera- tion of functional hybrid carbon nanomaterials of interest for energy storage, electronic, and multi-ferroic applica- tions. These may include hybrdi nanotube/graphen mate- rials, nanotube/nanowire heterojunctions, and nanotube composites with complex oxide thin films. 304

Page 305

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project Hybrid Nanostructures for Photoreduction of CO to Hydrocarbons 2 Hongwu Xu 20130787PRD2 Introduction be determined using high-temperature calorimetry. The The goal of this research is to develop novel hybrid goal is to determine the synthesis-structure-property- nanostructures for photocatalytic conversion of carbon stability relations, which will then be fed back to the dioxide plus water into hydrocarbons. Semiconductor- design of new nanostructures with enhanced properties based nanomaterials are promising materials that may and stability. enable this catalytic process, as they can couple single photon events with accumulation of multiple redox Benefit to National Security Missions equivalents upon CO2 reduction. However, most of the This project represents a research topic with great nanomaterials studied are single-component systems, scientific value and tremendous potential for techno- and their conversion efficiencies are too low for practical logical applications. Through controllable synthesis and application. In addition, these nanoparticles often suffer systematic characterization of core/shell nanomaterials, from instability due to light-induced anodic oxidation in the research is expected to lead to design and devel- aqueous solutions. Hybrid nanostructures composed of opment of a novel class of hybrid nanostructures for core/shell nanoparticles may have effectively separated efficient photocatalytic conversion of carbon dioxide charges and catalytic sites for enhanced conversion and plus water into hydrocarbon fuels and chemicals. The also possess good stability at relevant conditions. synthesis-structure-property-stability relationship to be determined in this research will lay the foundation for The core/shell nanoparticles contain a semiconductor the ultimate development of solar-to-fuel conversion core and a metal oxide shell, linked to a nanoparticle devices and have broad implications for rational design catalyst. An example is ZnTe/ZnO core/shell linked to Pt. of functional materials in general. This project is tied The core/shell nanostructure serves for initial photo-in- to missions in DOE office of science, fossil energy and duced charge separation, while the metallic component renewable energy as well as NSF, and to LANL’s Materials provides catalytic sites for subsequent chemical conver- Grand Challenge in the central themes of defects/inter- sion. We will take the advantage of high reducing ability faces and emergent phenomena. of photo-generated electrons in the semiconductor core to produce high hydrocarbons (e.g., ethylene & propyl- Progress ene) in high yields. Coating the nanoparticle surface with No progress at the time of data sheet request. The post- a metal oxide shell not only protects the nanoparticle doc is scheduled to hired on July 1, 2013. from dissolution during the photoreaction, but also facilitates charge separation by forming a barrier for Future Work controlled back electron transfer. During the first year of this project, we will design, syn- thesize and characterize several hybrid nanostructures The project involves controllable synthesis of hybrid of photoactive materials that have both effective charge nanostructures with wet-chemical methods, followed separation and high photo-driven reductive properties by characterization of their structures, morphology and for efficient photocatalytic CO2-to-hydrocarbon conver- compositions using XRD, SEM, and TEM. The charge sion. Our synthesis includes three steps: 1) A wet-chemi- separation and transfer properties will be measured cal synthetic approach will be employed to prepare ZnTe by transient visible-pump/visible probe and EPR spec- nanoparticles with various sizes, shapes and structures; troscopes, and photoreduction of CO2 with a photo- 2) A layer of ZnO or TiO2 will be coated on the surface of electrochemical system. Thermodynamic stability will 305

Page 306

ZnTe nanoparticles to form core/shell structures through chemical methods; and 3) The core/shell nanostructures will be linked to Pt or RuOx catalytic nanoparticles through a successive growth approach. The structure, morphology and composition of the obtained nanostructures will be characterized using powder X-ray diffraction, scanning and transmission electron microscopy. Transient absorption spectroscopy will be used to measure their charge separa- tion and transfer properties. Thermodynamic analysis will be conducted using high-temperature solution calorimetry and/or differential scanning calorimetry. The obtained results will allow revelation of the synthesis-structure- property-stability relationship, which will then be fed back to the design of new hybrid nanostructures with enhanced photocatalytic properties and stability. Conclusion Successful execution of this research will lead to design and development of a novel class of hybrid nanostructures for efficient photocatalytic conversion of carbon dioxide plus water into hydrocarbon fuels and chemicals. Through controllable synthesis and systematic characterization of the core/shell nanostructures, we will determine the synthesis-structure-property-stability relationship for this family of nanomaterials, which have implications for ratio- nal design of functional materials in general. The obtained results will lay the foundation for development of solar-to- fuel conversion devices and directly address LANL/DOE’s grand challenges in energy and environmental materials research. 306

Page 307

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project Alternating Positive-negative Charge Systems: New Compounds and Synthetic Routes David E. Chavez 20130788PRD2 Introduction • Attempt synthesis of compound 1 This project will focus on the development of alternating • If synthesis is successful, characterize compound 1 charge molecules for energetic materials applications. • Attempt synthesis of compound 2 These alternating charge molecules are predicted to have very promising energetic material performances • If synthesis is successful, characterize compound 2 properties, and may have very interesting safety proper- ties. There exists a paucity of examples of these types Conclusion of materials described in the literature, and therefore While the high thermal stability of APNC systems is the basic scientific understanding and principles for known, little is known about their mechanical sensitivi- synthesizing these materials required. The materials will ties and the other molecular features affecting same. A be very challenging to construct, however, due to their fundamental understanding of these properties must be predicted performance properties, there is high poten- developed in order to design explosives of high perfor- tial for opening up a new area of research in the area of mance and stability. The potentially unprecedented energetic materials. energetic performances of compounds of these types would lead to revolutionary effects in energetic systems Benefit to National Security Missions as a result of higher energetic yields from smaller-sized The project focuses on the development of novel ener- devices, as well as new fundamental insights into factors getic materials with unprecedented performance, and affecting molecular stability. potentially, safety. This topic is directly relevant to NNSA defense programs, science campaign 2 and to the DoD. Because the project will develop a fundamental under- standing of the proposed materials, it will have a direct impact on expanding fundamental chemistry concepts. Progress This project started in September 2013; therefore, there is no progress to date. Future Work The project will begin with the synthesis of novel tetra- zine dioxides. The chemistry and synthetics routes to these materials will be the focus of the first year of the project. The tasks for the first fiscal year include: • Propose two new tetrazine dioxide molecules for study • Perform predictive modeling calculations on these materials • Propose synthetic routes to the materials 307

Page 308

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project Graphene Quantum Dots for Carrier-multiplication-enhanced Solar Cells Victor I. Klimov 20130790PRD2 Introduction of graphene-based nanostructures, this project may lead Quantum dots (QDs) are nanometer-sized crystals of to a paradigm shift in research on advanced CM-based semiconductors that offer many unique optical and solar cells. electronic properties of great interest to use in solar cells. One such property is high-efficiency carrier mul- Benefit to National Security Missions tiplication (CM) whereby absorption of a single pho- This work is a fundamental study of a relatively new ma- ton produces multiple electron-hole pairs, potentially terial class that has been predicted to have properties of increasing the current produced in solar cells. To date, great relevance to optoelectronic devices in general, and the measured efficiencies of CM in inorganic QDs are not solar cells in particular. Specifically, we will produce and high enough to impact the power output of real devices. study nanometer-sized pieces of graphene. Graphene is In this project, we will explore the potential of graphene- a form of carbon consisting of individual sheets of gra- based nanomaterials. phitic carbon. Interest in graphene for use in electronic devices has exploded over the past decade, but studies Graphene is a very strong absorber of light, and is highly of nano-graphene pieces, or graphene quantum dots (in conductive. Furthermore, recent theoretical calcula- which the effects of quantum-confinement will produce tions have shown that CM yields in graphene can greatly very different properties) are still only at the starting exceed those in inorganic solids. However, ordinary point. Our focus will be on studying and optimizing those graphene cannot serve as an absorber in a solar cell, properties of nano-graphene of direct relevance to ap- because it is a “semimetal” that has no band gap, which plications in solar energy capture and energy efficient means it cannot produce a voltage in a device. How- solid-state lighting. In this way, our work is directly tied ever, it is possible to open the band gap by introducing to the LANL Science Mission in the Basic Understanding confinement effects similar to those found in inorganic of Materials, and in the process will involve exploration QDs, by making nanometer-sized graphene pieces, or of Fundamental Chemistry. As the work progresses into graphene-QDs. producing graphene devices, it will directly add to prog- ress in the LANL Energy Security Mission, particularly Although the production of ordinary graphene is a rap- in the area of Renewable Energy. As such, it will be of idly developing field, the exploration of graphene-QDs primary interest to various offices within DOE, including is in its relative infancy. Our effort will start with exist- especially the Office of Science. It can also be anticipated ing techniques for making graphene, and modify them that research into combining nano-graphene with other for the production of graphene-QDs, which will allow types of quantum dots will lead to new phenomena of us to perform the first high-level spectroscopic studies relevance to other types of devices, including in biosens- of these unique materials, including especially the first ing and radiation detection. Thus, there may be long- measurements of CM efficiencies. Correlating CM yields term benefits to LANL efforts and eventually to federal with the structure of the graphene-QDs will allow us to funding agencies in these areas. optimize the material by further modifying the growth conditions, with the goal of approaching the theoreti- Progress cally predicted record efficiencies. This project will commence when the Oppenheimer Fel- low, Dr. Shaojun Guo, is hired on July 15, 2013. Finally, we will explore the application of this material in proof-of-principle PV devices. By validating the promise 308

Page 309

Future Work This project will commence when the Oppenheimer Fel- low, Dr. Shaojun Guo, is hired on July 15. The remaining three months of Fiscal Year 2013 will be devoted to estab- lishing the grahene-production capability. This is a com- pletely new activity for the Nanotechnology and Advanced Spectroscopy (NanoTech) Team, and thus will take some amount of time, and a great deal of Dr. Guo’s effort and ingenuity to accomplish. It will involve Dr. Guo completing his relevant LANL training, working with Drs. Klimov and Pietryga (as mentors) to establish the proper Integrated Work Management (IWM) envelope for this work to pro- ceed safely and securely, and constructing the necessary apparatus through leveraging of existing equipment and some small purchases. Project Year 1 will then also include the first three quarters of Fiscal Year 2014, during which Dr. Guo will optimize the graphene production apparatus to achieve materials with reproducibly high quality, and start to modify his techniques to produce nanometer- sized pieces of graphene (the graphene quantum dots, or graphene-QDs). At this point, he will begin to interact with spectroscopists on the Nanotech team to determine the most promising means for measuring the carrier dynam- ics in excited graphene-QDs, including for establishing the efficiency of carrier multiplication in these materials. As the materials production progresses, Dr. Guo will also keep all the members of the NanoTech Team apprised of devel- opments through participation in weekly Team meetings, continuously looking for opportunities to use these materi- als in other studies. Conclusion We will adapt existing graphene fabrication techniques to produce high-quality graphene quantum dots. We will then perform high-level spectroscopic study of these materials, including the first measurements of graphene-based “car- rier multiplication” (CM), a phenomenon known in other quantum dots in which a single photon produces more than one electron-hole pair. With spectroscopic feedback, we will optimize the fabrication techniques further to achieve materials with the highest CM efficiencies. Finally, we will fabricate “proof-of-principle” solar cells, based on graphene quantum dots, to demonstrate that CM can have a measureable impact on the power output of real devices, a potentially field-changing revelation. 309

Page 310

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project Microstructured Biohybrid Synthesis of Photosynthetic Assemblies Gabriel A. Montano 20130796PRD2 Introduction Benefit to National Security Missions Photosynthesis ubiquitously occurs in membranes. The work to be performed directly addresses needs These lipid bilayers can be among the most crowded in- within DOE and energy-related efforts and fundamental terfaces in nature with up to ~70% of the total available biomaterials research design. Success of the proposed area occupied by proteins. Further, the effective con- work will result in necessary information in order to centration of photosynthetic pigments can be as high as develop a new class of energetic materials capable of 1 molar. Despite this, photosynthetic charge separation achieving a long-standing goal: harnessing the energy proceeds with unprecedented quantum efficiencies that of the sun with the efficiency of natural photosynthetic approach unity due to a high level of two- and three- assemblies and the stability of standard photovolta- dimensional organization. The energetics of the reaction ics. Secondarily, optimizing the bio/synthetic interface is sufficient to oxidize water and drive cellular processes will have an impact far beyond energy science effecting that ultimately reduce CO2 into carbohydrates. How do arenas such as bio/chemical materials development and these organisms capture the energy of light with such detection, sensor and prosthetic development and en- high efficiencies but more importantly, how can the de- vironmental remediation. Thus, the proposed research sign principles be exploited? While the former question will generate opportunities within the confines of DOE has been investigated over the past decades, the latter as well as provide avenues for development within agen- has been relatively unexplored. cies such as DOD, IC, NIH and potentially more. Utilizing soft materials capabilities developed by our Progress team along with micro- and nano-patterning technolo- The project started in August 2013; therefore, there is gies at CINT we will investigate programmed organi- no progress to date. zation of photosynthetic assemblies into two- and three-dimensional asssemblies. We will investigate Future Work parameters effecting biohybrid assembly such as packing Growth of green and purple bacteria for harvesting of density, distance and orientation of light-harvesting (LH) individual photosynthetic components. and reaction center (RC) complexes and optimize the biophysical phenomena of energy transfer and charge • Requires initial cell culture growth separation in robust, long-life assemblies. • Requires isolation and purification of individual pho- tosynthetic components The ability to intricately organize and design biohybrid • Design of artificial LH/RC complexes using polymer/ photosynthetic systems would have transformative lipid and porphyrin based supramolecular assembly impact across multiple fields. In the proposed research, • Development of initial 2D and 3D assemblies for controlled assembly of photosynthetic complexes would characterization allow for more efficient light harvesting into the system • Requires demonstration of isolated photosynthetic while also elucidating the spatial parameters required assemblies and artificial photosynthetic assemblies for optimal assembly functionality. Lastly, the ap- into soft materials architectures proaches described here can be extended any variety • Characterization of photosynthetic assemblies- i.e. of biomolecule (other proteins, DNA etc.) opening up stability, performance, communication (energy additional applications in bio-sensors, -materials and transfer, charge separation) research platforms. 310

Page 311

• Begin refinement of architectures to enhance stability/ performance • Begin expression and growth of modified LH Conclusion The specific aims of this project are to: 1) Genetically engi- neer photosynthetic proteins and potentially bio-inspired artificial photosynthetic contructs for programmed assem- bly onto functionalized and patterned 2D and 3D architec- tures; 2) Investigate parameters effecting the optimization of biohybrid assembly e.g. packing density, distance and orientation of light-harvesting and reaction center photo- synthetic complexes; 3) Systematically explore and opti- mize the biophysical phenomena of energy transfer/charge transfer in assemblies using an iterative design strategy. Results will have a tremendous impact in artificial photo- synthesis, energy applications and biohybrid design. 311

Page 312

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project Photoactive Energetic Materials for Quantum Optical Control Sergei Tretiak 20130804PRD3 Introduction Applications The project will involve the development of a combined By concerted experimental and theoretical character- quantum mechanical and molecular mechanical (QM/ ization of the dynamics of energy localization, chemi- MM) approach to characterize the dynamics of energy cal bond activation, and chemical reaction, insight can localization, chemical bond activation, and chemical be gained on how to manipulate electronic structure reaction of photoactive energetic materials. through synthesis in order to design photoactive ener- Programmatic needs exist for the development of getic materials with controllable optical functionality. emergent technology that can address future require- ments for enhanced safety and security of the nuclear I plan to identify specific vibrational degrees of freedom stockpile. For example, laser-based ignition of insensitive responsible for bond breakage, rapid decomposition, explosive materials would provide tremendous improve- and the onset of the exothermic chain reactions relevant ments in detonator safety. to explosive initiation processes in the surrounding material. The developed methodology will be used to design and propose controllable materials. I will perform Benefit to National Security Missions detailed numerical NA-ESMD simulations to investigate The proposed research aligns closely with national secu- photoinduced pathways, timescales, branching effects, rity missions of the Laboratory, in particular, high explo- and multiple product formation, which can be experi- sives science, detonator technologies, and Life Extension mentally validated by ultrafast spectroscopy capabilities Programs (LEPs) addressing stockpile safety and security in WX-9. I will assess possible photoinduced reaction issues. coordinates and electron-vibrational relaxation pathways to elucidate mechanisms for optical initiation. I plan to Progress use first principle electronic structure calculations to The project started July 2013; therefore, there is no quantitatively evaluate the nonlinear optical responses progress to date. to understand excited-state properties and to provide design strategy for new photoactive energetic materials. Future Work Methodology Conclusion I plan to develop a combined quantum mechanical and The research will extend simulations of the excited state molecular mechanical (QM/MM) approach to treat large dynamics to the solid state using Quantum Mechanics/ systems and include solvent and thermal bath effects. Molecular Mechanics (QM/MM) framework. In addition Our current NA-ESMD implementation only allows for to the myriad of basic science applications, this approach phenomenological treatment of the thermal bath. QM/ will be used to model photoexcitation dynamics of MM is a much more accurate approach in which a sys- energetic materials. This research will be tightly coupled tem is divided into a quantum mechanical region and a with ongoing spectroscopic studies, aiming to establish molecular mechanical region allowing realistic modeling a theory-guided strategy to design photoactive energetic of dielectric media. Depending on the molecular system, materials with controllable optical functionality. different force fields will be interfaced with our existing NA-ESMD code. 312

Page 313

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project Poly-triphenylamine-functionalized Graphene Quantum Dots for Flexible High- Performance Polymeric Memory Devices Hsing-Lin Wang 20130806PRD3 Introduction able energy and threat reduction. Unlocking the struc- The main focus of this project is to develop synthetic tural features and charge transfer mechanisms will allow strategies for making graphene quantum dots (GQDs)- a design and synthesis of next generation memory and small piece of grapheme with well defined molecular energy devices. The proposed work is in alignment with structure. These as-synthesized GQDs are highly pro- the central DOE missions and supports new program cessible and can be incorporated with poly(triphenyl development in Office of Science, DTRA, and DARPA. amine), PTPA, to render memory and energy devices. The synthesis of PTPA-GQDs and their incorporation in Progress memory and energy devices has never been attempted. The project started August 2013; therefore, there is no The challenge of this project lies in the fact that the progress to date. organic synthetic scheme for PTPA-GQD requires more than 20 steps. This is an extremely costly and time con- Future Work suming process. In the next fiscal year, we will focus on the synthesis of highly processible graphene quantum dots. Our design Although synthesis of simple GQDs have been demon- synthesis of graphene quantum dots employing multi- strated, synthesis of PTPA-GQD remains a challenge and step organic synthesis using Suzuki coupling reactions the final yield of this compound may be low, which could will led to polyphenylene dendritic precursor containing impede us from developing application using this materi- triphenyl amine group, which will then be exposed to als. Therefore, our challenges are not only in the synthe- an excess of FeCl3 in a dichloromethane/nitromethane sis of PTPA-GQDs, but also to find ways to increase the mixture, yielding the polytriphenylamine-based gra- yield of the final product. phene quantum dot (PTPA-GQD). The covalently bound TPA-GQD will be thoroughly characterized by X-ray We expect the as-prepared PTPA-GQD will reveal excel- photoelectron spectroscopy, MALDI-TOF MS, NMR, and lent memory properties. In addition, polymer memory IR spectroscopy. devices have potential to be flexible and easily integrat- ed with conventional devices to show added value and Conclusion complexity. The PTPA-GQD will also be tested as energy We expect to achieve synthesis of processible grapheme and sensing devices. Success in demonstrating deriva- quantum dots (GQDs). We expect the as-prepared PTPA- tized GQD in the above devices could have intellectual GQD will reveal excellent memory properties. In addi- property value and warrants publications in high impact tion, polymer memory devices have the potential to be journals. flexible and easily integrated with conventional devices to show added value and complexity. The PTPA-GQD Benefit to National Security Missions will be used as a memory layer and fabricated on plastic Success of this project will lead to the development poly(ethylene naphthalate) (PEN) substrates in a cross- of one or several stable, high performance electronic, point structure. We expect to demonstrate the reliable memory and energy devices. Success of this project will and reproducible performance of sandwiched PEN/Al/ have immediate commercial application and add to the PTPA-GQD/Al memory devices; the performance charac- scientific knowledge of the interface between materials, teristics will be analyzed statistically and under bending and nanotechnology crucial to the lab mission for renew- stress. 313

Page 314

Publications Chen, C., H. Yen, Y. Hu, and G. Liou. Novel Programmable Functional Polyimides: Preparation, Demonstrating Mechanism of CT Induced Memory, and Ambipolar Electrochromic Behavior. 2013. J. Mater. Chem. CJ. Mater. Chem. C. Chou, Y., H. Yen, C. Tsai, W. Lee, G. Liou, and W. Chen. Nonvolatile transistor memory devices using high dielectric constant polyimide electrets. 2013. J. Mater. Chem. CJ. Mater. Chem. C. 1 (19): 3235. Yen, H., J. Wu, W. Wang, and G. Liou. High-Efficiency Photoluminescence Wholly Aromatic Triarylamine- based Polyimide Nanofiber with Aggregation-Induced Emission Enhancement. 2013. Advanced Optical MaterialsAdvanced Optical Materials. 1 (9): 668–676. 314

Page 315

Chemistry and Material Sciences Postdoctoral Research and Development Continuing Project Understanding and Controlling Magnetism in Multiferroics with THz Pulses Rohit P. Prasankumar 20130812PRD3 Introduction coupling in multiferroic materials through selectively In ferromagnets, the magnetization can be controlled exciting the low-energy modes responsible for material with an external magnetic field, an idea widely used, for functionality and probing their effects on this coupling. example, in magnetic data storage. However, magnetic More generally, our effort to interface materials science fields have some drawbacks: there are limitations to how with ultrafast terahertz (THz) and optical probes repre- fast they can be switched, they are inconvenient to use, sents an essential element in the MaRIE strategy that and they consume extra energy. Much effort has thus connects the M4 facility to the Multi-Probe Diagnostic been devoted to multiferroic materials, in which electric Hall. This work directly addresses the LDRD Grand Chal- and magnetic polarizations exist simultaneously and lenge in Materials, which underpins all three Laboratory can be coupled, enabling control of magnetism with an mission areas. It also addresses several of the Grand electric field. Challenges for Basic Energy Sciences identified by the DOE Office of Science. We will work with the Program Recently, there have been significant advances in the Director for Basic Energy Sciences to explore future ability to synthesize these materials, which has led to funding opportunities within BES in the growing areas of much effort focused on increasing the magnetoelec- ultrafast materials science. tric (ME) coupling, especially near room temperature. However, the microscopic origin of this coupling can Progress vary across different classes of materials, and is not well Project started September 2013; therefore, there is no understood. Low energy excitations (i.e. soft modes, progress to date. magnons, etc.) are thought to play a key role in ME coupling, and therefore also offer a promising route to Future Work controlling it. Building on our previous work in this field, we will begin by optimizing our existing ultrafast optical and terahertz Here, we will use ultrashort optical and terahertz (THz) (THz) systems for the proposed experiments, with a pulses to shed light on the microscopic origin of mag- particular focus on generating high energy mid-to-far- netoelectric coupling in different multiferroic materials infrared pulses for photoexciting low energy modes in (building on our previous work in this area). In these multiferroics. Our initial focus will be on the canonical experiments, we will use these pulses to separately ma- multiferroic material, BiFeO3 (BFO), which has attracted nipulate and probe the magnetic and electric orders in a much interest due to its room temperature coexistence given material, enabling us to, e.g., modify the magnetic of magnetic and ferroelectric order, and the manganite order and probe the resulting changes in ferroelectric TbMnO3 (TMO), which is well known for its strong mag- order, or vice versa. This will provide much insight into netoelectric (ME) coupling, albeit at low temperatures. magnetoelectric coupling in multiferroics, which should We will also examine multiferroic heterostructures, con- extend our knowledge of their basic physics and also en- sisting of thin ferroelectric and magnetic films grown on able researchers to optimize them for applications. top of one another to maximize the coupling between them. Benefit to National Security Missions We will explore several experimental directions. In gen- The proposed experiments will provide LANL, as well as eral, we will use these pulses to separately manipulate CINT, with the capability to investigate magnetoelectric and probe the magnetic and electric orders in a given 315

Page 316

material, enabling us to, e.g., modify the magnetic order and probe the resulting changes in ferroelectric order, or vice versa. This can shed light on ME coupling in the materials described above. For example, in BFO, an impor- tant unresolved question is whether the soft mode phonon drives the ferroelectric (FE) transition, and how this links to ME coupling. We can address this by photoexciting this mode and monitoring the resulting changes in the FE and magnetic order with an optical probe pulse. Similarly, in TMO, we can test a recent theoretical prediction by us- ing intense THz pulses to photoexcite low energy modes, potentially enabling us to control magnetic order. Finally, in a multiferroic heterostructure, we can photoexcite low energy modes in a ferroelectric (FE) layer and examine the effect on ferromagnetic order in a proximal magnetic layer. These experiments should thus provide much insight into ME coupling in multiferroics, with wide ranging impact in physics and materials science. Conclusion Multiferroic materials, in which magnetic and ferroelec- tric order can be closely coupled, offer much promise for a variety of applications in data storage, novel logic ele- ments, and sensing. However, the mechanisms underlying this coupling are not well understood, limiting the poten- tial of these materials. Here, we will use ultrashort optical and terahertz (THz) pulses to shed light on the microscopic origin of magnetoelectric coupling in different multiferroic materials. This will provide much insight into magneto- electric coupling in multiferroics, which should extend our knowledge of their basic physics and also enable research- ers to optimize them for applications. 316

Page 317

Environmental and Biological Sciences

Page 318

Environmental and Biological Sciences Directed Research Continuing Project Dynamic Earthquake Triggering, Granular Physics and Earthquake Forecasting: Determining the Physical Controls Paul A. Johnson 20120007DR Introduction increased earthquake risk owing to triggering, and c) The March 11, 2011 M9.0 Tohoku-oki Earthquake, bridge laboratory scales to Earth scales. located ~100 km off the east cost of Honshu, was the fourth largest earthquake in recorded history, rupturing Benefit to National Security Missions a 300 km segment of the plate boundary. The earth- Large earthquakes pose serious risks to national energy quake was due to ongoing subduction of oceanic crust and economic security through their deleterious impacts beneath Japan. The slip magnitude between the two on infrastructure, a stated LANL mission area. This work plates was 30-40 meters resulting in abrupt uplift of the could have an enormous societal/infrastructure impact seafloor, creating a tsunami that inundated port cities in by forecasting times of increased earthquake hazard. eastern Honshu. A resultant nuclear disaster continues Triggering may also have significant import to Ground to unfold at nuclear power plants located in Fukushima. Based Nuclear Explosion Monitoring (GNEM)—clandes- Despite the relatively sparse population in this region, tine events could be better hidden in regions of trig- the resulting economic losses are predicted to be about gered event clustering. Our work will improve under- 4% of Japan’s GDP. standing of seismic events as well as the exploitation, analysis, and interpretation of faint seismic signatures. An earthquake of M8-9 in a densely populated region The work may have broad application to vibration- such as Seattle could have catastrophic effects on the induced failure in materials, including those relevant to US. Prior warning of increased earthquake probability the weapons program, aircraft, industrial machinery, and during a specific time interval could significantly mitigate shaking of infrastructure from strong ground motion. the potential economic disaster.It has recently been pos- This work will contribute to hazards analysis at LANL tulated that a majority of earthquakes may be dynami- (CMRR). Moreover the work improves interdisciplinary cally triggered. Our unique perspective is that character- capabilities at LANL by integrating efforts in seismology, izing triggering leads to fundamentally new means of geophysics, granular media, and continuum elastic- earthquake forecasting of intervals of increased hazard. plastic deformation that are underpinning science for a The Problem: Although earthquake forecasting is a diverse set of LANL priorities including MaRIE. We note highly complex, unsolved problem, significant advances the large costs to NNSA and nuclear power infrastruc- may be possible in predicting time intervals of increased ture that are driven by seismic requirements, particu- earthquake hazard. Institutes worldwide including the larly the largest credible events, which imply a need Working Group on California Earthquake Probabilities to quantify worst-case scenarios better than existing (WGCEP) currently do not include dynamic triggering in science can do. hazard estimates. Progress Our hypotheses are: a) Granular physics plays a key We have made significant progress toward our stated role in fault triggering in the laboratory and Earth. b) goals of characterizing the effects of seismic waves on a Isolating the triggering mechanism and characterizing fault system, with the ultimate goal of characterizing pe- its signatures will improve earthquake hazard analysis riods of increased earthquake hazard. We have placed by determining the duration of enhanced earthquake significant effort into the question of whether or not probability from triggering.Our project goals are to: a) large earthquakes cluster globally, meaning one earth- Characterize the granular physics that we posit enables quake may trigger another. In short, our work shows triggering, b) apply novel statistical approaches to infer 318

Page 319

tantalizing suggestion of clustering of large earthquakes; ratory and numerical simulation data that can be tested for however, as the earthquake catalog of large events is so scaling to the earth. limited the statistical confidence in this result is limited. One paper has been published on our analysis and the We are also studying the processes that lead to instability other is in press. We are now considering other statistical in the fault ‘core’ or fault ‘gauge’ (the ground-up material means by which to test for global clustering, applying very that lies between fault blocks) induced by seismic waves. small amounts of earthquake data. These studies include both numerical simulation and laboratory studies as well. From these studies we hope We have also been developing the means to characterize to learn what the causes are for hypothesized long-lived hypothesized seismic-induced perturbation of the material changes in the earth that are due to seismic waves from in regions of active faulting. We believe that this perturba- other earthquakes. This long-lived behavior is responsible tion leads to the periods of increased seismic hazard. Our for the times of increased earthquake hazard. work has been at the Parkfield segment of the San Andreas Fault, as stated in the proposed work, and also in Japan We are working progressively more with external institu- using their extremely good network of borehole stations. tions that are intrigued by our unique vision and approach. To date, we can state that there appears to be widespread These institutions include Pennsylvania State University, changes in Japan seismicity due to the recent Indian Ocean the United States Geological Survey, ETH Zurich (Switzer- earthquake of 2012, strongly suggesting earthquake trig- land), Bath University (UK), The ‘Ecole Normal’ and the In- gering and long lasting elastic effects. Currently we are stitute of Physics of the Globe in Paris (France), the Univer- also, (i) looking at Global Positioning System (GPS) data sity of Tokyo, Sendai University (Japan) and the University that can be applied to see if there is associated earth of Grenoble (France). surface displacement as we predict; (ii) looking for associ- ated seismic wave velocity changes applying what is known In summary, we are using a multi-pronged approach that as ‘passive noise imaging’ [the method employs use of includes seismic and other observations in the earth, ocean-generated noise that can be used for seismic imag- laboratory and numerical modeling studies, to address the ing]; and (iii) beginning work on other potential triggering earthquake triggering/earthquake hazard problem. We earthquakes such as the Great Chile earthquake (Maule) are certain to make tremendous progress on this problem of 2010. At Parkfield we are developing other statistical over the remaining period of time we have. methods that we will employ there and elsewhere, to de- termine if seismicity rates change due to potential trigger- Future Work ing earthquakes. Our goals for next year are, (1) to establish whether or not large earthquakes trigger large earthquakes worldwide Our supporting laboratory and numerical modeling work (preliminary work suggests that the answer is yes) and if is advancing as well. The new shearing experiment meant so, begin establishing a new model of how earthquakes to simulate earthquakes applying what is known as ‘pho- occur globally; (2) to begin work on what the implications toelasticity’ is now operational and the first data sets are of such a conclusion may have on evaluating earthquake being collected. Photoelasticity allows one to visually hazard; (3) to determine if mechanical property changes observe and interpret changes in the stresses on and off take place from seismic waves from distant earthquakes the model fault. Model earthquake triggering studies and to establish methods for monitoring such changes. A will begin soon, and we intend to compare these to the major task for the next fiscal year includes statistical analy- numerical simulations that are well advanced. The other ses of earthquake catalogs to address the global triggering laboratory experiment is opaque but is three-dimensional, question. A second task is honing our skills at the Parkfield in contrast to the photoelastic experiment. This experi- section of the San Andreas Fault (the Parkfield laboratory) ment is well advanced and we have a number of publica- in regards to determining if seismic waves from near and tions describing results. We are also comparing the results distant earthquakes perturb the mechanical properties of to three-dimensional simulations using the same numeri- the fault region, and if so, for what time duration. Methods cal simulation method and a number of publications have to be considered include using the interval between earth- resulted. In short the model and experimental methods quakes in the perturbed region that should change in their are being used to develop methods to characterize shear frequency for a period of time after perturbation based along faults as well as the effects of earthquake trigger- on laboratory and simulation studies. A second method is ing. We can statistically test whether or not the numerical monitoring velocity change in the perturbed region apply- simulations and experiments make sense in the context of ing a number of approaches, including monitoring passive real earthquakes, by looking at certain aspects of the labo- earth noise that can be applied to extract velocity changes 319

Page 320

in the earth’s crust. We will support the Parkfield studies probabilities as evidenced by two M>7.5 earthquakes. with laboratory experiments and discrete element model- 2013. Geophysical Research Letters. 40: L50933. ing. Gong, X., Q. Chen, Z. Peng, W. Wang, C. Wu, and J. Wu. Remotely triggered seismicity around the Fangshan Conclusion Pluton near Beijing following the 2010 Mw 8.8 Chile Our goal is to determine intervals of increased earthquake earthquake. To appear in Chinese Journal of Geophys- hazard in regions of high earthquake risk. In particular, ics. seismic considerations play large in the siting of energy and Griffa, M., B. Ferdowsi, E. Daub, R. Guyer, P. A. Johnson, C. weapons complex infrastructure. By further delineating Marone, and J. Carmeliet. Meso-mechanical analysis of the time interval of increased hazard, local, regional and deformation characteristics for dynamically triggered national hazard response officials could take appropriate slip in granular materials. 2012. Philosophical Maga- precautions in order to mitigate casualties and to address zine. : 1. pressing infrastructure issues. Further, our work will have an enormous impact on characterizing the physics of Griffa, M., B. Ferdowsi, R. A. Guyer, E. G. Daub, P. A. John- sheared granular media and how these materials transition son, C. Marone, and J. Carmeliet. Influence of vibration from locked, or jammed, to fluidized via nonlinear dynam- amplitude on dynamic triggering of slip in sheared granular layers. 2013. PHYSICAL REVIEW E. 87 (1): -. ics, a topic that we are currently advancing. Griffa, M., E. G. Daub, R. A. Guyer, P. A. Johnson, C. Ma- Publications rone, and J. Carmeliet. Vibration-induced slip in Ben-Naim, E., E. G. Daub, and P. A. Johnson. Recurrence Sheared Granular Layers and the Micromechanics of statistics of great earthquakes. 2013. GEOPHYSICAL RE- Dynamic Earthquake Triggering. 2011. Europhysics Let- SEARCH LETTERS. 40 (12): 3021. ters. 96 (1): 14001 (6 pp.). Ben-Naim, E., and P. L. Krapivsky. Statistics of superior re- Johnson, P.. Nonlinear acoustic/seismic waves in earth- cords. 2013. PHYSICAL REVIEW E. 88 (2): -. quake processes. 2012. In Nonlinear Acoustics State- of-the-art and Perspectives: 19th International Sympo- Daub, E., D. Shelly, R. Guyer, and P. Johnson. Brittle and sium on Nonlinear Acoustics ; 21-24 May 2012 ; Tokyo, ductile friction and the physics of tectonic tremor. Japan. Vol. 1474, p. 39. 2011. Geophysical Research Letters. 38 (10): L10301. Johnson, P. A., B. Carpenter, M. Knuth, B. M. Kaproth, P. -Y. Daub, E., E. Ben-Naim, R. Guyer, and P. Johnson. Are mega- Le Bas, E. G. Daub, and C. Marone. Nonlinear dynami- quakes clustered?. 2012. Geophysical Research Letters. cal triggering of slow slip on simulated earthquake 39 (6): L06308. faults with implications to Earth. 2012. JOURNAL OF GEOPHYSICAL RESEARCH-SOLID EARTH. 117: B04310. Elst, N. J. van der, E. E. Brodsky, P. Y. Le Bas, and P. A. John- son. Auto-acoustic compaction in steady shear flows: Johnson, P. A., B. Ferdowski, B. Kaproth, M. Scuderi, M. Experimental evidence for suppression of shear dilat- Griffa, J. Carmeliet, R. A. Guyer, P. Y. L. Bas, D. T. Trug- ancy by internal acoustic vibration. 2012. JOURNAL OF man, and C. Marone. Acoustic emission and microslip GEOPHYSICAL RESEARCH-SOLID EARTH. 117: -. precursors to stick-slip failure in sheared granular ma- terial. To appear in Geophysical Research Letters. Ferdowski, B., M. Griffa, R. A. Guyer, P. A. Johnson, C. Ma- rone, and J. Carmeliet. Microslips as precursors of large Miller, P. W., and E. Ben-Naim. Scaling exponent for incre- slip events in the stick-slip dynamics of sheared granu- mental records. To appear in Journal of Statistical Me- lar layers: A discrete element model analysis. 2013. chanics. Geophysical Research Letters. 40 (16): 4194. Trugman, D. T., E. G. Daub, R. A. Guyer, and P. A. Johnson. Ferdowski, B., M. Griffa, R. A. Guyer, P. A. Johnson, C. Ma- Modeling dynamic triggering of tectonic tremor using rone, and J. Carmeliet. 3D Discrete Element Modeling a brittle-ductile friction model. 2013. Geophysical Re- of triggered slip in sheared granular media. Physical search Letters. 40: L50981. Review E. Wu, C. Q., D. R. Shelly, J. Gomberg, Z. G. Peng, and P. John- Ferdowski, B., M. Griffa, R. A. Guyer, P. A. Johnson, and son. Long-term changes of earthquake inter-event J. Carmeliet. Effect of boundary vibration on the fric- times and low-frequency earthquake recurrence in tional behavior of a dense granular layer. To appear in central California. 2013. EARTH AND PLANETARY SCI- Acta Mechanica. ENCE LETTERS. 368: 144. Gomberg, J.. Permanently enhanced dynamic triggering Wu, C., J. Gomberg, E. Ben-Naim, and P. Johnson. Testing 320

Page 321

delayed triggering of repeating earthquakes in central California. Geophysical Research Letters. Wu, C., X. Meng, Z. Peng, and Y. Ben-Zion. Lack of spatio- temporal localization of foreshocks before the 1999 Mw 7.1 Duzce, Turkey earthquake. To appear in Bulle- tin of the Seismological Society of America. Wu, J., Z. G. Peng, W. J. Wang, X. Gong, Q. F. Chen, and C. Q. Wu. Comparisons of dynamic triggering near Bei- jing, China following recent large earthquakes in Suma- tra. 2012. GEOPHYSICAL RESEARCH LETTERS. 39: -. 321

Page 322

Environmental and Biological Sciences Directed Research Continuing Project Predicting Climate Impacts and Feedbacks in the Terrestrial Arctic Scott L. Painter 20120068DR Introduction Progress We will address one of the most critical questions in The Arctic Terrestrial Simulator (ATS) is being developed contemporary climate science: will a warming terrestrial to provide Arctic process modeling capability within the Arctic provide a net positive or negative feedback to framework of the highly parallel Amanzi code. Amanzi the climate system? The Arctic region has been a global provides a number of important features for the ATS, in- carbon sink for thousands of years and is currently the cluding a very general parallel unstructured mesh capa- largest terrestrial store of carbon, but is now warming bility, advanced discretization toolsets for these meshes, faster than the rest of the Earth. The scientific commu- and parallel output for modern visualization tools. The nity cannot say how much Arctic carbon will thaw, de- hydrological, chemical, and ecological processes relevant compose into greenhouse gases (GHG), and be released to the Arctic tundra models are implemented in ATS in a to the atmosphere or how fast it will happen because series of Process Kernels (PKs). In FY2013, development current modeling tools lack a critical process cascade: of key PKs progressed significantly. In addition, computa- thaw-induced subsidence, reorganization of topography tional strategies for developing terrain-following un- and drainage networks, and resulting redistribution of structured model grids based on available data sources soil moisture. We propose to address this uncertainty by were developed and successfully demonstrated. leveraging recent computational advances to develop a novel Arctic Terrestrial Simulator (ATS) that will incorpo- A new constitutive model for water partitioning between rate for the first time the complex interactions among ice and liquid phases in unsaturated frozen soils was thermal, mechanical and hydrologic permafrost pro- developed and successfully compared to existing labora- cesses that control GHG production and sequestration in tory data. The new model forms the basis of the freezing a thawing and topographically reorganizing landscape. soil hydrology PK, a key component of the permafrost Our overarching goals are to 1) provide the first reliable model. A journal article on that work was submitted quantitative estimates of GHG releases and uptake from to the Vadose Zone Journal. A second journal article a warming Arctic landscape undergoing rapid topograph- describing implementation of the freezing soil model ic reorganization, and 2) provide critical new capabilities was also prepared and submitted. Performance of the to further understand and refine predictions of GHG freezing soil hydrology PK was improved significantly releases from the Arctic landscape. through the use physics-based preconditioners, and is now sufficient for use in the fully coupled model. Benefit to National Security Missions Our work addresses one of the most significant gaps in A new and computationally tractable model for overland the Nation’s capability to understand and predict im- flow and heat transfer with freezing of ponded water pacts of climate change. The LDRD investment in ter- was developed and implemented. Performance of that restrial Arctic prediction will fill the last gap in current model was optimized and successfully coupled to the high latitude high-performance modeling capability by freezing soil model. The coupled model is the core com- providing high-resolution landscape modeling capability ponent of a permafrost hydrology model. for the Arctic to complement existing ocean, sea ice, ice Significant progress was made on implementing the sheet and atmosphere predictive capability. computational infrastructure required for modeling de- forming Arctic tundra topography. 322

Page 323

Characterization of permafrost dominated landscapes is Conclusion crucial for understanding and predicting the spatial and Critical processes related to permafrost degradation are temporal distribution of climate change related impacts not represented explicitly in the current generation of and feedbacks in the Arctic. We developed and evalu- climate models because they cannot be resolved on the ated an automated approach to identify and characterize scale of a grid cell. The new parallel simulation capability Arctic ice-wedge polygonal tundra landscape components, will enable mechanistic process understanding emerging such as troughs, ponds, river- and lake-like objects, which from field studies to be used in a high-resolution predic- constitute drainage networks, using high spatial resolu- tive framework at appropriate temporal and spatial scales. tion satellite imagery. Identification and characterization The Arctic Terrestrial Simulator will be used to model large of drainage network components is important for tracking swaths of the Arctic landscape, resulting in the first pro- Arctic tundra terrain evolution as well as for estimating jections of carbon release rates from the most vulnerable water, heat and carbon fluxes for use in climate models. A lowland tundra regions with high carbon stores, rapidly journal article on the approach was prepared and submit- degrading permafrost and evolving landscape. ted to Remote Sensing Letters. Publications Simulations to predict the future evolution of Arctic Coon, E., M. Berndt, J. D. Moulton, and S. Painter. The lowland tundra require high-resolution terrain-following Arctic Terrestrial Simulator: Developing a flexible multi- computational grids. A computational workflow for gen- physics simulator based on Amanzi. Presented at 2012 erating such grids using available high-resolution digital SIAM Annual Meeting . (Minneapolis, 9-13 July 2012). elevation maps and remote sensing data was developed and successfully demonstrated using data from the Barrow Frampton, A., G. Destouni, Y. Sjoberg, and S. Painter. Transient modeling of permafrost dynamics in chang- Environmental Observatory. ing climate scenarios. Presented at 2011 Institute of Electrical and Electronics Engineers 7th International A global sensitivity analysis for permafrost thawing simu- Conference on E-Science. (Stockholm, Sweden, 5-8 De- lations was conducted to help understand the relative cember). importance of several physical parameters in controlling depth to permafrost. The thermal conductivity, porosity Frampton, A., S. L. Painter, and G. Destouni. Permafrost and vegetation characteristics were determined to be the degradation and subsurface-flow changes caused by three most important factors contributing to the active surface warming trends. 2013. HYDROGEOLOGY JOUR- layer depth. NAL. 21 (1): 271. Frampton, A., S. L. Painter, and G. Destouni . Effects of Future Work hydrological inputs on the dynamics of permafrost Work on integrating newly developed thermal hydrologic system formation and degradation. Presented at Eu- process models in the Arctic Terrestrial Simulator (ATS) will ropean Geophysical Union General Assembly 2012. continue in FY2014. Optimization and fine-tuning of the (Vienna, Austria , April 7-12, 2012). permafrost thermal hydrology model and the overland Frampton, A., S. Painter, S. W. Lyon, Y. Sjoberg, and G. flow model, two of the four critical models, were complet- Destouni. Transient modeling of permafrost dynamics ed in fiscal year 2013. FY2014 model development work in a changing climate. Presented at American Geophys- will focus on a thaw-subsidence algorithm and a surface ical Union Fall Meeting. (San Francisco , 5-9 Dec. 2011). energy balance model that includes snow. Lewis, K. C., G. A. Zyvoloski, B. Travis, C. Wilson, and J. Row- We plan to initiate the first fully coupled simulations of land. Drainage subsidence associated with Arctic per- tundra evolution including evolving topography in early mafrost degradation. 2012. JOURNAL OF GEOPHYSICAL RESEARCH-EARTH SURFACE. 117: -. FY2014. Carbon cycle dynamics will then be added. Our goal is to have completed initial simulations of the future Painter, S. L., J. D. Moulton, and C. J. Wilson. Modeling carbon releases from Arctic tundra including the effects of challenges for predicting hydrologic response to de- evolving topography by end of FY2014. In parallel with the grading permafrost. 2013. HYDROGEOLOGY JOURNAL. ATS development and applications, we will continue de- 21 (1): 221. veloping journal articles on our process models and model applications. Painter, S. L., S. Karra, and P. C. Lichtner. Three-phase nu- merical model for subsurface hydrology in permafrost- affected regions. The Cryosphere. 323

Page 324

Skurikhin, A. N., C. Gangodagamage, J. C. Rowland, and C. J. Wilson. Arctic tundra ice-wedge landscape character- ization by active contours without edges and structural analysis using high-resolution satellite imagery. 2013. Remote Sensing Letters. 4 (11): 1077. 324

Page 325

Environmental and Biological Sciences Directed Research Continuing Project Elucidating Humankind’s Evolving Environment: From the Earth’s Core to the Cosmos Harald O. Dogliani 20120166DR Introduction focus and affects the particle background that is directly This project combine tasks from geology, atmospheric connected to understating the backgrounds in which sciences, space sciences, and astrophysics (normally very non-proliferation measurements are made. The astro- diverse fields) with the common theme that these fields physics research produce techniques to process large have an effect on mankind. Earthquakes probably have image-based data bases that build capability for space the most obvious effect on mankind. We will study the situational awareness. acoustical properties of rocks to understand dynamic triggering of seismic slips along major earthquakes Progress zones. To understand volcanoes better, we will model We have described how pyrocumulus clouds form during giant explosive eruptions, in particular, the interactions intense fires and can lead to previously unexpected rapid between gas-particulate mixtures and the surrounding spreading of fires. This occurred at a recent Los Alamos rock. Climate change has many drivers and one of the area fire. most important is how sea ice in the Arctic interacts with the lower atmosphere. We will build computer mod- An analytic program package was developed in support els of that interaction to understand the impact of the of HAWC, High Altitude Water Cherenkov Experiment, to diminishing southern ice cap. Our lives now depend on measure and understand very high energy cosmic rays. satellites and those satellites are susceptible to damage Cosmic rays irradiate the earth with a stream of high from space radiation. In particular, low earth orbit satel- energy nuclear particles that can effect communications lites are affected by the radiation that is determined by and biological systems. The detector system located the ring current in Earth’s magnetosphere. In turn, that in Mexico has achieved initial operating capabilities ring current is affected by the upper atmosphere. We and successfully mapped a portion of the sky, includ- will produce the first upper atmosphere - ring current ing where the earth’s moon cast a shadow because it computer simulation to learn how to predict the levels blocked cosmic rays from that direction. This is a first of radiation that could affect our satellites. Astrophysics time event. can be a laboratory that can teach us the fundamental nature of the universe. We will model dark energy in the This project created a Web site to encourage the popu- sun to see if there is a way to directly detect it on Earth lation at-large to report auroral observations, location, in laboratories. We will investigate how cosmic strings time and descriptions. These reports are then used as might be observable in large databases of variations that tip-off to the scientific community to make detailed ob- we see in the brightness of stars. servations of the phenomenology. Auroras are generally caused by streams of solar charged particles interacting Benefit to National Security Missions with the earth’s magnetic fields and reacting with the The basic science that we carry out has a breadth of mis- upper atmosphere where the tenuous and colorful glows sion impacts. Our work in climate and geology is related are observed at night. Increased solar and hence auroral to energy production and understanding the impact of activity is relevant to understanding space weather that climate change. The research will connect University can have deleterious effects on satellite performance. research in these areas to Los Alamos where they can be Los Alamos scientists are studying the role of ocean applied to our programmatic work. Ring currents in the eddy currents in the establishment of large-scale ocean magnetosphere will be studied under the space science circulation and climate. Indications are that such eddy 325

Page 326

currents can play a significant role in how major currents earthquakes zones on Earth helps prepare us for such flow and can have a significant effect on global climate handling events that would affect the entire country. Our dynamics. modelling of interactions between the upper atmosphere and the Earth’s magnetosphere will help predict how our In collaboration with New Mexico Tech, researchers assets in space will survive and function. Studying the pos- are demonstrating that seismic data can be filtered and sibilities of cosmic strings will probe at the fundamental processed to detect and quantify the intensity of ma- nature of the Universe. jor oceanic storms by detecting the impact of waves on land-water boundaries. They have detected increased Publications wave activity and storminess in the Southern Pacific North Abdo, A. A., A. U. Abeysekara, B. T. Allen, T. Aune, D. Ber- Atlantic. Decreasing secondary microseismic power trends ley, C. Chen, G. E. Christopher, T. DeYoung, B. L. Din- in the Northern Europe and the Arctic may be the result of gus, R. W. Ellsworth, M. M. Gonzalez, J. A. Goodman, reduced large-scale storm activity in the Arctic Ocean. J. Granot, E. Hays, C. M. Hoffman, P. H. Huntemeyer, B. E. Kolterman, J. T. Linnemann, J. E. McEnery, A. I. Mincer, T. Morgan, P. Nemethy, J. Pretz, E. Ramirez- Future Work Ruiz, J. M. Ryan, P. M. Parkinson, A. Shoup, G. Sinnis, A. We continuously negotiate with researchers specific tasks J. Smith, V. Vasileiou, G. P. Walker, D. A. Williams, and to improve understanding of catastrophic events: global G. B. Yodh. Constraints on the emission model of the climate, geoscience, solar system, cosmology and relevant “naked-eye burst” GRB 080319B. 2012. signatures. New task proposals are presently under evalu- Astrophysical Journal, Letters. 753 (2): L31 (5 pp.). ation for selection to begin in FY14. On going research Abdo, A. A., B. T. Allen, R. Atkins, T. Aune, W. Benbow, D. includes: Descriptions of how rocks and hot gases result in Berley, E. Blaufuss, E. Bonamente, J. Bussons, C. Chen, an explosive volcanic events. G. E. Christopher, D. G. Coyne, T. DeYoung, B. L. Dingus, D. E. Dorfan, R. W. Ellsworth, A. Falcone, L. Fleysher, R. Collect data on the transition zone between arctic glacial Fleysher, J. Galbraith-Frew, M. M. Gonzalez, J. A. Good- fields and tundra to understand evolution of natural plant man, T. J. Haines, E. Hays, C. M. Hoffman, P. H. Hunt- life, how it might survive or be replaced by other more emeyer, L. A. Kelley, B. E. Kolterman, C. P. Lansdell, J. T. temperate plants. Mathematically describe how ice ridges Linnemann, J. McCullough, J. E. McEnery, T. Morgan, A. form on sea ice to improve sea ice/climate computer mod- I. Mincer, M. F. Morales, P. Nemethy, D. Noyes, J. Pretz, eling programs. J. M. Ryan, F. W. Samuelson, P. M. Parkinson, A. Shoup, G. Sinnis, A. J. Smith, G. W. Sullivan, V. Vasileiou, G. P. Investigate how pyrocumulus clouds form and dissipate Walker, M. Wascko, D. A. Williams, S. Westerhoff, and G. B. Yodh. OBSERVATION AND SPECTRAL MEASURE-

• this is a special situation that when it occurs forest fires MENTS OF THE CRAB NEBULA WITH MILAGRO. 2012. expand much more rapidly and destroy vast areas well ASTROPHYSICAL JOURNAL. 750 (1): 63. beyond conventional predictions. Abdo, A. A., U. Abeysekara, B. T. Allen, T. Aune, D. Berley, E. Collect detailed data on glacial ice formation and evolu- Bonamente, G. E. Christopher, T. DeYoung, B. L. Dingus, tion. The purpose is to improve glacial ice computer mod- R. W. Ellsworth, J. G. Galbraith-Frew, M. M. Gonzalez, J. eling programs to better predict how glacial warming is af- A. Goodman, C. M. Hoffman, Hu&uml, C. M. Hui, B. E. fecting global warming. Study Greenland ice core samples Kolterman, J. T. Linnemann, J. E. McEnery, A. I. Mincer, to better understand previous climate cycles and compare T. Morgan, P. Nemethy, J. Pretz, J. M. Ryan, P. M. Par- kinson, A. Shoup, G. Sinnis, A. J. Smith, V. Vasileiou, with main ocean current circulation; major changes of G. P. Walker, D. A. Williams, and G. B. Yodh. Spectrum ocean circulation patterns have significant impact on and morphology of the two brightest Milagro sources global climate. For example Norway is relatively temperate in the Cygnus region: MGRO J2019+37 and MGRO because of a strong northward Atlantic current bringing J2031+41. 2012. Astrophysical Journal. 753 (2): 159 (8 warm water up from the more temperate south. pp.). Conclusion Abeysekara, A. U., J. A. Aguilar, S. Aguilar, R. Alfaro, E. Almaraz, C. Alvarez, J. de D. Alvarez-Romero, M. Al- This project will bring together LANL staff with Univer- varez, R. Arceo, J. C. Arteaga-Velazquez, C. Badillo, sity researchers and students from around the nation to A. Barber, B. M. Baughman, N. Bautista-Elivar, E. Bel- study the structure of the earth, space, and natural cli- mont, E. Benitez, S. Y. BenZvi, D. Berley, A. Bernal, E. mate change. Our study of the underlying physics behind Bonamente, J. Braun, R. Caballero-Lopez, I. Cabrera, dynamic triggering of seismic and aseismic slip along major A. Carraminana, L. Carrasco, M. Castillo, L. Cham- 326

Page 327

bers, R. Conde, P. Condreay, U. Cotti, J. Cotzomi, J. C. 2011. GEOPHYSICAL RESEARCH LETTERS. 38: L13704. D’Olivo, E. de la Fuente, C. De Leon, S. Delay, D. Del- epine, T. DeYoung, L. Diaz, L. Diaz-Cruz, B. L. Dingus, M. Chylek, P., C. Folland, L. Frankcombe, H. Dijkstra, G. Lesins, A. Duvernois, D. Edmunds, R. W. Ellsworth, B. Fick, D. and M. Dubey. Greenland ice core evidence for spatial W. Florino, A. Flandes, N. I. Fraija, A. Galindo, J. L. Gar- and temporal variability of the Atlantic Multidecadal cia-Luna, G. Garcia-Torales, F. Garfias, L. X. Gonzalez, Oscillation. 2012. Geophysical Research Letters. 39 (9): M. M. Gonzalez, J. A. Goodman, V. Grabski, M. Gussert, L09705. C. Guzman-Ceron, Z. Hampel-Arias, T. Harris, E. Hays, Chylek, P., C. Folland, L. Frankcombe, H. Dijkstra, G. Lesins, L. Hernandez-Cervantes, P. H. Huntemeyer, A. Imran, and M. Dubey. Greenland ice core evidence for spatial A. Iriarte, J. J. Jimenez, P. Karn, N. Kelley-Hoskins, D. and temporal variability of the Atlantic Multidecadal Kieda, R. Langarica, A. Lara, R. Lauer, W. H. Lee, E. C. Oscillation. 2012. GEOPHYSICAL RESEARCH LETTERS. Linares, J. T. Linnemann, M. Longo, R. Luna-Garcia, 39: L09705. H. Martinez, J. Martinez, L. A. Martinez, O. Marti- nez, J. Martinez-Castro, M. Martos, J. Matthews, J. E. Cowee, M. M., S. P. Gary, H. Y. Wei, R. L. Tokar, and C. T. McEnery, G. Medina-Tanco, J. E. Mendoza-Torres, P. A. Russell. An explanation for the lack of ion cyclotron Miranda-Romagnoli, T. Montaruli, E. Moreno, M. Mo- wave generation by pickup ions at Titan: 1-D hybrid stafa, M. Napsuciale, J. Nava, L. Nellen, M. Newbold, R. simulation results. 2010. JOURNAL OF GEOPHYSICAL Noriega-Papaqui, T. Oceguera-Becerra, A. Tapia, V. Oro- RESEARCH-SPACE PHYSICS. 115: A10224. zco, V. Perez, E. G. Perez-Perez, J. S. Perkins, J. Pretz, C. Ramirez, I. Ramirez, D. Rebello, A. Renteria, J. Reyes, D. Cowee, M. M., S. P. Gary, H. Y. Wei, R. L. Tokar, and C. T. Rosa-Gonzalez, A. Rosado, J. M. Ryan, J. R. Sacahui, H. Russell. An explanation for the lack of ion cyclotron Salazar, F. Salesa, A. Sandoval, E. Santos, M. Schneider, wave generation by pickup ions at Titan: 1-D hybrid A. Shoup, S. Silich, G. Sinnis, A. J. Smith, K. Sparks, W. simulation results. 2010. Journal of Geophysical Re- Springer, F. Suarez, N. Suarez, I. Taboada, A. F. Tellez, search - Part A - Space Physics. 115 (A10): A10224 (12 G. Tenorio-Tagle, A. Tepe, P. A. Toale, K. Tollefson, I. pp.). Torres, T. N. Ukwatta, J. Valdes-Galicia, P. Vanegas, V. Vasileiou, O. Vazquez, X. Vazquez, L. Villasenor, W. Dentz, M., and D. Tartakovsky. Probability density functions Wall, J. S. Walters, D. Warner, S. Westerhoff, I. G. Wish- for passive scalars dispersed in random velocity fields. er, J. Wood, G. B. Yodh, D. Zaborov, and A. Zepeda. On 2010. Geophysical Research Letters. 37 (24): L24406. the sensitivity of the HAWC observatory to gamma-ray bursts. 2012. ASTROPARTICLE PHYSICS. 35 (10): 641. Dentz, M., and D. Tartakovsky. Probability density functions for passive scalars dispersed in random velocity fields. Baldridge, S. W., J. Valdes, O. Nedorub, B. Phrampus, M. 2010. GEOPHYSICAL RESEARCH LETTERS. 37: L24406. Cording, L. W. Braile, J. F. Ferguson, and M. Benage. Seismic Investigations of an Accommodation Zone in English, N., N. McDowell, C. Allen, and C. Mora. The effects the Northern Rio Grande Riff, New Mexico. Invited pre- of alpha-cellulose extraction and blue-stain fungus on sentation at American Geophysical Conference. (San retrospective studies of carbon and oxygen isotope Francisco, CA, December 13-17, 2010). variation in live and dead trees. 2011. Rapid Communi- cations in Mass Spectrometry. 25 (20): 3083. Berryman, E. M., J. D. Marshall, T. Rahn, S. P. Cook, and M. Litvak. Adaptation of continuous-flow cavity ring-down English, N., N. McDowell, C. Allen, and C. Mora. The effects spectroscopy for batch analysis of delta C-13 of CO2 of alpha-cellulose extraction and blue-stain fungus on and comparison with isotope ratio mass spectrometry. retrospective studies of carbon and oxygen isotope 2011. RAPID COMMUNICATIONS IN MASS SPECTROM- variation in live and dead trees. 2011. RAPID COMMU- ETRY. 25 (16): 2355. NICATIONS IN MASS SPECTROMETRY. 25 (20): 3083. Chylek, P., C. Folland, G. Lesins, and M. Dubey. Twentieth Flowers, B. A., H. H. Powers, M. K. Dubey, and N. G. Mc- century bipolar seesaw of the Arctic and Antarctic sur- Dowell. Inter-comparison of two high-accuracy fast- face air temperatures. 2010. Geophysical Research Let- response spectroscopic sensors of carbon dioxide: ters. 37 (8): L08703. a case study. 2012. ATMOSPHERIC MEASUREMENT TECHNIQUES. 5 (5): 991. Chylek, P., C. Folland, H. Dijkstra, G. Lesins, and M. Dubey. Ice-core data evidence for a prominent near 20 year Keeling, R., A. Manning, and M. Dubey. The atmospheric time-scale of the Atlantic Multidecadal Oscillation. signature of carbon capture and storage. 2011. PHILO- 2011. GEOPHYSICAL RESEARCH LETTERS. 38: L13704. SOPHICAL TRANSACTIONS OF THE ROYAL SOCIETY A-MATHEMATICAL PHYSICAL AND ENGINEERING SCI- Chylek, P., C. Folland, H. Dijkstra, G. Lesins, and M. Dubey. ENCES. 369 (1943): 2113. Ice-core data evidence for a prominent near 20 year time-scale of the Atlantic Multidecadal Oscillation. Levanic, T., M. Cater, and N. McDowell. Associations be- 327

Page 328

tween growth, wood anatomy, carbon isotope discrim- active tracer in saturated alluvium described using a ination and mortality in a Quercus robur forest. 2011. three-component cation-exchange model. 2003. Jour- TREE PHYSIOLOGY. 31 (3): 298. nal of Contaminant Hydrology. 62-63: 675. Likhachev, E. R.. Dependence of water viscosity on tem- Welling, D. T., V. K. Jordanova, S. G. Zaharia, A. Glocer, and perature and pressure. 2003. TECHNICAL PHYSICS. 48 G. Toth. The effects of dynamic ionospheric outflow (4): 514. on the ring current. 2011. JOURNAL OF GEOPHYSICAL RESEARCH-SPACE PHYSICS. 116: A00J19. MacCarthy, J. K., B. Borchers, and R. C. Aster. Efficient stochastic estimation of the model resolution matrix Welling, D. T., V. K. Jordanova, S. G. Zaharia, A. Glocer, and diagonal and generalized cross-validation for large G. Toth. The effects of dynamic ionospheric outflow geophysical inverse problems. 2011. JOURNAL OF GEO- on the ring current. 2011. JOURNAL OF GEOPHYSICAL PHYSICAL RESEARCH-SOLID EARTH. 116: B10304. RESEARCH-SPACE PHYSICS. 116: A00J19. McCabe, M., P. Chylek, and M. Dubey. Detecting ice-sheet Wiens, R. C.. A tale of 2 Missions …and 2 Very Different melt area over western Greenland using MODIS and Landings. Invited presentation at IGPP External Advi- AMSR-E data for the summer periods of 2002-2006. sory Review Committee Dinner presentation. (Santa 2011. Remote Sensing Letters. 2 (2): 117. Fe, NM, July 17, 2012). Mesler, R., D. Whalen, N. Lloyd-Ronning, C. Fryer, and Pihlstro&uml. Gamma-ray Bursts in Circumstellar Shells: A Possible Explanation for Flares. 2012. Astro- physical Journal. 757 (2): 117 (11 pp.). Reimus, P., G. Pohll, T. Mihevc, J. Chapman, M. Haga, B. Lyles, S. Kosinski, R. Niswonger, and P. Sanders. Test- ing and parameterizing a conceptual model for solute transport in a fractured granite using multiple tracers in a forced-gradient test. 2003. Water Resources Re- search. 39 (12): SBH141. Reimus, null.. Laboratory testing and modeling to evaluate perfluorocarbon compounds as tracers in geothermal systems. 2011. DOE. Russell, C. T., D. R. Weimer, N. Omidi, L. K. Jian, J. G. Luh- mann, and R. J. Strangeway. Interplanetary field en- hancements travel at the solar wind speed. 2010. GEO- PHYSICAL RESEARCH LETTERS. 37: L07204. Russell, C. T., D. R. Weimer, N. Omidi, L. K. Jian, J. G. Luh- mann, and R. J. Strangeway. Interplanetary field en- hancements travel at the solar wind speed. 2010. GEO- PHYSICAL RESEARCH LETTERS. 37: L07204. Severino, G., D. M. Tartakovsky, G. Srinivasan, and H. Viswanathan. Lagrangian models of reactive transport in heterogeneous porous media with uncertain prop- erties. 2012. PROCEEDINGS OF THE ROYAL SOCIETY A-MATHEMATICAL PHYSICAL AND ENGINEERING SCI- ENCES. 468 (2140): 1154. Severino, G., D. M. Tartakovsky, G. Srinivasan, and H. Viswanathan. Lagrangian models of reactive transport in heterogeneous porous media with uncertain prop- erties. 2012. PROCEEDINGS OF THE ROYAL SOCIETY A-MATHEMATICAL PHYSICAL AND ENGINEERING SCI- ENCES. 468 (2140): 1154. Sullivan, E., P. Reimus, and D. Counce. Transport of a re- 328

Page 329

Environmental and Biological Sciences Directed Research Continuing Project Maximizing Flux Through Engineered Metabolic Pathways Clifford J. Unkefer 20130091DR Introduction Benefit to National Security Missions The long-term goal of synthetic biology is to engineer Synthetic biology is on the verge of producing practical microorganisms to produce high-value products for solutions in a wide variety of application areas (energy, biotechnological applications like biofuels production, sensing, medicine, materials). This DR will allow LANL environmental remediation and biosynthesis of phar- to become a major contributor to this rapidly growing maceuticals. To achieve this goal, it will be necessary to field. We will apply our new synthetic biology tool set to engineer functional multi step metabolic pathways in solve a key limitation in biofuel production. This project organisms, which will require a new approach to regu- will help sustain LANL leadership in algal bioenergy and late the expression of each of the enzymes in a meta- contribute directly to LANL’s mission in Energy Security. bolic pathway in order to optimize the flux through the At the same time there are many potential application pathway. of synthetic biology in health security. The riboregula- tors and synthetic biology tools developed here will We seek to develop a new paradigm in metabolic path- provide underlying science and technology necessary to way engineering by creating a simple approach to opti- apply synthetic biology in LANL health security missions mize the expression of individual enzyme in a pathway including developing countermeasures for emerging to maximize flux through the pathway. Because previous pathogens and biological weapons. This project will posi- attempts to improve metabolic flux by over-expressing tion LANL to compete as a strong science and technology individual enzymes have achieved only limited success, player for new synthetic biology calls from DOE/EERE we will simultaneously optimize the expression of each and DOE/BER in biofuels research, as well as DARPA and of the enzymes in the pathway. This will be done using NIH funding for health security. the novel strategy of directed evolution of riboregula- tors using in vivo selection. Riboregulators are recently Progress discovered class of RNA molecules that can be used to To optimize the expression of enzymes in a metabolic control the translation of a gene and it is possible to pathway, we are developing two-component riboregula- design an ensemble of riboregulators so one can indi- tors that will independently control the concentration of vidually control the translation of each of the genes that each enzyme in our engineered metabolic pathway. Our encode for a metabolic pathway. Riboregulators make cis-repressor is designed to form a helix in the untrans- pathway optimization possible because: 1) they allow lated region of the messenger RNA, which occludes the the expression of a particular gene to be tuned over a ribosome-binding site, blocking translation. Thus, mes- wide dynamic range; 2) a set of riboregulators can be senger RNAs for our target genes have the default state readily designed to independently regulate the expres- of gene expression turned off. Regulation of expression sion of multiple genes; and 3) they are very tractable for involves a second trans-activating RNA designed to directed evolution, because they can be finely tuned by hybridize with the messenger RNA in a way that frees varying only a few nucleotides. Our riboregulators ap- the ribosome-binding site, allowing ribosome binding proach to synthetic biology will have wide applicability. and protein synthesis. We have designed and tested We will demonstrate this novel approach by engineering the inherent dynamic range of our first generation of a more carbon- and energy-efficient pathway for lipid the two-component riboregulators. The cis-repressing biosynthesis in a photosynthetic organism, an essential sequence was designed to be universal with respect to goal of biofuel research. the regulated gene. This was accomplished by utilizing 329

Page 330

sequence only in the 5’-untranslated region to fold into a methanol utilizing bacterium in which PEPC also serves secondary structure (helices) having a predicted thermo- a central role in carbon assimilation. After expressing dynamic stability and blocking the ribosome-binding site. and purifying this PEPC, we developed an activity assay Two trans-activators were designed to modulate repres- coupled to malate dehydrogenase and measured kinetic sion by the cis-acting elements. Each trans-activator was parameters. We demonstrated that indeed malate and designed with sequence elements that afforded differen- aspartate do not inhibit the PEP carboxylase activity. Simi- tial stability of inter-molecular helices vs. intra-molecular lar studies are near completion for glyoxylate carboligase, helices formed with the cis-repressor. When the ther- and 2-hydroxy-3-oxopropionate reductase, and we have modynamic stability of the intra-molecular helices is far constructed expression systems for all the other enzymes greater than that of the productive inter-molecular helices, in our pathway. translation is all-off. The converse situation produces the all-on state. A single copy plasmid containing the trans- In preparation for future engineering in a phototroph, we acting riboregulators and their target cis-repressed gene have established protocols for growth and transformation was constructed using the Gibson assembly method, which of the cyanobacterial strain, Synechococcus elongatus allows for the efficient, simultaneous insertion of multiple PCC7942. In addition, we have demonstrated lipid produc- DNA fragments into a vector. The cis-repressed gene was tion from an engineered strain, provided by our collabora- the chloramphenicol acetyltransferase (CAT) gene, which tors, harboring two heterologous genes responsible for imparts resistance to the antibiotic chloramphenicol. The the biosynthesis of triacylglycerides (TAGs) using a neutral plasmid was used to transform E. coli, and transformants lipid dye with flow cytometry-based detection of the lipid were confirmed by DNA sequencing. Then individual colo- droplets. Using flow cytometry to screen for lipid produc- nies were plated on medium containing various concentra- tion will be essential for pathway optimization. tions of chloramphenicol. Growth differences between the all-off and all-on variants were apparent indicating that Future Work the trans-activators are able to increase the expression of We will continue to design and test a family of riboregula- the CAT imparting greater resistance to chloramphenicol. tors that can be used for translational regulation of the However, our analysis of the antibiotic resistance data expression of each of the enzymes in a complete metabolic indicated only ~2-fold increase in the expression of CAT in pathway. In FY13, we will complete trans-activator library the presence of the all-on vs. the all-off trans-activators. selection, verify and refine our ability to target multiple We carried out other experiments that proved that while genes simultaneously and construct an initial four-gene the basil level of expression of CAT was significant the pathway in E. coli. The final step of riboregulator construc- trans-activator did indeed increase the expression. Overall tion will demonstrate our ability to selectively trans-activa- these experiments were encouraging because we dem- tors to multiple genes, and our strategy of directed evolu- onstrated that the two-component riboregulators could tion of riboregulators through in vivo selection. To facilitate be used to control gene expression, but it is necessary this process, we will use two orthogonal fluorescent to design the cis-repressor with greater thermodynamic reporters and two antibiotic reporters. As outlined in the stability to minimize expression in the absence of to trans- proposal, the library of riboregulators will be composed of activator. As a result we have redesigned the cis-repressor pools of randomized trans-activators driving the expres- and corresponding trans-activators, increasing the size of sion of the cis-repressed fluorescent and antibiotic marker the helix-forming regions from 16 to 32 residues. We have genes. Targeting sequences will be tested simultaneously just completed the construction of this second-generation by incorporating two orthogonal trans-activators targeting two-component riboregulators. The dynamic range of this two proteins at different locations in the vector. Two dis- new system is now being evaluated using all-on and all-off tinct targeting sequences will be combined with two regu- trans-activators. We are confidant that we will soon have latory structures from either the high activity pool or low a library of selected riobregulators with an activity range activity pool. These trans-activators will allow us to verify spanning two orders of magnitude. our ability to target and differentially regulate two genes simultaneously. We will design and test 10 distinct target- To evaluate our progress toward pathway optimization, we ing sequences that preserve dynamic range and minimize will develop a full kinetic model of our engineered path- regulatory crosstalk for use in the construction of novel way. Toward this end we have initiated a project to ex- multigene pathways. Once we have finalized the design of press, purify and characterize the kinetics of the enzymes the riboregulators, we will start assembling the large DNA in our pathway. Essential to our success was a phospho- necessary to encode the metabolic pathway in both E. coli enolpyruvate carboxylase (PEPC) that is not inhibited by and Synechococcus elongates. Initially we will focus on the metabolites in our pathway. We selected a PEPC gene from first four enzymes in the pathway, which should recover 330

Page 331

growth in mutants unable to express pyruvate dehydro- Novikova, I. V., A. Dharap, S. P. Hennelly, and K. Y. Sanbon- genase. Finally, to evaluate our progress toward pathway matsu. Shotgun secondary structure determination of optimization, we will develop a full kinetic model of our long non-coding RNAs. 2013. Methods. 63: 170. engineered pathway. We will finish obtaining the necessary Novikova, I. V., S. P. Hennelly, and K. Y. Sanbonmatsu. Sizing kinetic parameters for the enzymes in the pathway. up long non-coding RNAs: Do lncRNAs have secondary and tertiary structure? . 2012. BioArchitecture. 2 (6): Conclusion 189. We will develop a family of riboregulators that can be used Novikova, I. V., S. P. Hennelly, and K. Y. Sanbonmatsu. . Ex- for translational regulation of the expression of each of the perimentally determined secondary structures of can- enzymes in a complete metabolic pathway. Demonstrate a cer-related long non-coding RNAs. 013. In 245th ACS strategy for the simultaneous directed evolution of ribo- National Meeting & Exposition. (New Orleans, 7-11 regulators to optimize the expression of the enzymes to April, 2013). , p. PHYS. Abstracts of Papers: ACS. maximize flux through a metabolic pathway. We will use known enzymes to engineer a more carbon- Novikova, I. V., S. P. Hennelly, and K. Y. Sanbonmatsu. Tack- and energy-efficient pathway for lipid biosynthesis es- ling structures of long noncoding RNAs. To appear in sential for biofuel production. In addition, we will use the International Journal of Molecular Science. directed evolution of riboregulators strategy developed Novikova, I. V., S. P. Hennely, C. S. Tung, and K. Y. San- above to optimize growth and lipid production in Syn- bonmatsu. Rise of the RNA machines: Exploring the echococcus elongates, a cyanobacterium. structure of long non-coding RNAs. 2013. Journal of Molecular Biology. 425: 3731. Publications Sanbonmatsu, K. Y.. Rise of the RNA Machines: Long Non- Hall, R. S., R. Martí-Arbona, S. P. Hennelly, T. S. Maity, F. Coding RNAs. Invited presentation at 10th Horizons Mu, J. M. Dunbar, C. J. Unkefer, and P. J. Unkefer. In- in Molecular Biology Conference International . (Max vitro Characterization of an L-Kynurenine-Responsive Planck Institute, Gottingen, Germany, Sept 9-12 2013). Transcription Regulator of the Oxidative Tryptophan Degradation Pathway in Burkholderia xenovorans. Sanbonmatsu, K. Y.. Integrating simulations and experi- 2013. Journal of Molecular Biology Research. 3 (1): 55. ments of the SAM-I riboswitch. Invited presentation at RNA Dynamics. (Telluride, Colorado, 22-26 July, 2013). Hennelly, S. P., I. V. Novikova, and K. Y. Sanbonmatsu. The expression platform and the aptamer: cooperativity Sanbonmatsu, K. Y.. Large-scale simulations of biomolecu- between Mg2+ and ligand in the SAM-I riboswitch. lar machines. Invited presentation at Biosupercomput- 2013. NUCLEIC ACIDS RESEARCH. 41 (3): 1922. ing Symposium. (Tokyo Japan, December 2012). Hennelly, S. P., I. V. Novikova, and K. Y. Sanbonmatsu. The Sanbonmatsu, K. Y.. Understanding the atomistic mecha- expression platform and the aptamer: cooperativity nism of magnesium effects of RNA dynamics. Invited between Mg2+ and ligand in the SAM-I riboswitch. presentation at Mini-symposium on Modeling, Simula- 2013. Nucleic Acids Research. 41 (3): 1922. tion and Function of Biomolecular Assemblies. (Tokyo University, Tokyo Japan, November 2012). Jarchow-Choy, S., D. Fox, E. Schmidt, A. Koppisch, R. Marti- Arbona, and C. Unkefer. Monitoring the distribution of isoprenoid products from the MEP pathway in algae and cyanobacteria. 2013. ABSTRACTS OF PAPERS OF THE AMERICAN CHEMICAL SOCIETY. 245: -. Martí-Arbona, R., T. S. Maity, J. M. Dunbar, C. J. Unkefer, and P. J. Unkefer. Discovery of a Choline-Responsive Transcriptional Regulator in Burkholderia xenovorans. 2013. Journal of Molecular Biology Research . 3 (1): 91. Nguyen, T. T., R. Martí-Arbona, R. S. Hall, T. Maity, Y. E. Val- dez, J. M. Dunbar, C. J. Unkefer, and P. J. Unkefer. Iden- tification and In-vivo Characterization of a Novel OhrR Transcriptional Regulator in Burkholderia xenovorans LB400. 2013. Journal of Molecular Biology Research. 3 (1): 37. 331

Page 332

Environmental and Biological Sciences Directed Research Final Report Terrestrial Vegetation, CO Emissions, and Climate Dynamics 2 Nathan G. Mcdowell 20110014DR Abstract ingly, emerging evidence of a global rise in widespread Vegetation mortality rates are increasing globally, appear vegetation mortality strongly supports these predic- to be driven by our changing climate, and are releasing tions. However, we do not have a system for monitoring large and sustained amounts of CO2 to the atmosphere. vegetation mortality remotely, which greatly limits our The critical urgency of forecasting climate impacts and ability to evaluate models and predict future mortality. feedbacks, coupled with the rising international focus Our objectives were to develop and evaluate a high on carbon management, make understanding, quan- precision, remotely sensed system for determining for- tifying, and predicting terrestrial carbon balance and est mortality, to improve models of vegetation mortal- subsequent climate impacts one of the greatest science ity during drought, and to couple these to assess the challenges currently facing America. We have built an ex- climate forcing impacts of mortality. tremely high resolution, highly accurate monitoring sys- tem for documenting forest mortality and for simulating Scientific Approach and Accomplishments the consequences of the mortality on climate forcing. We had three key tasks: 1) remote sensing detection of Through this effort, we have generated three separate vegetation stress and mortality, 2) estimating carbon predictions of the future of forests in western USA. All storage and fluxes, and 3) assessing climate consequenc- results suggest that southwestern USA will be conifer- es based on results of (1) and (2). Our goal is to develop free due to chronic drought by 2050 AD, with the Pacific a globally comprehensive tool; however, because the Northwestern forests lagging only slightly behind. Tests computational requirements for complete global cov- of this approach in other regions suggest it is robust to erage are beyond the capacity of a DR, we tested our assumptions associated with variation in forest type. approach utilizing data sets of vegetation mortality and The monitoring and prediction system we have devel- carbon storage, CO2 fluxes, and climate from multiple oped promises to continue yielding high impact, highly disparate locations. informative scientific results regarding the current and future changes to our global forests as climate continues Our accomplishments were many. Perhaps most impor- to warm. tantly, we have three independent results (Jiang et al. 2013; Williams et al. 2013; McDowell et al. unpublished Background and Research Objectives manuscript) showing that southwestern forests will Terrestrial vegetation buffers fossil fuel emissions be largely gone by 2050 AD. This shocking result has through storing CO2. Unfortunately, rapidly changing cli- gained a lot of press (BBC, NPR, NY Times, for example). mate and accelerating vegetation mortality pose a threat We have also presented this result to the UC-Board of to the long-term maintenance of the land CO2 sink. Pre- Regents, to the LANS-Board of Governors, and to Energy dictions from dynamic global vegetation models suggest Secretary Dr. Ernest Moniz. These results were the sum- that despite enhanced photosynthesis due to elevated mation of many technical advances, including successful atmospheric CO2, increased temperature and reduced development of a high precision mortality monitoring precipitation (drought) are expected to overwhelm this system (Garrity et al. 2012; Mu et al. 2013; Skurikhin et positive benefit and accelerate mortality, thus tipping al. 2013; Gangodoagamage et al. in review), and radical the terrestrial system from a negative to a positive improvements to the mechanistic underpinnings of mor- climate feedback in upcoming decades. The bulk of this tality simulations (Xu et al. 2012; 2013; McDowell et al. response is due to drought-induced mortality. Alarm- 2013). The science developed through this LDRD project 332

Page 333

is ongoing through a recently held workshop where we McDowell, N., D. Beerling, D. Breshears, R. Fisher, K. Raffa, brought together the world’s experts in remote sensing of and M. Stitt. The interdependence of mechanisms disturbance for a summary paper on the way forward, and underlying climate-driven vegetation mortality. 2011. TRENDS IN ECOLOGY & EVOLUTION. 26 (10): 523. through pending funding to DOE to continue the applica- tion of the integrated remote sensing and modeling. Mu, Q.. A remote sensed global terrestrial drought severity index. . 2013. Bulletin of the American Meteorological Impact on National Missions Society. 94: 83. The impact is through allowing documentation of current and prediction of future vegetation consequences of cli- Williams, P., and e. al. Forest response to drought in the mate change so that we can predict impacts on resources North American Southwest from 1000-2100 A.D. . 2012. Nature Climate Change. : 1. at home and globally. Furthermore, according to multiple New Mexico state senators and Secretary Moniz, we are Williams, P., and e. al. Who’s the new sheriff in town regu- doing the nation a favor by bringing these shocking results lating boreal forest growth?. 2012. Environmental Re- to the awareness of the public. search Letters. : 041104. Publications Xu, C., C. Liang, S. Wullschleger, C. Wilson, and N. McDow- Allen, C., A. Macalady, H. Chenchouni, D. Bachelet, N. Mc- ell. Importance of feedback loops between soil inor- Dowell, M. Vennetier, T. Kitzberger, A. Rigling, D. Bres- ganic nitrogen and microbial communities in the het- hears, E. H. Hogg, P. Gonzalez, R. Fensham, Z. Zhang, erotrophic soil respiration response to global warming. J. Castro, N. Demidova, J. Lim, G. Allard, S. Running, A. 2011. NATURE REVIEWS MICROBIOLOGY. 9 (3). Semerci, and N. Cobb. A global overview of drought Xu, C., N. G. McDowell, S. Sevanto, and R. Fisher. Our lim- and heat-induced tree mortality reveals emerging ited ability to predict vegetation responses to water climate change risks for forests. 2010. In Conference stress. To appear in New Phytologist. on Adaptation of Forests and Forest Management to Changing Climate with Emphasis on Forest Health ; Xu, C., and e. al. Toward a better understanding of nitro- 20080825 - 20080828 ; Umea, SWEDEN. Vol. 259, 4, SI gen-photosynthesis relationships based on nitrogen Edition, p. 660. trade-offs among light absorption, electron transport, carboxylation, and carbon sinks. . 2012. Plos One. 7: Craine, J., and N. McDowell. Resource limitation, tolerance, e37914. and the future of ecological plant classification. 2013. Frontiers in Functional Plant Biology. : fpls.2012.00246. Franks, P., and N. McDowell. Forest sensitivity to CO2. . 2013. New Phytologist. : nph.12104. Garrity, S., and e. al. Quantifying drought-induced tree mortality in a piñon-juniper woodland using multitem- poral high spatial resolution satellite imagery. 2013. Remote sensing of Environment. 129: 54. Jiang, X., S. Rauscher, T. Ringler, D. Lawrence, P. Williams, and M. Cai. Increasing risk of tree mortality in western North America towards the end of the 21st century. . 2013. J. of Climate. : JCLI. McDowell, N.. Mechanisms Linking Drought, Hydraulics, Carbon Metabolism, and Vegetation Mortality. 2011. PLANT PHYSIOLOGY. 155 (3): 1051. McDowell, N. G., M. G. Ryan, M. Zeppel, and D. Tissue. Im- proving our knowledge of drought-induced forest mor- tality through experiments, observations, and model- ing. To appear in New Phytologist. McDowell, N. G., and e. al. Evaluating theories of drought- induced vegetation mortality using a multi-model-ex- periment framework. To appear in New Phytologist. 333

Page 334

Environmental and Biological Sciences Directed Research Final Report Advanced Metagenomic Analysis to Understand Dynamics of Soil Microbial Community under Conditions of Climate Change Shunsheng Han 20110034DR Abstract facilitate the analysis processes as the second approach Microbes form the majority biomass on earth and are re- will provide more reference genomes from the same lated to almost all aspects of human life. The goal in this sample for accurate and unbiased analysis of data from project was to develop tools and techniques to probe the first shotgun metagenomic approach. Recent devel- the dynamics of microbial communities of biological soil opments in metagenomic sequencing make it possible crusts in arid deserts and to document their response to generate enough data for studying a relative complex to various climate change related perturbations. Our community. However, as only a small fraction of all bac- technologies improved performance in two previous teria have been cultivated in isolation, a pre-requisite for bottleneck areas. One is to sequence genomes of uncul- standard genome sequencing efforts, it is very difficult to turable bacterial species, and another is the analysis of sequence each member of microbial community sepa- the huge amount of data from metagenomic sequencing rately. This problem has been only partially addressed in a timely fashion. We applied these technologies in by DNA amplification of isolated genomes from a single analyzing soil microbiota in arid soils from Utah, one of cell (known as Single Cell Genomics, or SCG). Although our field sites for climate manipulation studies. sequences have been obtained from a number of dif- ferent bacteria at the single cell level, the current SCG Background and Research Objectives method results in hundreds to thousands of sequences Microbial communities are complex ’supra-organisms’ for a genome with poor coverage, rendering it difficult of bacteria, archaea, single-cell eukaryotes and viruses to analyze. This speaks to the need for a more robust that function together as a whole in order to survive and method to sequence genomes from single cells. Meth- thrive in their natural environmental niches. Microbes ods we developed are able to improve the coverage and make up the majority of the living biomass on Earth and to reduce the number of gaps to a near completion. as such, are major contributors to carbon cycling and to Existing methods for understanding uncultured microbial the biogenic production and destruction of other green- communities fall into three broad analysis techniques. house gases, including CO2, N2O and methane. Our None of these methods are adequate to fully analyze the present lack of understanding regarding the biochemical huge amount of metagenomic data. Targeted metage- fluxes of C and N within these microbial communities nomics uses universal primers to preferentially sequence constitutes a major source of uncertainty in our predic- one particular gene (usually 16S rRNA). This method tions of the consequences of climate change. cannot provide functional profiles. Fragment recruit- Characterizing the collective genomes of microbial ment (or nucleotide BLAST) is computationally expen- communities, a nascent field known as metagenomics, sive and requires closely related reference organisms. can help bridge this information gap by revealing how Protein-based BLAST is computationally expensive, fails climate change impacts the broad composition and with short reads (less than 100 bp) and confuses se- function of members within environmental microbial quence similarities arising from inheritance with similari- communities, and by quantifying the net effect of these ties arising from functional pressure (natural selection). changes on both the C and N cycles. There are two ways Because the next generation of DNA sequencing technol- to characterize metagenomes of microbial communi- ogy generates gigabytes data in short sequences, new ties, sequencing the community collectively (shotgun analysis techniques are needed. metagenome) or each member separately. A combina- Analysis methods based on local patterns offer enor- tion of the two approaches, as proposed in Figure 1, will 334

Page 335

mous advantages over methods based on global-similarity Lactobacillus acidophilus, a gut microbe that is also a measures such as BLAST. We based our method around significant constituent of yoghurt. This antibody was used signatures of amino-acid 10-mers. Our signature-based to assess the abundance of L. acidophilus in a number of method bypasses many paralog problems, uses signatures yoghurt samples, with the lowest being 0.3%. The genome that persist across larger phylogenetic distances than do of the amplified L. acidophilus purified from this yoghurt nucleic acids, produces matches that are highly significant, sample was then sequenced using Illumina sequencing, and is quite rapid to implement because patterns can be and the quality and extent of coverage, of a de novo as- organized and searched via index structures. Some 20 mil- sembly, was equivalent to that of a cultured organism. This lion orthogenomic amino-acid 10-mer signatures are suf- work, which demonstrated the power of this approach on ficient to cover all of current sequenced bacterial diversity a true mixed-culture sample, has been submitted for pub- down to the genus level while allowing detection of reads lication. The second set of experiments were carried out from never-before-sequenced phyla with reasonable likeli- on cultured soil bacteria. In this case two specific antibod- hood. ies were selected against Streptomyces species cultured from soil. These antibodies were able to specifically stain Scientific Approach and Accomplishments the cultured Streptomyces, but not related species. Most Artificial polyploidy significantly, as shown in Figure 3, these antibodies were The typical genome recovery rate from SCG is generally also able to label a small number of bacteria in soil, dem- less than 50% from environmental samples. The low recov- onstrating the power of the approach. ery rate is related to amplification bias and DNA breakage Sequedex – the bioinformatics tool for rapid analysis of due to only one copy of a genome. Artificial polyploidy is metagenomic data a technique that can be employed to increase the genome We developed a bio-informatics tool that leverages our copy number in individual cells by blocking cell division understanding of evolution to identify local peptide sig- while allowing for cell growth. The resulting cells are much natures to classify phylogenetically and functionally raw larger with multiple genome copies. Sequencing from poly- sequence reads (4). An important feature of the algorithm ploid cells significantly increases the coverage of genomic is its ability to extrapolate to organisms not present in the sequence. database. That feature is essential when applied to soil Gel microdroplet (GMD) metagenomes. Much of our effort focused in releasing GMD allows us to perform DNA amplification on hundreds software to the bio-informatics community that works on to thousands of copies of a bacterial genome. GMDs are all platforms (www.sequedex.lanl.gov). The methodology small spheres composed of an agarose matrix in which one offers unprecedented speed of analysis, and was awarded can grow colonies of bacteria starting from a single cell. an R&D 100 award in 2012. The documentation of the They are very permeable and thus allow for communica- released software is a substantial document that provides, tion among the microbial community members, which in addition to guidance on how to install and run the many bacteria require to complete some undetermined Sequedex, various demonstration of how to use the output aspect of their life cycle. Critically, the reproducing cells of our tool to answer various questions of interest. In a remain spatially separate so individual colonies can be ma- single step, the tool allows us to calculate phylogenetic and nipulated following group culture. In this study, we dem- functional profiles of a metagenome. These profiles can onstrate its utility for producing near-complete genomes be use to compare multiple samples, and to identify com- of environmental microorganisms by using human fecal mon determinants of an ecosystem (Figure 4, 5). and oral bacteria as a demonstration. We show that the The usefulness of Sequedex goes well beyond the profiles. resulting assemblies are vastly superior to those produced In particular, it provides a filter to classify raw sequence via single-cell sequencing and that GMDs also have the reads both in terms of phylogeny and functions. Once potential to produce genomic data for currently uncultured filtered, the reads can be assembled de novo, mapped to and unsequenced organisms. The near complete genomes identified near neighbors, and placed into phylogenetic in a single sample allow us to assess the genome diversity trees. As a result, Sequedex naturally fits into modern in environmental samples (Figure 2). pipelines for analyzing reads from next generation se- Phage antibody as tool for bacterial species isolation from quencers. environmental microbiota Studies on soil microbiota under conditions of climate Antibodies were selected using in vitro display libraries changes against two bacterial classes. In the first, proof of principle Biological soil crusts (hereafter referred to as biocrusts, experiment, a highly specific antibody was selected against colonize the interspace soils between patchily distributed 335

Page 336

plants in dryland ecosystems. Biocrusts have been identi- Impact on National Missions fied in dryland ecosystems on every continent on Earth and The project falls in the scope of two grand challenges for function in retaining water, fixing soil particles, and fixing science and engineering at Los Alamos: Complex Biological atmospheric nitrogen and carbon. Our first task in the Systems and Information, Science and Technology. The project was to assess the ability of the current available scientific goal of deep understanding of microbial commu- sequencing technologies, using the available tools, in their nities in arid lands will eventually facilitate the progress of ability to characterize microbial communities in a complex another grand challenge, Energy and Earth Systems. It soil environment and document their response to climate directly addresses key goals identified in the 2009 DOE change, such as elevated CO2. report “Carbon Cycling and Biosequestration: Integrating Biology and Climate Through System Science.” Studying We compared three different sequencing datasets, tar- microbial community with LANL’s advantageous metage- geted amplification and sequencing of bacterial 16S rRNA nomic technologies is one of the focuses of the grand genes, shotgun metagenome sequencing, and parsing challenge of Complex Biological Systems. These experi- rRNA genes out of the metagenomes. Our data indicated mental and computational tools developed in the projects shotgun metagenomes were not able to discriminate eco- has put LANL at the forefront of metagenomics, and logical differences as well as targeted methods. This was strengthened our leadership position in genomics and apparently due to a lack of reference genomes and com- metagenomics.Figure 1. Probing the dynamics of metagenome. putational ability to process sequence datasets, a critical need being addressed in the other section of the project. Our next objective was to assess the spatial heterogeneity of desert biocrust communities. Using a spatially nested sampling procedure we collected samples representing three different soil parent materials and different spatial scales, including a vertical dimension. We found that des- ert biocrusts were similar over large horizontal scales but differed significantly at small vertical scales. We then wanted to study biological soil crusts at a con- tinental scale to address whether soil crusts in different systems are unified by similar microbial populations. In this regard we collected soil crusts from Ward Hunt Island, Canada, the northern most terrestrial system on Earth . Soil crusts on Ward Hunt Island were similar in diversity Microbiome in soil crust was sequenced as a community, and composition to crusts from more temperate latitudes, metagenome, and individually as single cell or microcolonies suggesting a biocrusts share similar structures even at propagated from individual cells. Then the two sets of data were great spatial scales. analyzed together with Sequedex. Our final work involved documenting the response of soil communities to field manipulations simulating particular climate change and land use perturbations. In this vein we collected metagenomic samples to document changes in the microbial community structure and function as well as transcriptomes to characterize the specific functions microbes are employing under various climate change scenarios. We anticipate two to three more publications arising from this work in the near future. Besides studies on soil microbiome, we also participated in analysis of human microbiomes associated with Hu- man Microbiome Project and ocean microbiomes resulted in several publications in high-ranking scientific journals Figure 2. Genomes completed with GMD technology. The reddish including Nature. color indicates the differences in these almost finished genomes comparing to a reference genome. 336

Page 337

Publications A framework for human microbiome research. 2012. Na- ture. 486 (7402): 215. Structure, function and diversity of the healthy human mi- crobiome. 2012. Nature. 486 (7402): 207. Beckloff, N., S. Starkenburg, and P. Chain. Bacterial genome annotation. 2012. Methods Mol Biol. 881: 471. Berendzen, J., W. J. Bruno, J. D. Cohn, N. W. Hengartner, C. R. Kuske, B. H. McMahon, M. A. Wolinsky, and G. Xie. Accurate, sensitive, and rapid classification of short random fragments of bacterial DNA using phylogenetic signature peptide 10-mers. 2012. BMC Research Notes. 5: 460. Close, D. W., F. Ferrara, A. E. K. Dichosa, A. R. Daughton, H. E. Daligault, K. G. Reitenga, N. Velappan, T. C. Sanchez, S. Iyer, C. S. Han, and A. R. M. Bradbury. Phage anti- bodies to dissect microbiomes for complete de novo genome sequencing of rare microbes. BMC Microbiol- Figure 3. Labeling of bacterial cells from soil sample with phage ogy. antibody. Column 1 (left) phase contrast microscopy of individual soil bacteria; column 2: DNA stain; column 3: staining with Dichosa, A. E K, M. Fitzsimons, L. L Weston1, L. G. Preteska, antibodies; column 4: merging of column 2 and 3. The figure J. P. Snook, C. C. Lo, X. Zhang, W. Gu, K. McMurry, L. D. shows that a small number of soil bacteria were labeled with the Green, P. S. Chain, J. C. Detter, and C. S. Han. Artificial specifically selected antibody. polyploidy improves bacterial single cell genomics. 2012. PLOS One. 7 (5): e37387. Dichosa, A. E., A. R. Daughton, K. G. Reitenga, M. S. Fitzsi- mons, and C. S. Han. Capturing and cultivating single cells in gel microdroplets (GMDs) to obtain near com- plete genomes. To appear in NATURE PROTOCOL. Dunbar, J., S. Eichorst, L. Gallegos-Graves, G. Xie, N. Hen- gartner, R. Evans, B. Hungate, R. Jackson, J. Megonigal, C. Schadt, R. Vilgalys, D. Zak, and C. Kuske. Common bacterial responses in six ecosystems exposed to ten years of elevated atmospheric carbon dioxide. 2012. Figure 4. Phylogenetic profiles of 242 environmental metage- Environmental Microbiology 2012 May;14(5):1145-58. nomes. doi: 10.1111/j.1462-2920.2011.02695.x. Epub 2012 Jan 20. 14 (5): 1145. Fitzsimons, M. S., M. Novotny, C. C. Lo, A. E. K. Dichosa, J. L. Yee-Greenbaum, J. P. Snook, W. Gu, O. Chertkov, K. W. Davenport, K. McMurry, K. G. Reitenga, A. R. Daugh- ton, J. He, S. L. Johnson, C. D. Gleasner, P. L. Wills, B. Parson-Quintana, P. S. Chain, J. C. Detter, R. S. Lasken, and C. S. Han . Nearly finished genomes produced using gel microdroplet culturing reveals substantial in- traspecies genomic diversity within the human micro- biome. 2013. Genome Research. 23 (5): 878. Hu, B., G. Xie, C. C. Lo, S. R. Starkenburg, and P. S. Chain. Pathogen comparative genomics in the next-genera- Figure 5. Functional profiles of 400 human microbiota metage- tion sequencing era: genome alignments, pangenomics nomes. Microbiota from different body sites had similar function- and metagenomics. 2011. Brief Funct Genomics. 10 al profiles when analyzed at top level of functional categories, (6): 322. though their phylogenetic profile were distinct. 337

Page 338

Martinez-Garcia, M., D. M. Brazel, D. M. Brazel, B. K. Swan, C. Arnosti, P. S. Chain, K. G. Reitenga, G. Xie, N. J. Poul- ton, M. Lluesma Gomez, D. E. Masland, B. Thompson, W. K. Bellows, K. Ziervogel, C. C. Lo, S. Ahmed, C. D. Gleasner, C. J. Detter, and S. Stepanauskas. Capturing single cell genomes of active polysaccharide degraders: an unexpected contribution of Verrucomicrobia. 2012. PLoS One. 7 (4): e35314. Mason, O. U., T. C. Hazen, S. Borglin, P. S. Chain, E. A. Du- binsky, J. L. Fortney, J. Han, H. Y. Holman, J. Hultman, R. Lamendella, R. Mackelprang, S. Malfatti, L. M. Tom, S. G. Tringe, T. Woyke, J. Zhou, E. M. Rubin, and J. K. Jansson. Metagenome, metatranscriptome and single- cell sequencing reveal microbial response to Deepwa- ter Horizon oil spill. 2012. ISME J. 6 (9): 1715. Scholz, M. B., C. C. Lo, and P. S. Chain. Next generation se- quencing and bioinformatic bottlenecks: the current state of metagenomic data analysis. 2012. Curr Opin Biotechnol. 23 (1): 9. Steven, B., L. Gallegos-Graves, C. M. Yeager, D. Evans, J. Bel- nap, and C. R. Kuske. Dryland biological soil crust Cya- nobacteria show unexpected decreases in abundance under long-term elevated CO2. . 2012. Environmental Microbiology. 14: 3247. Steven, B., L. Gallegos-Graves, J. Belnap, and C. R. Kuske. Dryland soil bacterial communities display spatial bio- geographic patterns associated with soil depth and soil parent material. 2013. FEMS Microbiology Ecology. 86 (1): 101. Steven, B., L. Gallegos-Graves, S. R. Starkenburg, P. S. Chain, and C. R. Kuske. Targeted and shotgun metage- nomic approaches provide different descriptions of dryland soil microbial communities in a manipulated field study. . 2012. Environmental Microbiology Re- ports. 4: 248. Steven, B., M. Lionard, C. R. Kuske, and W. Vincent. High bacterial diversity of biological soil crusts in wa- ter tracks over permafrost in the high Arctic polar desert. 2011. PloS One. e71489: 10.1371/journal. pone.0071489. 338

Page 339

Environmental and Biological Sciences Directed Research Final Report Multi-scale Science Framework for Climate Treaty Verification: Attributing and Tracking Greenhouse Gas Fluxes Using Co-emitted Signatures Manvendra K. Dubey 20110081DR Abstract Our project has integrated world-class LANL expertise in As the consensus on the dangers of climate change and remote sensing, spectral inversion, in situ sensors and ocean acidification from uncontrolled anthropogenic atmospheric models to develop sound and innovative fossil energy emissions strengthens, international agree- monitoring strategies. We harnessed the unique breadth ments will demand stringent limits on greenhouse gas of Los Alamos expertise for an initial demonstration of emissions. The Copenhagen negotiations underscored verifying CO2 emissions over denied areas and quan- that a sound technical strategy to verify compliance with tifying CH4 fugitive leaks from oil and gas, both critical any such treaties is lacking, creating a major obstacle to issues for energy security. meaningful agreements. Existing accords that stipulate Background and Research Objectives concerted international action lack commitments to enforceable emissions reductions, primarily because of a Trace gas pollutants (Nitrogen oxides (NOx), sulfur failure to define verification protocols. Currently, carbon dioxide (SO2), and carbon monoxide (CO)) and carbon dioxide (CO2) emissions derived from energy consump- dioxide (CO2) emissions from anthropogenic fossil tion data are the basis of inventories, but these are energy production cause air pollution and global warm- uncertain and can be manipulated. Inventories must be ing, respectively. In the United States, pollutants (e.g. verifiable, especially over denied territories such as Chi- NOx, SO2, CO) and CO2 are considered a threat to public na and India. Monitoring national CO2 emissions from health and welfare under the Clean Air Act (CAA) and afar is challenging because of CO2’s large and variable are regulated by the Environmental Protection Agency background (390 ppm) and the dilution of anthropo- (EPA). While reporting requirements for air pollutants genic sources by atmospheric mixing. Quantification of are mature, they are under development for CO2. Re- anthropogenic CO2 emissions that are 3% of natural CO2 ported inventories convert the amount of fuel used in fluxes globally and apportioning them among nations is a specific activity to CO2 and pollutant gases produced, a daunting task. We developed a calibrated scientific using emission factors that depend on fuel composition, framework for climate treaty verification by reaching combustion efficiency, and scrubbing methods. These beyond isolated CO2 measurements to the monitoring bottom-up inventories are subject to significant uncer- of associated trace gases, in situ and remotely, to ap- tainties, manipulations, and also change with technol- portion and track emissions from distinct sources. We ogy [1]. Atmospheric observations offer an independent evaluated our method at Farmington, NM to verify the top-down method to verify reported emissions. Pollut- well calibrated in stack emissions reported by the Four ant trace gases have low atmospheric backgrounds and Corners and San Juan power plants. We compared our show large and distinct increases near various combus- regional scale remote sensing composition observations tion sources. In fact, satellite observations of NO2 have with forward model simulations to show that they can been used to evaluate regional and local emissions [2,3]. verify emissions with an accuracy of 2%. This finding will While satellite NO2 observations are useful for trend help design future verification satellites. Our remote analysis, they can underestimate the amount of NO2 observations can also discriminate the high NOx emit- because of their large observing footprint. In contrast, ting Four Corners power plant from the low NOx emit- CO2 has a large background and variability. Conse- ting San Juan. In addition, we discovered large-scale quently, source attribution and verification by direct CO2 enhancements of methane (CH4) a potent greenhouse measurements remains elusive [4], limiting our ability gas that attributed to large fugitive leaks from coal and to develop an effective global climate treaty or carbon- gas production that are missing in current inventories. trading scheme [5, 6, 7]. We postulated that measure- 339

Page 340

ments of co-emitted trace gases and isotopic composition CO2 emissions verification and fugitive CH4 leak detec- can be used to isolate anthropogenic CO2 emissions. tion [11]. To achieve this we also developed a customized Furthermore, contributions from specific sectors with dis- atmospheric model that uses the nested grid Weather tinct composition (trace gas to CO2 emission ratios, ERX = Forecast Model with chemistry (WRF-Chem). The model X/CO2, or isotopic ratio 13CO2/12CO2) can be delineated. assimilates large-scale meteorological fields from satellites We hypothesized that remote column trace gas measure- and balloon sondes (NCEP reanalysis) over the western US ments over large scales, which are less sensitive than in and downscales it to the Four Corners region. It simulates situ point surface measurements to small-scale variability the power plant plume and its regional evolution at a very from meteorology, can provide a more precise method to high resolution of 200m. LANL also used its specialized verify emissions [8, 9, 10]. We tested our hypothesis using High GRADient (HIGRAD) model to simulate the buoyant ground based in situ and remote observations at a site rise and dispersion of the hot power plant plumes into the with large power plant emissions of CO2 and pollutants, in ambient atmosphere at an ultra-high resolution of 10 m. the Four Corners, NM area. Real time in stack measurements of the CO2 and co-emit- ted pollutants made by Continuous Emission Monitoring The Four Corners region has also been a hub for conven- System in the power plants (mandated by EPA in the US) tional coal, oil and gas production and exploration for over were specified as input to the model. The results of these half a century. It is currently undergoing a shift to non- forward model simulations were then compared with our conventional hydraulic fracturing. There is concern that far field in-situ and remote atmospheric measurements to this transition to non-conventional extraction of natural evaluate our top down CO2 and NO2 verification methods. gas could lead to fugitive CH4 leaks that, if large enough, Emission ratios (e.g. ΔNOx/ΔCO2) are easier to verify since would offset the climate benefits of this clean and domes- the dispersion and dilution effects affect both species simi- tic fuel, since methane is 25 times more potent a green- larly- both are long-lived relative to the transport time. Fi- house gas than CO2. Our project is the first to provide nally, to evaluate the fugitive CH4 emissions we performed regional scale observations of fugitive leaks from conven- forward models with WRF-Chem using emissions from the tional mining that are a needed baseline to quantify the global Edgar CH4 inventory in the Four Corners Region. changes from non- conventional extraction in these oil and gas plays. Fingerprinting power plant signals: The in situ and regional columnar chemical and isotopic composition of the at- Scientific Approach and Accomplishments mosphere and its time evolution on a typical day when Our monitoring site is located in Farmington, NM (36.79°N our instruments sampled the power plant is illustrated in 108.48°W, at 1643 m above sea level), an arid region with Figure 2. It shows the comprehensive chemical detail with two large coal-fired power plants that emit approximately which we can quantify the plume [12]. We observe an 30 Mton/year and 80Kton/year of CO2 and NOx, respec- early morning column XCO2 (mixing ratio of CO2) of 398 tively. This region is the largest point source of pollution in ppm that is 5 ppm higher than clean air background of 393 North and South America. The Four Corners power plant ppm. This indicates enhancements from the power plant has high ΔNOx/ΔCO2 emission ratio and is located ~12 km from accumulation on the previous night. At 9:30 AM we South of our site. The San Juan power plant, which has see a clear rise in XCO2 that peaks at 407 ppm at 10:55 AM a low ΔNOx/ΔCO2 ratio due to environmental upgrades, (Figure 2a) and then subsides by 11.30. This time evolution is located about 3.7 km east of our site. LANL deployed a is closely tracked by all the other gases (Figures 2b-2e): comprehensively instrumented automated observatory X13CO2 increases by ~0.1 ppm, NO2 by 3.1 Dobson Units that monitors regional greenhouse (CO2, CH4, N2O) and (DU), and XCO by ~30 ppb. The in situ measurements of pollutant gases (CO, NO2, O3, C2H6) retrieved from three CO2, 13CO2, NO2, and CO for June 6, 2012 also shown in solar spectrometers (Bruker 125 HR and Pandora), in situ Figure 3, indicate synchronous but much larger increases measurements of CO2, CH4, CO, 13CH4, 13CO2, H2O and in all species at the surface that occurred between ~9 AM HOD made with Picarro laser cavity ring-down analyzer, in and 12 PM local time. situ measurements of air pollutants NO2, NOx, SO2 and We have analyzed regional column ∆NO2/∆CO2 ratios over O3 using standardized EPA sensors (Figure 1). Our high a 6-month period and compared them to in situ plume ra- frequency (seconds to minutes) measurements are con- tios to show that a large and stable fraction of the regional tinuous, with the solar spectrometers operating during atmosphere is polluted (70% to 75%) in the vicinity of large daytime. The data collection began March of 2011 and power plant sources in Four Corners [12]. We demon- ended in November 2013. We have assembled an exhaus- strate that our remote column ∆NO2/∆CO2 observations tive and unprecedented data on atmospheric composition can resolve variations from regional sources with distinct at multiple scales that spans over 2.5 years to enable both 340

Page 341

emission factors and apportion them using other signa- demonstrates that remote sensing of regional CO2 can tures like SO2 and CO. Our findings offer promise for future verify emission with an accuracy of about 2%, considerably satellite-based monitoring approaches that simultaneously greater than achievable by distributed in situ sensors [8, 9, measure column CO2 and NO2 at high spatial resolution 10, 15]. and determine emission factors from space [13]. We also Fugitive Methane Leaks from Conventional Fossil Fuel Min- show the value of high frequency measurements that can ing at Four Corners: A serendipitous discovery we made at be made from geostationary satellites and can better con- our site was a pattern of large CH4 increases in the region- strain effects of meteorological variability on the signals. al column in the morning, and the in situ data at night. We Long-term observations of atmospheric ∆NO2/∆CO2 above attributed this to fugitive emissions from fossil energy ex- power plants would allow the verification of ∆NO2/∆CO2 ploration, because ethane (C2H6) was also enhanced in the emission factor reductions from technology improvement regional column at about 1-2% of CH4. This composition is and provide reliable constraints to estimate CO2 emission consistent with the composition of thermogenic oil and gas [e.g. Vulcan; 14] reservoirs in the region [16]. We confirmed this with in situ Power plant signal and natural variability: Our data set, analysis of the isotopic 13CH4 content of ambient air and spanning 2 seasonal cycles, shows clearly the large-scale found it to heavier than biogenic sources (e.g. landfills or regional column increases in CO2 from the power plant ruminants) as well as the absence of co-emitted combus- emissions [Fig 3, 11]. We can use thresholds for co-emitted tion signatures like CO from fires [17]. Our regional CH4 pollutants like CO to filter out these plumes. This filtering enhancements are large (20-50 ppb) in the morning and allows us to resolve the natural CO2 seasonal cycle in the fall off in the afternoon. This behavior is consistent with US southwest - minima in summer due to photosynthetic the local meteorology, where the winds come from the sinks and maxima in winter from dormant vegetation. The SE in the morning bringing in air from the coal, oil and gas seasonal amplitude of the natural CO2 cycle of about 5 region, and shift to SW in the afternoon. To evaluate these ppm at Four Corners is consistent with published measure- “fugitive” CH4 emissions we performed model simulations ments in the region. The power plant signals we detect using the Edgar CH4 inventory and compared it with our remotely in the column range from 2 to over 10 ppm observations. The comparison of our data with archived are significant and frequent (Figure 3) with well-defined satellite results over the last decade show a CH4 hot spot distributions that are controlled by the meteorology. over the Four Corners region (Figure 5a). Furthermore, the Such large signals can be detected from space with ultra morning enhancements in column CH4 we measure are hyper-spectral sensors with sufficient spatial resolution for much larger than the forward WRF model simulation, im- verification. plying that observed leaks 3.3 times greater than in current Edgar emissions (Figure 5b) [18]. We find that missing Verification of reported power plant emissions: We evalu- fugitive CH4 leak in Four Corners could be contributing ate our hypothesis that regional-scale CO2 measurements about 10% to the total national US emissions of CH4 from are less sensitive than in situ point surface measurements the natural gas sector. Furthermore, since current San Juan to small-scale variability from meteorology, and better suit- basin mining is dominated by coal, coal bed methane, and ed for emission verification. We also performed vertically conventional oil and gas production, our finding demon- resolved retrievals of CO2 from our solar spectrometer to strates that the Edgar data are not suitable as a from which use the much larger power plant signals in the bottom, to assess changes from rising non-conventional fossil fuel well-mixed, boundary layer to constrain the reported emis- extraction. sions. Our observations are compared to forward model simulations for a particular period in Figure 4 [15]. Both Impact on National Missions models and observations reveal large CO2 (30-50ppm) In 2013, the President prioritized action on climate change. increases above the background in the boundary layer Recent extreme weather events like Sandy, Midwest tor- (bottom to 3-km). The fluctuations in the observed CO2 are nadoes, and the California drought underscore the need also very similar, although the model is unable to represent to predict, prepare and mitigate their impacts. Extremes in the high frequency features in the data. The linear regres- weather exacerbate regional and local conflicts amongst sion of the observed boundary layer regional CO2 with the peoples and nations and are projected to worsen with WRF-Chem simulated CO2 is extremely tight demonstrat- climate change. Verifying an enforceable future climate ing that regional scale boundary layer CO2 measurements treaty will be an important global-scale US-led action, can be used to verify emissions. The observations show much like the nuclear test ban treaty, that would help about 2 percent more CO2 than the model based on the mitigate these adverse climate change impacts on national in-stack emissions for this particular time (Figure 4). This security. 341

Page 342

Our nation is also undergoing an energy revolution that promises energy independence by harvesting domestic oil and gas profitably using non-conventional method. However, gaps in our knowledge of fugitive CH4 leaks from such activities and a national distribution system poses unknown risks to the environment and is limiting its widespread use and public acceptance. Our work is one of the first nationwide to quantify fugitive CH4 leaks using Figure 3. a) Observations of vertically resolved regional CO2 made with our solar spectrometer sequence by time (profile regional scale observations and modeling and will help the number) for March 15, 2011, b) Simulated regional CO2 profiles nation achieve energy independence by extracting natural using in stack measured CO2 in the WRF model as a function of gas efficiently, profitably, and responsibly. time (profile number( for the same day, c) Observed CO2 in the bottom layer versus simulated CO2 in the bottom layer show tight correlation and agree to within 2%. Figure 1. a) Picture of our Four Corners multi-scale monitoring site with remote and in situ instruments on 3 different sea-tain- ers, the site is robotic and instruments run automated, b) Picture of the San Juan power plant plume next our site and c) The LANL plume model simulates the dispersion of greenhouse gases and Figure 4. About 2.5 years of column regional CO2 data measured pollutants from the two power plants and results are compared at 4 corners orange line is the data and red line is derived by fil- with observations at our site to verify reported emissions from tering out the power plant signals using CO above background as the in stack monitors. a marker, b) Daily time series of the power plant signal estimated as the difference between the maximum and minimum observed for each day. Figure 5. a) Satellite (European Space Agency, SCIAMACHY) ob- servations of mean (2003-2009) excess column methane above background show very high (30 ppb) enhancements over Four Corners, b) Daily variation in column methane excess measured at our site (black) , simulated column methane excess using current Edgar emissions (blue) and the same scaled by 3.3 (red) Figure 2. Example of the exquisite detail with which our multi- that brings the mean model in line with the measurements. The scale observations measure the plume on June 6, 2012 in panels shaded are is the variance in the observations and the box and a) through e) that note the specific gases and isotopes. The left whisker shows the mean and standard deviation for the observed scale is for the in situ sensors at the site that are sampling air at column methane enhancement at our site. about 10m above the surface, and all the in situ data are in cyan. The right axis is for the regional scale column measurements References made with our solar spectrometers. Panel f) displays the chang- es wind directions and speeds at our site for June 6, 2012 that 1. Guan, D., Z. Liu, Y. Geng, S. Lindner, and K. Hubacek. explain with the time dependence of our plume observations. The gigatonne gap in China’s carbon dioxide invento- ries. 2012. Nature Climate Change. 2: 672. 2. Richter, A., J. P. Burrows, H. Nuss, C. Granier, and U. Niemeier. Increase in tropospheric nitrogen dioxide 342

Page 343

over China observed from space. 2005. NATURE. 437 12. Lindemaier, R., M. K. Dubey, B. G. Henderson, Z. But- (7055): 129. terfield, J. R. Herman, T. Rahn, and T. Hertel. Remote observations of power plant plume composition for 3. Kim, S. W., A. Heckel, G. J. Frost, A. Richter, J. Gleason, emission factor verification. Atmospheric Chemistry J. P. Burrows, S. McKeen, E. Y. Hsie, C. Granier, and M. and Physics Discussions. Trainer. NO2 columns in the western United States observed from space and simulated by a regional 13. Bovensmann, H., M. Buchwitz, J. P. Burrows, M. Reuter, chemistry model and their implications for NOx emis- T. Krings, K. Gerilowski, O. Schneising, J. Heymann, A. sions. 2009. JOURNAL OF GEOPHYSICAL RESEARCH- Tretner, and J. Erzinger. A remote sensing technique for ATMOSPHERES. 114: -. global monitoring of power plant CO2 emissions from space and related applications. 2010. ATMOSPHERIC 4. Marquis, M., and P. Tans. Climate change - Carbon MEASUREMENT TECHNIQUES. 3 (4): 781. crucible. 2008. SCIENCE. 320 (5875): 460. 14. Gurney, K. R., D. L. Mendoza, Y. Y. Zhou, M. L. Fischer, 5. Dimotakis, P. E., B. Walker, K. K. Jonietz, and D. A. Roth- C. C. Miller, S. Geethakumar, and S. D. Du Can. High man. A greenhouse gas information system: monitor- Resolution Fossil Fuel Combustion CO2 Emission Fluxes ing and validating emissions reporting and mitigation. for the United States. 2009. ENVIRONMENTAL SCIENCE 2011. A report prepared for the DOE Office of Science, & TECHNOLOGY. 43 (14): 5535. , Sept. 26, 2011. 15. Lindenmaier, R., M. K. Dubey, F. Hase, S. H. Lee, K. R. 6. Pacala, S.. Verifying greenhouse gas emissions: meth- Costigan, and B. G. Henderson. Vertical profile retriev- ods to support international climate agreements. 2010. als of CO2 from solar Fourier Transform Spectrometry: National Research Council. National Academies Press, Top Down Verification of Power Plant Emissions at Washington D.C., 2010. Four Corners, NM. Atmospheric Chemistry and Physics Discussions. 7. Eggleston, H. S., L. Buendia, K. Miwa, T. Ngara, and K. Tanabe. 2006 IPCC Guidelines for National Greenhouse 16. Lindenmaier, R., M. K. Dubey, D. Wunch, B. Henderson, Gas Inventories. 2006. Institute for Global Environmen- and K. Costigan. Regional Column Methane and Ethane tal Strategies, Hayama, Japan. Prepared by the Nation- Measurements by Solar Fourier Transform Spectros- al Greenhouse Gas Inventories Program.. copy: Attributing Fugitive Leaks from Fossil Fuels at Four Corners, NM. Atmospheric Chemistry and Physics 8. McKain, K., S. C. Wofsy, T. Nehrkorn, J. Eluszkiewicz, Discussions. J. R. Ehleringer, and B. B. Stephens. Assessment of ground-based atmospheric observations for verifi- 17. Arata, C., M. K. Dubey, and T. Rahn. Detection and at- cation of greenhouse gas emissions from an urban tribution of methane sources in Four Corners, NM by region. 2012. In PROCEEDINGS OF THE NATIONAL a new 13CH4 cavity ring-down spectrometer. Atmo- ACADEMY OF SCIENCES OF THE UNITED STATES OF spheric Chemistry and Physics Discussions. AMERICA. Vol. 109, 22 Edition, p. 8423. 18. Kort, E. A., C. Frankenberg, K. R. Costigan, R. Linden- 9. Wunch, D., P. O. Wennberg, G. C. Toon, G. Keppel- maier, and M. K. Dubey. Four Corners: North America’s Aleks, and Y. G. Yavin. Emissions of greenhouse gases largest methane anomaly viewed from space. To ap- from a North American megacity. 2009. GEOPHYSICAL pear in Nature. RESEARCH LETTERS. 36: -. Publications 10. Costigan, K. R., S. H. Lee, and M. K. Dubey. Simulated Arata, C., M. K. Dubey, and T. Rahn. Detection and attribu- CO2 column concentrations of the Los Angeles Basin. tion of methane sources in Four Corners, NM by a new Atmospheric Chemistry and Physics Discussions. 13CH4 cavity ring-down spectrometer. Atmospheric Chemistry and Physics Discussions. 11. Dubey, M. K., R. Lindenmaier, B. Henderson, S. P. Love, J. R. Herman, Z. Butterfield, and T. A. Rahn. Multiscale Chylek, P., M. K. Dubey, G. Lesins, J. Li, and N. Hengartner. Greenhouse Gas, Pollutant and Water Monitoring at Imprint of Atlantic multi-decadal oscillation and Pacific Four Corners, NM: Emission Verification, Climate and decadal oscillation on southwestern US Climate: Past, Fugitive Leak Detection. Atmospheric Chemistry and present and future . To appear in Climate Dynamics. Physics Discussions. Costigan, K. R., M. K. Dubey, P. Chylek, B. Flowers, J. Reis- ner, A. Aiken, and L. Zhang. Combining Measurements 343

Page 344

and Modeling to Quantify Power Plant Contributions San Francisco: AGU, EOS. to Atmospheric NO2 and CO2. 2011. In American Me- teorological Society. (Seattle, WA, 22-27 Jan. 2011). , p. Flowers, B. A., H. H. Powers, M. K. Dubey, and N. G. Mc- 250, 1. Seattle: AMS. Dowell. Field inter-comparison of two high-accuracy fast-response spectroscopic sensors of carbon dioxide. Costigan, K. R., M. K. Dubey, P. Chylek, S. P. Love, B. G. Hen- 2011. Atmospheric Measurement Techniques Discus- derson, B. A. Flowers, J. M. Reisner, T. Rahn, and C. R. sions. 4 (1): 5837. Quick. Multi-Scale Science Framework for Attributing and Tracking Greenhouse Gas Fluxes at LANL’s Four Keeling, R., A. Manning, and M. Dubey. The atmospheric Corners New Mexico Test Bed. 2010. In American Geo- signature of carbon capture and storage. 2011. Philo- physical Union Annual Meeting. (San Francisco, 0-13 sophical Transactions of the Royal Society A - Math- Dec. 2010). , p. GC13D. San Francisco: AGU. ematical, Physical, and Engineering Sciences. 369 (1943): 2113. Costigan, K. R., S. H. Lee, M. K. Dubey, and D. Wunch. Simu- lated and observed CO2 column concentrations of the Kort, E. A., C. Frankenberg, K. R. Costigan, R. Lindenmaier, Los Angeles Basin. Atmospheric Chemistry and Physics and M. K. Dubey. Four Corners: North America’s larg- Discussions. est methane anomaly viewed from space. To appear in Nature. Costigan, K. R., S. Lee, J. Reisner, M. K. Dubey, S. P. Love, B. G. Henderson, and P. Chylek. Muli-scale Modeling of Lindenmaier, R., M. K. Dubey, B. G. Henderson, Z. But- Power Plant Plume Emissions and Comparisons with terfield, J. R. Herman, T. Rahn, and T. Hertel. Remote Observations. 2011. In American Geophysical Union observations of power plant plume composition for Annual Meeting. (San Francisco, CA, 5-9 Dec. 2011). , emission factor verification. Atmospheric Chemistry p. A41B. San Francisco: AGU EOS. and Physics Discussions. Dubey, M. K.. Solar Fourier Transform Spectrometry in mid- Lindenmaier, R., M. K. Dubey, F. Hase, S. H. Lee, K. R. Costi- near infrared and visible-ultraviolet to monitor green- gan, and B. Henderson. Vertical profile retrievals of house gas and co-emitted pollutant emissions from CO2 from solar Fourier Transform Spectrometry: Top Four Corners, NM power plants. 2011. In European Down Verification of Power Plant Emissions at Four Geosciences Union General Assembly 2011. (Vienna, Corners, NM. Atmospheric Chemistry and Physics Dis- 3-8 April 2011). Vol. 13, p. EGU2011. Vienna: Geophysi- cussions. cal Research Abstracts. Pratola, M. T., J. Reisner, D. Higdon, and M. K. Dubey. Dubey, M. K.. Multi-scale Measurements and Modeling to Quantifying Uncertainty in CO2 Emissions with a Re- Verify CO2 and Pollutant Emissions from Four Corners, stricted Number of Remote Sensors: A Comparison of NM Power Plants. 2012. In Global Emissions Initiative: Model Calibration and Kalman Filtering Techniques. Emissions to Address Science Policy Needs . (Toulouse, 2011. In American Geophysical Union Annual Meeting. France, 11-13 June 2012). , p. Web based. Boulder, CO: (San Francisco, CA, 5-9 Dec. 2011). , p. A31B. San Fran- GEIA. cisco: AGU, EOS. Dubey, M. K.. Multi-scale Measurements and Modeling to Verify Greenhouse Gas Emissions from Four Corners Power Plants. 2012. In 8th International Workshop on Greenhouse Gas Measurements from Space. (Pasade- na, CA, 18-20 June 2012). , p. 1. Caltech: NASA-JPL. Dubey, M. K., R. Lindenmaier, B. Henderson, S. P. Love, J. R. Herman, Z. Butterfield, and T. Rahn. Multiscale Greenhouse Gas, Pollutant and Water Monitoring at Four Corners, NM: Emission Verification, Climate and Fugitive Leak Detection. Atmospheric Chemistry and Physics Discussions. Dubey, M. K., S. P. Love, B. G. Henderson, S. Lee, K. R. Costi- gan, J. Reisner, B. A. Flowers, and P. Chylek. Multi-scale Measurements and Modeling to Verify and Attribute Carbon Dioxide Emissions from Four Corners Power Plants. 2011. In American Geophysical Union Annual Meeting. (San Francisco, CA, 5-9 Dec 2011). , p. A41H. 344

Page 345

Environmental and Biological Sciences Directed Research Final Report Multi-scale Dynamics of Biological Systems Robert E. Ecke 20110435DR Abstract and able to function as flexible and controllable nano- Biological systems involve dynamics in a fundamental machines instead of as rigid particles with fixed chemical way, spanning many orders-of-magnitude in time with behavior. Natural bio-molecules evolve with small biases corresponding multi-scale spatial structures. We helped in the landscape that can make their thermodynamic elucidate the complex interactions of temporal and behavior well defined while maintaining the functional spatial degrees of freedom by using the modern tools of benefits of plasticity. A grand challenge in molecular biol- physical and mathematical sciences to make quantitative ogy has been to characterize and predict structure-func- predictions about biological systems. Our work spans tion relations for bio-molecules. Our goal is to address length scales from nanometers to millimeters, ranging this challenge, but also to understand the role of dynam- from individual bio-molecules through complex phenom- ics in the control of bio-molecular function. Our specific ena within a single cell to the behavior of group of cells. goals at the macromolecular level were to characterize At the molecular level, we contributed to the under- allosteric (non-local) regulation, multidimensionality and standing of how the diversity of biological molecules intrinsic disorder in proteins using networks models and such as DNA, RNA, or proteins arises from competing to investigate protein-protein, protein-carbohydrate, deterministic energy scales associated with chemical protein/carbohydrate-membrane interactions. bonds, mechanical bending/breaking, and electro- Cells use networks of interacting bio-molecules to under- magnetic interactions and stochastic thermal energy. We go differentiation, to maintain homeostasis, and to sense have applied this insight to model proteins and mem- and respond to environmental conditions. Interactions in brane functions with applications to bio-fuel production. these networks span the 0.1-100 μm length scale, from At the cellular level, we performed averaging over parts local regions to the whole cell. The specific functions of the complex cellular bio-chemical kinetics to produce of cellular networks are diverse: metabolic networks reduced models of gene regulation, expression and dif- control the transformation of molecular building blocks ferentiation. We have also developed phenomenological and energy storage/consumption; trafficking networks models of cell regulatory processes that allow effective control the transport of bio-molecules across compart- descriptions of cell functions without precise correspon- mental boundaries; cell signaling networks couple tran- dence to microscopic detail. sient environmental and intracellular signals to longer- lasting covalent modifications of molecular targets; and Background and Research Objectives genetic regulatory networks control production of mRNA Molecules consisting of many atomic units, i.e., poly- and proteins using interactions among proteins, DNA, mers, proteins, DNA, RNA, are the building blocks of and the cell’s transcriptional and translational machin- biological systems. Not only do the competing effects of ery. Several large networks may be coupled together to chemical bond distortions, van der Waals interactions, give rise to quantifiable input-output behaviors. Com- and electrostatic interactions have approximately equal pared to the complexity of the underlying networks, the energies, these energies are comparable to the entropic input-output behaviors often have a simpler representa- free energies at biological temperatures. Thus, the mol- tion. Such cases motivate the development of accurate ecules that define biological phenomena on the 1-100 coarse-grained models that will be a major focus of our nm scale are frustrated systems that live on a relatively efforts. One of our goals represents a major challenge shallow energy landscape. On the one hand, this shal- in modern biology: using a combination of quantitative low landscape makes individual bio-molecules relatively experiments and theoretical/computational tools to unstable. On the other hand, it makes them plastic 345

Page 346

develop models of cellular networks that predict cellular dwindling supply of antibiotics. Current work in this area behavior under new conditions. Specifically, our goals were is hindered by lack of a complete all-atom structure of the to improve rule based bio-chemical reaction networks protein complexes that enable drug resistance in bacte- using novel methods and approaches from applied mathe- ria. We developed an experimental data-driven modeling matics and statistical physics and to develop bio-informatic pipeline which facilitates the construction of a wide range methods based on the physical sciences, e.g., prediction of of these complexes, a first step in a broader multi-scale promoters based on DNA melting properties or transcrip- modeling of antibiotic drug resistance. tion factor binding sites. We explored antibody-viruses interactions. An effective Cells act collectively to realize macroscopic biological func- vaccine for a virus such as HIV should generate antibodies tion. At this level of description and analysis, one needs that inactivate or neutralize the virus. Two major ob- to reduce microscopic detail and focus on cell models that stacles to vaccine development are immune evasion and incorporate the features necessary to characterize the co- extreme genetic diversity. By using network analysis we operative behavior. In this area, our goals were to develop have identified pathways within the complicated antibody models that connect molecular states of cell signaling sys- protein structures that may help understand how the virus tems in cells to emergent cellular behavior, to model viral can escape the immune response of the vaccine induced infection and evolution of multi-cellular complexes, and to antibody proteins (Sethi et al, PLoS Comp. Bio. 2013). Also investigate higher level motion driven by surface character- in the area of understanding HIV immune response, we istics or asymmetries in individual bacterial movement. performed theoretical calculations of the electrostatic sur- face potential (ESP) for HIV envelope (Env) proteins (Stieh, Scientific Approach and Accomplishments Retrovirology 2013). The Env proteins are responsible for In the molecular area we considered the properties of initiating viral transmission by binding to T-cell surface pro- disordered proteins. Many proteins fold to well-ordered teins, and this process may be impacted by environmental structures that perform the intricate biological functions factors such as salt concentration and pH. The impact that make life possible. Some proteins, however, do not of such factors is largely ignored in current HIV vaccine fold to an ordered state but instead are intrinsically dis- development efforts despite their potential significance. ordered. These proteins are less well studied but impor- We performed ESP calculations that predict that binding tant nevertheless. We developed an ideal chain model would be enhanced or “primed” over a narrow range of of intrinsically ordered proteins and demonstrated that pH, results that led us to hypothesize that virus transmit- one can divide the whole protein up into smaller peptides ted from donors to recipients have more effectively primed that can be independently investigated. This reduces the ENV proteins (compared to the rest of the donor viral computational demands on creating an effective under- population). standing of protein dynamics for disordered proteins (Sethi We examined the conversion of biomass to a biofuel, a et al, Biophys. J. 2012). We also modeled and simulated process that may provide an alternative and economi- using molecular dynamics the enzyme, organophosphorus cal energy source. A challenge for this conversion is the anhydrase (OPAA), which is known to neutralize certain degradation of the robust cellulose structures that give chemical nerve agents. The OPAA metalloprotein is a plants their structure. We have used multiple approaches promising vehicle for therapeutic or sanitation purposes to understand and improve this degradation process. due to ease of production, but shows only limited activity In one approach, we used a coarse-grained hydrogen to several V-type nerve agents. Molecular simulations of bound network model to study the stability of cellulose OPAA and several OPAA mutants could suggest mutations (Asztalos et al, Biotech. Biofuels 2012). We combined this that would enhance activity for the V-type agents. We have with all-atom molecular dynamics simulations to develop a developed all-atom models of OPAA using state-of-the-art suite of tools to consider the multi-scale structural pro- parameterization approaches. The results suggest that cesses of cellulosic break down by both biochemical and the mutant OPAA protein shows enhanced activity to the thermochemical conversion (Bellesia et al, J. Phys. Chem V-type agents, a result that was confirmed by our experi- 2011, 2012). Finally, we used comparative genomic analy- mental collaborators. sis to detect and resolve inconsistencies in publicly avail- Using coarse-graining molecular dynamics, we extended able microbial genomes (Wall et al, PLoS Comp. Bio 2011). this work to the molecular modeling of multi-protein This approach may lead to better gene prediction by using complexes that confer antibiotic resistance in bacterial multiple genomes for comparison. pathogens. Understanding the molecular basis of this drug In the area of cellular processes, we used single molecule resistance may help revitalize the effectiveness of our experiments (Shepherd et al, SPIE Proc. 2012) and random- 346

Page 347

process models (Munsky et al, Science 2012, Neuert et al, differentially depending on exclusion radius and on details Science 2013) to characterize the gene regulation process of the flocking interaction. In particular, a model of flock- whereby the proteins that are created from the DNA code ing particles, it is possible to create a ratchet effect using act to regulate which further proteins are produced. This a substrate in the form of an asymmetric barrier. In the process is noisy, i.e., stochastic, and may involve small RNA ratchet effect, the particles preferentially translate in a as well as messenger RNA. The single molecule methods particular direction even though the motive force of the allow certain tagged molecules, illustrated in Figure 1, to individual particles has no net directionality as illustrated be counted and their temporal variation to be determined. in Figure 3. We specifically demonstrated that for differ- The models to account for these biological processes ent species of flocking particles, it is possible to adjust require this detailed data inputs but are already showing the ratchet potential such that one species moves in one great promise as a tool for understanding biological func- direction while the other species moves in the oppo- tion at the cellular level. We also developed a model that site direction. This could be a very powerful method for reproduces the temporal variation of certain types of cel- sorting different particle species, and could also lead to lular vesicles under a variety of conditions. Cells can form an understanding of how certain organisms can navigate these vesicles to surround and degrade cellular intercellu- complex environments. Finally, we investigated cell-to-cell lar components. The model describes accurately a mecha- communication between antiobiotic-resistant and antio- nism of cell function that strongly impacts cell survival. biotic sensitive cells. We learned that quorum signaling molecules control the dissemination of antibiotic resis- Hepatitis C viral (HCV) infection is a major cause of chronic tance at the population scale, suggesting a mechanism for liver disease and affects nearly 170 million people world- high antibiotic resistance of certain bacteria (Chatterjee et wide. In Figure 2, we show a schematic representation of al, PNAS 2013). the infection mechanism of HCV. Whereas the previous standard of care had modest effectiveness, the recent In the area of evolutionary biology, we used mathematical approval of two highly potent protease inhibitors and the models, particularly stochastic processes, to study multi- ongoing development of dozens of direct-acting antivi- level selection. For example, a pathogen may be subject to ral agents (DAAs) constitute a major milestone for HCV selection within its host for faster replication. If faster repli- therapy. Mathematical modeling of viral kinetics under cation is, however, associated with decreased host mobil- treatment has played an instrumental role in improving ity and therefore lower transmission, it is unclear what our understanding of virus pathogenesis and in guiding the overall evolutionary trajectory of the pathogen would drug development. Viral kinetic models have proved useful be. Using tools from stochastic processes and differential to characterize treatment effectiveness during HCV therapy equations, we developed a mathematical formulation to with interferon (IFN), but there are challenges associated understand the dynamics of such with the standard model when fitting viral kinetic models to data. Understanding the combined effects of pharma- Impact on National Missions cokinetic and viral kinetics on the performance of DAA’s Our quantitative approach to biological dynamics at the could provide valuable insights into its use of DAA’s either molecular, cellular and systems level may have significant individually or in combination with other DAAs in order to impact on applications in bio-security, energy, and health optimize future therapies. We developed a pharmacokinet- by ultimately allowing for cellular engineering, i.e., the ic/viral kinetic (PK/VK) model to describe HCV RNA kinetics manipulation of cellular behaviors for useful purposes. In during treatment with DAAs. Additionally, we have devel- bio-security, our work may contribute to the mitigation oped intracellular models of HCV viral kinetics to further and detection of pathomic agents and infectious diseases. understand the effect of DAAs on HCV dynamics (Chatter- For example, such approaches may be key to the rapid jee et al, Antiviral Therapy 2012). development of drug therapies or vaccines against emerg- ing bio-threats or pandemics. Another area of important At the extracellular level that includes the behavior of impact is design of new bio-fuel processes for a sustainable populations, we developed a biogeochemical model of energy economy. Our current work on theory and model- phytoplankton populations that leads to surprising nutri- ing of cellulosic degradation has made important contribu- ent- dependent food uptake. This model has important tions to this important national need. In the longer term, implications for carbon cycle models. We also created a our multi-scale modeling from molecules to intercellular model to investigate the influence of the environment interactions may have far reaching impact on public health, on systems of flocking bacteria (Drocco et al, PNAS 2012) bio-security, and fundamental understanding of biological with potential application in disrupting biofilms and nano- systems at a quantitative and predictive level. Our ap- sorting devices. We discovered that barriers sort particles proach and research efforts will contribute substantively 347

Page 348

to that agenda in service of the LANL national security Publications mission. Akimov, A., and N. A. Sinitsyn. Sensitivity field for nonau- tonomous molecular rotor. 2011. Journal of Chemical Physics. 135 (22): 224105. Alexandrov, B. S., V. Gelev, A. R. Bishop, A. Usheva, and K. O. Rasmussen. DNA breathing dynamics in the pres- ence of a terahertz field. 2010. Physics Letters A. 374 (10): 1214. Asztalos, A., M. Daniels, A. Sethi, T. Shen, P. Langan, A. Re- dondo, and S. Gnanakaran. A Coarse-grained Model for Synergistic Action of Multiple Enzymes on Cellulose. 2012. Biotechnology for Biofuels. 5 (1): 55. Barua, D., W. S. Hlavacek, and T. Lipniacki. A computational model for early events in B cell antigen receptor signal- ing: analysis of the roles of Lyn and Fyn. 2012. Journal of Immunology. 189 (2): 646. Barua, D., and B. Goldstein. A Mechanistic Model of Early Figure 1. Fluorescent markers of biological expression of mes- Fc?RI Signaling: Lipid Rafts and the Question of Protec- sHenCgeVr R NliAf e(m cRyNAc)l eus ing single cell optical imaging. The green tion from Dephosphorylation. 2012. PLoS ONE. 7 (12). dots indicate the presence of mRNA in the cell - multiple cells are imaged. Bellesia, G., A. Asztalos, T. Y. Shen, P. Langan, A. Redondo, and S. Gnanakaran. In silico studies of crystalline cel- lulose and its degredation by enzymes. 2011. Acta Crystallographica Section D-Biological Crystallography. 201 (66): 1184. Bellesia, G., A. I. Jewett, and J. E. Shea. Relative stability of de novo four-helix bundle proteins: Insights from coarse grained molecular simulations. 2011. Protein Science. 20 (5): 818. Bellesia, G., S. P. Chundawat, P. Langan, B. E. Dale, and S. Gnanakaran. Probing the Early Events Associated with Figure 2. Schematic illustration of the life cycle of HCV in a host Liquid Ammonia Pretreatment of Native Crystalline liver cell illustrating the mechanisms of viral infection and propa- Cellulose. 2011. Journal of Physical Chemistry B. 115 gation. (32): 9782. Bellesia, G., S. P. S. Chundawat, P. Langan, A. Redondo, B. E. Dale, and S. Gnanakaran. Coarse-Grained Model for the Interconversion Between Native and Liquid Ammonia-Treated Crystalline Cellulose. 2012. Journal of Physical Chemistry B. 116 (28): 8031. Chatterjee, A., J. Gudej, and A. S. Perelson. Mathematical Modeling of HCV Infection: What Can It Teach Us in the Direct Antiviral Agents Era?. 2012. Antiviral Therapy. 17: 1171. Figure 3. Images of flocking particles (bacteria) in a container Chatterjee, A., L. C. Cook, C. Shu, Y. Chen, D. Manias, D. with an array of funnel shapes. (a) Initial positions with an equal Ramkrishna, G. Dunny, and W. Hu. Antagonistic self- number of particles on either side of the barrier array. (b) After sensing and mate-sensing signaling controls antibiotic- a period of time, flocking formations occur and the particles con- resistance transfer. 2013. Proceedings of the National centrate in the top chamber of the container. This results sug- Academy of Sciences. 110 (17): 7086. gests that the controlled motion of certain types of active matter systems can be achieved with a fabricated substrate. Chatterjee, A., P. Smith, and A. Perelson. Hepatitis C Viral 348

Page 349

Kinetics: The Past, Present, and Future. 2013. Clinics in F. Haynes, M. S. Cohen, G. M. Shaw, M. S. Seaman, A. Liver Disease. 17 (1): 13. Kumar, F. Gao, D. C. Montefiori, and B. Korber. Genetic Signatures in the Envelope Glycoproteins of HIV-1 that Chundawat, S. P., G. Bellesia, N. Uppugundla, L. D. Sousa, Associate with Broadly Neutralizing Antibodies. 2010. D. H. Gao, A. M. Cheh, U. P. Agarwal, C. M. Bianchetti, PLoS Computational Biology. 6 (10): e1000955. G. N. Phillips, P. Langan, V. Balan, S. Gnanakaran, and B. E. Dale. Restructuring the Crystalline Cellulose Hy- Gnanakaran, S., S. Bhattacharya, M. G. Daniels, B. F. Keele, drogen Bond Network Enhances Its Depolymerization P. T. Hraber, A. S. Lapedes, T. Y. Shen, B. Gaschen, M. Rate. 2011. Journal of the American Chemical Society. Krishnamoorthy, H. Li, J. M. Decker, J. F. Salazar-Gon- 133 (29): 11163. zalez, S. Wang, C. Jiang, F. Gao, R. Swanstrom, J. A. An- derson, L. H. Ping, M. S. Cohen, M. Markowitz, P. A. Go- Drocco, J. A., C. Reichhardt, C. J. Olson Reichhardt, and A. epfert, M. S. Saag, J. J. Eron, C. B. Hicks, W. A. Blattner, R. Bishop. Statics and Dynamics of Vortex Matter with G. D. Tomaras, M. Asmal, N. L. Letvin, P. B. Gilbert, A. Competing Repulsive and Attractive Interactions in the C. Decamp, C. A. Magaret, W. R. Schief, Y. E. Ban, M. Presence of Periodic and Random Pinning. 2013. Jour- Zhang, K. A. Soderberg, J. G. Sodroski, B. F. Haynes, G. nal of Physics: Condensed Matter. 25: 345703. M. Shaw, B. H. Hahn, and B. Korber. Recurrent signa- ture patterns in HIV-1 B clade envelope glycoproteins Drocco, J. A., E. F. Wieschaus, and D. W. Tank. The syn- associated with either early or chronic infections. thesis–diffusion–degradation model explains Bicoid 2011. PLoS Pathogens. 7 (9): e1002209. gradient formation in unfertilized eggs. 2012. Physical Biology. 9 (5). Jin, M., J. Chen, X. - H. Zhang, M. Zhang, H. - Y. Li, W. - X. Cheng, N. Liu, M. Tan, T. Jiang, and Z. - J. Duan. Genetic Drocco, J., C. J. Olson, and C. Reichhardt. Bidirectional Diversity of Noroviruses in Chinese Adults: Potential sorting of flocking particles on asymmetric substrates. Recombination Hotspots and GII-4/Den Haag-specific 2012. Proceedings of the National Academy of Sci- Mutations at a Putative Epitope. 2011. Infection, Ge- ences of the USA. 85 (5): 056102. netics and Evolution. 11 (7): 1716. Drocco, J., C. Olson Reichhardt, and C. Reichhardt. Charac- Jung, J., A. Sethi, T. Gaiotto, J. Han, T. Jeoh, S. Gnanakaran, terizing Plastic Depinning Dynamics with the Fluctua- and P. Goodwin. Binding and Movement of Individual tion Theorem. 2011. The European Physical Journal E: Cel7A Cellobiohydrolases on Crystalline Cellulose Sur- Soft Matter and Biological Physics. 34 (10): 1. faces Revealed by Single-molecule Fluorescence Imag- ing. 2013. Journal of Biological Chemistry. 288 (33): Drocco, J., C. Reichhardt, C. J. Olson Reichhardt, and A. R. 24164. Bishop. Statics and Dynamics of Wetting-Dewetting Transitions for Particles with Attractive Interactions on Kozer, N., D. Barua, S. Orchard, E. Nice, A. Burgess, W. Periodic Substrates. 2012. In Optical Trapping and Opti- Hlavacek, and A. A. Clayton. Exploring higher-order cal Manipulation IX. , p. 84581j. EGFR oligomerisation and phosphorylation—a com- bined experimental and theoretical approach. 2013. Drocco, J., L. M. Lopatina, C. Reichhardt, and C. J. Olson Molecular BioSystems. 9 (7): 1849. Reichhardt. Dynamics of Self-Driven and Flocking Par- ticles on Periodic Arrays. 2012. In Optical Trapping and Lemons, N. W., B. Hu, and W. S. Hlavacek. Hierarchical Optical Manupulation IX. , p. 84581i. graphs for rule-based modeling of biochemical sys- tems. 2011. BMC Bioinformatics. 12: 45. Dunbar, J., J. Cohn, and M. E. Wall. Consistency of Gene Starts Among Burkholderia Genomes. 2011. BMC Lou, C., B. Stanton, Y. C. Chen, B. Munsky, and C. A. Voigt. Geonmics. 12 (1): 125. Ribozyme-based “insulator parts” buffer synthetic circuits from genetic context. 2012. Nature Biotechnol- Gao, D., S. S. Chundawat, A. Sethi, V. Balan, S. Gnanakaran, ogy. 30 (11): 1137. and B. Dale. Increased enzyme binding to substrate is not necessary for more efficient cellulose hydrolysis. Lynch, R. M., R. Rong, S. Boliar, A. Sethi, B. Li, J. Mulenga, S. 2013. Proceedings of the National Academy of Sci- Allen, J. E. Robinson, S. Gnanakaran, and C. A. Derdeyn. ences. 110 (27): 10922. The B Cell Response Is Redundant and Highly Focused on V1V2 during Early Subtype C Infection in a Zambian Gnanakaran, G., J. Tian, and A. Sethi. Conformational study Seroconverter. 2011. Journal of Virology. 85 (2): 905. of quaternary epitope region of V1/V2 loop: influence of disulfide bonds and glycosylations. 2012. Retrovirol- Maienschein-Cline, M., A. R. Dinner, W. S. Hlavacek, and F. ogy. 9 (Suppl 2). Mu. Improved Predictions of Transcription Factor Bind- ing Sites Using Physicochemical Features of DNA. 2012. Gnanakaran, S., M. G. Daniels, T. Bhattacharya, A. S. Lape- Nucleic Acids Research. : 771. des, A. Sethi, M. Li, H. L. Tang, K. Greene, H. M. Gao, B. 349

Page 350

Martin, K., D. Barua, A. Kauffman, L. Westrate, R. Posner, Photonics. 8907: 80970A. W. Hlavacek, and J. MacKeigan. Computational model for autophagic vesicle dynamics in single cells. 2013. Parthasarathi, R., G. Bellesia, S. P. Chundawat, P. Langan, B. Autophagy. 9 (1): 70. E. Dale, and S. Gnanakaran. New insights into hydrogen bonding and stacking interactions in cellulose. 2011. Morriss-Andrews, A., G. Bellesia, and J. -E. Shea. β-Sheet Journal of Physical Chemistry A. 115 (49): 14191. Propensity Controls the Kinetic Pathways and Morphol- ogies of Seeded Peptide Aggregation. 2012. J Chemical Sargsyan, O.. Analytical Framework for Identifying and Dif- Physics. 137 (145104). ferentiating Recent Hitchhiking and Severe Bottleneck Effects from Multi-Locus DNA Sequence Data. 2012. Morriss-Andrews, A., G. Bellesia, and J. E. Shea. Effect of PLoS ONE. 7 (5): e37588. surface interactions on peptide aggregate morphology. 2011. Journal of Chemical Physics. 135 (8): 085102. Sethi, A., B. Goldstein, and S. Gnanakaran. Quantifying Intramolecular Binding in Multivalent Interactions: Munsky, B., G. Neuert, and A. van Oudenaarden. Using A Structure-Based Synergistic Study on Grb2-Sos1 gene expression to understand gene regulation. 2012. Complex. 2011. PLoS Computational Biology. 7 (10): Science. 336 (6078): 183. e1002192. Munsky, B., I. Nemenman, and G. Bel. Erratum: ‘Specific- Sethi, A., C. Deredyn, B. Korber, and S. Gnanakaran. Ap- ity and Completion Time Distributions of Biochemical plication of Molecular Dynamics and Network Analysis Processes’ [J. Chem. Phys. 131, 235103 (2009)]. 2012. for Mechanistic Understanding of Allosteric Immune Journal of Chemical Physics. 136 (20): 209902. Escape Pathways in HIV-1. 2013. PLoS Computational Biology. : e1003046. Murphy, M. K., L. Yue, R. Pan, S. Boliar, A. Sethi, E. Karita, S. A. Allen, E. Cormier, J. E. Robinson, S. Gnanakaran, Sethi, A., D. Anunciado, J. Tian, D. Vu, and S. Gnanakaran. E. Hunter, X. Kong, and C. A. Derdeyn. Sequential ex- Deducing conformational variability of intrinsically posure to specific antibody escape mutations may disordered proteins from infrared spectroscopy with program neutralization breadth during subtype A HIV-1 Bayesian statistics. 2013. Chemical Physics. 422: 143. infection. 2012. Retrovirology. 9 (Suppl 2). Sethi, A., D. Vu, and S. Gnanakaran. Conformational and Murphy, M., L. Yue, R. Pan, S. Boliar, A. Sethi, J. Tian, K. Spectroscopic Characterization of Intrinsically Disor- Pfafferot, E. Karita, S. Allen, E. Cormier, P. Goepfert, dered Regions in Proteins. 2011. Biophysical Journal. P. Borrow, J. Robinson, S. Gnanakaran, E. Hunter, X. 100 (3): 13a. Kong, and C. Derdeyn. Viral Escape from Neutralizing Antibodies in Early Subtype A HIV-1 Infection Drives an Sethi, A., J. Tian, D. Vu, and S. Gnanakaran. Identification of Increase in Autologous Neutralization Breadth. 2013. minimally interacting modules in an intrinsically disor- PLoS Pathogens. 9 (2). dered protein. 2012. Biophysical Journal. 103 (4): 748. Nemenman, I., J. R. Faeder, W. S. Hlavacek, Y. Jiang, M. E. Shepherd, D. P., J. B. Sambur, Y. - Q. Liang, B. A. Parkinson, Wall, and A. Zilman. Selected papers from the fourth and A. Van Orden. In Situ Studies of Photolumines- annual q-bio conference on cellular information pro- cence Quenching and Photocurrent Yield in Quantum cessing. 2011. Physical Biology. 8 (5): 050301. Dot Sensitized Single Crystal TiO2 and ZnO Electrodes. 2012. Journal of Physical Chemistry C. 116 (39): 20169. Nemenman, I., S. Gnanakaran, W. S. Hlavacek, Y. Jiang, B. Munsky, M. E. Wall, and J. R. Faeder. The Fifth Annual Shepherd, D. P., N. Li, E. Hong-Geller, B. Munsky, and J. H. q-bio Conference on Cellular Information Processing. Werner. New Tools for Discovering the Role sRNA Plays 2012. Physical Biology. 9 (5). in Cellular Regulation. 2012. SPIE Proceedings. 8228: 822808. Nemenman, I., W. S. Hlavacek, Y. Jiang, M. E. Wall, and A. Zilman. The third q-bio conference on cellular informa- Skar, H., R. N. Gutenkunst, K. Wilbe, A. Alaeus, J. Albert, tion processing. 2010. IET Systems Biology. 4 (6): 331. and T. Leitner. Daily sampling of an HIV-1 patient with slowly progressing disease displays persistence of mul- Neuert, G., B. Munsky, R. Z. Tan, L. Teytelman, M. Kham- tiple env subpopulations consistent with neutrality. mash, and A. v. Oudenaarden. Systematic Identification 2011. PLoS One. 6: e21747. of Signal-Activated Stochastic Gene Regulation. 2013. Science. 339 (6119): 584. Stieh, D. J., J. L. Phillips, P. M. Rogers, D. F. King, G. C. Cianci, S. A. Jeffs, S. Gnanakaran, and R. J. Shattock. Dynamic Olson, C. J., J. Drocco, T. Mai, M. B. Wan, and C. Reich- electrophoretic fingerprinting of the HIV-1 envelope hardt. Active matter on asymmetric substrates. 2011. glycoprotein. 2013. Retrovirology. 10. Proceedings of the International Society for Optics and 350

Page 351

Tian, J., A. Sethi, B. Swanson, B. Goldstein, and S. Gnana- karan. Taste of Sugar at the Membrane: Thermody- namics and Kinetics of the Interaction of a Disaccha- ride with Lipid Bilayers. 2013. Biophysical Journal. 104 (3): 622. Tian, J., A. Sethi, D. Anunciado, D. Vu, and S. Gnanakaran. A-Synuclein Interactions with Sodium Dodecyl Sulfates: A molecular dynamics study of the self-assembly sys- tem. 2012. Journal of Chemical Physics. 116 (15): 4417. Wada, M., Y. Nishiyama, G. Bellesia, T. Forsyth, S. Gnana- karan, and P. Langan. Neutron crystallographic and molecular dynamics studies of the structure of ammo- nia-cellulose I: rearrangement of hydrogen bonding during the treatment of cellulose with ammonia. 2011. Cellulose. 18 (2): 191. Wall, M. E.. Structure-function relations are subtle in ge- netic regulatory networks. 2011. Mathematical Biosci- ences. 231 (1): 61. Wall, M. E., S. Raghavan, J. D. Cohn, and J. Dunbar. Genome majority vote improves gene predictions. 2011. PLoS Computational Biology. 7 (11): e1002284. Whitney, J. B., P. T. Hraber, C. Ludemann, E. E. Giorgi, M. G. Daniels, T. Bhattacharya, and S. S. Rao. Genital Tract Sequestration of SIV Following Acute Infection. 2011. PLOS Pathology. 7 (2): e1001293. Yang, J., X. Meng, and W. S. Hlavacek. Rule-based model- ling and simulation of biochemical systems with mo- lecular finite automata. 2010. IET Systems Biology. 4 (6): 453. Yang, J., and W. S. Hlavacek. Scaffold-mediated nucleation of protein signaling complexes: Elementary principles. 2011. Mathematical Biosciences. 232 (2): 164. Yang, J., and W. S. Hlavacek. Efficiency of reactant site sam- pling in network-free simulation of rule-based models for biochemical systems. 2011. Physical Biology. 8 (5): 055009. Zhang, M., B. Foley, A. K. Schultz, J. P. Macke, I. Bulla, M. Stanke, B. Morgensterm, B. Korber, and T. Leitner. The role of recombination in the emergence of a complex and dynamic HIV epidemic. 2010. Retrovirology. 7: 25. Zhuang, Z. Y., A. I. Jewett, S. Kuttimalai, G. Bellesia, S. Gnanakaran, and J. E. Shea. Assisted Peptide Folding by Surface Pattern Recognition. 2011. Biophysical Journal. 100 (5): 1306. Zilman, A., V. V. Ganusov, and A. S. Perelson. Stochastic Models of Lymphocyte Proliferation and Death. 2010. PLoS One. 5 (9): e12775. 351

Page 352

Environmental and Biological Sciences Directed Research Final Report Discovery Science of Hydraulic-fracturing: Developing Innovative Concepts for Novel Working Fluids and Their Interactions with Rocks, Fractures, and High Value Hydrocarbons Hari S. Viswanathan 20130150DR Abstract Background and Research Objectives Shale gas is an unconventional fossil energy resource Shale gas is an unconventional fossil energy resource profoundly impacting US energy independence and is projected to last for nearly 100 years[1]. The increased projected to last for at least 100 years. Production of availability of shale gas (i.e., methane), which produces methane and other hydrocarbons from low permeability 50% less carbon dioxide (CO2) than coal, is primarily re- shale involves hydrofracturing of rock, establishing frac- sponsible for US CO2 emissions in 2011 dropping to their ture connectivity, and multiphase fluid-flow and reac- lowest levels in 20 years. Production of methane and tion processes all of which are poorly understood. The other hydrocarbons from low permeability shale involves result is inefficient extraction with many environmental hydrofracturing of rock, establishing fracture connectiv- concerns. These phenomena are part of a broader class ity, and multiphase fluid-flow and reaction processes, of problems involving coupled fluid flow and fractures all of which are poorly understood. The result is often that are critical to other energy security areas such as inefficient extraction with many environmental con- shale oil, geothermal, carbon sequestration, and nuclear cerns [1]. The fracking industry is motivated to reduce waste disposal as well as crack propagation in weapons their 70 to 140 billion gallon per year water demand applications. A science-based capability is required to due to droughts in the west, a lack of deep wastewater quantify the governing mesoscale fluid-solid interac- disposal wells in the east, and possible forthcoming tions, including microstructural control of fracture regulations[1]. Our objective was to demonstrate unique patterns and the interaction of engineered fluids with microfluidic and triaxial core flood experiments and inte- hydrocarbon flow. These interactions depend on coupled grate these with state-of-the-art numerical simulation. A thermo-hydro-mechanical-chemical (THMC) processes secondary objective was to perform an initial evaluation over scales from microns to tens of meters. Determin- of CO2-based fracturing fluids and techniques that could ing the key mechanisms in subsurface THMC systems enhance production, greatly reduce water usage, and has been impeded due to the lack of sophisticated sequester CO2. experimental methods to measure fracture aperture Fracking phenomena involve fluid-solid interactions and connectivity, multiphase permeability, and chemical embedded within coupled thermo-hydro-mechanical- exchange capacities at the high temperature, pressure, chemical (THMC) processes over scales from microns to and stresses present in the subsurface. In this feasibility tens of meters [2]. Feedbacks between processes result study, we developed and prototyped the microfluidic in complex dynamics that must be unraveled if one is and triaxial core flood experiments required to reveal to predict and, in the case of unconventional resources, the fundamental dynamics of fracture-fluid interactions. facilitate fracture propagation, fluid flow, and interfacial This will enable transformation of hydrofracturing from transport processes. These issues are part of a broader present ad hoc approaches to science-based strategies class of complex systems involving coupled fluid flow while safely enhancing production. Specifically, we have and fractures that are critical to energy security, such as demonstrated an integrated experimental/modeling ap- shale oil, geothermal, carbon sequestration, and nuclear proach that allows for a comprehensive characterization waste disposal, as well as, crack propagation in weap- of fluid-solid interactions and develop models that can ons and materials applications. In materials, suppress- be used to determine the reservoir operating conditions ing crack propagation or limiting subsequent fluid flow necessary to gain a degree of control over fracture gen- may be the objective. However, while crack propagation eration, fluid flow, and interfacial processes over a range in dry materials has been studied the role of fluids in of subsurface conditions. 352

Page 353

propagating fractures is not well understood. Predicting Mesoscale simulation of crack-tip propagation dynamics: and controlling fracture propagation due to fluid-solid in- We identified how to extend our CartaBlanca [11-12] code teraction would be transformative, with significant impact (R&D100 winner) to resolve short timescale crack-growth beyond just the hydraulic fracturing of rock. dynamics due to fluid-structure interactions. We have demonstrated the capability of modeling water fracking Scientific Approach and Accomplishments through a simplified geomechanical model. We can now We experimentally measured and simulated a subset develop and calibrate models that consider rock hetero- of the multi-scale, coupled processes from each of the geneity and interactions among microcracks. We can also research task areas described below. This involved experi- now study potential enhanced fracturing due to thermal ments and models across the meso, core and reservoir stress at crack tips following rapid expansion of CO2. A scale. Specifically, capabilities developed and tested in this well-calibrated physics-based geomechanical model ca- feasibility study include: pable of predicting scCO2 or water fracking will be a signifi- cant development in the energy industry. Several industrial Mesoscale high-pressure microfluidics in geomaterial sub- collaborators have expressed strong interest in licensing strates etched at CINT: We have conducted high P micro- such a capability. fluidics with standard silica wafers and demonstrated use of etched geomaterial (shale) at low P. Both capabilities are Core scale finite-element/discrete-element modeling unique. Our experiment in Figure 1a shows super critical (FDEM) coupled with fluid flow: We have added the effect (sc) CO2 penetrating into a simple dead-end fracture to of instantaneous fluid pressure to FDEM for accurate simu- extract hydrocarbon at 1250 psi (water could not invade lation of how fluids and rock interact to generate fractures the dead-end fracture). Figure 1b shows penetration of oil [13]. For follow-on work we can incorporate a fluid solver (in red) in a complex fracture geometry following preferen- to characterize dynamic multiphase fluid effects. Current tial pathways. We are now positioned to quantify benefits methods do not accurately propagate fractures across of scCO2 in extracting hydrocarbons from fracture-etched material interfaces or due to fluid pressure, so if success- shale wafers (that will have realistic wetting, fracture and ful, this will constitute the next generation fracture propa- rock properties) at in situ pressure conditions. gation model. These models will be validated using the tomography experiments. Mesoscale Lattice Boltzmann modeling (LBM) of multi- phase flow and particle transport: Our internationally Reservoir scale THMC fracture modeling of reservoir-scale recognized LBM code was enhanced to account for mul- hydrocarbon extraction: We have recently developed the tiphase flow [3-5], multi-component chemistry [6], and capability to embed discrete fracture networks of arbitrary phase transitions [7]. These unique LBM enhancements complexity into the massively parallel multiphase flow are required to interpret microfluidic experiments of multi- code PFLOTRAN [14]. Fracture networks based on upscaled phase flow, proppant transport [8], and chemical exchange FDEM results can now be used to investigate communi- kinetics [6] in mesoscale fracture systems. LBM will be fur- cation between generated and natural fractures linked ther enhanced to simulate processes at the fracture-matrix to sweet spots in production. Through discussions with interface and in nanoscale pores. The powerful combina- industry, we now have an extraordinary opportunity to use tion of microfluidics and LBM sets the stage for transfor- fracking data obtained from Apache Corp.’s science wells mative discoveries. to benchmark our next generation models. Core scale high-pressure tomographic triaxial coreflood We have used these capabilities to define the type of stud- experiments: We successfully demonstrated a unique ies we can now conduct at the pore, core and reservoir apparatus for in situ generation and characterization of scales. fractures. This allows direct measurement of fracture Pore scale: Lattice Boltzmann Models (LBM) are ideally apertures, fluid distribution, and hydrocarbon extraction suited to simulate pore-scale processes since they capture efficiency and involves a combination of x-ray (AET) and/ intra-pore geometries, complex flows, and all relevant or neutron tomography (LANSCE) and high-pressure fluid- physicochemical processes with high computational ef- flow experiments on shale cores subject to differential axial ficiency. Our models represent the state-of-the-art in and confining pressure [9-10]. We have fractured dry and existing mesoscale simulations by considering multiphase wet rocks under compressive loading during this feasibil- flow, multi-component chemistry, and phase transitions. In ity study and conducted tomography ex-situ. We can now future research, LBM can be further enhanced and utilized begin development of hydraulic fracturing experiments in to investigate the flow and phase behavior of natural gas the lab with measurement of stress, strain, and permeabil- within the matrix and at matrix-fracture interfaces. Fig- ity while imaging 3D flow under in situ conditions. 353

Page 354

ure 2a shows a microfluidic experiment in which a simple behavior. This capability is within reach through identified fishbone fracture pattern has been etched into Marcellus enhancements to PFLOTRAN: a competitive adsorption/ shale. A LBM of the experiment captures the fingering as desorption model of gas from shale; multicomponent, the invading immiscible water displaces hydrocarbon, but multiphase compositions of CO2/water/oil/gas; and cou- bypassing the hydrocarbon in dead end fractures result- pling of flow equations with stress equations. PFLOTRAN ing in poor sweep (Figure 2b). At first glance, this example could then be used to estimate the amount of hydrocarbon may appear simplistic. However, the finger width is con- extracted using aqueous and CO2-based fracturing fluids trolled by flow rate and fluid viscosity ratios. Also, network over a wide range of T, P, stress, and fracture network geometry affects the finger width since fluid from the side properties. It will be used to optimize operating condi- channel narrows the finger. In proposed work, we will test tions that maximize hydrocarbon extraction rates, while the hypothesis that miscible scCO2 sweeps this network minimizing water usage. Preliminary calculations also show more efficiently (Figure 1a). This will require development that it should be possible to simultaneously enhance pro- of high-pressure geomaterial wafers that we will combine duction and trap CO2. with our current high-pressure microfluidics capability. In addition, we will refine the etching process to create sub- Impact on National Missions micron (micron is the current limit) fracture apertures that Natural gas from shale is a tremendous resource that has will allow study of smaller scale interfacial dynamics. We only recently been recognized and could lead to US en- can directly measure the interfacial area between phases ergy independence3. The proposed work has the poten- that is critical to formulating relationships for multi-phase tial to reshape fracking processes of the future making flow simulation. We can also observe phase transition them safer and more efficient. DOE is formulating calls for behavior to benchmark EOS we will develop. Improved advanced fracking and companies such as Apache, Hess, constitutive relationships and equations of state can also Chesapeake and Chevron appear eager to collaborate. The be tested at the continuum scale. Nation’s fossil energy focus may soon transition from coal/ CO2 sequestration to natural gas thus development and Core Scale: During this study, we have shown how fluids demonstration of new experimental and computational impact fracture patterns. In a triaxial experiment, dry capabilities in this area, as was accomplished in this feasi- anhydrite core fractured with multiple bifurcations (Figure bility study will undoubtedly prove important in facilitating 3a). A FDEM simulation of anhydrite results in qualitatively that transition. similar bifurcations (Figure 3b). When fluid pressure is used to initiate the fracture, enhanced propagation occurs highlighting the new ability of FDEM to capture fluid ef- fects (Figure 3c). The simulations allow us to explore the impact of changing fluid properties and stress conditions on the same heterogeneous rock sample allowing analysis of the relation of particular fracture patterns to different working fluids. By integrating tomographic characteriza- tion of heterogeneities in the rock prior to fracturing with a detailed computational representation, we will develop more detailed and realistic FDEM results. For future work, we can further extend the FDEM to simulate multiphase fluids by coupling to a full fluid solver to create reservoir scale fracture networks. Reservoir Scale: Figure 4 shows one workflow plan we developed in which FDEM will be validated at the core scale against triaxial experiments. The dominant fracture patterns from the FDEM technique will be used to deter- mine the large fracture network at the reservoir scale. Micro-fracture data from FDEM will be used to obtain the effective matrix permeability at the reservoir scale. We are unaware of any published codes or papers that present full THMC5 coupled simulations of shale gas with Figure 1. a) high pressure wafer, b) vascular network. explicit resolution of discrete fractures and complex phase 354

Page 355

5. Middleton, R., G. Keating, P. Stauffer, A. Jordan, H. Viswanathan, Q. Kang, and J. W. Carey. The cross-scale science of CO2 capture and storage: from pore scale to regional scale. 2012. Energy and Environmental Sci- ences. 5: 7328.
6. Kang, Q., P. C. Lichtner, H. Viswanathan, and A. I. Abdel-Fattah. Pore scale modeling of reactive transport involved in geologic CO2 sequestration. 2010. Trans- port in Porous Media. 82: 197.
7. Chen, L., and Q. Kang. Pore-scale modeling of multi- Figure 2. Displacement of hydrocarbon by water in a shale wafer. phase reactive transport with phase transition and dissolution-precipitation in closed systems. . To appear in Physical Review E..
8. Connington, K., Q. Kang, H. S. Viswanathan, A. I. Abdel- Fattah, and S. Chen. Peristaltic particle transport using the lattice Boltzmann method. 2009. Physics of Fluids. 21: 301. Figure 3. Triaxial experiments and FDEM models of fracture propagation in anhydrite. Colors show velocity of rock matrix 9. Carey, J. W., R. Svec, R. Grigg, J. Zhang, and W. Crow. movement (blue=low velocity and red=high velocity) due to frac- Experimental investigation of wellbore integrity and ture propagation. CO2-brine flow along the casing-cement microannulus.
9. International Journal of Greenhouse Gas Con- trol. 4: 272.
10. Newell, D. L., and J. W. Carey. Experimental evaluation of wellbore integrity along the cement-rock boundary.
11. Environmental Science and Technology. 47: 276.
12. Yang, D., R. Currier, and D. Z. Zhang. Ensemble phase averaged equations for multiphase flow in porous me- dia, part 1: the bundle-of-tubes model. 2009. Interna- tional Journal of Multiphase Flow. 35: 628. Figure 4. Work flow for benchmarking FDEM against triaxial experiment and then upscaling fracturing propagation statis-
13. Zhang, D. Z.. Ensemble phase averaged equations for tics to planar discrete fracture networks and rock matrix at the multiphase flows in porous media, part 2: a general reservoir scale. theory. 2009. International Journal of Multiphase Flow. 35: 640. References
14. Moniz, E. J., H. D. Jacoby, and A. J. Meggs. The future 13. Zubelewicz, A.. Liquid-liquid-solid transitions in visco- of natural gas. 2012. An Interdisciplinary MIT Study. elastic liquids. 2013. Sci. Reports. 3: 320.
15. Study of the potential impacts of hydraulic fracturing 14. Hammond, G. E., P. C. Lichtner, C. Lu, and R. T. Mills. on drinking water resources progress report. 2012. EPA PFLOTRAN: reactive flow and transport code for use on 601/R. laptops to leadership-class supercomputers. 2012. In Groundwater reactive transport models. By Zhang, F.,
16. Porter, M. L., E. T. Coon, Q. Kang, J. D. Moulton, and G. T. Yeh, and J. C. Parker. , p. 141. New York: Bentham J. W. Carey. Multicomponent interparticle-potential Books. lattice Boltzmann model for fluids with large viscosity ratios. 2012. Physical Review E.. 86: 1231. Publications Chaudhuri, A., H. Rajaram, and H. Viswanathan. Early-stage
17. Liu, H., A. J. Valocchi, and Q. Kang. A three-dimensional Hypogene Karstification in a Mountain Hydrologic lattice Boltzmann model for immiscible two-phase flow System: A Coupled Thermo-Hydro-Chemical Model simulations. 2012. Physical Review E. 85: 246. 355

Page 356

Incorporating Buoyant Convection. To appear in Water Resources Research. Zubelewicz, A.. Liquid-liquid-solid transitions in viscoelastic liquids. 2013. Scientific Reports. 3: 234. 356

Page 357

Environmental and Biological Sciences Directed Research Final Report Defeating Bacterial Multi-drug Resistance: An Experimentally Driven Multi-scale Study of Efflux Machinery Sandrasegaram Gnanakaran 20130183DR Abstract mechanisms include escape mutations on the target Bacterial multi-drug resistance efflux pumps are complex gene or acquisition of resistant genes by horizontal gene molecular machines that expel multiple drugs and antibi- transfer. Physical processes include bacterial membrane otics. They also influence important genetic and cellular structures or the impenetrable layer of biomolecules processes to confer additional drug resistance. We de- around a consortium of bacteria called biofilms. The veloped an initial mathematical framework to integrate MDR efflux pumps represent a special case in which structural, genetic, and cellular processes to understand physical processes and genetic mechanisms are coupled. how efflux pumps work. To test our model, we carried It appears that the MDR efflux pumps not only pump out measurements on the time-kill behavior of wild-type out antibiotics but also genetically regulate the forma- Pseudomonas aeruginosa as well a mutant strain lacking tion of biofilms. We will focus on a class of efflux pumps efflux pumps. The integrated model predicts the same (the resistance nodulation cell division or RND pumps), quantitative behavior as the measurements and suggests which are primarily responsible for multidrug resistance specific experimentally testable mechanisms that are in gram-negative bacteria. These tripartite efflux pumps important for drug efflux. During this feasibility study, we are assembled from three proteins: an inner-membrane have also determined the crystal structure of the inner protein, an outer-membrane protein, and a periplasmic membrane E. coli efflux protein and constructed a three- protein that connects the two. The three genes en- dimensional structural model of the entire efflux pump coding these proteins form a cluster with an upstream machinery of multiple organisms. switch acting as a regulator. Bacteria detect the presence of an antibiotic inside the Background and Research Objectives cell. Responding to this signal, they increase the produc- At present we rely heavily on antibiotics for the treat- tion of efflux pumps that expel the drug [2] and decrease ment of bacterial infections encountered in public health the production of porins that allow entry of the drug and bio-threat scenarios. However the rapid emergence [3]. The production of efflux pumps is controlled both by of antibiotic-resistant bacteria poses a major hurdle in local regulator proteins that turn on expression of the the treatment of such infections. Membrane protein pump genes, and by global regulator proteins that affect complexes, called the multi-drug resistance (MDR) efflux expression of many genes including porins. The detailed pumps, are the most important machinery that bacteria mechanisms of how drugs interact with and control utilize to pump out almost all types of antibiotics before drug efflux systems will allow an understanding of how they can act on their targets [1]. Although all bacteria the bacteria respond to the threat of antibiotics. Efflux express these MDR efflux pumps, very little is known pumps can also transport quorum-sensing molecules about how they work. In this project, we develop a which, when exported from one cell, induce neighboring multi-scale model to integrate structural, genetic, and cells to produce a new set of proteins for biofilm produc- cellular processes that underlie the functions of MDR tion allowing bacterial growth, virulence, and survival in efflux pumps in antibiotic resistance (Figure 1). This the presence of antibiotics [5-6]. It is our hypothesis and knowledge will provide us with important clues for de- the research objective that an experiment-based model signing drugs that block the function of the MDR efflux that integrates structural, genetic, and cellular mecha- pumps and rescue the activities of current drugs. nisms to predict drug uptake and survival will provide a Bacteria utilize both genetic mechanisms and physical quantitative understanding of how efflux pumps and the properties to resist the action of antibiotics. Genetic associated genetic and cellular systems defend bacteria 357

Page 358

against drugs and antibiotics. In this feasibility study, the periplasmic cavity. The residues 384-387 from each AcrB research goal is to develop such a quantitative integrated protomer are close to the bound Linezolid. The struc- framework for a model organism, Pseudomonas aerugi- ture suggests that drug binding causes a conformational nosa, based on published data and new measurements. change in residues 670-675 (Figure 3). The drug binding and associated conformational changes we have identi- Scientific Approach and Accomplishments fied will guide MD simulation studies for determining the We have developed a multi-scale model of a bacterial MDR mechanism of drug efflux through the tripartite complex. efflux system using coupled ordinary differential equations Preliminary MD simulations are underway that consider that integrate structural, genetic, and cellular processes. key mutations in this loop, seeking to find how they affect The model describes the dynamics of cell state, growth, accessibility and binding. and death; antibiotic uptake; and efflux regulation. Each of We have constructed a structural model [9] of the P. these behaviors is precisely quantified using variables that aeruginosa tripartite MexAB-OprM complex based upon correspond to experimental observables: live cells, dead the crystal structures of the individual components MexA, cells, and antibiotic levels using fluorescence assays, and MexB, and OprM, energetic calculations of the membrane- pump levels using mRNA sequencing and affinity reagents. bound complex, and distance constraints between com- In this way, the model is firmly grounded in the experimen- ponents derived from crosslinking studies. As shown in tal data. Our model is a precise mathematical representa- Figure 4, the three-dimensional structure of the complex tion of a set of hypotheses or assumptions concerning spans the inner-membrane, periplasm, and outer-mem- the MexAB RND efflux system of P. aeruginosa. The model brane of the bacteria. The three-dimensional structure of is designed to enable iterative refinement and may be the tripartite complex sheds insight into molecular mecha- adapted to predict any of the observed system behaviors nisms of drug binding and export illustrated in Figure 1, for wild type or mutant cells. and enables us to perform molecular dynamics simulations To test our model, we carried out measurements on the to quantify and develop more complex hypotheses. death of wild-type Pseudomonas aeruginosa as well a mutant strain with a deficient efflux system, in the pres- Impact on National Missions ence of an antibiotic (ciprofloxacin) that blocks a bacte- Multi-drug resistance poses a major hurdle in management rial DNA gyrase during replication [6]. P. aeruginosa has of bacterial diseases encountered in both bio-threat and several tripartite efflux pumps. Of particular interest is the public health scenarios. Development of a capability in un- MexAB-OprM pump because it has been shown to trans- derstanding and countering multi-drug resistance in bacte- port both ciprofloxacin and quorum sensing acetyl homo- ria will be directly relevant to the US and LANL missions in serine lactones [7]. Figures 2A and 2B show fits (lines) of bio-security. This project is directed at the development of our integrated model to the measurements of intracellular an integrated modeling capability that relies on appropri- antibiotic (Figure 2A) and number of dead P. aeruginosa ate and relevant experimental measurements to address (Figure 2B) as a function of time. Figure 2A shows that the multi-drug resistance. We have been able to measure and bacteria with an intact system (PAO1) took up antibiotic model the functional mechanism, genetic regulation, and rapidly (red points) and then pumped it out rapidly. This cellular effects of the multi-drug resistance efflux pumps is modeled with an intrinsic antibiotic uptake rate and an that offer resistance against all clinically approved antibiot- antibiotic-induced efflux rate that rapidly increases after ics. antibiotic enters the cells (green curve). A mutant lack- In this project, we developed an initial framework to ing the pumps (MexAB-OprM) reveals two differences; a couple processes at multiple length and time scales: the lower intrinsic uptake rate and much slower efflux of the nanometer molecular machine that pumps out drugs in antibiotic (blue points and purple curve). There is good seconds to minutes; the induction of gene clusters of mul- agreement between model and experiment, indicating that tiple genes spanning hundreds of nanometers and their our model provides insight into MDR efflux mechanisms in- induction occurring over hours; and the formation of bio- volved in the dynamics of cell death and antibiotic uptake film covering several microns over days. Such a capability for P. aeruginosa. for integrated modeling of a complex adaptive molecular Also as part of our feasibility project, we determined the transport system will have broad applications in biosecu- crystal structure of the inner membrane E. coli efflux rity, biofuel production, and clearance of toxic materials by protein, AcrB, in complex with the antibiotic Linezolid (an both microbial and human cells. inhibitor of bacterial protein synthesis) [8]. We found that three Linezolid molecules bind to the AcrB trimer in the 358

Page 359

Figure 4. (A) Experimentally determined structures of different Figure 1. Schematic representation of the structural, genetic, and components of efflux pump. (B) The structure of the entire efflux cellular processes associated with the MDR efflux pump pump machinery modeled using a geometry based computa- tional method that takes into account known constraints from measurements. References

1. Nikaido, H., and J. M. Pages. Broad-specificity efflux pumps and their role in multidrug resistance of Gram- negative bacteria. 2012. FEMS MICROBIOLOGY RE- VIEWS. 36 (2): 340.
2. Martins, A., C. Iversen, L. Rodrigues, G. Spengler, J. Ra- mos, W. V. Kern, I. Couto, M. Viveiros, S. Fanning, J. M. Pages, and L. Amaral. An AcrAB-mediated multidrug- Figure 2. Fit of the theoretical curve with experimental data resistant phenotype is maintained following restora- for normal P. aeruginosa and an efflux mutant: (A) intracellular tion of wild-type activities by efflux pump genes and level of ciprofloxacin; (B) mortality curve as a function of time their regulators. 2009. INTERNATIONAL JOURNAL OF ANTIMICROBIAL AGENTS. 34 (6): 602.
3. Davin-Regli, A., J. M. Bolla, C. E. James, J. P. Lavigne, J. Chevalier, E. Garnotel, A. Molitor, and J. M. Pages. Membrane Permeability and Regulation of Drug “Influx and Efflux” in Enterobacterial Pathogens. 2008. CUR- RENT DRUG TARGETS. 9 (9): 750.
4. Hoiby, N., T. Bjarnsholt, M. Givskov, S. Molin, and O. Ciofu. Antibiotic resistance of bacterial biofilms.
5. INTERNATIONAL JOURNAL OF ANTIMICROBIAL AGENTS. 35 (4): 322.
6. Minagawa, S., H. Inami, T. Kato, S. Sawada, T. Yasuki, S. Miyairi, M. Horikawa, J. Okuda, and N. Gotoh. RND Figure 3. Crystal structure of the E. coli AcrB trimer bound to 3 type efflux pump system MexAB-OprM of pseudomo- molecules of Linezolid. The binding domain for the outer-mem- brane TolC is located at the top. The drug binding cavity is on the nas aeruginosa selects bacterial languages, 3-oxo-acyl- periplasmic side just at the end of the trans-membrane helices. homoserine lactones, for cell-to-cell communication. The porter domain connects the TolC binding domain and the 2012. BMC MICROBIOLOGY. 12: -. drug binding cavity. The location of the 670-675 \ loops that undergo conformation change is shown 6. Bruchmann, S., A. Dotsch, B. Nouri, I. F. Chaberny, and 359

Page 360

S. Haussler. Quantitative Contributions of Target Altera- tion and Decreased Drug Accumulation to Pseudo- monas aeruginosa Fluoroquinolone Resistance. 2013. ANTIMICROBIAL AGENTS AND CHEMOTHERAPY. 57 (3): 1361. 7. Venturi, V.. Regulation of quorum sensing in Pseudo- monas. 2006. FEMS MICROBIOLOGY REVIEWS. 30 (2): 274. 8. Hung, L. W., H. Kim, S. Murakami, G. G. Gupta, C. Y. Kim, and T. Terwilliger. Crystal structure of AcrB com- plexed with linezolid at 3.5 Å resolution . 2013. J. Struct. Funct. Genomics. 14 (2): 71. 9. Phillips, J., and G. S. Gnanakaran. Coarse-Grained Simu- lations of the MexAB-OprM Multidrug Resistance Efflux Pump. 2013. BIOPHYSICAL JOURNAL. 104 (2): 286A. Publications Hung, L. W., H. Kim, S. Murakami, G. G. Gupta, C. Y. Kim, and T. C. Terwilliger. Crystal structure of AcrB com- plexed with linezolid at 3.5 Å resolution . 2013. Journal of Structural and Functional Genomics . 14 (2): 71. Phillips, J., and G. S. Gnanakaran. Coarse-Grained Simula- tions of the MexAB-OprM Multidrug Resistance Efflux Pump. 2013. BIOPHYSICAL JOURNAL. 104 (2): 286A. 360

Page 361

Environmental and Biological Sciences Early Career Research Continuing Project Identifying, Creating, and Controlling Functional Hotspots in DNA Boian Alexandrov 20110516ECR Introduction biomaterials of the future, and for environmental and Transcription, i.e. the decryption of genetic code, is the safety concerns. Our tools for predicting promoters and biological function that governs the production of the transcription factor binding sites as well as for control- biomolecules necessary to the cell. The ability to un- ling genomic functional elements will be a timely and derstand and control transcription yield: new biotech- impactful contribution to the rapidly growing field of nological methods for pharmaceutical drugs, nutrients, biotechnology and functional genomics. The capabili- bioenergy, biomaterials – in short, control over living ties provided by our project will provide unique as- matter. Transcription itself is controlled by various pro- sets and give LANL a significant advantage in genomic, teins (transcription factors), whose binding to the DNA biomaterials, and bioenergy science. This development double-helix enhance or suppress the process. Our prior will enhance LANL’s capability to be at the forefront of work has already established that transcriptional start these fields, and to harness vast databases of emerging sites (TSSs) and transcription-factor-binding sites (TFBSs) genomic data, and gain control of biological function in possess enhanced local DNA breathing dynamics, where the service of national missions in security, energy, and the two strands of DNA move back and forth, opening health. temporary gaps. When DNA local transient openings (bubbles) fall at the TSS or TFBS regions, we call them Progress functional hotspots. Importantly, the patterns of the We completed our transcription factors binding sites hotspots are strongly interconnected with the function (TFBS) recognition protocol and developed a technology of the TSS or TFBS. We have demonstrated this inter- for transcription start sites (TSS) and TFBS modifications, connection theoretically using computer simulations to based on modeling of DNA vibrational hotspots. The identify hotspots for several TSSs and TFBSs, and experi- technology has a novel capability to predict effects of mentally by: i) designing DNA single-point mutations various DNA base modifications. We applied the recog- to silence TSS dynamics (i.e. by destroying functional nition protocol and the modification technology to the hotspots), which led to suppressed transcription without genomic binding sites of two well-known transcription affecting transcription-factor binding; ii) introducing an factors. artificial mismatch opening at a TSS site (i.e. by creating functional hotspots), which led to bidirectional transcrip- Specifically, we investigated in detail the relation be- tion initiation in the absence of basal transcription fac- tween DNA breathing and the binding of Fis protein, tors. Additionally, we have demonstrated theoretically which is known to bind at highly variable DNA sequences that exposure to terahertz (THz) radiation can resonantly in Escherichia coli and Salmonella typhimurium. Leverag- induce DNA hotspots, potentially enabling control of ing our modification-prediction technology, we de- the local breathing dynamics, and hence control of gene signed various Fis binding sites, using various nucleotide expression. modifications, such as: mismatches; methylations; and substitutions in DNA segments with known interaction Benefit to National Security Missions with Fis protein. In each case, our intention was to make Los Alamos National Laboratory has a large interest in the DNA hotspots either closer to or farther from the genomic studies, development of modern statistical pattern of a typical hotspot of a “strong” Fis binding site, learning computational tools, and advanced knowledge derived by our technology. Our predictions have been on cell biological functions in connection with bioenergy, validated experimentally. 361

Page 362

We applied our computational framework to recognize DNA breathing. We propose that THz radiation has poten- YY1 binding sites in vivo, by deriving a specific DNA breath- tial for non-contact control of cellular gene expression. The ing profile of the transcription factor YY1 binding sites. By results were generalized in the Nature Magazine. computer simulations we indentified genomic YY1-binding sites and found that the genomic flanking sequence Future Work variations and single nucleotide polymorphisms, outside We will finish our analysis of the annotation of the whole the YY1 binding motif, are capable of exerting long-range genome of the new phage DANTE, sequenced recently effects on DNA breathing and hence predetermine the in Brown University. Next, with our collaborators from YY1 binding. Our conclusions and predictions have been Harvard Medical School we will experimentally validate confirmed in the experiments reported in a joint-work with some of the annotations predicted by our simulations. The collaborators from Harvard Medical School. The research validation experiments will verify and establish our tech- resulted in an article published in Nucleic Acid Research. nology. Further, we plan to apply the developed genomic method for prediction of transcriptional start sites and Led by our research on the interrelation between tran- transcription factor binding sites, to another whole ge- scription factors binding and DNA dynamics, we prepared nome of the phage SmallFry, recently sequenced in Brown an Invention Disclosure in the LANL/IDEA. Our proposal University. The two whole phage genomes annotations will was approved and an Invention Disclosure Record has ascertain the ability of our method for functional phage been created in the laboratory docketing system. predictions based on DNA hotspots dynamics calculations. Additionally, we will apply our new computational tech- We used our novel TSS and TFBS recognition protocol to nology for design of transcription factor binding sites with annotate the regulatory sites in the whole genome of a various strengths, to another transcription factor, v-Myb, novel phage, DANTE, sequenced by collaborators from which is important in plants, algae, and mammalian gene Brown University. Our work was reported at, National expression. Specifically, we will design DNA segments with scientific competition at HHMI. various modified binding affinities based on computer simulations. We will use base pairs substitutions and nucle- We finished a detailed analysis of our second large set of otide methylations. Our new predictions will be verified by terahertz (THz) irradiation experiments of mouse stem measuring the binding affinities to the designed DNA se- cells at MPA-CINT and NHMFL/LANL, and of the biological quences via EMSA experiments in Hong-Geller laboratory experiments conducted at Harvard Medical School and at LANL. Additionally, we will perform statistical analysis of at UNMKUGR Genomic Center. Specifically we analyzed the available genomic data for Myb-DNA binding in order the gene expression data from 18 gene chip micro-assays, to verify the prediction of our simulations in vivo. and additionally the RT-PCR data obtained for 23 selected genes, from the irradiation of cellular cultures of mouse Conclusion mesenchymal stem cells irradiated by two different tera- We expect to develop functional hotspot recognition pro- hertz sources - one with single frequency and the other tocols, which allow TSS and TFBS prediction significantly with broadband frequency, and with ultrafast pulses beam. out-performing the available state-of-the-art tools, and thus help us understand how DNA does its job. Further, we We demonstrated that the insignificant differential ex- expect to develop a methodology that will enable the de- pression of heat shock and stress related genes as well as sign of hotspots so we can control gene expression. Finally, our temperature measurements implied a non-thermal we expect to validate the hypothesis that THz irradiation response. The microarray survey and RT-PCR experiments has a frequency-specific effect on DNA functional hotspots demonstrate that at different irradiation conditions distinct that can be used for gene expression control. groups of genes are activated. Stem cells irradiated for 12 hours with the broadband THz source exhibit an acceler- Publications ated differentiation toward adipose phenotype, while the Alexandrov, B. S., A. R. Bishop, L. Kalaydjieva, I. S. Kohane, 2-hour (broadband or SF) irradiation affects genes tran- K. O. Rasmussen, A. Usheva, and V. I. Valtchinov. DNA scriptionally active in pluripotent stem cells. Phenotypic Conformation Dynamics and Human Diseases. 2012. and gene expression differences suggest that the THz LANL ADTSC Science Highlights 2012, LA-UR 12-20429. effect depends on irradiation parameters such as duration, frequency, and type of THz source, and additionally on the Alexandrov, B. S., A. R. Bishop, N. Zahariev, and I. level of stem cell differentiation. Computer simulations of Kostadinov. Dispersed stable states spectrum of the the core promoters of two pluripotency markers reveal wave equation with space-time periodic potential. association between gene upregulation and propensity for 2013. Europhysics Letters. 103: 50001. 362

Page 363

Alexandrov, B. S., K. O. Rasmussen, A. R. Bishop, A. Usheva, No.: e15806. L. B. Alexandrov, S. Chong, Y. Dagon, L. G. Booshehri, C. H. Mielke, M. L. Phipps, J. S. Martinez, H. T. Chen, Jablensky, A., D. Angelicheva, G. J. Donohoe, M. and G. Rodriguez. Non-thermal effects of terahertz Cruickshank , D. N. Azmanov, W. D. Morris, C. S. radiation on gene expression in mouse stem cells. Weickert, K. W. Carter, D. Chandler, B. S. Alexandrov, 2011. Biomedical Optics Express . 2 (9): 2679. A. Usheva, B. Morar, P. L. Verbrugghe, A. Filipovska, O. Rackham, A. R. Bishop, K. O. Rasmussen, M. Dragovic, Alexandrov, B. S., K. O. Rasmussen, A. R. Bishop, and M. Cooper, M. Phillips, J. Badcock, E. Bramon-Bosch, O. A. Usheva. Protein-DNA binding and breathing P. Almeida, L. Flicker, M. Gill, A. Corvin, S. Macgregor, dynamics of DNA. 2013. JOURNAL OF BIOMOLECULAR and L. Kalaydjieva. Promoter Polymorphisms in two STRUCTURE & DYNAMICS. 31: 49. Overlapping 6p25 Genes Implicate Mitochondrial Proteins in Cognitive Deficit in Schizophrenia. 2012. Alexandrov, B. S., M. L. Phipps, L. B. Alexandrov, L. G. Mol Psychiatry. 17 (12): 1328. Booshehri, A. Erat, J. Zabolotny, C. H. Mielke, H. T. Chen, G. Rodriguez, K. O. Rasmussen, J. S. Martinez, A. Maniadis, P., B. S. Alexandrov, A. R. Bishop, and K. O. R. Bishop, and A. Usheva. Specificity and Heterogeneity Rasmussen. Feigenbaum cascade of discrete breathers of Terahertz Radiation Effect on Gene Expression in in a model of DNA. 2011. PHYSICAL REVIEW E. 83 (1, Mouse Mesenchymal Stem Cells. 2013. SCIENTIFIC 1): 011904. REPORTS. 3: -. Nowak-Lovato, K., L. B. Alexandrov, A. Banisadr, A. L. Bauer, Alexandrov, B. S., V. G. Stanev, A. R. Bishop, and K. O. A. R. Bishop, A. Usheva, F. P. Mu, E. Hong-Geller, K. Rasmussen. Anharmonic dynamics of intramolecular O. Rasmussen, W. S. Hlavacek, and B. S. Alexandrov. hydrogen bonds driven by DNA breathing. 2012. Binding of Nucleoid-Associated Protein Fis to DNA Is PHYSICAL REVIEW E. 86 (6): -. Regulated by DNA Breathing Dynamics. 2013. PLOS COMPUTATIONAL BIOLOGY. 9 (1): -. Alexandrov, B. S., Y. Fukuyo, M. Lange, N. Horikoshi, V. Gelev, K. O. Rasmussen, and A. R. Bishop. DNA Breathing Dynamics Distinguish Binding from Nonbinding Consensus Sites for Transcription Factor YY1 in cells. 2012. Nucleic Acids Research. 40 (20): 10116. Alexandrov, B., V. Valtchinov, L. Alexandrov, V. Gelev, Y. Dagon, J. Bock, I. Kohane, K. Rasmussen, A. Bishop, and A. Usheva. DNA Dynamics Is Likely to Be a Factor in the Genomic Nucleotide Repeats Expansions Related to Diseases. 2011. PLOS ONE. 6 (5): e19800. Alexandrov, Boian S., Amy L. Bauer, Alan R. Bishop, F. Mu, E. Hong-Geller, K. Nowak-Lovato, Kim Ø. Rasmussen, and Ludmil B. Alexandrov. New Insights into Protein- DNA Binding from Simulations of DNA Breathing Dynamics. 2013. Los Alamos National Laboratory Associate Directorate for Theory, Simulation, and Computation (ADTSC) LA-UR 13-20839. Bishop, A. R., K. O. Rasmussen, A. Usheva, and B. S. Alexandrov. Entropy–Driven Conformations Controlling DNA Functions. 2012. In Disorder and Strain-Induced Complexity in Functional Materials, Springer Series in Materials Science. Edited by Kakeshita, T., T. Fukuda, A. Saxena, and A. Planes. Vol. 148, 1 Edition, p. 273. Berlin: Springer. Bock, J., Y. Fukuyo, S. Kang, M. Lisa. Phipps, L. Alexandrov, K. Rasmussen, A. Bishop, E. Rosen, J. Martinez, H. Chen, G. Rodriguez, B. Alexandrov, and A. Usheva. Mammalian Stem Cells Reprogramming in Response to Terahertz Radiation. 2010. PLoS One. 5 (12): Article 363

Page 364

Environmental and Biological Sciences Early Career Research Continuing Project Construction and Functionalization of Novel Biomimetic Human Liver Jun Gao 20120716ECR Introduction we can use these complete in vitro 3D tissue/organs ca- The main objective of this project is to develop a fully pabilities to develop proposals targeting environmental functional in vitro human liver organ, as opposed to and workplace exposure paradigms. simple in vitro liver tissue, using cutting-edge advance- ments in tissue engineering and material technologies. Progress In order to best replicate optimal liver organ perfor- The main objective of this project is to develop a fully mance and derive liver that can function normally, the functional in vitro human liver organ. Now we have liver tissue must contact essential body fluids, which successfully accomplished the specific aim I, and de- transport immune cells and allow for nutrient and gas veloped small-scale hepatic microvascular system in in exchange, Thus, an ideal tissue would be permeated vitro liver tissue (microfluidic platform). Different hol- with structures that represent the vascular system to low fiber membranes (PP, ME, PVDF, PS, mPES and PES) mimic such an exchange. Here, we propose to use the were evaluated for vascular system development. We CellMax® Bioreactor coupled with a biocompatible hol- demonstrated that PP, ME and PVDF can be served as low fiber network to grow an artificial liver organ while bio-compatible scaffolds for integral monolayer vascular simulating the necessary vasculature and flow-through endothelium development. The porosity of the hollow of immune cells. The internal lumen of hollow-fibers will fiber membranes allowed the cross-diffusion of low mo- be used as a scaffold to support vascular endothelial cell lecular weight secretory factors for exchange of nutrients growth and will facilitate endothelium formation. The and biological signaling molecules. We observed that the porosity of the hollow fibers can be engineered to allow hepatic vascular system was fully differentiated under for the diffusion of low molecular weight secretory fac- biological relevant shear stress. The differentiation bio- tors while permitting exchange of nutrients and waste, markers were monitored by real-time PCR. Now we are but also preventing the cross-diffusion of immune/liver reconstructing large-scale 3D liver system with function- cells. The external space of the hollow fibers will provide alized vascular system as descried in Aim 2. a three-dimensional (3D) matrix for hepatic cell growth, while allowing for migration, organization, prolifera- During this study, we developed both in vitro tissue en- tion, and differentiation of hepatic cells resulting in fully gineering and microfluidic capabilities. The cutting-edge developed, functional liver organ. microfluidic capability facilitated the development of novel genetic analysis program. We developed one DR, Benefit to National Security Missions 2 ERs and one DTRA proposals by using this microfluidic LANL has significant expertise in cutting edge Bioassess- capability. This project also helped the development of ment development. This novel liver organ platform will some other novel ideas. We have submitted one IDEA be an important component in the arsenal of in vitro (#13-00100 “Novel in vitro neuromuscular junction tis- tissue/organ assessment platforms that already includes sue platform on microfluidic for predictive toxicological 3D in vitro skin and lung tissue systems. Success in de- study”). veloping the 3D in vitro liver organ will find immediate application to CBWA risk analysis and mitigation, sup- We are in the process of preparing the manuscript for porting the LANL threat reduction mission and creating the small-scale liver tissue platform with functional viable external funding opportunities, such as NIH and vascular system. During the past one year project per- DTRA. Combining with our skin & lung tissue platform, formance, we also hired one postdoc and one student, 364

Page 365

and trained them on tissue engineering and microfluidic technologies. They have presented their studies on Post- doc Research Day and at the SULI student forum. This proj- ect provides me a great opportunity to collaborate with both internal (Chemistry, Material and Physics science) and external (UNM and Stanford University) senior scientists in this field. These collaborations will be extremely important for future career development. Future Work Construct bioreactor modular microsystems to represent complex physiology and biology of human liver. We will develop biologically relevant architectures of human liver and microvascular scaffolds using hollow fiber and micro- fluidic technology. We will integrate human primary liver cells and microvascular cells into the bioreactor modular, and develop a complete, surrogate small scale human liver organ. We will validate the system by identifying the liver specific development biomarkers. We will introduce hepa- totoxic compounds into the small scale system and validate the toxicological responses. Conclusion We anticipate that we will successfully develop a complete, surrogate in vitro 3D human liver organ that will be imple- mented into diverse range of clinical, biomedical, pharma- ceutical and toxicological applications. This platform can be used for long-term functional toxicological evaluation of chemical and biological warfare agents (CBWA). Fur- ther, this method of artificial organ construction can also be used to develop other human organs, such as kidney, pancreas and nerve tissue. Publications Gao, J., and S. Burchiel. Polycyclic Aromatic Hydrocarbons and the Immune System. Encyclopedic Reference of Immu- notoxicology. Edited by Vohr, H.. Gao, J., and S. Burchiel. Genotoxic Mechanisms of PAH- induced Immunotixicity. Molecular Immunotocicology. Edited by Corsini, E.. 365

Page 366

Environmental and Biological Sciences Early Career Research Continuing Project The World’s First Drought and Insect Caused Global Tree Mortality Monitoring System Chonggang Xu 20130733ECR Introduction Benefit to National Security Missions The critical urgency of forecasting climate impacts The development of the world’s first insect-caused tree and feedbacks makes understanding, quantifying, and mortality monitoring system will represent an enormous predicting terrestrial carbon balance and subsequent step forward in our understanding of terrestrial carbon climate impacts one of the greatest science challenges feedback to atmosphere, which is a key area of climatic currently facing the world. The real-time monitoring change research in LANL’s mission to mitigate the im- systems for dominant types of disturbances will provide pacts of global energy demand. It can help DOE Office of a key foundation for our understanding of global carbon Science build a global tree mortality database used for balance. We currently already have fire-burned area dynamic global vegetation model benchmarking, a key monitoring system and comprehensive land-use change area that DOE Office of Science is developing through database; however, there is no global monitoring system the International Land Model Benchmarking Project. of drought/insect-caused vegetation mortality, which The data generated through this project can be used to could be at similar magnitude of fire-caused tree mortal- substantially improve tree mortality prediction in the ity. DOE-sponsored Community Land Model (CLM) and thus improve global climate predictions. Armed with the world’s leading capability in dynamic vegetation modeling and tree mortality research, we Progress propose to develop the world’s first automated global We have made the following important progress that drought/insect-caused tree mortality monitoring system. direct us favorably toward our final deliverable. The mortality monitoring system is developed based on First, we have hired a new postdoc, Jordan Muss. He the fusion of different sources of information including started on May 20th and he will be mainly responsible real-time mortality signal from remote sensing imagery, for data collection and implementation of our monitor- vegetation change information simulated from a vegeta- ing system in the SW. tion dynamics model, radiative transfer and reflectance information from a forest reflectance model, and dif- Second, we have developed a make file so that the ED- ferent sources of background information from forest FRT code is able to run on super computers. This is will inventory and remote sensing products. The fusion of be a great step toward our final delivery for the whole different sources of information makes it feasible for the southwest. first time in the world to accurately quantify tree mortal- ity using Moderate Resolution Imaging Spectroradiom- Third, we have written a LANL high performance com- eter (MODIS) imagery from NASA, which is a common puting proposal for this project. We have been awarded remote sensing tool for monitoring earth system pro- with 50K CPU-hours under the project name of w13_ cesses globally. This detailed tree mortality quantifica- treemort. tion has never been possible by analyzing MODIS image alone. The successful development of our monitoring Fourth, we have improved the ED model to explicit system will represent an enormous leap forward in our improve the water storage function, which helps ED to understanding of terrestrial carbon feedback to atmo- have an improved simulation of tree mortality under sphere, which is a key area of climatic change re-search drought. This could potentially improve the ED-FRT per- in LANL’s mission to understand and predict the impacts formance under drought. of global energy demand. 366

Page 367

Fifth, we have compiled the benchmarking datasets for Publications five sites in New Mexico, Colorado, Western Australia and Adam, H., P. William, C. Xu, S. Rausher, and N. McDowell. Texas. This will provide the necessary data for our first Empirical and process-based approaches to climate- methodology paper. induced forest mortality models. To appear in Frontiers in Functional Plant Ecology. Sixth, we have purchased and installed necessary soft- Xu, C.. Decoupling correlated and uncorrelated uncertainty ware to support this project including the remote sensing contributions for nonlinear models. To appear in software (ENVI) and programming software (Intel Fortran Applied Mathematical Modelling. Compiler). Xu, C., R. Fisher, J. Muss, and N. McDowell. Uncertainty and Future Work Sensitivity Analysis for Process-based Tree Mortality Modelling. Invited presentation at ECOLOGICAL Our goal is to fully develop and test of the proposed tree ASSOCIATION OF AMERICA 2013. (Minneapolis, 8-13 mortality monitory system at fine scales. Specifically, we AGU 2013). will refine and test the drought and insect caused tree mortality monitoring system for representative locations Xu, C., R. Fisher, N. McDowell, and S. Sevanto. Our Limited with high resolution images at Colorado, New Mexico and Ability to Predict Vegetation Responses to Water Western Australia. Each location is comprised of ~100 MO- Stress. 2013. New Phytologist. 200 (2): 8. DIS pixels. The goal includes the following tasks: Xu, C., and M. Chen. Application of Remote Sensing in Ecosystem and Landscape Modeling. To appear in • Climate data, forest types, forest height and allometry Remote Sensing of Natural Resources. Edited by Wang, information will be collected for different sites; G.. • Simulating a browning phase using foliar chlorophyll data collected from an experimental site provided by our collaborator Dr. Marcy Litvak at UNM (email: mlit- [email protected]; phone:505-277-5580), where piñon trees were girdled to mimic tree mortality • At each site, proposed tree mortality monitory system will be applied to estimate amount of tree mortal- ity, which is then evaluated against the classification results from the high-resolution images; • A data assimilation approach will be developed to correct initial total plant density for a good fitting between simulated and observed reflectance signals from remote sensing imagery for the pre-mortality period Conclusion The success of the project will generate two critical prod- ucts for global carbon cycle assessment and Earth system model simulation. First, we will fully develop and test world’s first insect/drought-caused global tree mortal- ity monitoring system through the coupling of a dynamic vegetation model, a light reflectance and transmittance model, and real-time remote sensing observations. Sec- ond, tree mortality database for build for New Mexico, Colorado and Western Australia with estimated carbon loss during the 2000-2012 period will be developed using the our developed monitoring system, which will be used for assessing regional simulations using the DOE-sponsored Community Land Model. 367

Page 368

Environmental and Biological Sciences Early Career Research Continuing Project From Troposphere to Ionosphere: How Much do Thunderstorms Disturb the Total Electron Distribution? Erin H. Lay 20130737ECR Introduction (EMP) that are a key signature of nuclear detonation. The ionosphere is a high-density layer of plasma through Better understanding of ionospheric disturbances near which all satellite-detected natural and man-made elec- thunderstorms will improve EMP location accuracy and tromagnetic signals from the Earth’s surface must pass. discrimination between various impulsive sources. Similar to dispersion of light by transparent solids, signif- icant perturbations in the ionospheric plasma electrons Contribution to Space Science: We will empirically study disrupt electromagnetic signals. Variation in ionospheric a new phenomenon for the first time in which atmo- plasma has traditionally been attributed to changes in spheric pressure waves from thunderstorms can signifi- geomagnetic activity (space weather). Only recently, cantly disturb the electron distribution in the ionosphere the ionospheric community has begun to realize that above the storms. Variation in ionospheric plasma has tropospheric thunderstorms (below ~12 km) could have traditionally been attributed to changes in geomagnetic a significant effect on the ionospheric plasma by the fol- activity (space weather). Only recently, the ionospheric lowing mechanism: when the updraft in a thunderstorm community has begun to realize that tropospheric thun- lifts the top of the cloud above the tropopause (altitude derstorms (below ~12 km) could have a significant effect of a temperature inversion), the cloud can begin oscil- on ionospheric plasma. This study will provide basic lating about that altitude for tens of minutes, producing scientific evidence of the relative importance of thunder- a neutral pressure wave, called an atmospheric gravity storm effects on the ionospheric behavior. wave (AGW), with periods of tens of minutes to hours. Other contributions: Understanding and predicting The goal of this project is to quantify and characterize ionospheric variations is of importance in understand- the ionospheric response to thunderstorm activity, and ing disruption of satellite communications, GPS (Global to determine the mechanisms and magnitude of such Positioning System) geolocation signals, and electro- coupling. This study is the first empirical analysis com- magnetic signals such as EMP. paring ionospheric fluctuations in the peak plasma layer to underlying thunderstorms. For the first time, we will Progress be able to quantify and interpret the spatial extent and We have developed data analysis tools for freely-avail- dynamic evolution F-layer ionospheric disturbances due able measurements of total electron content made by to thunderstorm pressure waves. This is of extreme im- Global Positioning System ground receivers. These can portance for understanding and predicting ionospheric now be used to download and analyze many days of disruption of satellite communications, GPS geolocation data at one time. Now that these tools are in place, we signals, and electromagnetic signals such as EMP that can begin a larger-scale data analysis project to observe are an important part of the Lab’s national security mis- total electron content (TEC) fluctuations near thunder- sion. storms. Benefit to National Security Missions Our paper entitled “Variation in total electron content Contribution to Remote Sensing for Nuclear Nonpro- above large thunderstorms” by E.H. Lay, X-M. Shao, and liferation and Counterproliferation: Significant pertur- C.S. Carrano was published in Geophysical Research bations in the ionospheric plasma electrons disrupt Letters. This paper published the first account of TEC electromagnetic signals, such as electromagnetic pulses fluctuations observed corresponding to large mesoscale 368

Page 369

thunderstorms in time and space. It also made note of a by thunderstorms on a regional scale, and how they very high frequency perturbation (4 minute-period) associ- vary with thunderstorm properties (i.e. size, height, ated with the thunderstorm. lightning activity). • Characterize the magnitude and spatial extent of tro- Work has begun on identifying additional storms in the posphere/ionosphere coupling on a global scale as a same region (United States Midwest) that also exhibit function of geomagnetic latitude. perturbations vs ones that do not, in order to begin deter- mining the type of thunderstorm that contributes to these Impact ionospheric fluctuations. To date, we have discovered one The findings from this study will be used to test and evalu- additional storm with high frequency fluctuations associ- ate existing atmospheric gravity wave models, enabling us ated with another large mesoscale thunderstorm. to distill and discover the physical processes responsible for energy and mass transport between the D-layer and Future Work F-layer, and to better now-cast and forecast ionospheric Main task disturbances. Characterize and classify the upper ionospheric response to lower ionospheric disturbances detected using a region- Publications al lightning detection network. Lay, E. H., X. M. Shao, and C. S. Carrano. Variation in total electron content above large thunderstorms. 2013. Sub-tasks Geophysical Research Letters. 40: 1945. • Develop data analysis tools for freely-available mea- surements of total electron content made by Global Positioning System ground receivers. These tools will be useful to the lab for any future ionospheric projects involving total electron content measurements. • Identify storms in the Los Alamos Sferic Array lightning database with significant lower ionospheric perturba- tions. The technique to calculate perturbations already exists, but it must be applied to many storms to find which storms have significant perturbations. • Determine the properties of the thunderstorms with significant lower ionospheric perturbations (i.e. spatial size, height, lightning activity). • Perform correlational study between upper ionospher- ic fluctuations and lower ionospheric perturbations associated with thunderstorm activity. Goals • Begin to address thunderstorm/ionospheric coupling mechanisms. • Submit 1-2 journal articles addressing the question of “What are the spatial size, magnitude, and temporal evolution of upper ionospheric perturbations related to nearby underlying thunderstorms on a regional scale, and how do they vary with thunderstorm prop- erties (i.e. spatial size, height, lightning activity)?” Conclusion Goals • Characterize spatial size, magnitude, and temporal evolution of ionospheric perturbations related to near- 369

Page 370

Environmental and Biological Sciences Postdoctoral Research Final Report Mechanistic Studies Toward an Improved Carbonic Anhydrase for Biofuel Production Suzanne Z. Fisher 20110535ECR Abstract site water network of CA between different pH. Carbonic anhydrases (CA) are enzymes that can capture CO2 is the main source of carbon for most algae. A major and regulate carbon dioxide by interconverting it with bi- problem for the DOE algal biofuels mission is the slow carbonate. These enzymes are stable and extremely fast, growth rate of algae (days or weeks). The uptake of CO2 but work best over narrow ranges of pH and tempera- into cells as bicarbonate is the rate-limiting step in bio- ture. Our main goal was to characterize the structural mass production while solubility of CO2 is also an issue. and catalytic properties of CA and then to exploit that It has been shown that adding an efficient and stable CA information to re-engineer CAs that can be used over to the algal growth medium can significantly increase the broad range of process conditions, specifically those and regulate the amount of bicarbonate available for relevant to algae growth. Obtaining the unique infor- uptake, thus improving the growth of algae for the mation about the structure-function of CA will require production of biofuels and other value added products. combining for the first time several different experimen- Adding CA is an elegant and finely tunable way to deliver tal techniques, some available only at LANL. We propose the right amount of dissolved CO2 or bicarbonate to the to combine static and dynamic information on hydrogen cells. Before CA can be effectively used in these areas, atom positions in the active site of CA by combining the strong pH-dependence of activity has to be altered, neutron crystallography and NMR spectroscopy. We will as many algae grow best at pH > 8. also co-optimize and select for interdependent catalytic and pH stability properties of CA. Using this approach we Our approach led to many new insights into the CA ac- were able to construct a set of very interesting mutants tive site and overall structure. We were able to make with enhanced thermal stability and better kinetics com- a background mutant with enhanced thermal stability pared to wild type CA. and then used our explicit knowledge of the active to engineer a set of CAs with enhanced kinetics compared Background and Research Objectives to wild type, no small feat when you consider that this is Carbonic anhydrases are abundant enzymes, found in one of the fastest enzymes every characterized. Figure 2 all organisms. Human CA II (which we refer to here as shows a surface representation of the thermally stable just CA) is one of the fastest known enzymes, is easy CA and the mutations that confer stability. and cheap to produce, making it an attractive target for industrial development. Carbonic anhydrases activity Scientific Approach and Accomplishments has strong temperature and pH dependence with a peak Neutron crystallography was used to locate protons at pH 7. Catalysis occurs in 2 steps: 1) hydration of CO2 along the CA proton pathway at several pHs, providing bound at the active site to produce HCO3- and H+, and snapshots of water H-bonding networks and amino acid 2) rate limiting proton transfer away from the active site residue protonation states. We expect proton transport through a network of water molecules and amino acid to depend crucially on the ability of key amino acids to residues out into bulk solvent, to regenerate the active bind and release protons i.e. their pKa. Although the site ready for the next reaction with CO2. The slower presence of protons on amino acid residues at specific proton transfer step is the least understood part of the pH will be provided by our neutron studies, pKa (a dy- mechanism, mostly it has not been possible to visual- namic measure of proton binding and release) can only ize the location and movement of protons in CA using be provided by NMR titration. We also labeled key amino conventional X-ray and NMR structural measurement acids with 13C so that their individual pKa values could techniques. Figure 1 shows the differences in the active 370

Page 371

be determined. Combining static and dynamic information in this way will allow us to study the mechanism of proton transport and how it changes with pH. Site-directed muta- genesis allowed us to confirm the roles of specific amino acid residues. Our new mechanistic insights were used for rational site-directed mutagenesis to optimize fitness of CA for algal processes. In summary, our new approach combines neutron crystal- lography, which reveals the location of protons in snap shots at different pH, with 13C-NMR titration, which can provide dynamic information about how the locations and accessibility of those protons change continuously with pH. Combining static and dynamic information on H movement in enzymes in this way is innovative and has broad applica- bility to many biological problems. Figure 1. Active site water network hydrogen bond changes upon increasing pH. Left panel shows pH 7.8 and right panel shows pH This work generated several high impact papers on the 10. enhanced kinetics and thermal stabilization of CA (Fisher, S.Z et al. (2012) Kinetic and structural characterization of thermostabilized human carbonic anhydrase II, Prot. Eng. Des. Sel. 25, p.347-355). This work was selected to feature as a cover article. We also published the neutral pH structure by neutron diffraction of CA (Fisher et al. (2011) Neutron structure of human carbonic anhydrase II: A hydrogen bonded water network “switch” is observed between pH 7.8 and 10.0, Biochemistry 50, 9421-9423) and a few others, including a review and a JACS paper that was mentioned on over 240 news outlets (Fisher et al. (2012) Neutron diffraction of acetazolamide-bound human carbonic anhydrase II reveals atomic details of drug bind- ing, J Am Chem Soc. 134, p.14726-9). Impact on National Missions Figure 2. Surface representation of CA with the location of stabi- This research directly supports LANL, DOE, and national lizing mutations shown in yellow. missions in renewable energy, biofuels, and carbon man- agement by providing a robust product that will enhance Publications the production of renewable biofuels. Furthermore, the Fisher, Z., A. Kovalevsky, M. Mustyakimov, D. Silverman, R. development of a new measurement capability based on McKenna, and P. Langan. Neutron structure of human combining dynamic and static information from neutrons carbonic anhydrase II: A hydrogen bonded water net- and NMR, and a new engineering capability based on work “switch” is observed between pH 7.8 and 10.0. co-optimizing enzymes with complementary engineering 2011. Biochemistry. : 1. steps, will be of broad significance for the study of a large Fisher, Z., C. Boone, S. Biswas, B. Venkatakrishnan, M. Ag- range of biological problems. garwal, C. Tu, M. Agbandje-McKenna, D. Silverman, and R. McKenna. Kinetic and structural characteriza- tion of thermostabilized mutants of human carbonic anhydrase II . 2012. PROTEIN ENGINEERING DESIGN & SELECTION . 25 (7): 347. Fisher, Z., M. Aggarwal, A. Kovalevsky, D. Silverman, and R. McKenna. Neutron diffraction of acetazolamide-bound human carbonic anhydrase II reveals atomic details of drug binding. 2012. JOURNAL OF THE AMERICAN CHEMICAL SOCIETY. 134 (36): 14726. 371

Page 372

Environmental and Biological Sciences Postdoctoral Research Final Report Low-Frequency Acoustic Interferometry for Probing the Stratosphere Stephen J. Arrowsmith 20110556ECR Abstract for the interaction for mass and momentum with the This project has successfully demonstrated that passive troposphere. A series of recent papers have argued that acoustic sensors can be used to measure winds at high future improvements in numerical weather and climate altitudes. The basis of our innovation is that the differ- prediction lie in “raising the roof” of existing numerical ences in arrival times of continuous ocean noise re- models into the stratosphere [3,4]. Although the strato- corded by ground-based arrays separated by several 10s sphere is important, direct measurements of wind are of kilometers are only sensitive to atmospheric effects limited to discrete sampling and/or very specialized in- between the arrays, thus it is not necessary to simulate struments (Figure 1). Currently, winds in the middle and the wave from source to receiver to image the local upper stratosphere are derived from the thermal wind stratospheric atmosphere. Our novel approach has been equation, sporadic rocket soundings, and recently by accepted for publication in Geophysical Journal Inter- quasi-continuous Doppler Rayleigh lidar. Doppler lidars national and we are preparing on two additional pub- are frequently used for remote sensing of wind in the lications based on this work. While other national and aerosol loaded planetary boundary layer or parts of the international groups working in this area have developed troposphere. Due to the decreasing aerosol and molecu- frameworks that require ground-truth (known) sources, lar density, wind measurements by lidar in the strato- our work provides the necessary framework for continu- sphere require a complicated setup of large aperture ous measurement of winds at high altitudes using ocean telescopes and powerful lasers. Such facilities are expen- noise. Continuous measurements of high altitude winds sive to build and operate, and only provide wind profiles would provide a missing source of data for numerical at a particular location. Aside from Doppler lidar facili- weather and climate prediction models that has been ties, our current best estimates of winds in the strato- identified as a major limitation in predicting the weather sphere are derived from satellite-based measurements and climate [1,2,3]. Currently, winds in the middle and of temperature. There are two fundamental limitations upper stratosphere are derived from the thermal wind with deriving winds from temperature soundings in the equation, an equation that is known to fail in the trop- stratosphere: (1) uncertainties in temperature measure- ics. There is also a range of defense applications from ments amplify when converted to wind [3], and (2) the warfighter support, nuclear explosion monitoring, and thermal wind equation is known to be inadequate for predicting the dispersion of aerosols that would benefit estimating wind in the tropics [4]. from continuous wind measurements at high altitudes. Scientific Approach and Accomplishments Background and Research Objectives Research under this project has proved that it is theo- Atmospheric observations and modeling in the last retically possible to invert for winds in the stratosphere decade have stressed the necessity of a better under- using low frequency acoustic background noise or micro- standing of the stratosphere: 1) to improve weather baroms (acoustic energy from colliding ocean storms). forecast and climate prediction for short and seasonal The basis of our innovation is that the differences in time scales [1], and 2) to understand the dynamics of arrival times of microbaroms recorded by ground-based the troposphere. The challenge is that it is difficult to arrays separated by several 10s km are only sensitive to make direct measurements in this section of the atmo- atmospheric effects between the arrays, thus it is not sphere. The stratosphere lies between 10 and 50 km and necessary to simulate the wave from source to receiver features an increase in temperature due to the increase to image the local stratospheric atmosphere (Figures 2 in concentration of ozone; this layer plays a crucial role and 3). This work has been accepted for publication in 372

Page 373

Geophysical Journal International [5]. Our method, which in atmospheric dynamics). Furthermore, the opportunity we have validated numerically, is unique because it does of direct wind measurements in the tropics may be par- not require knowledge of the source location and can be ticularly important, where the thermal wind equation is applied to continuous infrasound from ocean storms. We known to work poorly. Our results also have implications applied our preliminary inversion framework to infrasound for the NASA program and warfighter support and we from a large bolide that was clearly detected by seven ar- have discussed with Reiner Friedel (ISR-1) how improved rays in Utah. Our results [5] demonstrate that the observed specifications of stratospheric winds has potential implica- data clearly favor a real-time model over a climatological tion for improving GPS accuracy through improved cor- averaged model for the time of arrival of infrasound at the rection for stratospheric effects. Improved specification of Utah network. In particular, the results suggest that the atmospheric winds also has an important implication for zonal wind speed in the climatological averaged model is Global Security (LANL POC: Dale Anderson, EES-17): since too weak, but that the zonal wind speed in the real-time knowledge of stratospheric winds affects the propagation model is more consistent with our observations, although of infrasound signals our results can improve infrasonic still too weak. Following this research, we have identified location and yield estimation of nuclear tests and improve three main shortcomings of existing infrasound networks, the prediction of dispersion of radioactive aerosols, which such as the Utah network in Figure 1, for measuring strato- can reach the stratosphere. spheric winds using microbaroms: 1) the arrays typically target mining and other impulsive sources and are too small to properly measure microbaroms, 2) the sensors used for such arrays have peak sensitivity at frequencies higher than the microbarom spectrum (between 0.2 and 0.3 Hz), and 3) the distances between arrays in existing networks are too far to calculate stratospheric celerity. Within the past two months we have deployed infrasound instruments in New Mexico to determine the optimum configuration for obtaining time-delay measurements from microbarom signals. We found that an appropriately designed network of ~1 km aperture arrays, separated by distances of ~60 km, would provide the necessary dataset for an inversion using microbarom noise. We are finishing a second journal article detailing our New Mexico micro- barom study. Impact on National Missions The ability to continuously measure winds at high altitudes Figure 1. Direct measurements of wind at high altitudes is cur- using low-cost acoustic sensors has implications for a rently only possible with snapshot measurements and expensive number of national missions in weather/climate prediction LIDAR installations (limited to a handful of locations on Earth). and defense. The joint DOE/NSF Community Earth System Model (CESM) is one of three US IPCC-class climate models in which LANL personnel in T-3 and CCS-2 currently play an active research role. The CESM model includes a whole at- mosphere configuration that goes up to the thermosphere to account for mass and momentum coupling between different layers. Future work with existing personnel at LANL, including Phil Jones (T-3), should investigate how our measurements can be incorporated into such models. Joan Alexander at Northwest Research Associates in Boulder, Colorado has also expressed interest in collaborating with us on quantifying benefits to the weather/climate predic- tion community. Of particular interest for the stratosphere dynamics are effects on the ozone layer, which plays a key role in radiative interactions as well as gravity wave drag (a sub-grid scale phenomena that plays an important role 373

Page 374

3. Kalnay, E., J. C. Jusem, and J. Pfaendtner. The rela- tive importance of mass and wind data in the present observing system. Presented at NASA Symp. on Global Wind Measurements. (Columbia, MD, 1 Jan. 1985).
4. Zagar, N.. Assimilation of Equitorial Waves by Line-of- Sight Wind Observations. 2004. Journal of the Atmo- spheric Sciences. 61: 1877.
5. Arrowsmith, S. J., O. Marcillo, and D. P. Drob. A frame- work for estimating stratospheric wind Speeds from unknown sources and application to the December 25, 2010 bolide. To appear in Geophysical Journal Interna- tional. Publications Arrowsmith, S. J., O. E. Marcillo, and D. P. Drob. A frame- work for estimating Stratospheric wind speeds from infrasound noise. Presented at Infrasound Technology Workshop 2012. (Daejeon, South Korea, 8-12 Aug. 2012). Arrowsmith, S. J., O. Marcillo, and D. P. Drob. “A frame- work for estimating stratospheric wind speeds from unknown sources and application to the December 25, Figure 2. Using an acoustic sensor network in Utah, we written a 2010 bolide. To appear in Geophysical Journal Interna- paper, which is currently in review, that demonstrates how delay tional. times between pairs of sensors can provide constraints on high- altitude winds. Figure 3. Our technique iteratively improves a starting model of the East-West wind component, obtained from state-of-the-art models, by modeling sound rays between sensors and minimizing the difference between the observed and predicted delay times. References
6. Gerber, E. P.. Assessing and understanding the impact of stratospheric dynamics and variability on the earth system. 2012. Bulletin of the American Meteorological Society. 93 (6): 845.
7. Shaw, T. A., and T. G. Shepherd. Raising the roof. 2008. Nature Geoscience. 1: 12. 374

Page 375

Environmental and Biological Sciences Exploratory Research Continuing Project Separation of Time Scales at High Latitudes: Fundamental Science and Numerical Method Development Robert E. Ecke 20120206ER Introduction Progress This project builds on a new, fundamental discovery ared Whitehead and Beth Wingate submitted the first — a new regime in fluid dynamics, relevant to the high- paper on the emergence of slow dynamics in the limit of latitude oceans (Wingate et al 2011) but with a broader fast rotation to the Journal of Fluid Mechanics. He and impact on climate model development. The impact of Terry Haut have also made progress on understanding this discovery can effect our understanding of important the mathematics of the 3 limit system. This will be a sec- climate-related science questions such as: why are there ond paper. The implications of this work is a better, more columnar vortices propagating in the deep Arctic and general, understanding of time-scale-separated flows do they contribute to the formation of deep water that which are ubiquitous for climate models. These physical can shut down the meridional overturning circulation?; results are expected to impact our numerical results de- and are there impacts due to oceanic heat transport scribed in the next section. Jared’s work has been so suc- related to in the reduction of Arctic sea ice? Further, cessful he has taken a faculty position at Brigham Young the theoretical underpinnings of this work give us new University where he’ll begin as an Assistant Professor in insight into how to best solve the dynamical equations December of 2013. on new, heterogeneous computer architectures, which could lead to new numerical algorithms, the kind of Terry Haut and Beth Wingate submitted the first paper algorithms that would allow climate models to possibly on the asymptotic parallel-in-time (also called parareal) reach the human scale, rather than the global scale, in method. This new method could address important our lifetime. We therefore have two key objectives: 1) computation issues that will arise in exascale comput- understand the fundamental fluid theory and its con- ing for climate models and have potential impact on nection to new observations in the Arctic and 2) use the other physics problem of interest to LANL. This is a very foundational principals in the theory to develop new, important paper that could not only be highly cited, but proof-of-principle, exascale-friendly, numerical methods lead to greater understanding of how climate models for next generation ocean models. will play out on new heterogeneous computing architec- tures. There is another paper in progress that describes Benefit to National Security Missions a highly parallel technique for solving linear propgators, a necessary key to the asymptotic-parareal method that This work directly impacts LANL’s mission for Mitigating uses general methods for the horizontal discretization. Impacts of Energy Demand Growth under the umbrella This work is a high priority for LANL and the DOE. As a of the Energy and Climate Impacts (ECI) program de- consequence of this work, and the paper, we have been velopment project. That work is focused on issues of invited to present a paper at the DOE ASCR Examath high latitude climate interests. The first objective of workshop, a workshop that will determine the content this project brings capability in Arctic ocean science of next-generation DOE ASCR mathematics work. CCS2 is and fundamental fluids. It also directly impacts LANL’s working on converting Terry Haut to staff. IS&T mission by bringing new applied mathematics and computing capabilities in the development of fast/slow Beth Wingate was the principle organizer of a major in- time splitting for ocean models on heterogeneous archi- ternational conference — the 2013 CNLS Annual Confer- tectures and will contribute toward the development of ence — on Ocean Turbulence. This conference brought models that will one day reach the human scale. people in mathematics and ocean physics together to understand the major issues in ocean turbulence. Beth 375

Page 376

Wingate and Geoffrey Vallis, the 2013 CNLS Ulam Scholar, were invited to write a 6 page paper for Nature by the Nature Editor for Climate, Michael White. Further, there is going to be a special issues of the Journal of Physical Oceanography on Ocean Turbulence for people to contrib- ute papers to. And lastly, we believe we may have created a new community. The conference was so productive that I believe it will be reconvened in three years, perhaps in Santa Fe again. Future Work Study the relationship between the 3 known time- separated dynamical regimes in climate models • Task 1: Finish the peer review process for the first paper • Task 2: Complete a second paper on the mathematics of the 3 regimes • Task 3: Use an experimental configuration to under- stand the relationship between the three limits Continue working on the parallel-in-time method for exascale • Task 1: Finish the peer review process for the first paper • Task 2: Finish the second paper on linear propagators Conclusion We expect this work will lead to new understanding of the remarkable type of dynamics that happens in some areas of the world’s oceans where there is weak stratifi- cation. We also expect this understanding to help in the development of proof-of-principle concepts for solving the dynamical equations used in the ocean on next genera- tion computers, contributing to the ability to bring climate modeling to the human scale. Publications Haut, T. S., and B. A. Wingate. An asymptotic parallel-in- time method for highly oscillatory PDEs. To appear in SIAM Journal of Scientific Computing. Whitehead, J. P., and B. A. Wingate. The influence of fast waves and fluctuations on the evolution of the ‘slow manifold’. Journal of Fluid Mechanics. 376

Page 377

Environmental and Biological Sciences Exploratory Research Continuing Project Time-gated Super Resolution Imaging James H. Werner 20120227ER Introduction methods to increase the speed of image acquisition and We are building instrumentation that significantly to reduce photo-damage to the material during excita- advances the state of the art in optical microscopy. This tion. We note that unlike confocal excitation/detection, new instrument design will have be approximately 50 only a single XY plane in the sample is illuminated at a times better than the state of the art in terms of resolu- time, with the spatial extent of the excitation light pro- tion and functional working depth. Our first goal is to viding Z-sectioning. extend the effective Z-range of newly developed super- resolution methods to provide nanometer resolution The simplest method for excitation in selective plane il- throughout 3D volumes over ten microns cubed, at lumination is to use a cylindrical lens to focus a standard depths up to 100 microns. This represents an increase Gaussian-shaped laser beam into a thin sheet of light. of better than an order of magnitude over the state of However, for tightly focused beams, the light rapidly the art. This substantial increase in imaging depth will be diffracts/diverges. To overcome these problems, people enabled by two advanced optical methods (two-photon have recently introduced Bessel beam excitation for se- excitation and time-gated imaging). Our second research lective plane illumination. Unlike a Gaussian laser focus goal is to use these advances in the extension in Z-depth (which diverges rapidly due to diffraction), Bessel beams to examine structure and defects of important nano- can be collimated at very narrow beam waists (0.3 mi- structured materials and polymer scaffolds over func- crons) for nearly 100 microns. While Bessel beams can tionally relevant length-scales. Our third research goal is have orders of magnitude larger Rayleigh ranges than to use these advances to image the cytoskeletal network Gaussian laser beams, a Bessel beam has side lobes/ throughout an entire mammalian cell. bands of intensity that can also excite the sample, lead- ing to photo-damage or out-of focus fluorescence. Two photon excitation can be used to minimize the impact Benefit to National Security Missions of these side-bands, but high-power two photon lasers Specifically, we will use this new instrumentation to im- are substantially more expensive than one-photon laser age polymer materials and biological cells. The applica- sources (e.g. a Ti:Saph laser used for two photon excita- tion of super-resolution methods to biological materials tion is ~200K, whereas a diode laser used for 1 photon can lead to the preliminary data needed for competitive excitation is ~0.1 K). Spatially filtering the Bessel beam proposal submissions to DARPA, DTRA, or the NIH. The emission was demonstrated by one group, but the set- soft-materials science aspects of this project may be sup- up required 3 synchronized galvo mirrors. ported by DOE-BER or DOE-BES. We expect the results on the soft materials may affect laboratory programs in In the past year, our lab has recently developed a micro- energy storage or light harvesting. scope design that can perform 1-photon Bessel beam scanning that only uses a single galvo mirror (hence Progress lower cost, lower maintenance, and no synchronization In the past year we have made substantial progress in issues.) In our design, the same galvo mirror (GM) that both instrument hardware (design) and in writing con- is used to scan the Bessel beam de-scans the fluores- trol software for the new microscope. In particular, while cence emission through a slit that acts as a spatial filter. we were making progress on point scanning by confocal A retro-reflection from a spherical conjugate mirror (CM) excitation and detection, we decided this year to begin passes the spatially filtered light through the slit, with testing line-scanning (or selective plane illumination) this spatially filtered light scanned across an imaging 377

Page 378

CCD or CMOS camera by the same galvo mirror used to scan the excitation. While we are still optimizing param- eters (including scan areas, optimal excitation powers), we have demonstrated proof of principle measurements on dye-filled beads adhered to a coverslip and we have characterized the Bessel beam shape used for fluorescence excitation. Future work includes combining this excita- tion method with single-molecule based super-resolution microscopy for rapid 3D volumetric imaging of whole cells and cell colonies at near 10 nm resolution. We submitted a LANL invention disclosure (S133062/L2013059) based upon this design. Future Work In the next fiscal year, we will have accomplished the fol- lowing research goals: • Fully automate data acquisition in the Bessel beam excitation geometry • Demonstrate the advantages of Bessel beam excitation for test 3D systems (e.g. beads in polymer matrix) • Fully characterize diblock copolymer samples by super- resolution imaging and AFM (separately) • Acquire 3D images of diblock copolymer morphologies over functionally relevant length scales. • Begin imaging 3D cell structures (in particular the actin cytoskeleton or mRNA spatial distributions) Conclusion We are building instrumentation that significantly ad- vances the state of the art in optical microscopy. This new instrument design will have be approximately 50 times better than the state of the art in terms of resolution and functional working depth. These improvements in resolu- tion and working depth will be then used to study defects and structures in important polymer materials. These materials are useful scaffolds for energy harvesting. In addition, we will study changes in cell structure that result from stimulation by external agents. This work will greatly impact material and biological research. Publications Han, J., A. Shreve, and J. Werner. Super-Resolution Opti- cal Microscopy. 2012. In Characterization of Materials (Volume 2). Edited by Kaufmann, E.. , p. 1026. New York: Wiley. Han, J., Y. Kunde, E. Hong-Geller, and J. Werner. Single Mol- ecule Localization Microscopy of Actin Restructuring during Salmonella typhimurium Infection. Journal of Biomedical Optics. 378

Page 379

Environmental and Biological Sciences Exploratory Research Continuing Project Biofuel Enzymes by Design William M. Challacombe 20120256ER Introduction to be rate limiting. This is the difficult step to fin- The research will develop novel, rational design capabili- ish; however, with the other steps in hand, we can ties for multi-property optimization of biofuel enzymes, draw this conclusion without completion of the last overcoming multi-scale challenges through integration step. (b) Formation of the salt pair, LYS183-H+ and of emergent theoretical and experimental technologies ASP287 -> LYS183 and ASP287H+. (c) This salt pair with complementary strengths: First, O(N) quantum explains a conserved glycine residue, which mutant chemical (QC) methods will be used to overcome length studies have incorrectly attributed to other factors. scale barriers with a cost that increases only linearly observation of an unproductive, non-Michaelis fork with number of atoms N in the QC model. Second, joint in the reaction path, corroborating an early conjec- X-ray/neutron (XN) protein crystallography provides ture of Petsko et al. snapshots of stationary points along a com-plex reaction • Collaborator Kovalevsky has carried out kinetic mu- path, including the all important active site proton- and tant studies, publication pending. water-networks, guiding the reaction pathway between • New work on Carbonic Anhydrase (as discussed with these experimental constraints. Innovation includes:(1) W. Priedhorsky and Fischer) has begun, examin- a unique combination of emergent technologies and (2) ing the role of TYR7 in deprotinating the catalytic an unprecedented scale of thermo-chemical calculations hydroxide anion. (1000+ atoms). In addition to a more efficient XI, the proposed research will strongly impact the basic science backing DOE’s biofuels mission, establishing rational Future Work design as a rigorous testbed for developing our under- • Bonding and electrostatic analyses on the XI reac- standing of enzyme biochemistry. tion mechanism • Calculations on putative mutants, develop correla- Benefit to National Security Missions tions between computed and observed reaction This project will enable program development broadly, rates from biofuels to health, materials and national security. • Calculations on Carbonic Anhydrase TYR7 pKa We will target DOE-BER for program development and stand ready for opportunities to bring center level fund- ing to LANL for biofuels research. Further opportunities Conclusion will be sought under the NSF CBET program Catalysis The scientific effort will attempt to increase the ef- and Biocatalysis and under DTRA and DHS for chemical ficiency of cellulosic biofuel production through the warfare agent degrading enzymes. engineering of more efficient enzymes, in particular Xylose Isomerase (XI). Our goals are to (1) advance Progress our understanding of the basic biological science that • Entire five-step reaction mechanism of XI com- underpins the biofuels enzyme XI through the design of pleted. Four of five transition state calculations enhanced mutants, (2) establish technical leadership in have been completed. It will be possible to publish rational biomolecular design, enabling LANL to attack a incrementally starting now, or soon after completion broad class of biofuel and energy security problems and, of fifth calculation. Novel results include: (a) The ring (3) explore the limits of this new capability through the opening rather than the isomerization step appears stringent feedback between theory and experiment. 379

Page 380

Publications Kovalevsky, A. Y., B. L. Hanson, S. A. Mason, T. Yoshida, S. Z. Fisher, M. Mustyakimov, V. T. Forsyth, M. P. Blake- ley, D. A. Keen, and P. Langan. Identification of the Elusive Hydronium Ion Exchanging Roles with a Proton in an Enzyme at Lower pH Values. 2011. Angewandte Chemie. 123 (33): 7662–7665. Waltman, M. J., Z. K. Yang, P. Langan, D. Graham, and A. Kovalevsky. Engineering Acidic Streptomyces rubigi- nosus D-Xylose Isomerase by Rational Enzyme Design. Protein Engineering Design and Selection. 380

Page 381

Environmental and Biological Sciences Exploratory Research Continuing Project Metabolic Engineering of an Algal Lipid Bioswitch Chris M. Yeager 20120400ER Introduction in algae. Currently, we continue in to design and study The ability to control or monitor gene expression in the individual components that will allow the successful algae, via an engineered “bioswitch”, would be a tre- implementation of a lipid-responsive bioswitch in algae. mendous advance in the state of the science and aid in Characterization of the FadR regulon in E. coli. During the development of engineered algal strains for large- year one we discovered that the FadR regulon in E. coli scale biofuel production systems. In this project, we will likely consists of more many more genes/operons than develop a bioswitch/sensor for the unicellular model mi- has been reported. Since that time, we have discovered croalga, Chlamydomonas reinhardtii. We propose to: 1) that the FadR consensus operon (ANTGGTCNGACCAG) is develop a FadR-based biosensor for fatty acids in algae very similar to the operator consensus sequence (ATTG- and 2) demonstrate a functional transcription factor- GTAANACCAT) for another transcriptional regulator, based bioswitch that can be used to control gene expres- PdhR. To our knowledge a single operator sequence that sion in algae. For goal #1, we will refine the FadR system binds more than one transcription factor has not been to recognize different fatty acyl CoA species (different reported; the paradigm is that transcriptional regulators chain lengths, saturation) and possibly free fatty acids. recognize distinct operator sequences. Binding of FadR Additionally, we will pursue the development of other to PdhR operator sequences had been confirmed by transcription factor-based biosensors to measure free gel shift assays. Reverse transcriptase qPCR assays have fatty acid and fatty acyl-ACP levels in algae. Bacterial and been developed for nine E. coli genes that we believe plant transcription factors that recognize these classes could be dually controlled by FadR and PdhR. Currently, of fatty acids have been identified. For goal #2, we will we are comparing expression of these 9 target genes in develop a FadR bioswitch to control the expression of wild type and fadR- and pdhR- E. coli. heterologous genes engineered into algae to reroute or fine-tune fatty acid metabolism. We will pursue the use FadR binding to FadO and the palmitoyl-CoA ligand of alternative transcription factors to act as a bioswitch, We have developed and optimized an experimental if FadR proves unsuccessful. assay applying fluorescence anisotropy to measure the binding constant of GST-FadR for its palmitoyl-CoA ligand and DNA operator sequence (FadO). Under the experi- Benefit to National Security Missions mental conditions we obtained a Kd value of 2 nM of The project builds capability in engineering cellular FadR at 1 nM labeled-FadO and 600 nM for the palmito- factories and specifically aims to reduce the cost of yl-CoA ligand at 10 nM of FadR at 1 nM labeled-FadO. biofuel production by simplifying oil extraction. Success- ful proof-of-principle demonstration in this project will Site Directed Mutagenesis of FadR position us to acquire external funding from the biofuels We have constructed the L102A, L165N, L102A/L165N, industry, from DOE EERE, or the DoD for further optimi- R213M, R213N and R213A mutants of FadR (target zation, expansion, and commercialization of our concept. mutations based on computational analysis performed in year 1) using site directed mutagenesis. The mutants Progress at the L102/L165 positions were constructed with the Following the FY12 LDRD review, we were encouraged to purpose of expanding the binding pocket of the lipid confirm the effects of the predicted FadR mutations as ligand to accommodate longer lipids. The A213 of FadR outlined in the proposal, and to develop an alternative is directly involved in the binding of the CoA moiety of strategy move their bioswitch into bioenergy application fatty acids; we have included mutations at this position 381

Page 382

in an attempt to decrease FadR’s affinity for the CoA moi- reinhardtii that expresses E. coli FadR (and select FadR ety of the ligand. The mutant protein containing the R213A variants identified in goal #1) and harbors a luciferase re- mutation was purified and exhibited drastically decreased porter under control of an algal promoter integrated with binding for palmitoyl-CoA ligand (Kd >5 µM; wild type 600 the FadR consensus operator sequence, 4) evaluate and nM). In contrast, the L102A mutant protein exhibited in- publish the utility of the FadR construct (or related vari- creased affinity for palmitoyl-CoA (Kd 390 nM). The mutant ants) to serve as a cellular sensor for fatty acyl-CoA levels proteins have proven difficult to purify. Expression of the in algae, 5) demonstrate the utility of the FadR construct as mutant L165A has proven challenging, and the R213M pro- a bioswitch to control heterologous gene expression in C. tein was not stable following purification. Ongoing efforts reinhardtii. are yielding positive results, and analysis of R213N and the L102A/L165N double mutant should be accomplished in Conclusion the coming weeks. The preliminary results from the char- Successful completion of the project will greatly advance acterization of the R213A mutant suggest that mutations our ability to manipulate and monitor gene expression at this position affect (decrease) the binding of the CoA- in algae. The development and refinement of the FadR lipids (which is what we intended) while the L102A mutant system as a biosensor to measure fatty acyl-CoA levels increased the binding of the CoA-C12-lipid, which may be in bacterial or algal cells will be a significant technologi- in line with our interest of improving the binding and of cal advancement. Apart from its specific contribution to CoA-lipids with longer chains (C18). the algal biofuel field, the proposed research will develop capabilities to create novel bioswitches to manipulate cel- Algae transformations lular processes in other biotechnology applications. First, a bioswitch protein (engineered FadR) needs to be constructed and inserted into an algal host strain. We have obtained the chloroplast PB155 vector from Dr. Richards Sayre’s laboratory for this task and have inserted the codon optimized E. coli FadR. Preliminary (two) attempts to transform the PB155/CO-FadR plasmid into Chlamydo- monas reinhardtii have failed. Second, a reporter gene under the control of the bioswitch protein (i.e. FadR) needs to be inserted into an algal host strain. To achieve this we have obtained a vector (pSE 3HB Kan D2 gfp D1; from Dr. Stephen Mayfield, UCSD) that utilizes the psbD promoter to control the expression of algal-optimized GFP. We plan to achieve dual control and bioswitch activity of this GFP reporter by inserting FadO operator sequence within the psbD promoter, without disrupting its constitutive promot- er function. Four variants of the psbD promoter contain- ing the FadO operator sequence were constructed. The preliminary attempt (one) to transform these 4 constructs into C. reinhardtii resulted in the incorporation of one of them into the algae’s chloroplast DNA. We are currently as- sessing GFP expression by this strain in the absence of the FadR bioswitch. Future Work The focus of our research for FY14 will be to: 1) complete testing of the binding affinity of the FadR variants that were computationally identified in year one of the project to the FadR operator and the response of the FadR variants different free fatty acids and fatty acyl-CoAs, 2) publish 2 manuscripts, one detailing the characterization of the FadR mutants discussed above, and a second describing the overlap between the FadR and PdhR regulons in E. coli, 3) successfully introduce a construct(s) into Chlamydomonas 382

Page 383

Environmental and Biological Sciences Exploratory Research Continuing Project A New Class of Antibody-like Molecules that Signal Upon Binding Geoffrey S. Waldo 20120449ER Introduction Progress Modern biology needs tools to specifically label target GFP (green fluorescent protein) 11-beta stranded proteins in living cells so that their movements and protein activities can be watched and better understood. Our GFP strands 1 to 10 (GFP 1-10) can be engineered to long-term objective is to make a new class of antibody- bind and fold with a peptide corresponding to strand 11. like proteins based on fluorescent proteins that become Our goal is to create new GFP 1-10 capable of recogniz- fluorescent when they bind to a specific peptide or ing strand 11 sequences chosen to match disordered protein surface. Unlike conventional antibodies, these regions in antigens (pathogen, human, etc) so the new reagents would be straightforward to make, and would GFP 1-10 can act as a signaling antibody, turning green in be expressed functionally in living cells. They would give the presence of the new target protein. The project has a signal only when they bound their intended target. a computational part (in silico evaluation of effects of The goals include: (1) making signaling antibodies that mutating strand 11 and the best mutations to place into recognize linear epitopes by structural complementation GFP 1-10) and an experimental part (screening libraries with a peptide sequence, restoring folding and creating of these mutations to find those that work in practice). fluorescence; (2) making signaling antibodies that bind Progress on Experimental Part: Our previous experi- to surfaces of folded proteins, restoring fluorescence by ments on diversified libraries of every single residues of shielding the chromophore of the fluorescent protein. GFP 11 showed that changing residues pointing inside the barrel reduce folding of the GFP, those outside are Benefit to National Security Missions more tolerant to mutations. A Goal was to evolve GFP 1-10 to recognize a disordered region of CFP-10 protein, The capability established by the successful comple- a secreted antigen from TB. Mutating M218E, L220Q and tion of this project will enable two major future mission E222Q to match CFP-10 mutations by themselves abol- areas, the first is the specific detection and monitor- ished complementation of GFP 1-10. Despite the enor- ing of protein targets in living cells, and the second the mous challenge, after several rounds of directed evolu- detection and analysis of protein trafficking and expres- tion and diversified libraries of GFP 1-10 we were able to sion during physiological processes including baselines select for GFP 1-10 recognizing a new GFP 11 sequence and responses to environmental factors or disease states containing 7 residues similar to CFP-10 disordered tail. such as cancer. This also supports the high-content high- Importantly, we were able to include the two most throughput cell assays needed for the DOE User Facility disturbing mutations: M218E and L220Q in the new currently in project development with B Division. The sequence of strand 11. Achieving this Goal scheduled for live cell assays and signaling properties of these reagents this FY13 was a key milestone. it indicates that even in support DOD and DHS interest in platform technologies this worst-case scenario, GFP 1-10 can be engineered to for tracking physiological responses to chemical and recognize very challenging GFP 11 proxy peptides. And it biological agents in real time. New antibody-like tech- paves the way to developing GFP 1-10 as a signaling an- nologies for cell biology are a priority in NIH, where it tibody for TB diagnostics and biological studies. We also is recognized that current research in live-cell biology is completed our work on optimizing the CalGreen (a GFP limited by a lack of effective antibody reagents. from another species of jellyfish) and obtained crystal structures for Rosetta in silico modeling. 383

Page 384

Our Computational goal is to predict which sequence to experimental evidence of disorder. Where possible, variants of GFP 1-10 have the most stability with alternate we will identify targets that have conventional antibod- sequences of S11 such as CFP-10 to reduce the number of ies as controls. clones to experimentally screen. To model and constrain • Engineer GFP 1-10 to recognize a subset (top 12) of the GFP chromophore to calculate reliable Rosetta ener- these disordered tags. gies, out Goal was to compare experimental data for the • Generate GFP 1-10 capable of signaling binding to effects of single point mutations of S11, and the stability commonly used epitope tags (to help distribute the of the GFP 1-10 WT and GFP 1-10 mutants vs. the CFP-10 technology to research community). mutants. We hoped that the better-performing experimen- tally validated GFP 1-10 mutants would correspond to the • Experimentally determine the stability of specific S11 better Rosetta energies. and GFP 1-10 mutants where they contact in the struc- ture, and compare to Rosetta stability calculations. Progress on Computational Part • Generate stable versions of CalGrn 1-10 for in vitro and GFP 1-10 and S11 sequences were modeled on crystal in vivo use as binders, that are more soluble that exist- structure 2B3P. We developed controlled physically-model ing versions, to provide an orthogonal tag for multi- constraints on the GFP chromophore during the simu- plexing with GFP 1-10 binders. lations, with small perturbation of the rigid body and sampling of side chain poses. We were able to calculate Conclusion the difference between the binding energy of the GFP 1-10 and S11 mutants for all 240 possible 20 amino acid vari- Developing effective treatment modalities for cellular dis- ants across the 12 amino acids of S11. Three methods of orders requires understanding where and when proteins running simulations were developed (Packing and minimi- are expressed and their movements in living cells. Current zation at interface; including soft repulsion to model bulky antibodies used to study cell processes need labels, are residues; and packing and full backbone relaxation). We hard to get into cells, and have high background. Our pro- see similar trends in all three methods of computational posed ‘signaling’ antibodies work in living cells and light up predictions and are generally consistent with the experi- when they bind to their targets. These signaling antibodies mental results. Importantly, we were able to accurately can take advantage of computer design to recognize new predict the experimental stability ranking of the various target molecules. With these tools researchers can speed CalGreen protein mutants. This means we have a basis set translational medicine by spying on the movements of bio- of new scaffolds to make signaling antibodies from. We did logical molecules in living cells under different conditions discover that E222 is quite sensitive to side chain position- and during drug candidate screens. ing, and our model based on static crystal structure 2B3P may not be flexible enough. We think determination of the Publications crystal structures for some of the GFP 1-10 + S11 mutants Cabantous, S., H. B. Nguyen, J. D. Pedelacq, F. Koraichi, A. can help us decide if our Rosetta simulation constraints Chaudhary, K. Ganguly, M. A. Lockard, G. Favre, T. C. Terwil- need to be adjusted. Future Work. We will adapt GFP 1-10 liger, and G. S. Waldo. A New Protein-Protein Interaction to the fully mutated S11 so that it recognizes the CFP-10 Sensor Based on Tripartite Split-GFP Association. 2013. disordered sequence, and to screen GFP 1-10 optima Scientific Reports. : 1. against the S11 single point libraries to test specificity and promiscuity of the GFP 1-10, and solve a crystal struc- ture of the final complex. We will work with our GFP and CalGreen mutants too. We will refine our Rosetta meth- ods using our experimentally-validated GFP and CalGreen structures and mutants. Future Work • Continue to work to establish a triage and ranking sys- tem for selecting disordered domains that are available for binding to GFP 1-10. Working with the Bradbury lab and members of B6 and the structural biology com- munity on a national scale, we will annotate our list of disordered proteins for biological relevance in addition 384

Page 385

Environmental and Biological Sciences Exploratory Research Continuing Project Energy Efficient, Inducible Harvesting of Microalgae for Biofuels Production Richard T. Sayre 20120535ER Introduction Progress A significant constraint for the large-scale commercial- Over the past year, we have successfully genetically en- ization of the emerging algal biofuels industry is the gineered A. protothecoides with the hyaluronic synthase financially and energetically costly process of harvesting with and without UDP-glucose dehydrogenase expres- biomass from ponds. There are two constraints that limit sion, and we have shown sucessful mRNA expression of the efficiency of microalgae harvesting. These include, hyaluronic synthase in Auxenochlorella protothecoides the small size (2-10 um) of microalgal cells, and their with quantitative PCR. This project has also resulted in very low densities in ponds ranging from 0.1-0.5% of the the creation of reliable molecular tools for the genetic total mass. As much as 30% of the cost of producing al- engineering of A. protothecoides. After selecting 5 gal biofuels is associated with the concentration of cells clones and development of a hyaluronic acid (HA) assay to a level that is workable for oil extraction (3, 4, 6). We to visualize HA on the cell surface using fluorescence propose to take advantage of the biology of the algae to microscopy, we find few cells with surface-exposed HA. facilitate harvesting by expressing surface-exposed, self- Based on recently published reports of HA expression aggregating biomolecules under the control of inducible in plants (Rakkhumkaew et al. Biotechnol Bioeng. 2013 gene switches. These molecules were chosen for their Apr;110(4):1174-9), we believe HA may be building up ability to induce auto-flocculation of algae into macro- within the algal cell and unable to exit the cell wall. We scale aggregates that can be cheaply harvested using will examine this hypothesis by DEEM visualization with low energy, high throughput, and commercially scalable our collaborators and compound detection over the technologies (acoustic focusing, straining, settling). coming year. This work falls under the Designated Unclassified Subject Area of Biological Science. This work falls under the Furthermore, we have successfully identified extended definition of applied research related to the function of cell surface proteins and selected a candidate for tag- biological systems and falls under the sub discipline of ging. By synchronizing cell cultures in photobioreactors, cellular, molecular and systems biology. we were able to narrow the selection of candidate pep- tides. Based on BLAST comparison to Genbank proteins, Benefit to National Security Missions A. protothecoides transcriptome and genomic data, amino acid content and hydropathy analysis, we identi- This project will address a critical national need and fied a proline-rich protein (PRP) with an extended rod- contribute directly to LANL’s energy security mission like structure with homology to mammalian and plant in alternative energy. Lipid producing algae have the cell wall proteins that was expressed in A. protothecoi- potential to displace petroleum for a large fraction of des. Due to the length of time for nanobody production, our transportation fuel needs. Economical methods for purification, and selection (>9 months), we utilized previ- harvesting essential for implementation of algae for ous reports of the cell wall content of A. protothecoi- biofuel production. Our objective is to develop a low en- des (high in chitin), to develop a novel cell-cell binding ergy, low cost means to aggregate algal biomass that will system in which we tag the C-terminus (hydrophilic por- also allow the culture media and water to be effectively tion) of the PRP cell surface protein with a chitin-binding recycled enhancing the sustainability of algal biofuel protein (CBP). In order to test the binding of CBP to cells, production systems. Controlled bioflocculation of algae we expressed and purified a GFP-CBP fusion protein in E. will complement and enhance the existing and emerging coli (utilizing a His-tag and a Ni column) and confirmed algal biomass harvesting technologies being developed the binding of this soluble protein to A. protothecoides. globally and at LANL. 385

Page 386

As a negative control, we used C. reinhardtii (no chitin in the cell wall). We then created the PRP-CBP fusion in the A. protothecoides expression vectors we designed last year for microalgal protein expression. This vector has been transformed into A. protothecoides, and we will character- ize mutants in the coming year. Also, we have been identifying inducible systems of expres- sion in order to “turn-on” gene expression in A. protothe- coides. We tested the A. protothecoides promoter region of cyc6 (a Ni-induced system in C. reinhardtii) for Ni-induc- tion, but this showed a low-level response to Ni, and the features responsible for Ni-induction in C. reinhardtii were not present in the promoter region. In the coming year, we will engineer the C. reinhardtii Ni-induced promoter in A. protothecoides and test expression, first using an antibiot- ic-resistance gene for proof-of-principle. As of June, 2013, the work from this project will have been presented at two conferences and recognized by an a Los Alamos Outstanding Poster Award honorable mention. Future Work We have completed construction of transformation vectors to introduce critical genes involved in hyaluronic acid (HA) synthesis into the single-celled algae, Chlamydomonas and Chlorella. We will be expressing these genes and charac- terizing their phenotypes over the next year. In addition, we will build multi-gene constructs that we predict would enhance HA synthesis. In addition, we have identified cell surface proteins on Chlorella and their appropriate genes. We will be generating nanobodies against some of these peptide fragments and cloning the corresponding nanobody-encoding gene as a fusion to genes encoding other cell-surface displayed proteins with the objective of inducing autoflocculation. These constructs will be tested over the next year. Conclusion We expect to develop: 1) innovative technologies to con- trol gene expression in Algae, 2) Expression of novel bio- molecules that will efficiently flocculate algal biomass with virtually no energy costs. These results will lead to and energy efficient, economical methods for harvesting algae. Publications Subramanian, S., A. N. Barry, S. Pieris, and R. T. Sayre. Com- parative energetics and kinetics of autotrophic lipid and starch metabolism in chlorophytic microalgae: implications for biomass and biofuel production. 2013. BIOTECHNOL- OGY for BIOFUELS. 6 (1): 150. 386

Page 387

Environmental and Biological Sciences Exploratory Research Continuing Project Label-Free Measurement of Single Cells by Impedance Cytometry in a Microfluidic Device Babetta L. Marrone 20130239ER Introduction medical diagnostics, biomedical research, pharmaceuti- Single cell analysis is needed for characterizing hetero- cal screening, toxicology, and environmental science. geneous populations, which is not only important for most applications in the biosciences, but is particularly Benefit to National Security Missions critical for emergent industries like algae biofuels and We will develop a new measurement modality and new global public health. Single cell analysis systems require set of signatures for investigating and identifying biologi- integration of technologies for counting, positioning, cal particles in real time for high-impact applications separating, sorting, characterizing, and identifying single in biofuel production, as well as medical diagnostics, cells in a high-throughput manner. Typically, these func- drug discovery and screening, toxicology testing, and tions are accomplished using flow cytometry and sorting biomedical research. The greatest impact of this simple, in a laboratory setting, which requires extensive sample label-free, nondestructive approach is that it can be preparation and fluorescent labeling for sensitive detec- implemented on small, low cost platforms, thus enabling tion. Although very effective and highly specific, these introduction of real-time sensing and measurement approaches do not lend themselves to modern needs technologies into new fields of bioenergy and health in bioenergy or global public health where cost, robust- care. The work has immediate applications to DOE-EERE ness, speed, and sample throughput are as important programs in renewable energy, replacing petroleum- as the measurements being performed. Microfluidics based fuels with fuel made from algae feedstocks. The “lab on a chip” technologies hold great promise for the work will support automation of algae cultivation sys- future of single cell analysis in the field. However, as tems for biofuel production. Automation will save oper- instruments “shrink” it will also be essential to simplify ating costs, optimizing growth and harvesting schedules measurements. to reduce the overall cost of biofuels production. The capability to design and test lab-on-chip devices directly We will develop a new approach for single cell analysis addresses mission needs of the National Institutes of using impedance sensing on a microfluidic device. We Health (NIH) and the Department of Defense. Both agen- will: 1) Design and construct a microfluidic-scale, imped- cies are actively seeking to develop miniaturized sensor ance sensing device for detection and manipulation of devices for biomedical and biosecurity applications in large numbers of single biological particles based on support of National and Global Security mission chal- intrinsic properties, without the use of specific labels. 2) lenges. Specifically, the work directly addresses transla- Develop new signatures for distinguishing specific cells tional medicine and global public health missions of NIH in context, based on intrinsic properties obtained by by providing a new approach for disease diagnostics that impedance sensing; and investigate the cellular sources is rapid, simple, and affordable. of these responses. 3) Apply the new approach to solve real-world problems in biological detection using the Progress model system of direct quantitation of biomass and lipid This project has two main parts: Signature Development productivity during algae cultivation for biofuel produc- and Device Development. Our initial approach is to focus tion. The expected outcome of this project is a simpli- on extracting key intrinsic (unlabelled) features from fied measurement modality and new set of signatures cells that can be used as signatures. We will also build a for investigating and identifying biological particles in device that can detect these signatures. Our hypothesis real time. High-impact applications include monitoring is that measurement of cell impedance will enable us to biofuel production, and other potential applications in 387

Page 388

distinguish key cell features that differ between popula- eration device with a Surface Acoustic Wave (SAW) cell- tions of interest. focusing feature. The device was constructed from silicone rubber on a lithium niobate substrate. SAW producing Signature Development finger electrodes were electrodeposited onto the substrate A primary goal is to develop precise new classifiers of at CINT. Microchannels were molded into PDMS rub- complex cell populations based not on the properties of in- ber sheets, which were bonded onto the lithium niobate dividual cells but on the multi-dimensional distributions of substrate. Feed and product flow occurred through holes their collective statistics. Toward this end we have used the drilled through the PDMS rubber sheets prior to assembly. laboratory’s Amnis ImageStreamX imaging flow cytometer. Initial bulk impedance measurements of cellular solutions This instrument can detect twelve channels: two bright have been noisy, running into the detection limits of our field, side scatter, and nine fluorescence wavelengths. For Agilent impedance analyzer, a fact that confirms the need each channel an image of each particle in flow is obtained. to measure cellular impedance on a microfluidic platform. The high information content that we can obtain on popu- The microfluidic platform with SAW focusing, gives us lations of single cells makes it ideal for identifying intrinsic the ability to precisely align each cell in flow. This precise properties of cells that could be used as classifiers and alignment and the comparatively high cell to orifice ratio attributed to specific cell properties. allows for a high signal to noise ratio. Additionally, the mircofluidic platform allows for high throughput while still Initial experiments were focused on comparison of algae measuring on the single cell scale. First generation micro- populations with low and high lipid content. Using several fluidic SAW-impedance measurement devices have been different combinations of laser excitations, fluorescence designed, built and are in the initial testing stage. filters, and lipid staining protocols, we have collected a few hundred different features (related to cell fluorescence, After completing a literature search, the team reached a size, shape, etc.) for each of approximately 10k cells in consensus on the best methodology to investigate novel each of two replica populations of the high and low lipid impedance signatures of cells. The necessary equipment to populations. These data, as well as simulated mixtures of implement this technology was researched and ordered. the data, are being used to train preliminary classifiers, We purchased a digital microscope camera with analysis based upon non-parametric mixture models. We have software and a National Instruments control system. This first considered classifiers based upon the distribution of hardware is essential for development of the impedance each individual feature, and we are now extending to ex- cytometer system. plore the classification of populations based upon several features at once. Our preliminary goal is to develop an Future Work approach to quantify a priori the accuracy with which we • Build the next version of a microfluidic device combin- can identify the precise mixture of two different cellular ing SAW (for focusing) with impedance sensing of cell populations (e.g., high and low lipid content), particularly properties. Apply numerical modeling to guide and when the heterogeneity within each population has the optimize the design of the device. The next version same or greater order of magnitude as the average dif- prototype will be built during the 2Q of FY14. ference between the two populations. To test each of our • Continue studies to identify intrinsic features of cells existing and future classification strategies, we intend to that can be used as candidate signatures in biological run a series of double blind tests in which one member of applications, using established methods (e.g. imaging the team will prepare several unknown mixtures of cells flow cytometry). During FY14 we will focus on distin- and the algorithms will be asked to identify, and estimate guishing cancer cells from non-cancer cells; and cancer uncertainty bounds for, the mixture. We envision that the cells from background blood. computational classification tools being developed in this project (in Matlab and more recently in Python) will apply • Use a special chamber for making single cell measure- not only to the Amnis device but also to measurements ments of impedance responses off-line from the main taken through the acoustic focusing, impedance device microfluidic device to guide us in development of mod- under development as well as to other existing flow cytom- els needed to relate impedance responses to physical etry capabilities at the laboratory. cell properties. • Use the microfluidic device to study and validate the Device Development impedance sensing and develop signatures by attribut- Cell focusing is an essential feature of the impedance cy- ing the measurements to specific cell properties. Apply tometer device that we plan to build as part of this project. computational approaches to model and predict the Within the first quarter of FY13, we fabricated a first gen- 388

Page 389

effect of physical properties on measured impedance signals and verify the model by performing impedance measurements of cells with known physical properties. • Develop the sorting feature of the microfluidic device, in 4Q of FY14. Sort cells based on size and another fea- ture as measured by impedance analysis. Confirm the impedance measurements by comparison to estab- lished methods. • Continue to demonstrate the general applicability of impedance sensing in new applications. Conclusion We will build a microfluidic device to measure and sort cells, based on impedance signals; and will demonstrate quantitative, measurement performance of this device to analyze cell size, lipid content, viability, and other intrinsic features. We will also determine the physical cell features that result in the impedance differences between single cells. The overall result of the project will pave the way for simple, label-free, single cell analysis devices that can be used for a variety of applications, e.g. to automate real- time cell analysis for algae biofuel production; or to bring cell diagnostic capabilities into resource-limited areas of the world. Publications Ai, Y., C. K. Sanders, and B. L. Marrone. Separation of Esch- erichia coli bacteria from Peripheral Blood Mononucle- ar Cells (PBMC) using Standing Surface Acoustic Waves (SSAW). 2013. Analytical Chemistry. 85 (19): 9126. 389

Page 390

Environmental and Biological Sciences Exploratory Research Continuing Project Multidisciplinary Studies of Long Non-coding RNAs: Towards a Structural Basis for RNA in Epigenetics Karissa Y. Sanbonmatsu 20130319ER Introduction NIH’s epigenomics program for epigenetic mark discover, RNA (ribonucleic acid) is DNA’s molecular cousin. RNA is including the role of noncoding RNAs in regulation of similar to DNA in composition; however it often per- transcription (Jerome Garcia). The project is directly forms a function in the cell, unlike DNA, which is consid- related to DARPAs CLIO Memory thrust, including “basic ered solely an information carrier. In all living systems, research towards use of epigenetic systems to report including humans and bacteria, the information for the environmental events” (Cathy Cleland). It is indirectly re- blue print of the organism resides in DNA. The blue print lated to DOE BER Biological Systems Science “Low dose is implemented by transferring the genetic informa- radiobiology effects on epigenetic regulation.” tion from DNA to RNA. Next, the genetic information is converted from the RNA into proteins, which make up Progress the structures of the cells and perform the biochemistry. In this fiscal year, we accomplished many of our goals RNA has long been thought to mainly serve as a transfer- and made tremendous progress. We were able to ring medium, moving information from DNA to proteins. complete the first secondary structure of an intact long However, in the past five years, it has been found that non-coding RNA, resulting in several publications and the vast majority of the human genome (more than invited talks. We employed our novel shotgun second- 90%) is converted into RNA, but never gets converted ary structure (3S) method, which did garner significant into proteins. In the past 2 years, it has been found attention in the RNA community (especially at the recent that much of this RNA is in the form of long non-coding RNA 2013 conference last week). In this method we RNA (lncRNA, or ‘link’ RNA). These gigantic RNAs do not perform several rounds of structural chemical probing. ‘code’ for protein, but instead are thought have a func- In chemical probing, a special reagent is added that only tion themselves: turning on and off other genes. The reacts with highly mobile nucleotides, allowing us to RNAs are often associated with epigenetics. Epigenetics map which nucleotides are base paired in double helices has exploded in the past 5 years and is closely related and which nucleotides are single stranded. Thus, we to stem cell programming. In epigenetics, Larmarkian- perform 1 round of chemical probing on the entire long like effects occur, where the environment modifies, but non-coding RNA. We then perform successive rounds does not mutate DNA. These modifications are passed on fragments of the RNA. If we obtain a match between down to future generations of organisms. For example, signals for the full RNA and its fragments, this dem- babies undergoing extreme stress shortly after birth onstrates that the fragments have a modular fold and have an altered stress response, which is passed down comprise modular subdomains. By probing smaller and to the grandchildren. While many long RNAs have been smaller fragments, we are able to determine the com- found to be critical for cancer, hereditary disease, brain plete fold of the long non-coding RNAs, a hard problem, function and development, the mechanism of long RNA which has been long sought after. To verify the fold, we action is not understood. We will perform the first ever perform comparative species analysis, which enables us structural study of long RNAs to help determine their to identify conserved base pairs within the Watson-Crick mechanism. double helices. Prior to our results, it was not known if long non-coding RNAs were completely disordered or Benefit to National Security Missions if they were highly structured. Our study demonstrated We will produce preliminary data for specific NIH calls that they can be highly structured, generating intense in cancer-related non-coding RNAs and several calls for interest in the RNA structural biology community. 390

Page 391

Future Work Novikova, I. V., S. P. Hennelly, and K. Y. Sanbonmatsu. Sizing up long non-coding RNAs: Do lncRNAs have secondary In this fiscal year, we will apply our structure technique and tertiary structure?. 2012. Bioarchitecture. 2 (6): (shotgun secondary structure) to several more long non- 189. coding RNA systems. The first system is COOLAIR, a canoni- cal epigenetic switch in plants that allows plants to flower Novikova, I. V., S. P. Hennelly, and K. Y. Sanbonmatsu. Tack- only after prolonged exposure to cold. The second system ling structures of long noncoding RNAs. To appear in is Braveheart, a long non-coding RNA that plays a key role International Journal of Molecular Science. in heart cell development. The third system is Gas5, which Novikova, I. V., S. P. Hennelly, and K. Y. Sanbonmatsu. is critical for hormone signaling. To accomplish this, we will Experimentally determined secondary structures of transcribe the RNA for each system, denature the RNA, cancer-related long non-coding RNAs. Invited presenta- fold the RNA and chemically probe the RNA. In chemi- tion at 245th ACS National Meeting & Exposition. (New cal probing, a special reagent is added that only reacts Orleans, 7-11 April, 2013). with highly mobile nucleotides, allowing us to map which nucleotides are base paired in double helices and which Sanbonmatsu, K. Y.. Rise of the RNA Machines: Long Non- nucleotides are single stranded. In this way, we are able to Coding RNAs. Invited presentation at 10th Horizons determine the secondary structure of the long non-coding in Molecular Biology Conference International. (Max RNA. Planck Institute, Gottingen, Germany, 9-12 Sept. 2013). Sanbonmatsu, K. Y.. Integrating simulations and experi- Conclusion ments of the SAM-I riboswitch. Invited presentation at Our goal is to perform the first structural study of long RNA Dynamics. (Telluride, CO, 22-26, July 2013). non-coding RNAs (ribonucleic acids). The only example of a Sanbonmatsu, K. Y.. Large-scale simulations of biomolecu- large RNA complex whose 3-D structure has been solved is lar machines. Invited presentation at Biosupercomput- the ribosome, which took 3 decades of work. We will take ing. (Tokyo, Japan, Dec. 2012). the first step in structural studies of long non-coding RNAs by determining the 2-D structure. For large systems, 2-D Sanbonmatsu, K. Y.. Understanding the atomistic mecha- structures are impossible to accurately determine compu- nism of magnesium effects of RNA dynamics. Invited tationally due to the large number of permutations and presentation at Mini-symposium on Modeling, Simula- possible structures. We have devised a novel experimental tion and Function of Biomolecular Assemblies.. (Uni- technique called SHOT-GUN structure determination. We versity of Tokyo, Tokyo, Japan, Nov. 2012). will use this to obtain the 2-D structure experimentally. Publications Hennelly, S. P., I. V. Novikova, and K. Y. Sanbonmatsu. The expression platform and the aptamer: cooperativity between Mg2+ and ligand in the SAM-I riboswitch. 2013. NUCLEIC ACIDS RESEARCH. 41 (3): 1922. Novikova, I. V., A. Dharap, S. P. Hennelly, and K. Y. Sanbon- matsu. Shotgun secondary structure determination of long non-coding RNAs. 2013. Methods. 63 (170): 1. Novikova, I. V., S. P. Hennelly, C. S. Tung, and K. Y. San- bonmatsu. Rise of the RNA machines: Exploring the structure of long non-coding RNAs. 2013. Journal of Molecular Biology. 425 (19): 3731. Novikova, I. V., S. P. Hennelly, and K. Y. Sanbonmatsu. Do lncRNAs have secondary and tertiary structure?. 2012. RNA BIOLOGY. 9 (12): 1398. Novikova, I. V., S. P. Hennelly, and K. Y. Sanbonmatsu. Struc- tural architecture of the human long non-coding RNA, steroid receptor RNA activator. 2012. NUCLEIC ACIDS RESEARCH. 40 (11): 5034. 391

Page 392

Environmental and Biological Sciences Exploratory Research Continuing Project How Trees Die: Multi-scale Studies of Carbon Starvation and Hydraulic Failure During Drought Sanna A. Sevanto 20130442ER Introduction as well as communication. Mechanistic understanding of In this project we will develop Ultra Low Field (ULF) vegetation function during drought and mortality is criti- Nuclear Magnetic Resonance (NMR) and Magnetic Reso- cal fundamental knowledge for prediction of progress of nance Imaging (MRI) to simultaneously detect changes global climate change via the interactions of vegetation in both water and carbon use of trees and use this novel with climate. technique in targeted drought stress experiments for building a comprehensive multi-scale view of the role of Forests store ~30% of fossil fuel emissions, and forest carbon and water in plant mortality. NMR and MRI are mortality may release large amounts of carbon dioxide ubiquitous tools for non-invasive studies of soft-tissue to the atmosphere affecting predictions of atmospheric anatomy and function in medicine and biology and their carbon storage and effects of fossil fuel use. Forest mor- suitability to studying water dynamics inside plants has tality and increasing drought will also impact restoration been demonstrated. The traditional techniques, how- of natural environments. Currently, understanding the ever are difficult to use in studies of trees because they mechanisms of vegetation mortality relies on unproven require permanent magnets of high magnetic field that hypotheses because we lack non-invasive, in vivo meth- limit detectable sample size and shape restricting in ods to study plant water and carbon dynamics. The in- vivo applications. The major challenge of studies of tree teraction of water and carbon transport in plants is one physiology and mortality is a lack non-invasive methods of the fundamental unresolved topics of plant biology. to detect water and carbon dynamics. These substances Improving our understanding of carbon-water interac- travel through the plant along pressure gradients and tions inside plants is critical for understanding controls any invasive method will immediately change teh state of productivity (photosynthesis) and material alloca- of the system when applied. Therefore, development of tion as well as plant water use. In addition to trees, the non-invasive methods capable of in vivo measurements methods developed in this project will be applicable to of plant function is essential for advancing our under- other type of plants from crops to biofuel crops, and the standing of plant function and responses to climate. fundamental science results can be applied to improve renewable food and energy resources. ULF NMR/MRI uses simple pulsed electromagnets and optimized detection coils that operate in room tem- Progress perature using very low magnetic fields; three orders of Our goal for this year was to produce a robust Ultra Low magnitude lower than traditional NMR/MRI making this Field (ULF) Nuclear Magnetic Resonance (NMR) system technique adaptable to natural plant environment. The for measuring changes in water content inside living ULF NMR/MRI system is also very light (~10lb) and coil trees in greenhouse conditions, and evaluate whether size is adaptable to tree size enabling measurements at carbon could be detected with this system as well. To several locations on a tree simultaneously. achieve this we have performed temperature sensitiv- ity tests with water bottles and tree branches and made Benefit to National Security Missions substantial progress in understanding the temperature Understanding global climate change impacts and feed- systematics. The error we currently expect from tem- backs with terrestrial ecosystems is a critical mission for perature variation is ~1-2%. We have standardized the DOE-SC-BER. Climate change and weather will affect na- measurement setup and were able to operate without tional security impacting also trade and transportantion a shield when using a slightly higher fixed magnetic field 392

Page 393

instead of pre-polarization. This will enable multiple ULF goal is to use ULF NMR/MRI to determine the roles and NMR systems on one tree. We have run calibration mea- interactions of water and carbon cycling in tree drought surements to determine the relationship between NMR mortality. Currently plant physiology and mortality studies signal and water content of different tree species and lack non-invasive, in vivo methods. Our new technique will found out that NMR signal changes very linearly with wa- provide the key parameters and insights for mechanistic ter content in all species, but the overall signal strength de- modeling of plant mortality, which is critical for under- pends on the species. To get a better understanding of the standing climate impacts and feedbacks on vegetation. source of this variation we have conducted x-ray tomogra- phy and NMR measurements on the same branches and Publications are analyzing the data for linking wood structure with NMR Cruz, N., J. Yoder, and M. Espy. Active Q-switching in NMR signal strength. We have also performed in vivo water con- applications. Presented at 13th Annual Student Sym- tent measurements on an apple tree. The results from that posium, Los Alamos National Laboratory . (Los Alamos, experiment show that we can detect long term changes in NM, July 26-27). water content very reliably (drought and re-watering), but Sevanto, S.. Phloem transport and drought. 2013. In 3rd interpretation of the diurnal variations in the NMR signal International Conference on Plant Vascular Biology. require good understanding of the temperature effects. (Helsinki, Finland, 26-30 Aug, 2013). , p. 56. Helsinki: Helsinki University Press. We are currently repeating these measurements on an aspen tree with our improved coil temperature detection Sevanto, S.. Phloem transport and drought. Journal of Ex- system which allows correction of the signal for tempera- perimental Botany. ture effects, and preparing a methods publication of our Yoder, J.. Measuring absolute water content of trees in vivo data. As of June 2013, we have not tested potential carbon with low field NMR. Invited presentation at Physics injection methods for carbon detection, but theoretically Division Summer Seminar PostDoc Series. (Los Alamos, we should be able to use either uptake of labeled carbon NM, June 26, 2013). isotopes via photosynthetic pathway, spraying of labeled sugar water on the leaves or injection of sugar-solution Yoder, J., M. Espy, N. McDowell, J. Resnick, and S. Sevanto. into the conduits. The carbon signal should be high enough Measuring absolute water content of trees in vivo for detection with ULF NMR at least on the twigs close to by low field NMR in an uncontrolled environment. the location of injection. Presented at 54th Experimental Nuclear Magentic Resonance Conference. (Pacific Grove, CA, 14-19 April, 2013). Future Work The goal of the next fiscal year is to 1) perform first experi- ments on progress of hydraulic failure in trees using Ultra Low Field (ULF) Nuclear Magnetic Resonance (NMR) in a greenhouse environment, and 2) continue development of Magnetic Resonance Imaging (MRI) of water content changes and NMR detection of carbohydrates. In the experiments we will use manipulative methods to cause hydraulic failure in the conductive tissues and follow its progress with ULF-NMR. We will try to determine the link between hydraulic failure and carbohydrate availability by using surfactants (water loss -based hydraulic failure) and phloem girdling (carbohydrate loss -based hydraulic failure) as manipulations. The experiments will be supported by well-established plant physiological methods (gas ex- change, sap flow, tissue water content). Conclusion Our technical goal is to develop Ultra Low Field (ULF) Nu- clear Magnetic Resonance (NMR) and Magnetic Resonance Imaging (MRI) to accurately quantify water and carbon dynamics inside a tree in natural enviroments. Our science 393

Page 394

Environmental and Biological Sciences Exploratory Research Continuing Project Pyrocumulus Collapse: Unpredicted Wildfire Dangers Rodman R. Linn 20130487ER Introduction larger coupled wildfire simulations and enable wildfire The Las Conchas wildfire surprised everyone when it un- risk assessment across the DOE complex. This work expectedly burned over 35,000 acres in less than seven increases LANL’s unique capacity to study coupled fire/ hours during its first night even though it was burn- atmosphere phenomena in realistically complex environ- ing downhill in sparse vegetation and under mild wind ments and supports the USDA Forest Service (USDA FS) conditions. One theory for this peculiar behavior is that a wildfire research and management missions. This work particulate-laden portion of the fire-induced pyrocumu- will position LANL to serve a larger research role and to lus (pyro-cu) column collapsed, and unleashed a symbi- assume direct relevance to the fire management sector otic density current/fire front that incinerated everything of the USDA FS. Pyro-cumulus columns inject significant in its path. However the precipitating phenomenology of amounts of particulate high in the atmosphere and thus these events is not understood. have significant impacts on the atmosphere’s radiation balance. LANL’s coupled fire/atmosphere model, FIRETEC, will be used in three tandem efforts and will focus on funda- Understanding pyro-cumulus dynamics, conditions mental process and then merge these findings into com- leading to collapse, and the effects of resulting density bined analysis as the project progresses. These initial currents on fire behavior is crucial for elucidating the foci are: 1) assuring that the coupled fire/atmosphere fire-related effects from an urban nuclear event (on US code, HIGRAD/FIRETEC adequately captures the interac- soil or elsewhere). In this setting, density currents would tion between winds and plumes, 2) deciphering onset be channeled by street canyons, thereby focusing the of a pyro-cu collapse and density current formation, and kinetic and thermal energy and potentially increasing 3) understanding the interactions between a pyro-cu damage by an order of magnitude. LANL has ongo- density current and a fire front. ing activities in support of DHS, STRATCOM, and DTRA to understand damage and lives at risk from an urban The final goal is to reconstruct key aspects of the cou- conflagration that could likely result in a pyro-cumulus pled fire/atmosphere interaction that occurred in the collapse. first night of the Las Conchas fire and identify if it was possible that the combined dynamics of the fire and Progress ambient conditions could have lead to a partial pyro- A set of equations was derived to represent combustion cu collapse and whether such a collapse and resulting products that could be tracked in the LANL-developed, density current could have been responsible for the atmospheric dynamics model HIGRAD/FIRETEC. The observed fire behavior. Such an effort has never before fundamental concept behind this model is to adjust the been possible due to computational cost and complex- buoyancy term in the momentum equation to account ity of process interactions, but LANL’s R&D 100 award- for fire-produced particulate matter. When combustion winning HIGRAD/FIRETEC is being reformulated to take products are injected into a volume of air, they increase advantage of LANL’s new High Performance Computing its mass thereby reducing is buoyancy or even caus- (HPC) machines. ing it to become negatively buoyant. This is analogous, but opposite to the way that water vapor loading in the Benefit to National Security Missions atmosphere decreases the density of air and can causes This work will establish the capability to perform much a parcel to rise. Combustion-produced particulate mat- 394

Page 395

ter results in more dense air than its surroundings and this we hope to show that salient characteristics of the less products-laden air will have a tendency to descend. thoroughly studied B-JICF can be directly related to well This model was a fairly simple representation of soot as known features of the more comprehensively studied NB- Fullerene (C60). In a large, crowning wildfire, there is a JICF. A publication on the relationship between these two range in the size of debris that is released. The size of scenarios will result pending simulations of the B-JICF that these particulates can be small, such as molecular soot. are currently underway. But, it can range in size up to ash, pine needles and larger debris. The soot/water-vapor model consists of a new Future Work equation of state, including modified specific heats and The research will continue along parallel foci that will gas constant. It also includes a source term to the mass coalesce into a combined study in later years: 1) coupled conservation equation and requires additional conserva- fire/plume/atmosphere/topography interactions leading tion equations for soot and water vapor. to pyro-cu collapse; and 2) the interaction of the resulting density current with wildfire behavior. The theoretical soot/water vapor model described above has been incorporated into HIGRAD, in-order to test the In order to have confidence in the models ability to cap- plume collapse hypothesis. Indeed, preliminary results ture essential atmospheric processes, we will complete show that a soot-laden plume could collapse under its own verification of FIRETEC’s ability to capture critical plume weight under certain atmospheric and wildfire conditions. interactions with shear environment. A second ongoing effort to study this phenomenon in- volves a simple analytical model that describes the fluid To capture the coupled fire/atmosphere behavior at high dynamics resulting from the presence of both a vertical resolution (2–5 m), represent realistic length-scales (100s temperature and a vertical soot gradient. Together, these of meters to km), and avoid adverse impacts of numerical gradients result in a vertical density gradient that could be boundaries, one of the initial tasks will be to reformulate stable or unstable. From this model, a parameter space the Monte Carlo radiation and parallel input/output data can be extracted to quantify the effects of a large range exchange so that fully coupled fire/atmosphere simulations of soot and temperature profiles within the confines of a can be performed on the large domains required for this wildfire plume. This parameter space can then be used to study. theoretically predict the collapse potential for plume and atmospheric conditions. New models will continue to be implemented and evalu- ated in FIRETEC to include soot and water effects on buoy- Additionally, initial-phase work efforts have been aimed at ancy. This formulations are currently expected to include confirming HIGRAD/FIRETEC’s ability to capture critical as- both Lagragian and Eulerian approaches. pects of a cross flow interacting with a column of rising air. Thus far, we have focused on simulated results of the jet The effects of a range of soot concentrations will continue in a crossflow (JICF). The JICF has two sub-catagories, the to be evaluated in order to determine their implications non-bouyant JICF (NB-JICF) and the buoyant JICF (B-JICF). and importance. In the event that plume dynamics are In the NB-JICF, mass and vertical momentum are injected highly sensitive to soot concentration, this important find- from the bottom of a rectangular domain into a uniform ing would dictate a critical need for new instrumentation crossflow, whereas the B-JICF involves a heated jet issu- to accurately assess this hazard. ing from the surface. This work not only validates model accuracy, but has also yielded insightful results suggesting Initial wildfire acceleration and intensification as the lead- non-dimensional parameter regimes wherein plume col- ing edge of the density current reaches and passes through lapse may be more or less likely. the fire will be studied over flat, sloping, and canyon topographies. These simulations will highlight the impor- To date we have performed numerous simulations and tance of: 1) the pressure gradient across the leading edge compared model results against experimental data and of the density current; 2) the uplift-ing vertical velocities theoretical formulations for the NB-JICF jet-to-crossflow ahead of the current; and 3) the sustained accelerated velocity ratios (Vr) and crossflow Reynolds number (Re) winds within the current. regimes as obtained through non-dimensionalisation of the governing equations. Similarly, it can be shown that Conclusion the non-dimensional parameter Vr of the NB-JICF is analo- Fire behavior experts have suggested one viable hypoth- gous to a convective Froude number (Fr ) obtained for the esis for the fire behavior during the first night of the Las B-JICF governing equations. Armed with this new discovery 395

Page 396

Conchas fire: a partial collapse of the soot-laden pyrocu- mulus column (pyro-cu) that towered above the fire, caus- ing a sustained density current carrying fire at high rates of speed. The proposed research will test this hypothesis, and decipher the combination of environmental conditions and fire behavior dynamics that enabled such an event. The insight provided by this research will help fire managers to identify pyro-cu collapse risk scenarios and to implement possible life-saving emergency actions if one occurs. 396

Page 397

Environmental and Biological Sciences Exploratory Research Continuing Project Biocatalysts for Remediation of Uranium Wastes Francisca N. Rein 20130590ER Introduction Benefit to National Security Missions A strategic priority of the U.S. Department of Energy is This research supports the above missions by providing the cleanup of a multitude of sites across the Nation advancements toward the environmental decontamina- that contain subsurface contamination with radionu- tion and cleanup of nuclear waste. A key component of clides from legacy waste accumulated over five decades this research that underpins core missions in national of nuclear activities during the Cold War. Given the security and stewardship is the development of new daunting volume scales of soil and groundwater con- capabilities and knowledge about the physicochemical tamination with uranium and toxic heavy metals at low behavior of actinides/radionuclides. This project also ties concentrations, in-situ bioremediation by microorgan- directly into the LANL/LDRD Grand Challenge “Complex isms has been greatly regarded as a promising approach Biological Systems” (specifically “to resolve national to cost-effective, viable technologies for environmental challenges in energy, health and environment” as well as decontamination. This biological approach is based on “biological interactions and metabolic patterns that con- the established fact that some bacteria contain specific trol ecosystem functions”). The importance and urgent multi-heme, c-type cytochromes that can catalyze the needs for basic science to enable bioremediation tech- reduction of uranium from its soluble form in oxidation nologies have been recognized and stimulated by several state U(VI) to its insoluble U(IV) state as uraninite (UO2). DOE initiatives, especially through the Office of Biologi- However, this reaction takes weeks to reach equilibrium cal and Environmental Research. In addition to DOE, this and leads to incomplete reductions of uranium(VI) in research has high relevance to NNSA, DoD, and DHHS contaminated sediments. Moreover, the use of sacrificial (NIH) programs. With underlying themes related to (bio) electron donors in bioremediation is impractical and pro- sensing/detection of radionuclides or toxic substances, motes deactivation due to the competing re-oxidation by nuclear forensics, molecular bioelectronics, and bioen- the reversible activity of sacrificial agents. ergy, the outcomes of our proposed R&D may naturally lead to ties into programs within DHS, Intelligence Agen- A major bottleneck for further advances in the deriva- cies, DoD (DARPA, DTRA, ONR), and DOE (BES, ARPA-E) tion of effective technologies to address this grand among some of the identified agencies/programs. challenge is the missing knowledge about the reaction mechanisms and their specific metal-binding interac- Progress tions. Therefore, the overarching goal of our proposed In these first nine months since the project start, our research is the development of improved, bioengineered efforts have been primarily focused on establishing the enzymes as biocatalysts toward the efficient and selec- preparative methodologies and (bio)analytical tech- tive bioremediation of nuclear waste. niques for the isolation and characterization of our model biocatalyst for uranium electroreduction - the We will achieve this high-impact objective through an tetra-heme protein known as D. desulfuricans G20 cyto- interdisciplinary application of our combined capabilities chrome c3 (Cyt c3). From our initial approach, express- in biological and physicochemical sciences. Our research ing this protein into Escherichia coli (E. coli) proved to efforts were designed to address a critical environmen- be nontrivial. Multi-heme cytochromes are generally tal/health challenge of immediate national concern as more difficult to express correctly because co-factors well as to create new capabilities of relevance to current are needed to enable the proteins to obtain their native and emerging Laboratory missions in this direction. fold. In our case, Cyt c3 was no exception. Its four hemes 397

Page 398

must be covalently attached (via thioether linkages) to ture/function-property relationships as a function of the the polypeptide chain at the CXX(XX)CH binding sites, and multiple oxidation states of Cyt c3 but also to investigate the correct distal axial ligands must be connected to the the mechanistic details of the reactions involving uranium metal center (ferric ion) in the nascent protein. For these electroreduction by the biocatalyst. In preparation for the critical interactions to form properly during E. coli recombi- next phase of our planned work, we have also developed nant expression, specific molecular assembly proteins are and obtained approvals for all the required laboratory-re- needed. These helper proteins are collectively known as lated documentation (IWD, WPF, etc) to cover the radio- the Cyt c maturation proteins Ccm A-H (ccm cluster), which logical/chemical safety scope of our activities involving are responsible for the correct ligation of the hemes to the electrocatalytic testings with radionuclides (uranium-238) Cyt c3 apoprotein while it is translocated to the periplas- in our designated low-level rad laboratory. mic space. To help promote the expression of Cyt c3, we therefore co-expressed the pC3 plasmid (which carries the Future Work Cyt c3 gene) with the pEC86 plasmid (a pACYC derivative With the successful implementation of our preparative carrying the ccm gene cluster). methods and analytical techniques for the expression, iso- lation, purification, and structural characterization of our Despite the significant improvements from the use of model uranium-reductase in this initial phase (as described helper maturation proteins, the expression and growth of in the Progress section), we can next focus on 1) the struc- Cyt c3 required further optimizations due to our low yields ture/function-property relationships as a function of the obtained overall. To that end, we have incorporated sev- multiple redox states of Cyt c3 in aqueous solutions and eral changes to our strategy. Various systematic attempts immobilized as electrode biofilms, and 2) the elucidation have led us to find that the most successful approaches of reaction mechanisms for the multielectron uranium-re- involved auto-induction and temperature effects. After ini- duction catalysis by Cyt c3. Of critical application into these tially adopting methods that relied on using isopropylthio- tasks have been our capabilities developed for interfacial β-galactoside (IPTG) to induce the expression of Cyt c3 bioelectrochemistry, optical spectroscopies, and in-situ under control of the T7 promoter, we have now found that spectroelectrochemical methods. auto-induction processes are superior because of control over amounts of media and supplements (e.g. glucose, Our next step will also require the generation of Cyt c3 mu- lactose, phosphate, sulfate, and trace metals) to allow for tants. Site-directed mutagenesis in association with struc- bacteria growth to high densities prior to starting protein tural and biochemical characterization of the protein and expression. its mutants will lead to identification of structural features that are essential to U(VI)/U(IV) reduction. We plan to Another key finding was the reduction of the typical initially replace the axial His with Met to elucidate the role growth temperature from 37 °C to 30 °C. While reducing of bis-histidinyl axial ligation and cooperative redox behav- the temperature caused E. coli to grow slower, it presum- ior in the tetra-heme architecture. The cysteines from the ably allowed the protein more time to properly fold into CXXCH and CXXXXCH heme-binding motifs will also be re- its native form. Following expression of Cyt c3, standard placed with alanines to perturb the binding of the porphy- protocols have been successfully applied to its extraction, rin ring from the polypeptide chain. Because aromatic rings purification (ion-exchange chromatographic columns), and in the vicinity of porphyrin ring substituents may potential- analysis (gel electrophoresis, absorption spectroscopy, and ly play some role in electron transfer and redox regulation, Coomassie/heme staining). Upon incorporating the above we also plan to replace conserved aromatic residues with adaptations into the development of our methodology, alanines (e.g. Phe19 and Tyr64). Replacement of Lys14 and we have now successfully obtained Cyt c3 with yield levels Lys56 with alanine will also demonstrate whether these that are highly improved and viable to our project goals. residues are essential to binding and reduction of uranium as predicted. Parallel to our preparative emphasis of this initial phase of the project, our efforts have also involved implementing Combined with the synergistic tasks involving structure- the approaches to electrode surface functionalizations as function relationships and mechanistic redox characteriza- well as the analytical techniques needed for the bioelec- tion of Cyt c3 and its mutants, the ability to finely manipu- trochemical and spectroscopic characterizations of the late the protein environment at electrochemical interfaces redox protein, both in aqueous solutions and immobilized will allow us not only to understand but also to re-design as electrode films. Initial studies have already shown that the protein with improved binding efficiency and redox our in-situ spectroelectrochemical techniques will be a reactivity toward multielectron uranium-reductase activity. particularly powerful approach not only to probe struc- Conclusion 398

Page 399

By seeking the understanding and application of chemical principles underlying the multi-electron reductase activity of cytochromes c3, our research can lead to high-impact contributions to the permanent environmental bioreme- diation of radionuclides and toxic metals. Through inter- disciplinary application of biological and physicochemical capabilities, our work is aimed at advances of fundamental and practical importance for enabling the development of improved, bioengineered radionuclide-reducing systems to efficiently address the problem of incomplete reduction and recovery of uranium(VI) wastes. In addition to health/ environment sciences and radionuclide (bio)chemistry, this project has a high potential for impact in areas such as nuclear forensics, (bio)sensing/detection, and bioenergy. Publications Rein, F. N., D. M. Vu, and R. C. Rocha. Multi-Electron Re- ductive Bioelectrocatalysis by a Tetraheme Cytochrome c3. . 399

Page 400

Environmental and Biological Sciences Exploratory Research Continuing Project Structure Determination of Large and Membrane-bound Proteins by Nuclear Magnetic Resonance (NMR) Spectroscopy Ryszard Michalczyk 20130620ER Introduction Benefit to National Security Missions This project addresses our current lack of knowledge of Knowledge of the structures and mechanisms of both structure and function of membrane proteins and large large proteins and their complexes and membrane- protein assemblies by developing and applying novel bound proteins is critical to our understanding of and Nuclear Magnetic Resonance (NMR) and stable isotope intervention in processes as varied as antibiotic resis- labeling methods to determine their solution structures. tance and biosynthesis, biofuels production, and disease The limited spectral resolution of traditional multi-di- development (Parkinson’s, Alzheimer’s, diabetes, etc.). mensional NMR experiments, coupled to magnetization This project is directly related to LANL’s missions in Glob- losses due to fast relaxation in large proteins, presently al Security, Biothreat Reduction, and Energy Security and limits the size of biomolecules suitable for NMR spec- success of this project will open the structural biology troscopy to <50 kDa for high resolution structures. field to the detailed mechanistic studies of large protein complexes and membrane-bound proteins involved in We will address these critical shortcomings by develop- the above as well as many other processes placing LANL ing new solution NMR methods and isotope labeling at the forefront of the field. Leveraging the results ob- schemes to produce the largest de novo NMR solution tained here with LANL’s synthetic stable isotope capabili- structures to date of a large, flexible protein complex ties, will open funding areas from NIH, DTRA and DOE. (132 kDa) and a membrane-embedded system (74 kDa). Program managers in Global Security (Cathy Cleland) The innovative aspect of this research lies in the interac- and DOE (BES/BER) will be informed of the results of our tive use of specific, site-selective isotope labeling and research so that future funding for these efforts can be optimization of NMR methods for this specific labeling developed. scheme. We will improve existing and develop new NMR methods applicable to solution structure determination Results obtained from this project will allow intervention of large protein complexes and membrane proteins by into systems related to antibiotic resistance and toxin selective stable isotope labeling. We will design selec- synthesis, leading to further research into development tive isotope labeling patterns and synthesize selectively of new antimicrobials, inhibitors of efflux pumps and labeled amino acids and metabolic precursors for pro- mechanism of membrane protein action relevant to tein labeling in vitro and in vivo. These methods will be multiple disease states. This relates directly to Biologi- applied to obtain the NMR solution structures of a 132 cal Weapons and Defense and Pathogen Detection and kDa protein complex and a membrane-inserted perme- Countermeasures. As such, this research is of potential ase of 74 kDa. These will be the largest structures ever high relevance to DOE, DOD and DHHS. It will also lead determined by NMR and will place LANL at the forefront to discovery and innovation in Fundamental Chemistry of modern structural biology. (syntheses with stable isotopes) and Fundamental Bio- science. The results will provide new, unique, and exciting infor- mation about large, complex systems involved in syn- Progress thesis of bioactive compounds (antibiotics, anti-cancer During the first nine months of the project, we began drugs, virulence factors, toxins, etc.) and development of work on development of new NMR (Nuclear Magnetic antibiotic resistance, leading to their manipulation and Resonance) pulse sequences and implementation of subsequent production of new therapeutics and drugs. data handling protocols. We have created parameter The potential impact of this research is vast. 400

Page 401

sets we expect to be optimal for specifically labeled pro- to the sultam for further reactions. teins based on expected heteronuclear coupling constants in the NMR. Once specifically labeled amino acids are We have also explored alternatives to thiomethyl protec- synthesized and incorporated into target proteins, these tion of glycine. To this end, we have synthesized labeled parameters will be further evaluated and optimized. glycine protected with diphenylmethylene group. Reac- For data acquisition and handling we have imported and tion was high yielding and resulted in a product that is tested Non-Uniform Sampling (NUS) schemes based on stable, less volatile and, significantly, less pungent than maximum entropy and Iterative Soft Threshold (IST) recon- the dimethylsulfide protected glycine. H/D exchange on struction algorithms. We have initially tested both meth- this compound also proceeded well. This will be our new ods and found IST to be superior. The method will allow protection scheme for glycine. us to reduce number of spectra taken 4-fold for 3-dimen- sional NMR experiments and 10-20-fold for 4-dimensional Towards synthesis of amino acids with desired labeling pat- experiments. Final tests will be performed on selectively terns, we have reacted [2-13C, 15N, 2H]-glycinate sultam labeled large proteins once they are produced from la- with bromoacetate, which after deprotection yielded 2g beled amino acids. of [2-13C, 15N, 2H]-aspartic acid. Similarly, reaction of the glycinate sultam with 2H3-methyl iodide yielded [2-13C, We have also made significant progress towards selective 15N, 2H]-alanine. Thus we have fully synthesized three amino acid labeling for large proteins. We have devised of the required amino acids and prepared an intermedi- a pattern of selective isotopic substitutions for each of ate for syntheses of remaining compounds. We have also the twenty natural amino acids and identified required explored experimentally syntheses of phenylalanine and synthetic protocols based on literature data and previ- tyrosine side chains starting from pyranone and triethyl ous developments at LANL’s Stable Isotope Resource. orthoformate for isotopic labeling. These syntheses are For optimal NMR performance, all amino acids will have currently being optimized. Overall, chemical synthesis of amino nitrogens labeled with N-15 and α-carbons labeled amino acids is well on track. with C-13 and duterated. The carbonyl carbons will be unlabeled and the side chains will have C-13 labeled and For metabolic labeling of proteins, we have cloned the protonated carbons alternating with unlabeled deuterated gene for the small protein, ubiquitin (76 amino acids) carbons. A common precursors to these amino acids is [2- into E. coli hosts. Expression of this protein under various 13C, 15N, 2H]-glycine attached to a chiral auxiliary synthon labeling conditions will allow us to monitor incorporation (2,10-camphorsultam). To this end, we have synthesized of isotopes at specific positions in amino acids, since NMR [2-13C, 15N]-glycine from 2-13C-bromoacetic acid (which spectra for ubiquitin are well resolved and characterized. we had available in the lab) and 15N-potasium succin- So far, we have performed growth in the presence of la- imide. The reaction was performed on 0.5 mole scale and beled glycerol and labeled pyruvate. Analysis of produced proceeded with good yield. The purified glycine was subse- ubiquitin for isotope incorporation is in progress. quently protected by esterification with methanol on the carboxyl group and with thiomethyl groups on the amino Future Work functionality. During first year of the project, new NMR pulse sequences and data handling protocols will be developed and opti- Since all amino acids will be deuterated on C-alpha posi- mized; synthesis of labeled amino acids and precursors will tion, we attempted H/D exchange on glycine, protected begin with a target completion by the end of the first 18 glycine and glycinate sultam to establish the most efficient months. Tasks to be completed include: deuteration protocol. Deuteration attempt on glycinate sultam resulted in cleavage of glycine from the sultam • Modify and optimize recently developed 3D TROSY- and was abandoned. Deuteration of glycine proceeded hNCAnH and Time-Shared 3D HSQC-NOESY NMR very slowly at high temperatures (>160oC) in a microwave experiments by optimizing sequence pulses and delays reactor and could only be performed on small scale (<2g for selectively labeled amino acids. To dramatically at a time). However, deuteration of amino- and carboxyl- reduce acquisition time, we will incorporate non- protected glycine proceeded at 25-40oC with catalytic uniform sampling (NUS) and a newly developed data amounts of sodium carbonate in D2O/methanol-d and re- reconstruction algorithms that only require a reduced sulted in 98%+ duteration and pure product with minimal number of data points to be recorded without loss workup (extraction). This procedure was followed for [2- of resolution and sensitivity. Experimental time will 13C, 15N]-labeled protected glycine to obtain 0.1 mole of be reduced by an order of magnitude, with improved the protected deuterated glycine. This was then attached sensitivity. 401

Page 402

The Time-Shared 3D-HSQC-NOESY will be extended to a 4D pulse sequence with 13C phase shifts allow- ing for unambiguous identification of cross-peaks and determination of distance constraints. Here too we will use and optimize the NUS scheme in combination with coupled read-out and excitation TROSY for maximum sensitivity and resolution. This will reduce the required experiment time, increase spectral resolution, and will also simplify resonance assignments and allow accu- rate determination of distance constraints for high- resolution structure determination. • Devise new stable isotope labeling schemes for se- lected amino acids to simplify NMR studies of large and membrane inserted biomolecules. The isotopic labels will be incorporated at specific positions to optimize NMR signal and limit relaxation losses during experiments. Co-optimization of sample labeling and NMR methods development will improve the spectral resolution and signal quality allowing very large and membrane-inserted proteins to become accessible by NMR spectroscopy for the first time. The selectively la- beled amino acids will be prepared by direct chemical synthesis and/or by using labeled metabolic precursors (glycerol, pyruvate) for in vivo protein production to allow very specific and selective isotopic enrichment. Conclusion We will improve existing and develop new NMR methods generally applicable to solution structure determination of large protein complexes and membrane proteins by selective stable isotope labeling. We will design selective isotope labeling patterns and synthesize selectively labeled amino acids and metabolic precursors for protein labeling in vitro and in vivo. These methods will be applied to ob- tain the solution structures of a 132 kDa protein complex and a membrane-inserted permease of 74 kDa. These will be the largest structures ever determined by NMR and will place LANL at the forefront of modern structural biology. 402

Page 403

Environmental and Biological Sciences Exploratory Research Continuing Project Integrated Biosurveillance Benjamin H. Mcmahon 20130810ER Introduction HIV, malaria, Salmonella, Staphylococcus, and tubercu- Our pilot study will apply three state of the art diagnos- losis combine to cause 1) poor outcome (30% mortality tic technologies to a high disease burden population in rate, at age 5), 2) emergence of antibiotic resistant forms western Kenya. We will also perform the multiple types and 3) presence of novel pathogens. of epidemiological analyses necessary to derive under- standing from the novel data. The samples are all de- This project will lay the foundations to achieve our long- rived from co-morbidities (tuberculosis, Salmonella, and term technical goals of situational awareness for global Staphylococcus) associated with a 12 year longitudinal pathogen circulation and emergence. study performed in a pediatric malaria clinic in the Saiya District of Kenya by our collaborator, Prof. DJ Perkins Benefit to National Security Missions of the University of New Mexico. The three diagnostic The primary relevance of this work to basic health re- technologies are: search will be from improved diagnostic capabilities de- veloped, as well as the direct understanding of specific • Rapid Biomarker Detection: Approaches will be co-morbidities to patient outcome in the Siaya District of developed to detect proteins, lipids, and carbohy- Kenya. drates from human or animal samples. Reagents for tuberculosis, Staphylococcus, and Salmonella will be The primary relevance of this work to Fundamental applied and assessed. Bioscience will be from improved understanding of the evolution and emergence of pathogens from an immu- • Oligonucleotide-based diagnostics for pathogen no-compromised population the a high disease burden characterization: This will include 48x or 96x mul- area - the Siaya District of Kenya. tiplexed, PCR-based nucleic acid assays to species- strain- and antibiotic resistance- type for Salmonella The secondary relevance of this work to Information Sci- and Staphylococcus. ence and Technology lies in the integration of disparate • Sequencing of non-host RNA: Novel sample prepa- information types which will occur in the epidemiologi- rations will be developed and automated to provide cal modeling portion of this project. 10 million pathogen RNA sequences from each sample. These data will be analyzed to observe Progress unexpected pathogens, expression level of virulence This Integrated Biosurveillance project was a three factors, validity of nucleotide diagnostics from previ- month effort, designed to establish a collaboration and ous task, and strain-type relative database and other initiate a pilot study. A significant accomplishment of samples. this effort was a series of presentations by the task leads to an external review on September 11 and 12, 2013, Our epidemiological model will predict outcomes for during the Biosurveillance Deep Dive, which was well re- both individual patients and the population of Siaya ceived. The technical accomplishments of this work can (800,000 residents), as a function of investments, treat- be grouped into four LANL-funded portions and efforts ments, and mitigations. Our focus in the modeling will of our UNM collaborator to produce project success. All be to examine how the widespread co-morbidities of five areas performed well against proposed tasks. 403

Page 404

The sequencing task received 19 pure culture samples Future Work from DJ Perkins’ culture collection, of known antibiotic The overarching goal is to put us in a position to be in ‘pro- resistance profiles and phenotypically identified as Sal- duction mode’ for National Security-driven biosurveillance monella and Staphylococcus species. They are currently of a high disease burden population. (late September) being sequenced. For the RNA sequenc- ing task a round of oligonucleotides have been designed, Specifically, in this project, we will: ordered, and received, and our DJ Perkins has identified 1. SRS-sequence RNA preps derived from human serum a set of six whole-blood samples to be shipped to LANL to (a) detection & ID bacterial and viral species, (b) according to protocols jointly designed and approved by identify genetic markers for drugs and mapping to both the Human Subjects Review Board and the Bio-safety assemblies in (1), (c) identifying phylogenetic markers committee. Extensive interactions with the other tasks has and virulence factors of Staphylococcus and Salmo- occurred to design the workflow downstream of this se- nella. quencing effort, to identify known virulence and antibiotic 2. Develop methods to substantially deplete samples of resistance markers. host and ribosomal RNA, to increase detection and characterization of pathogen RNA. The PCR diagnostics task has performed a literature search and an assay design calculation based on the contents of 3. Apply and iteratively improve oligonucleotide designs Genbank to design assays for genus and species identifica- to species- strain- and AB-resistance type Salmonella, tion of Salmonella and Staphylococcus isolates, and known Staph, and TB in pediatric blood and serum samples. drug-resistance markers of these species. Reagents emerg- 4. Use diagnostics deployed in (3) to constrain epidemio- ing from this design were ordered and will be applied logical model (see below). to the DNA extraction preparations used for sequencing 5. Continue to validate and apply biomarker assays for described above, in the sequencing task. Salmonella and Staphylococcus and test on pure cul- ture and samples from (2). Assess sensitivity. The biomarkers task has performed a literature search and ordered antibodies to detect Salmonella and Staphylococ- 6. Benchmark Mass spec on samples from (1) and (2) cus directly from serum samples. This task also interacted 7. Demonstrate diagnosis of tuberculosis in pediatric with DJ Perkins to identify eight pediatric samples in the blood samples with biomarker assay. archive most-likely to test positive for tuberculosis, and they will be sent to LANL in the near future. Sources of 8. Develop statistical methods to constrain our baseline biomarkers for testing the antibody assays were identified epidemiology model with patient data from Cohorts and ordered. Preliminary evaluation of Staphylococcus 1&2 (~1,500 patients, 10% positive for bacteremia). antibodies against a commercial lysate has occurred. 9. Develop visualization and publication of epidemiology The Epidemiology task has created a baseline capability model of Saiya District, Kenya. in the R programming language for both an agent-based model and a deterministic, compartmentalized model Conclusion capable of modeling disease progression of co-morbidities, effectiveness of treatment both early and late in disease We will obtain draft genomes of twenty cultured samples progression, and the competition between strains nec- of Staphylococcus and Salmonella obtained from the essary for the emergence of drug-resistant strains. The laboratory of DJ Perkins at UNM. We will obtain data from land-use, roads, and rivers of the study area (Siaya District three types of diagnostics: sequencing of messenger RNA, in Kenya) was abstracted into the geographically explicit a panel of either 48 or 96 nucleotide diagnostics, and a model, and 1200 patients were mapped onto the study panel of biomarkers. For the information integration, we area with GoogleEarth. will extend the regression analysis to identify risk factors of clinical outcome for 1100 patients enrolled in Dr. Perkins’ Our UNM collaborator has provided extensive guidance clinical trial, produce a baseline epidemiological model to our effort in terms of formulating the scope, providing of co-morbidity in western Kenya, and develop the data samples, and initiating activities necessary for the second structures to integrate information from our novel diagnos- portion of this project. In particular, he has expended tic data. significant effort into collating Cohort 2 of patients (n=850) with the already-published data from Cohort 2 (N=750). 404

Page 405

Environmental and Biological Sciences Exploratory Research Final Report Grayscale Flow Cytometry - Multicoil NMR Sensors for Portable Flow Cytometry Pulak Nath 20110166ER Abstract method, namely NMR (Nuclear Magnetic Resonance) The long term research goal of this project was to relaxometry, to develop capabilities for flow cytometry. develop a portable flow cytometer based on magnetic We chose magnetic technology because of the recent detection. The immediate goal was to conduct the proof popularity of magnetic separation in cell/molecular biol- of principle study for a novel NMR sensor applicable to ogy. Since most biological materials are diamagnetic or flow cytometry. Several new capabilities were developed only weakly magnetic, labeling target cells with magnetic to address the immediate goal. Two unique permanent micro /nanoparticles makes them available to magnetic magnet designs were developed that simplify the fab- manipulation using simple permanent magnets. This rication of complex permanent magnet based circuits. method is very specific to the target cells or molecules. New and flexible techniques to develop microfluidic It is low cost, easy to operate and a large selection of platforms were developed that are low cost and com- magnetic micro/nanoparticles is commercially available. patible with mass-manufacturing methods. Preliminary However, there are limited developments of magnetic data were obtained to detect Bacillus spores using NMR sensors that are suitable for flow cytometry. Conse- relaxometry - suggesting a novel method to spore detec- quently, we will investigate NMR (Nuclear Magnetic tion without using conventional biochemical methods. Resonance) based detection of magnetically labeled A compact, multichannel, integrated circuit based NMR cells (with magnetic nanoparticles) in a hydrodynamic spectrometer console was also developed through col- configuration similar to flow cytometry. Furthermore, laboration with the Texas A&M University. Combination the NMR sensor proposed here will bear small footprints of these technologies has built the foundation to a new making it suitable for miniaturization and eventually, LANL capability: Compact, permanent magnet based a portable platform for flow cytometry. Portable flow NMR technology for portable applications. Using these cytometers will make several medical diagnostics (e.g. capabilities, we have demonstrated the proof of prin- Complete Blood Count tests, CD4 Testing for AIDS, Cir- ciple which can lead to a portable flow cytometer based culating Tumor Counts testing for cancer etc.) available on magnetic detection. The current project supports at bed-side and in resource poor settings such as rural several aspects of our national science initiatives includ- areas in the third world countries. Personalized diagnos- ing advanced manufacturing, bio threat reduction and tics, such as measuring patient specific responses from public health. antiviral therapy for HIV patients by counting CD4+ve cells or Chemotherapy for cancer patients by counting Background and Research Objectives circulating tumor cells, will be available to patients at Measuring cell specific information in high throughput bed-side reaching far beyond the capabilities of sophisti- platforms such as flow cytometers have found several cated healthcare facilities. The device can also be useful clinical applications ranging from complete blood count for in-the-field ‘label-free’ detection of anthrax spores to cancer diagnosis [1]. However, flow cytometers suffer (magnetic due to accumulation of Manganese from the from a fundamental limitation. They require significant, culture media) or malaria infected cells (magnetic due difficult-to-automate sample processing steps to ana- to the presence of paramagnetic crystals excreted by the lyze turbid samples like blood. They are also sensitive malaria pathogen). to cell autofluorescence – a common trait in cells that Although inspired by flow cytometry, this project does reduces sensitivity of measurements. These limitations not aim to reinvent or replace current flow cytometers. are fundamental to the optical methods utilized in flow Rather, we provide an innovative solution where the cytometry. Therefore, we will investigate and alternate 405

Page 406

conventional methods have struggled in spite of thirty shift measurement, the homogeneity required for magnet years of research and development. The long term re- is <1 ppm. search goal of this project is to develop a portable flow cy- In this work, our objective was to primarily design a mag- tometer based on magnetic detection. The immediate goal net that is suitable for NMR relaxometry with a homo- is to conduct the proof of principle study for a novel NMR geneity <100 ppm over a microliter of sample. There are sensor applicable to flow cytometry. This goal is divided different ways to obtain the magnetic field required for into the two following specific aims: NMR experiments. We chose permanent magnets due Specific aim 1: To design and fabricate a magnetic sensor to their ability to obtain moderate fields with a compact, based on NMR microcoils for flow based measurements. low cost, robust, and maintenance-free design. However, the static magnetic field from a single magnet depends Specific aim 2: To demonstrate flow based measurements on the shape of the magnet and offers minimal flexibility of cellular/molecular information of magnetically tagged for configuration. Therefore, magnetic circuits containing cells in turbid samples. multiple magnets and/or ferromagnetic components (e.g., flux guides, pole-pieces) are often utilized to configure Scientific Approach and Accomplishments a static magnetic field. However, the presence of strong The principal objective of this project was to demonstrate magnetic force between magnets and/or ferromagnetic the feasibility of measuring NMR relaxometry data from components leads to complicated fabrication methods cells in flow. To enable the objective, our plan was to place for a given magnetic circuit. The majority of permanent multiple RF pickup coils in series along the flow path to magnet circuits are, therefore, fabricated manually requir- pick up part of the signal at different time intervals and uti- ing bonding materials, specialized equipment, and experi- lize the signal data points from each coil to extrapolate the ence. We have developed two unique magnet designs that Free Induction Decay. In order to achieve the goal, we have are capable of providing homogeneous magnetic fields for divided our approach into two main tasks: (1) Platform NMR relaxometry. development; (2) Development and integration of NMR pulse sequences that can be applied in a multi-coil setting Shim-a-ring magnet: This design is termed as the ‘Shim-a- for flow based NMR relaxometry of cells. ring’ magnet. The design consists of a diametrically mag- netized ring-shaped, cylindrical permanent magnet (‘ring’) Platform Development placed inside a concentric ferromagnetic ring (‘shim’). There are three major components of a NMR spectrom- In addition to providing uniform field, the ‘Shim-a-ring’ eter: (1) the magnet – which provides a magnetic field in design has several unique characteristics suitable for NMR which nuclear magnetic resonance takes place; (2) RF coil and other applications. For example: – which is used to manipulate proton spins of the liquid sample and collect signals emitted due to spin relaxation; Ease of fabrication: ‘Shim-a-ring’ design is composed of and (3) a spectrometer – which is a hardware/software only two components. The ring magnet and the shim interface to control the RF pulses and data acquisition. De- self-assemble due to magnetic attraction between the velopment of a NMR based sensor for flow cytometry can two components when they are co-axially placed close to therefore be categorized in three different sections: each other. This method eliminates the need for rigorous alignment of multiple components or the use of bonding Magnet design and development: There are critical design materials to hold the different components in place. requirements for any magnet to be used for NMR applica- tions. A highly homogeneous field is required for successful Configurable magnetic fields: The field strength can be var- NMR experiments. Homogeneity of a magnetic field is de- ied by changing the thickness of the shim, while the nature fined parts per million (ppm). 1 ppm means a millionth of of the magnetic field can be varied (uniform and gradient the ratio (change in magnetic fields)/mean magnetic field field) by changing the shape of the bore hole of the ring over a given volume. Different types of measurements can magnet. By varying the shim diameter, the magnetic field be carried out using NMR phenomena. NMR relaxometry can be tuned to high precision requirements typically re- involves the measurement of spin-lattice relaxation (T1) quired in NMR applications. and spin-spin relaxation (T2), which are physical character- Higher uniform field: The uniform volume obtained from istics of a given solution. NMR spectroscopy, on the other a ‘Shim-a-ring’ design has been found to be ~30 times hand, can measure chemical shifts due to relaxation behav- greater than that of a comparable (with respect to magnet ior of a given sample. The uniformity of the magnetic fields volume and mean magnetic flux densities) Halbach design. required to measure NMR relaxation parameters is moder- It should be noted that, Halbach designs are one of the ate compared to measuring chemical shifts. For chemical 406

Page 407

most prominent permanent magnet design used for NMR In addition to using capillary tube based microfluidic applications. platforms, planar microfluidic platforms were also investi- gated. The most common fabrication techniques of pla- Ease of miniaturization: Since the ‘Shim-a-ring’ design is nar microfluidic platforms rely on photo-lithography and capable of self-aligning, it is also suitable for miniatur- PDMS (Polydimethylsiloxane) based replication techniques. ization. Miniaturization of other magnetic circuits (e.g. Although very popular at academic scale, PDMS based Halbach cylinders) can be complex since they require align- microfluidics techniques are usually not very suitable for ment of multiple magnets against their natural alignment mass production. To this end, we have chosen to investi- tendency. gate alternate methods to fabricate microfluidic channels. Our approach involved the use of off-the-shelf, finished or The uniformity obtained with a ‘Shim-a-ring’ magnet semi-finished materials (e.g. adhesive tapes, Kapton films, was in the order of 50 ppm over 1 microliter volume. The plastic sheets, etc.) and low cost readily available fabrica- maximum magnetic field was ~0.54 Tesla. The magnet was tion tools (e.g. laser cutters, lamination tools, press, etc.) a 50 mm in height and 80 mm diameter, and weight about to fabricate microfluidic platforms such that the fabrication 2.5 Kg. The design of a ‘Shim-a-ring’ magnet is presented method is easily transferrable to mass manufacturing. We in Figure 1(a). have developed two types of fabrication. The first method was based on the lamination of CO2 laser patterned ad- Automated Halbach Assembly: The maximum field obtain- hesive transfer tapes and plastic sheets. We have demon- able with the ‘Shim-a-ring’ design for any given configu- strated excellent biocompatibility of these platforms by ration was in the order of 0.55 Tesla. Although, this was demonstrating Polymerase Chain Reaction (PCR). To fur- sufficient for NMR relaxometry applications – a higher ther enable mass compatibility with mass manufacturing, magnetic field could provide higher signal to noise ra- we have developed a second technique that utilizes fab- tio. With Halbach designs, it is possible to obtain higher rications methods and materials common to the Flexible magnetic fields up to 2.0 Tesla. Therefore, we also investi- Circuit Industries. Due to recent progresses in the portable gated the fabrication of Halbach cylinders. Halbach cylin- electronics, a wide range of films have become available. ders are composed of multiple magnets (e.g. 8) and their Some these films are based on Polyimide and Teflon – two fabrication requires the alignment of these components in materials well known for their temperature resistance, specific orientations that are different form their natural chemical compatibility, and biocompatibility. Functional alignment tendency. Typically, special tools and fixtures microfluidic devices are often expected to be integrated are utilized to align and fix one magnet at a time to ob- with electrical components such as heaters, electrodes, tain Halbach cylinders. In this part of the work, we have coils etc. Since flexible circuits are designed to make multi- demonstrated a novel idea to align multiple magnets into ple layers of electrical circuits, our objective was to investi- a Halbach orientation. Our approach utilized the magnetic gate if it is possible to integrate microfluidic channels with field form a larger magnet to automatically align multiple the flexible circuit manufacturing methods. To this end, we magnets into a Halbach assembly. Use of multiple Halbach have worked directly with a flexible circuit company, Rigid- rings assembled in this way could provide a higher magnet- Flex (Santa Ana, CA), and developed the ability to integrate ic flux density inside the bore. Pictures showing automati- microfluidic channels with the flexible circuit technology. It cally aligned Halbach ring is presented in Figure 1(b). is possible to cut small channels (down to features with di- RF coils and integrated microfluidic probe development: mensions of 25 microns) using UV laser based laser cutters. Solenoid shaped RF coils were utilized in conjunction with However, a key challenge was to develop suitable bonding the Shim-a-ring magnet design to demonstrate NMR relax- methods to form enclosed microchannels. Typical lamina- ometry. A single channel commercial NMR spectrometer tion techniques used in the flexible circuits industry rely on (iSpin NMR console, Spincore technologies, Gainesville, FL) high pressure during bonding – such pressure caused the was utilized for data acquisition. RF microcoils were first laminating films to bend into the channels resulting into fabricated by winding thin gauge wire around 3 mm diam- deformation and sometimes, closing the entire channels eter commercial NMR tubes. To develop NMR microcoils, restricting flow. Therefore, we have developed an innova- thin gauge wires were wound around capillary tubing such tive solution to attain bonding that can avoid bending of that the sample volume was in the order of 1 microliter. the laminating films into the channels. We investigated the Initial ability to obtain NMR Relaxometry data was carried use of thin films that are coated with thin heat sensitive out on common liquid samples such as DI water and DI adhesives on both sides. One side faced the laser cut chan- water doped with CuSO4. Figure 2(a) shows a photograph nel and the other side was placed against a copper film – our experimental setup and NMR relaxometry data. such that during the bonding process the laminating films (that form the top and bottom layer of the microchannels) 407

Page 408

bonds with the copper. When the top/bottom laminating adopting droplet based microfluidics to our flow geom- film bonds with the copper – the mechanical strength of etry. Each droplet flows past coils D1 … Dn. At each time the copper film limits the laminating films form deform- samples are located below the coils, a NMR scan is trig- ing into the channel. After bonding, the copper layers can gered. The NMR scan is simultaneous for each coil. The be completely removed by a commercial etchant. In case, NMR scan consists of application of a radio-frequency it is desired to have electrical circuits integrated with the pulse with small flip-angle α (e.g. α=30-45˚), followed by channel – the copper layer can be selectively removed by data acquisition. The time between scans is tr. Therefore, common flexible circuit fabrication methods (e.g. pattern it will be possible to find T1 by curve fitting the intensities exposure and etching). Figure 2(b) shows the difference in from each coil. bonding outcome using conventional Flex Circuit fabrica- Demonstration of T1 measurement using Biological sam- tion method and the bonding methods developed at LANL. ples: In order to demonstrate the ability obtain T1 form Using these fabrication techniques we have developed biological samples with our magnet assembly, we have microfluidic droplet generators that can be used to syn- chosen two types of biological samples. First we chose chronously deliver samples into the NMR RF coil region for to measure sterilized sheep blood and serum samples. measurement in flow. The presence of red blood cells in the serum can cause a reduction in the T1 relaxation time due to the effect of the Multichannel NMR spectrometer development: Conven- iron laden protein hemoglobin in the red blood cell. We tional NMR spectrometers are large and not suitable for have prepared different concentrations of red blood cells portable applications. However, due to progress in micro- by adding different volumes of sterilized sheep blood in electronics, new generations of spectrometers have been serum and have measured the T1. The T1 increased with developed in recent years that are based on integrated cir- increasing dilution of the blood. cuits such as Field-programmable Gate Array (FPGA). FPGA based NMR spectrometers are sufficiently small to become As a novel application of this technology we also investi- portable. However, a multi-channel FPGA based spectrom- gated the effect of spore concentration of the T1 relaxation eter is not available commercially. In order to implement parameters. Bacillus spores are known to contain manga- multiple RF coils in series along the flow path to perform nese oxides in their spore-coats [2]. Manganese oxides are NMR relaxometry in flow – we developed we developed an known Magnetic Resonance Imaging (MRI) contrast agents eight-channel FPGA based NMR spectrometer that is small that can reduce the T1 relaxation time of a sample. There- enough to become portable. This part of the work was fore, we hypothesized that the T1 of a concentrated spore carried out in collaboration with Dr. Christian Hilty at the sample should increase with increasing dilution. We mea- Texas A&M University. The relaxometry console was a NMR sured T1 relaxation parameters for different concentration spectrometer for simultaneous acquisition of 8 signals of spores. Figure 3(a) shows a picture of the complete after transmitting a common excitation pulse sequence. system developed under this project. The T1 dependence The console is intended for relaxometry, but can be used of blood cell concentration and spore concentrations are for other, general NMR spectroscopy applications as well. presented in Figure 3(b). The console hardware is based on an Altera Cyclone IV FPGA with the DE0nano evaluation board, which interfaces Impact on National Missions to specific custom electronics. In addition to the hardware, This project supports Los Alamos National Laboratory’s a control software was developed for NMR data acquisition mission by developing science and engineering capabilities and processing. for the Science of signature pillar. Several unique capabili- ties have been developed to strengthen our core capabili- Development and integration of NMR pulse sequences ties in Bioscience, Chemical Science, and Sensors, Instru- for flow based NMR mentation Systems. We have developed novel permanent NMR pulse sequence for flow based NMR: In typical NMR magnet designs, new microfabrication methods for mass relaxometry the sample sits under the RF coil while the manufacturing of microfluidic platforms, demonstrated RF coil is used to perform a series of operations, which the feasibility of developing NMR relaxometry based flow involves the application of a series of polarization pulse cytometers, and have shown the novel ability to detect followed by acquisition signals due to the electromagnetic Bacillus spores using NMR relaxometry. These capabilities radiation emitted by the spin decay. This is not practical will support several applications beyond the scope of the for flow based NMR relaxometry. To enable detection in current project. For example the simple fabrication of com- flow, first we will have to discretize the sample into small plex permanent magnet circuits and microfluidic platforms droplets and introduce them into a series of RF coils in a can support the national initiatives on advance manufac- synchronous fashion. This capability was demonstrated by 408

Page 409

turing. The ability to detect spores with NMR can support Figure 2. (a)(i) Experimental set up showing early single coil NMR our missions in biosecurity, public health and simplifies relaxometry. (ii) T1 measurements of DI water averaged over quality control in the sterilization industry. This work also 4 scans. (b) Microscope photograph showing the difference in bonding outcome using (i) conventional Flex Circuit fabrication fits the themes of the National Institutes of Health (NIH) method and (ii) the bonding methods developed at LANL. roadmap. The DOD Congressionally Directed Medical Re- search Programs also seeks new technologies for medical diagnostics applications. Finally, this new capability can be the basis for seeking external funding from NIH or DOD to develop programs based on magnetic sensing for biological analysis. (i) (ii) y (-5, 5) (0,5) (-5,0) (0,0) (5,0) x (-5,-5) (0,-5) (iii) 1.5 1 0.5 0 -5.0 -2.5 0.0 2.5 5.0 Figure 1 (a): (i) Schematic diagram showing the different regions of a “Shim-a-ring” magnet design. (ii) Geometric parameters of the different cross-section shape of the air gap; (iii) Graphs showing magnetic flux densities along x-axis for different cross-section shapes of the air gap. Figure 1. (a) (i) Schematic diagram showing the different regions of a “Shim-a-ring” magnet design. (ii) Geometric parameters of the different cross-section shape of the air gap; (iii) Graphs showing magnetic flux densities along x-axis for different cross- section shapes of the air gap. (b) Photograph showing the different stages of the ‘two-step’ automatic alignment: (a) in the first step interleaved sections were aligned and fixed in one holder (b) in the second step remaining sections were aligned and fixed in a different holder (c) The complete prototype after the two holders were assembled. Figure 3. (a) Experimental set up showing the final device for flow based NMR Relaxometry (b) T1 relaxation data for different concentration of cells (i) Blood; (ii) Spore References

1. Ling, J., U. Wiederkehr, S. Cabiness, K. R. Shroyer, and J. P. Robinson. Application of flow cytometry for biomarker-based cervical cancer cells detection. 2008. DIAGNOSTIC CYTOPATHOLOGY. 36 (2): 76. 409 alseT ,ytisned xulf citengaM (i) (ii) y (-5, 5) (0,5) (-5,0) (0,0) (5,0) x (-5,-5) (0,-5) (iii) 1.5 1 0.5 0 -5.0 -2.5 0.0 2.5 5.0 Circle Figure 1 (a): (i) Schematic diagram showing the different regions of a “Shim-a-ring” mRaecgtnanegtl e design. (ii) Geometric parameters of the different cross-section shape of the air gap; (iiTir)i aGngrlaephs showing magnetic flux densities along x-axis for different cross-section shapes of the air gap. Distance, mm Figure 1 (b): Photograph showing the different stages of the ‘two-step’ automatic alignment: (a) in the first step interleaved sections were aligned and fixed in one holder (b) in the second step remaining sections were aligned and fixed in a different holder (c) The complete prototype after the two holders were assembled. alseT ,ytisned xulf citengaM Circle Rectangle Triangle Distance, mm Figure 1 (b): Photograph showing the different stages of the ‘two-step’ automatic alignment: (a) in the first step interleaved sections were aligned and fixed in one holder (b) in the second step remaining sections were aligned and fixed in a different holder (c) The complete prototype after the two holders were assembled. 6.0x106 Scan Data Fitted exponential 4.0x106 2.0x106 0.0 -2.0x106 -4.0x106 0 1000 2000 3000 4000 5000 6000 Figure 2 (a): (i) Experimental set up showing early single coil NMR relaxometry. (ii) T1 measurements of DI water averaged over 4 scans. )sitnu yrartibra( thgieh DIF 6.0x106 Scan Data Fitted exponential 4.0x106 2.0x106 0.0 -2.0x106 -4.0x106 0 1000 2000 E3qu0a0t0ion y4 0= 0A01exp(5-x0/t010) + y06000 Adj. R-Squ0.98955 are Value Error y0 6.9E+06 1.9E+05 Figure 2 (a): (i) Experimental set up showing early single coil NMR relaxomeA tt1 1ry. (ii-)1 .0TE 2 + 110 8 7 2 1.6E+ 1 0 1 5 6 measurements of DI water averaged over 4 scans. Pulse delay (ms) Top and bottom layer Top and bottom layer bending into channel NOT bending into channel (i) (ii) Figure 2 (b): Microscope photograph showing the difference in bonding outcome using (i) conventional Flex Circuit fabrication method and (ii) the bonding methods developed at LANL. )sitnu yrartibra( thgieh DIF Figure 3 (a): Experimental set up showing the final device for flow based NMR Relaxometry (i) Figure 3 (a): Experimental set up showing the final device for flow based NMR Relaxometry (i) (ii) (ii) Equation y = A1exp(-x/t1) + y0 Adj. R-Squ0.98955 are Value Error y0 6.9E+06 1.9E+05 A1 -1.0E+07 1.6E+05 t1 2182 116 Pulse delay (ms) Figure 3 (b): T1 relaxation data for different concentration of cells (i) Blood; (ii) Spore Top and bottom layer Top and bottom layer bending into channel NOT bending F int i o g ch u an r n e el 3 (b): T1 relaxation data for different concentration of cells (i) Blood; (ii) Spore (i) (ii) Figure 2 (b): Microscope photograph showing the difference in bonding outcome using (i) conventional Flex Circuit fabrication method and (ii) the bonding methods developed at LANL.

Page 410

2. MANN, S., N. H. C. SPARKS, G. H. E. SCOTT, and E. W. DEVRINDDEJONG. OXIDATION OF MANGANESE AND FORMATION OF MN3O4 (HAUSMANNITE) BY SPORE COATS OF A MARINE BACILLUS SP. 1988. APPLIED AND ENVIRONMENTAL MICROBIOLOGY. 54 (8): 2140. Publications Chandrana, C., D. Dunkerley, J. A. Neal, E. Eigenbrodt, B. Morgan, D. Platts, and P. Nath. Application of a simple, easy to fabricate, magnet for Nuclear Magnetic Reso- nance Relaxometry. Journal of Magnetism and Mag- netic Materials. Chandrana, C., J. Neal, B. Morgan, and P. Nath. Automated Fabrication of Halbach Rings for NMR Applications.
3. In 54th ENC (Experimental Nuclear magnetic resonance Conference). (Monterey, 14-19 April, 2013). , p. 80. Monterey, CA: ENC. Nath, P., C. Chandrana, J. A. Neal, B. Morgan, and D. Platts. Microfluidics integrated NMR coils on flexible circuits.
4. In 54th ENC (Experimental Nuclear Magnetic Resonance Conference). (Monterey, 14-19 April, 2013). , p. 1. Monetery, CA: ENC. Nath, P., C. K. Chandrana, D. Dunkerley, J. A. Neal, and D. Platts . The “Shim-a-ring” magnet: Configurable static magnetic fields using a ring magnet with a concentric ferromagnetic shim. 2013. Appl. Phys. Lett.. 102 (20): 202409 (4 pages). Nath, P., T. S. Maity , F. Pettersson, J. Resnick, Y. Kunde, N. Kraus, and N. Castano. Polymerase chain reaction com- patibility of adhesive transfer tape based microfluidic platforms. 2013. Microsystem Technologies. : 1. 410

Page 411

Environmental and Biological Sciences Exploratory Research Final Report Biofuel from Magnetic Algae Pulak Nath 20110168ER Abstract tain magnetic materials. There are bacteria and algae The long term goal of this project is to develop cost ef- in the nature that can uptake iron from their habitat fective and scalable technology to harvest biofuel from to form magnetite nanoparticles (known as magneto- algae. In spite of significant research and developments, somes) [1]. Several genes responsible for the formation sustainable biofuel from algae remains elusive. Accord- of these particles have been identified. Our objective is ing to the 2012 report on “Sustainable Development to express one of these genes ‘MagA’ in algae to form of Algal Biofuels in the United States” by the National biogenic nanoparticles that make the cells susceptible to Research Council (NRC), “Algal biofuels have the poten- magnetic separation. Following a recent precedent with tial to contribute to improving the sustainability of the human cells [2], we will introduce the MagA gene into transportation sector, but the potential is not yet real- algal cells enabling them to form intracellular magnetic ized. Additional innovations that require research and nanoparticles from soluble iron. The presence of these development are needed to realize the full potential of nanoparticles will make them susceptible to low cost algal biofuels”. This project addresses this concern by magnetic separation for harvesting. The performance providing innovative solutions such as ‘magnetic algae’. of this technology in a continuous flow system will be Several new capabilities have been developed under quantified to assess feasibility and scaling requirements this project. The ability to genetically engineer algae to for technology transfer to industry. produce algae that are susceptible to magnetic separa- Scientific Approach and Accomplishments tion without the need for any labels is unique to LANL. Different types of continuous flow magnetic separators The principal objective of this project was to demon- have been also developed which can find additional strate the feasibility of genetically engineering algae applications in bioscience and chemical science. There such that they become susceptible to permanent mag- may also be opportunities to harness the potential of net based separation. To enable the objective, our plan magnetic algae to make magnetic nano-particles, which was to express an iron transporter gene from magneto- can be a valuable co-product for the algal biofuel indus- tactic bacteria into a model algae system and investigate try. This project supports Los Alamos National Labora- the ability to obtain ‘magnetic’ algae. In order to achieve tory’s Energy security mission by developing science and this goal, we have divided our scientific approach into engineering capabilities for the biofuel program and the two main tasks: (1) Development of ‘Magnetic’ algae; (2) science pillar on the experimental science focused on Development of flow-through, scalable magnetic separa- materials for the future. tion techniques suitable for algae harvesting. Development of ‘Magnetic’ algae Background and Research Objectives Genetic Transformation: We have investigated two The long term goal of this project is to develop cost ef- genetic transformation methods to express MagA genes fective and scalable technology to harvest biofuel from in Chlamydomonas reinhardtii: (i) Agrobacterium based algae. High production costs limit algae based biofuel transformation; and (ii) Bolistic gene gun based transfor- from becoming a viable replacement for fossil fuels. mation. MagA was derived from Magnetospirillum mag- Current technologies for algae harvesting are either neticum (a magnetotactic bacterium). These bacteria are too expensive or only feasible at a laboratory scale. In known to form magnetic nano-particles inside the cells. this project, we will use magnetism to transform algal The gene sequence of MagA was optimized using codon harvesting and oil extraction. A key challenge in using equivalency to match the Chlamydomonas GC-rich magnetics is that the object of interest needs to con- 411

Page 412

genome. Protocols have been developed to insert MagA tory magnetic sorters. A special magnet with high gradient into two different vectors: (i) pSL18 and (ii) PSAD. These was developed to analyze whether the cells were becom- transformation constructs contained the MagA sequence ing magnetic. The first generation of the magnet was and a paramomycin resistance gene. The paramomycin analogous to commercially available laboratory magnetic resistance gene was inserted so that the transformed cells sorters. A glass cuvette containing the cells was placed on can be selected using their antibiotic resistance. Agrobac- the magnet such that the cuvette was positioned verti- terium based transformation protocols were developed cally. Due to gravity the algae are expected to settle to the using the vector pSL18, whereas, Gene gun based transfor- bottom of the cuvette. However, if magnetic algae were mation technique was developed using the vector PSAD. present they would be attracted to the magnet and their Figure 1 shows a schematic diagram of the steps involved deposition on the magnet will be visible through the naked for the PSAD based genetic transformation. eye (Figure 2(a)). Further detail on the magnet design will be discussed in a later section. Using this method, we were able to identify whether a transformation to obtain mag- netic algae was successful. (a) (b) Figure 2. (a) Photograph showing the magnetic deposition Figure 2: (a) Photograph showing the magnetic deposition platform for magnetic algae. The two platfdoarrkm gr efeonr l imne aingsindee ttihec cauvlgetatee in. dTichatees t twhe opr edsaenrcke ogfr meaegnne ltiicn aelg iane.s i(db)e T EM images the cuvette indicsahotweisn gt hmeag pnerteics neannocpear toicfl em foarmgend einti Mc aagnlgetaice a.l g a(eb.) TEM images showing magnetic nanoparticle formed in Magnetic algae. Figure 1. Schematic showing different steps involved in the genetic engineering of Chlamydomonas to obtain magnetic The magnetic properties of the transformed Chlamydomo- Figurea 1lg: aSech.ematic showing different steps involved in the genetic engineering of nas were tested using our custom high gradient magnetic Chlamydomonas to obtain magnetic algae. sorter within 24 hours of transformation and selection. The Gene gun based transformation with PSAD vector Magnetic responses of the cells were qualitatively exam- was more successful than the agrobacterium based ap- ined by observing the accumulation of the transformed proach. Following transformation, the cells were grown Algae near the high gradient locations of the magnet. Wild in antibiotic solutions. Only the successfully transformed type algae were used as positive controls for comparison. cells were able to grow in this medium. The colonies were PSAD is a known strong, constitutive promoter native to then isolated and grown in iron rich medium. These cells Chlamydomonas [3]. Therefore, the Gene gun/PSAD based were not susceptible to commercially available labora- 412

Page 413

transformed cells showed a stronger response to the phate concentrations. These aggregates were susceptible external magnetic field. The effect of iron (Fe+2) concen- to magnetic fields and therefore, were separable with a tration (10 micromols – 200 micromols) and the types of permanent magnet (Figure 3). These observations were iron salts (Ferrous sulfate vs ferrous citrate) in the culture probably not due to the formation of intracellular iron. media was also investigated. Approximately 100 micromol The aggregates and magnetic susceptibility developed of ferrous citrate was found to be optimum for the produc- immediately after the addition of the ferrous salt (Figure tion of magnetotactic algae. The presence of MagA gene in 3). Microscopic investigation showed that the aggregates the transformed population was confirmed by polymerase composed of the cells and other materials surrounding the chain reactions. Positive response to magnetic fields cells. It is well known that ferrous sulphates exhibit para- indicated MagA gene expression/activity (i.e. magnetic magnetic properties and ferrous salts have the tendency to nanoparticle production) in the transformed cells. flocculate cells. Based on this observation, a new tech- nique harvesting technique ‘magnetic flocculation’ may be Electron microscopy analysis: While it was evident from developed which can be applied to non GMO (genetically the magnet test the transformed Chlamydomonas were modified organisms) Tetraselmis. indeed becoming magnetic, to ascertain the reason behind their acquired magnetism, we have chosen to investi- gate the presence of intracellular iron using Transmission Electron Microscopy (TEM). Cells were harvested, freeze- dried, and fixed on copper grids for transmission electron microscopy. However, it was not possible to obtain clear images of inside the intact cells. Therefore, sample prepa- ration techniques were developed to enable TEM im- ages of cellular cross-section. The freeze dried cells were stained with TEM sensitive dye, embedded in wax and FigureF i3gu.r eP 3i:c Pticuturrees ssh sowhinogw maignngeti cm deapogsintioen tiofc w idlde typpeo Tseitrtiasoelmni so dfu ew to itlhde a tddyitpioen of ferrous sulphate–a possible new technique to separate algae –magnetic flocculation. sliced using an ultra-microtome. The sample preparation Tetraselmis due to the addition of ferrous sulphate – a possible allowed us to successfully image the cell cross-section and new technique to separate algae – magnetic flocculation. investigate the presence of electron dense regions inside Rapid/low cost analysis of algae growth: To determine the cells. Electron dense regions will indicate the presence the efficiency of magnetic harvesting and cell capture, we of metal nanoparticles inside the cells. Non-magnetic cells have identified a rapid method for determining cell con- were compared to magnetic cells. Images revealed elec- centrations and total cell numbers. The technique uses tron dense regions (Dark spots) only in the magnetic cells low cost, portable, off-the-shelf, automated cells counters. (Figure 2(b)). The dark sports were predominantly present The cell counters offers very simple procedure compared inside the vacuoles of the cells. None of these dense spots to the use of flow cytometers. They were able to provide were noted in the non-magnetic cells. The fact that the cell concentrations and cell size information. Cell size can cells are magnetic and electron dense regions are seen in change with each algal species dependent on physiological TEM images – we can assume that the presence of MagA state. We have determined growth characteristics, aver- causes increased influx of iron into the cells. The cells have age cell size, optical density, and cell numbers for the algal found a mechanism to store the extra iron into the vacu- species Tetraselmis striata, Nannochloropsis salina, and oles to isolate the additional iron from potential active Chlamydomonas reinhardtii. These techniques can be used metabolic processes. Dark spots were in the size orders to determine the efficiency of cell capture dependent on of 10 nm or less. The TEM images suggest that iron oxide magnetic harvesting system design and algae cell magnetic nanoparticles are being formed inside the magnetic algae, potentials. which make them susceptible to permanent magnets. Development of flow-through, scalable magnetic Effect of Ferrous rich media on other algae: We have also separation investigated the effect of high iron concentration of other Magnetic deposition assay: Since the anticipated magnetic types of algae, namely Tetraselmis and Nannochloropsis. properties were unknown for the ‘Magnetic’ Algae – we Interestingly, both the type of algae showed attraction have developed a permanent magnet circuit to obtain towards permanent magnets – even without any genetic very high magnetic gradient. Several designs were consid- modification. Tetraselmis showed greater response to ered. The optimized design was able to produce ~450T/m high concentration of iron than Nannochloropsis. Further gradient over a desired thickness of the channel (Figure 4). investigation with Tetraselmis revealed that the cells were This is an order of magnitude higher than common labo- forming aggregated in the presence of high ferrous sul- ratory based magnetic separators. The magnet assembly 413

Page 414

was built using block magnets such that the design can be channel was designed that can be mounted on the deposi- scaled to obtain larger units for high throughput separa- tion magnet. Algae solution is passed at a given flow rate tion. However, the first objective was to design the magnet over the high gradient zones of the magnet such that the such that it can be used to determine whether the algae cells can deposit due to magnetic attraction. When the are magnetic or not. To this end, the design consisted of zones are saturated with the magnetic cells, a bubble is in- two very high gradient locations where the algae could de- troduced in the flow such that the bubble volume occupies posit and form a band –visible to the naked eye. A cuvette the width and depth of the channel. As the bubble passes filled with algae is placed next to the magnet. If the cells the high gradient zone the cells are scraped off from the are magnetic, they accumulated in the high gradient loca- deposited area (Figure 5(a)) and are carried to the outlet. tion and thus, the magnetic deposition assay was devel- The concentrated cells can then be collected by opening a oped to distinguish between magnetic and non-magnetic valve at the outlet. algae. Concept Experiment Magnet Figure 5: (a) Collection of deposited algae in a semi-continuous process using gas-liquid interface; (b-i) Conventional high gradient magnetic deposition; (b-ii) Use of oil-water interface to continuously recover magnetic particles in high gradient separation. Figure 4. (i) Design of the high gradient deposition magnet; (ii) Magnetic flux densities in the one deposition zone. 2.2.2 Continuous Magnetic separation While the magnetic deposition assay was useful to iden- tify magnetic algae – it was not suitable for continuous collection of the separated algae. For algae harvesting application, it would be necessary to collect the algae in a continuous fashion. Therefore we have developed two new techniques for collecting weakly magnetic cells after separation in a high gradient magnetic separator. The first method is a semi-continuous process by which the de- posited cells can be removed periodically by applying air bubbles. The surface tension arising from the gas liquid interface was stronger than the attraction of the magnet to the weakly magnetic cells. In this method, a flow through 414 wolF Magnetic particles air‐water interface Concept Experiment Magnet wolF Particles Wall Magnet Water Grad B Magnetic particles wolF retaW liO Deposition zones – air‐water interface high magnetic gradient Ferro‐magnets Magnets (a) Concept Experiment (i) Inlet 1 Inlet 2 Particles Interface Grad B Outlet 1 Outlet 2 Magnetic particles Oil‐Water interface (b-i) (b-ii) Figure 5. (a) Collection of deposited algae in a semi-continuous process using gas-liquid interface; (b-i) Conventional high gradient magnetic deposition; (b-ii) Use of oil-water interface to continuously recover magnetic particles in high gradient (ii) separation. Figure 4: (i) Design of the high gradient deposition magnet; (ii) Magnetic flux densities in the one deposition zone. This method allowed us to collect concentrated algae by magnetic separation. However, this is a semi-continuous method. Therefore, we also investigated a continuous method of separation. Typically, continuous flow separa- tion of magnetic cells can be achieved by deflecting the magnetic cells in flow with a carefully designed permanent magnet (Figure 5(b-i)). However, in this case magnetic algae are weakly magnetic and require very high gradient from separation. Therefore, we devised a new method that could allow continuous collection of magnetically sepa- rated cells. The proof of concept was demonstrated using magnetic particles and a fluidic channel. The fluidic chan- nel consisted of two inlets and two outlets. High gradient magnet is placed on one side of the channel. If the channel was filled with flowing magnetic targets – the high gradi- ent would cause the targets to accumulate on the wall

Page 415

close to the magnet. However, we introduce a stream of cal Science, and Sensors, Instrumentation Systems. Finally, immiscible fluid such that, under laminar flow condition, this new capability can be the basis for seeking external the two streams (oil phase and aqueous phase containing funding from DOE or DOD to develop programs based on the magnetic targets) can proceed side by side (Figure 5(b- magnetic separation and biogenic materials production. ii)). Due to the presence of the oil phase the magnetically separated targets cannot reach the wall. The surface ten- References sion between the oil and aqueous phase limits the magnet- 1. Komelli, A.. Molecular mechanisms of magnetosome ic particles to cross the interface and reach the wall. As a formation. 2007. ANNUAL REVIEW OF BIOCHEMISTRY. result, the separated particles proceed towards the outlets 76: 351. as a concentrated solution. The proof of this concept was demonstrated using magnetic microparticles. 2. Zurkiya, O., A. W. S. Chan, and X. P. Hu. MagA is suffi- cient for producing magnetic nanoparticles in mamma- A perspective on cost: Additional iron is needed to grow lian cells, making it an MRI reporter. 2008. MAGNETIC magnetic algae. However, the cost of ferrous salt addition RESONANCE IN MEDICINE. 59 (6): 1225. will only be 0.0052 – 0.052$/gallon, if we add between 100 -1000 mol of ferrous salt. 3. Fischer, N., and J. D. Rochaix. The flanking regions of PsaD drive efficient gene expression in the nucleus of Impact on National Missions the green alga Chlamydomonas reinhardtii. 2001. MO- LECULAR GENETICS AND GENOMICS. 265 (5): 888. In spite of significant research and developments, sustain- able biofuel from algae remains elusive. According to the Publications 2012 report on “Sustainable Development of Algal Biofuels Nath, P.. ‘Magnetic’ Algae. 2013. In ARPAe Innovation Sum- in the United States” by the National Research Council mit. (Washington D.C., 25-27 february, 2013). , p. 126. (NRC), “Algal biofuels have the potential to contribute to Washington D.C.: ARPA-e. improving the sustainability of the transportation sector, but the potential is not yet realized. Additional innova- tions that require research and development are needed to realize the full potential of algal biofuels”. This project addresses this concern by providing innovative solutions such as ‘magnetic Algae’. We have demonstrated initial feasibility to address a major bottleneck in the Algal bio- fuel technology – microalgae harvesting. The Department of Energy plans to produce >1 billion gallons per year of cost-competitive algal biofuels by the year 2022. Significant challenge lies within the SCALE at which the algae would have be harvested. Permanent magnets for separation re- quire no energy input for separation. Commercial perma- nent magnet based separators capable of processing up to 1000 liters/min (~250 gallons/min) are already available to deliver dry ferrous contaminants from fluids and coolants. Although at a nascent stage, this project’s ability to pro- duce ‘magnetic’ algae has the potential to support these initiatives by DOE. Furthermore, the ‘magnetic’ algae pro- duce magnetic nanoparticles inside the cells – which may be utilized as a valuable co-product. The ability to produce magnetic nanoparticles in algae can also support other na- tional initiatives such as the ‘Living foundries’ program at DARPA (Defense Advanced Research Projects Agency). This project supports Los Alamos National Laboratory’s Energy security mission by developing science and engineering capabilities for the biofuel program and the science pillar on the experimental science focused on materials for the future. Several unique capabilities have been developed to strengthen our core capabilities in Bioscience, Chemi- 415

Page 416

Environmental and Biological Sciences Exploratory Research Final Report Understanding Thunderstorm Effects on the Ionosphere: A New Approach to Investigate Possible Convective and Electrical Coupling Mechanisms Xuan-Min Shao 20110184ER Abstract mechanisms of the troposphere storms. First, far-ultravi- The goal of this project is to investigate and characterize olet (FUV) observations onboard the IMAGE satellite dis- the possible mechanical and electrical coupling mecha- covered ionospheric variations that were closely coupled nisms from thunderstorms to the ionospheric D-layer, with tropospheric weather systems [1, 2]. Immel et al. and to quantify the magnitude of the electron density [1, 2] suggested the variations were driven by atmo- perturbations due to the various coupling mechanisms spheric tidal waves. On the other hand the proposers in space and time relative to the storm, by assimilating of this project found a close correspondence between archived lightning observations and model simulations. locations of peak FUV brightness with the global-scale Understanding how thunderstorms affect the D-layer is lightning activity [http://wwlln.net], suggesting that important in forecasting radio frequency (RF) communi- thunderstorms and their lightning could be the origin of cation and electromagnetic pulse (EMP) detection within the variations. Second, VLF lightning observations made the earth-ionosphere waveguide for ground-based sys- by a Duke University team have shown that intense tems, and in forecasting transionospheric communica- lightning discharges enhanced the D-layer electron tion and detection for satellite systems. Various thun- density by 108m-3 [6, 7, 8]. This VLF study was based on derstorm-ionosphere coupling mechanisms have been a single station measurement and used VLF waveguide proposed but very limited evidence has been obtained propagation model for the data analysis, and therefore to quantify each of the theories. The unique approach of was only capable of providing a 1D, source-to-receiver this project is to use lightning-produced EMPs captured path-averaged variation. It is very likely that the actual from an array of receivers (LASA) to map the spatial and electron enhancement was substantially underestimated temporal perturbations of the D-layer electron density, by this technique. The Duke team [6, 7] concluded that and to compare the perturbations with existing and the observed enhancement was due to transient light- new electrical and convective models to understand the ning EMP. Third, Inan et al., [9] measured narrowband physical mechanisms. VLF disturbances while storms were on the path from a CW transmitter to their receiver, found a fast VLF change Background and Research Objectives followed by a long, 20-200s decay, and argued that the D-layer disturbances over lightning storms have been parent storms would continuously “heat” the iono- sporadically reported and various thunder-storm-iono- sphere, as suggested by Pasko et al. [10]. However, the sphere coupling mechanisms have been proposed. But heating was not observed by Duke’s technique. Finally, the nature of the disturbances in either time or space Johnson and Davis [3, 4] monitored the sporadic E-layer has been poorly understood due to the lack of proper with a localized ionosonde, and reported that AGWs observational methods. It is not uncommon that conflict- and transient lightning EMP may both contribute to the ing claims are reported in the scientific articles. By using variations. They derived their result from 5142 lightning the new technique further developed in this project, we events across several years by cumulatively averaging have provided, for the first time, spatial and temporal the possible variations, and it is not clear how reliable information regarding ionospheric disturbances over and their claims are. around the thunderstorms, and characterized the effects A critical component necessary to distinguish the differ- of the various proposed mechanisms. ent coupling mechanism hypotheses is the horizontal Ionospheric variations recently have been hypothesized extent of the variation. The other equally important to relate to various convective and electrical coupling component is the temporal evolution of the variation, 416

Page 417

which will show how the variation propagates to other discovered. In papers we reported the new techniques regions of the ionosphere. Before this project, neither of along with fluctuation observations from a single sta- these components had been available for the bottom of tion measurement. It was clearly demonstrated that the the ionosphere. With the unique LASA dataset and our re- apparent reflectivity and height exhibit significant varia- cently developed full-wave VLF/LF propagation models [11, tion on spatial scales of tens of kilometers and over time 12], these questions will be answered in this project. Un- periods of hours. The range of the reflectivity variation was like the waveguide model that includes all possible bounc- as large as 100% away from the averaged reflectivity, and es between the ground and ionosphere along the entire the height varies by as much as 5% (4 km). The time scales path, this model only concerns the single, first-hop reflec- and propagation velocities (50-100 m/s) of the fluctuations tion off the ionosphere of the radiated lightning pulse, and appear to be consistent with signatures of atmospheric therefore only probes a small region of the D-layer at the gravity waves (AGW) at D-layer altitudes, and the direction reflection point with each measurement. With waveforms of the fluctuation propagation suggests that the gravity from the distributed LASA sensors, a spatially-resolved waves were originated from the storm. In some localized (rather than a path-averaged) map can be generated. ionosphere regions, we discovered some apparent splitting of the D-layer by 2 – 4 km that lasted a short time period Scientific Approach and Accomplishments of about 10 minutes. The research started with an analysis of the archived LASA Following the initial analysis, we extended the analysis to data with an array of 6 sensors over the Great Plains. A multiple-station measurement and to different storms. We LANL developed full-wave VLF/LF propagation model [11, confirmed that the AGW-like fluctuations were indeed orig- 12], which simulates the time-domain ground and iono- inated from the storm, by looking the D-layer from three spheric waves, has been further developed and used to different directions. For the analysis, fluctuation maps in compare with the lightning waveforms to determine the D- space and time were produced, and the movement of the layer height and electron density profile for each measure- fluctuations was examined based on the space-time maps. ment. These parameters at the reflection region directly As proposed, we also examined the D-layer with two radio determine the received waveforms and can be conversely frequency bands (<30 kHz and 30-60 kHz) by digitally filter- obtained by fitting the model with the measurement. A ing the original waveforms and found that the high-band typical storm over the Great Plains lasts many hours and signal penetrated deeper into the ionosphere by 2-4 km, produces many thousands of lightning flashes over a range and produced more detailed features of the fluctuations. of hundreds of km. Together with the array of sensors, a time-evolving 2D map can be obtained. In order to understand the D-layer properties behind the measurements, we improved the full wave, ground- The time-evolving 2D map will be examined against the dif- ionosphere VLF/LF propagation model by increasing the ferent coupling mechanisms to understand their respective frequency and incident angular resolutions in the model effects, based on their predicted different space/time na- computation. For the specific purpose of this project, a tures. EMP coupling mechanism can be readily compared new technique was developed to coherently combing the with the measurement. The theory for QE heating is based ground and ionospheric-reflected waves at the sensor loca- on the same physics as for EMP and its predicated effects tions. The output of the model predication was compared have been reported in a series of research papers [9, 10]. with the measurement, and some outstanding agreements Unrelated to the ionosphere community, AGW models on the time waveforms were obtained. In the paper, we for neutral atmosphere that simulate con-vective energy/ showed that reliable lightning return stroke waveforms moment transportations from troposphere to mesosphere can be reproduced at different observing distances by the have advanced significantly in recent years [13, 14]. And model simulation. Applying the model results to actual ob- more importantly, OH airglow observations of AGW at the servations we found that additional and stationary ioniza- D-layer height that coincided in time and region with LASA tion at the lower-boundary of the D-layer was introduced data have been reported in recent months [15], and can be by the storm. This newly discovered phenomenon might directly used for this study. be due to static electrical heating from the storm. We are First year accomplishments in the process of further examining this new result. In the area of data analysis, we confirmed that our tech- The other new scientific result came out from this project nique provides a much higher time and spatial resolution in FY11 is a new understanding of the discharge physics of measurement of the D-layer fluctuation as compared to the lightning stroke to ground. In an attempt to obtain a previously reported frequency domain techniques. As “standard” return stroke time waveform from the observa- a result, a series of new D-layer phenomena have been tions for the VLF/LF propagation simulation, we found that 417

Page 418

a statistically significant feature in the waveform after ac- Third year accomplishments cumulated thousands of individual waveforms. Any existing We realized that thunderstorms not only affects the return strokes models could not explain this small but per- ionosphere’s lower boundary (D-layer) but also introduce sistent feature. Instead, the feature can only be explained fluctuations deep in the ionosphere’s F-layer (300-400 km by a dispersive discharge process along the lightning altitude). By examining the total electron content (TEC) channel. This new result has a fundamental impact on the observations with ground-based GPS receivers, we found understanding of the dynamics of lightning discharge, and that thunderstorms are responsible for AGW and infra- we are in the process of submitting the result for a publica- sound fluctuations in the TEC measurement. This result tion. has been published in J. Geophysics Res. During the first ten months of this project, one paper was Following the Nature-Geoscience paper, we applied the published in JGR, two papers were presented in the XIV advanced techniques to mesoscale storms, and find that International Conference on Atmospheric Electricity, two static electrical field atop storms actually enhances the papers to GRL and JGR are in the process of submission, electron density as compared the EMP effects that reduces and one presentation was given in the AGU fall meeting, as the electron density. This work has been submitted to J. listed below. Geophys. Res. For publication. Second year accomplishments We further find that a specific infrasound signal at frequen- We discovered that the electron density in the nighttime cies 3-5 mHz was sometimes related to thunderstorms. lower ionosphere is reduced in response to a small thun- We have developed a new technique by using the phase derstorm, and the extent of the reduction is closely related information in the fluctuations to geolocate the sources in both time and space to the lightning activity of the of the signals. This work in still in its initial stage, but at- underlying thunderstorm. This work has been published in tracted attentions from multiple government agencies for Nature-Geoscience. programmatic applications. We found that thunderstorms produced wave-like fluctua- An updated VLF/LF propagation code has been delivered to tions (atmospheric gravity wave, AGW) at the D-layer. From AFTAC. simultaneous multi-station observations, we determined the speed and direction of the AGW propagation and con- Impact on National Missions firmed that these fluctuations are indeed originated from The proposed work has a direct impact on LANL’s new the underlying storm. This work has been published in thrust area of Space Situation Aware-ness (Jeffrey Bloch, Geophysical Research Letter. [email protected]), which depends on understanding of space weather conditions for reliable space-ground com- A new lightning return stroke model has been developed munication and persistent and accurate global surveil- to explain both the general and detailed electric radiation lance. The success of this project will form the foundation waveforms. Return stroked has been studied for decades, for a new LANL program supported by DOD. Through this but our new model is the first to reveal the general be- project, our understanding on time-domain, long-distance havior of electric current along the lightning channel. This VLF/LF propagation will be significantly improved, and this work has been published in J. Geophysical Research. will be directly applied to a new Air Force mission that uses ground-based RF measurement for long-range nuclear Significant improvement on VLF/LF propagation model, explosion detection (Tomas Carey, [email protected]). In ad- that makes the accurate analysis possible, and leads to the dition, the result of this project will be used to understand discovery of the reduction of the electron density in the fundamental ionosphere anomalies, and will be used to nighttime lower ionosphere. The improvement including improve the performance of LANL’s space-based nuclear the examinations of geomagnetic effects on VLF/LF propa- sensors (W, V sensors). gation. This part of the work has been published in Radio Science. References Our VLF/LF propagation code has been delivered to AFTAC 1. Immel, T. J., S. B. Mende, M. E. Hagan, P. M. Kintner, for programmatic missions. and S. L. England. Evi-dence if tropospheric effects on the ionosphere. 2009. Eos Trans. AGU. 90 (9): 69. Found evidence of thunderstorm effect on ionospheric F- layer by comparing thunderstorm activity with GPS-based 2. Immel, T. J., E. Sagawa, S. L. England, S. B. Henderson, TEC measurement. This work is still in progress. M. E. Hagan, S. B. Mende, H. U. Frey, C. M. Swenson, and L. J. Paxton. Control of equatorial ionospheric 418

Page 419

morphology by atmospheric tides. 2006. GEOPHYSICAL 13. Vadas, S. L., and D. C. Fritts. Reconstruction of the grav- RESEARCH LETTERS. 33 (15): -. ity wave field from convective plumes via ray tracing. 2009. ANNALES GEOPHYSICAE. 27 (1): 147. 3. Johnson, C. G., and C. J. Davis. The location of lightning affecting the ionospheric sporadic-E layer as evidence 14. Yu, Y. H., M. P. Hickey, and Y. F. Liu. A numerical model for multiple enhancement mechanisms. 2006. GEO- characterizing internal gravity wave propagation into PHYSICAL RESEARCH LETTERS. 33 (7): -. the upper atmosphere. 2009. ADVANCES IN SPACE RESEARCH. 44 (7): 836. 4. Davis, C. J., and C. G. Johnson. Lightning-induced in- tensification of the ionospheric sporadic E layer. 2005. 15. Yue, J., S. L. Vadas, C. Y. She, T. Nakamura, S. C. Reising, NATURE. 435 (7043): 799. H. L. Liu, P. Stamus, D. A. Krueger, W. Lyons, and T. Li. Concentric gravity waves in the mesosphere generated 5. Shao, X. M., M. Stanley, A. Regan, J. Harlin, M. Pon- by deep convective plumes in the lower atmosphere gratz, and M. Stock. Total lightning observations with near Fort Collins, Colorado. 2009. JOURNAL OF GEO- the new and improved Los Alamos Sferic Array (LASA). PHYSICAL RESEARCH-ATMOSPHERES. 114: -. 2006. JOURNAL OF ATMOSPHERIC AND OCEANIC TECHNOLOGY. 23 (10): 1273. Publications Jacobson, A. R., X. M. Shao, and E. H. Lay. Full- -wave model 6. Cheng, Z., S. A. Cummer, H. T. Su, and R. R. Hsu. Broad- of D- region upward VLF coupling to whistlers in the band very low frequency measurement of D region plasmasphere. Presented at AMERICAN GEOPHYSICAL ionospheric perturbations caused by lightning elec- UNION FALL MEETING. (San Francisco, CA, 3-7 Dec. tromagnetic pulses. 2007. JOURNAL OF GEOPHYSICAL 2012). RESEARCH-SPACE PHYSICS. 112 (A6): -. Jacobson, A. R., X. Shao, R. H. Holzworth, and E. H. Lay. 7. Cheng, Z., and S. A. Cummer. Broadband VLF measure- Magnetic-azimuth dependence of D-layer radio re- ments of lightning-induced ionospheric perturbations. flectivity, using lightning sferics as radio transmitters. 2011. In AMERICAN GEOPHYSICAL UNION FALL MEET- 2004. GEOPHYSICAL RESEARCH LETTER . 32 (L08804): ING . (San Francisco, CA, 5-9 Dec. 2011). , p. AE21B. 1. Washington DC : AGU. 8. Cummer, S. A., U. S. Inan, and T. F. Bell. Ionospheric D Jacobson, A., X. Shao, and E. Lay. Time domain waveform region remote sensing using VLF radio atmospherics. and azimuth variation of ionospherically reflected VLF/ 1998. RADIO SCIENCES. 33: 1781. LF radio emissions from lightning. 2012. Radio Science. 47 (4): RS4001. 9. Inan, U. S., V. P. Pasko, and T. F. Bell. Sustained heating of the ionosphere above thunderstorms as evidenced Lay, E. H., X. M. Shao, and A. R. Jacobson. High temporal in the early/fast VLF events. 1996. GEOPHYSICAL RE- and spatial-resolution detection of atmospheric gravity SEARCH LETTER. 23: 1067. wave effects on D-layer electron density. 2012. In 2012 COUPLING, ENERGETICS, AND DYNAMICS OF ATMO- 10. Pasko, V. P., U. S. Inan, Y. N. Taranenko, and T. F. Bell. SPHERIC REGIONS (CEDAR) WORKSHOP. (Santa Fe, New Heating, ionization and upward discharges in the me- Mexico, 24-29 Jun. 2012). , p. MLTG. Boulder: UCAR. sosphere due to intense quasi-electrostatic thunder- Lay, E. H., X. M. Shao, and A. R. Jacobson. D-layer electron cloud fields. 1995. GEOPHYSICAL RESEARCH LETTER. profiles observed with substantial spatial and temporal 22: 365. change near thunderstorms. JOURNAL OF GEOPHYSI- CAL RESEARCH. 11. Jacobson, A. R., X. M. Shao, and R. Holzworth. Full- wave reflection of lightning long-wave radio pulses Lay, E. H., X. M. Shao, and A. R. Jacobson. Electrical modifi- from the ionospheric D region: Numerical model. cation of D-layer electron distribution by local thunder- 2009. JOURNAL OF GEOPHYSICAL RESEARCH-SPACE storm. Presented at AMERICAN GEOPHYSICAL UNION PHYSICS. 114: -. FALL MEETING. (San Francisco, CA, 3-7 Dec. 2012). 12. Shao, X. M., and A. R. Jacobson. Model Simulation of Lay, E. H., X. M. Shao, and C. S. Carrano . Variation in total Very Low-Frequency and Low-Frequency Lightning electron content above large thunderstorms. 2013. Signal Propagation Over Intermediate Ranges. 2009. GEOPHYSICAL RESEARCH LETTERS. 40: 1. IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPAT- Lay, E. H., and X. -M. Shao. Multi-station probing of IBILITY. 51 (3): 519. thunderstorm-generated D-layer fluctuations by using 419

Page 420

time-domain lightning waveforms. 2011. Geophysical PLING, ENERGETICS, AND DYNAMICS OF ATMOSPHER- Research Letters. 38 (23): L23806. IC REGIONS (CEDAR) WORKSHOP . (Santa Fe, New Mexico, 24-29 Jun. 2012). , p. SPRT. Boulder: UCAR. Lay, E. H., and X. M. Shao. High temporal and spatial- resolution detection of D-layer fluctuations be us- Shao, X. M., E. H. Lay, and A. R. Jacobson. Using lightning ing time-domain lightning waveforms. 2011. JOUR- waveforms to probe thunderstorm effects on the iono- NAL OF GEOPHYSICAL RESEARCH. 116 (A01317): spheric D-layer . 2012. HIGH ENERGY DENSITY PLAS- doi:1029/2010A016018. MAS AND FLUIDS CAPABILITY REVIEW . Lay, E. H., and X. M. Shao. High temporal and spatial- Shao, X. M., E. H. Lay, and A. R. Jacobson. Reduction of resolution detection of D-layer fluctuations by using electron density in the night-time lower ionosphere in time-domain lightning waveforms. 2010. In AMERICAN response to a thunderstorm. 2013. NATURE GEOSCI- GEOPHYSICAL UNION 2010 FALL MEETING. (San Fran- ENCE. 6 (1): 29. cisco, CA, 13-17 Dec. 2010). , p. AE21B. San Francisco, CA: American Geophysical Union. Shao, X. M., E. H. Lay, and A. R. Jacobson. Model simulation and remote electric field change measurement show Lay, E. H., and X. M. Shao. High temporal and spatial- return stroke current travels in a dispersive and lossy resolution detection of D-layer fluctuations by using manner. Presented at AMERICAN GEOPHYSICAL UNION time-domain lightning waveforms. 2011. In XIV INTER- FALL MEETING. (San Francisco, CA, 3-7 Dec. 2012). NATIONAL CONFERENCE ON ATMOSPHERIC ELECTRIC- ITY. (Rio de Janeiro, Brazil, 8-12 Aug. 2011). , p. 1. Rio Shao, X., E. H. Lay, and A. R. Jacobson. Using lightning de Janeiro, Brazil: International Commision on Atmo- waveforms to probe thunderstorm’s electromagnetic spheric Electricity. effects on the ionospheric d-layer. 2011. In AMERICAN GEOPHYSICAL UNION FALL MEETING. (San Francisco, Lay, E. H., and X. M. Shao. Detection of D-layer perturba- CA, 5-9 Dec. 2011). , p. AE21A. Washington DC: AGU. tions with high time and spatial resolution. Invited pre- sentation at CEDAR. (Boulder, CO, 1-2 JUN. 2010). Lay, E. H., and X. Shao. D-layer fluctuations directly above a thunderstorm region detected by using time-domain LF/VLF waveforms. 2011. In AMERICAN GEOPHYSICAL UNION FALL MEETING. (San Francisco, CA, 5-9 Dec. 2011). , p. AE21A. Washington DC: AGU. Shao, X. M., A. R. Jacobson, and E. H. Lay. On the be- havior of return stroke current and the remotely detected electric field change waveform. 2012. JOUR- NAL OF GEOPHYSICAL RESEARCH. 117 (D07105): doi:10.1029/2011JD017210. Shao, X. M., A. R. Jacobson, and E. H. Lay. Study of iono- spheric D-layer fluctuation by comparing VLF/LF propa- gation model simulation with remotely detected light- ning time waveforms. 2011. In XIV INTERNATIONAL CONFERENCE ON ATMOSPHERIC ELECTRICITY. (Rio de Janeiro, Brazil, 8-12 Aug. 2011). , p. 1. Rio de Janeiro, Brazil: International Commision on Atmospheric Elec- tricity. Shao, X. M., E. H. Lay, A. R. Jacobson, and C. S. Carrano. Changes of electron density in the ionosphere in response to lightning discharges and their parental thunderstorms. Invited presentation at AMERICAN GEOPHYSICAL UNION FALL MEETING. (San Francisco, CA, 9-13 Dec. 2013). Shao, X. M., E. H. Lay, and A. R. Jacobson. Observations of electron density changes in ionospheric D-layer above tropospheric thunderstorms . 2012. In 2012 COU- 420

Page 421

Environmental and Biological Sciences Exploratory Research Final Report High Density Neuronal Recording Using Nanowire Capacitor Sensors John S. George 20110264ER Abstract Capabilities developed and demonstrated by this work Understanding the neural basis of the processes of will support future external proposals to NIH, DARPA, sensory perception, motor control, learning and mem- DOD and other agencies, who all project increased and ory, and even consciousness, is the final frontier of the targeted investment in this area. biological sciences. This task requires us to measure and Background and Research Objectives ultimately manipulate the dynamic function of large populations of individual neurons. We proposed and Understanding the neural basis of the processes of have demonstrated a novel scheme for in-vivo neuronal sensory perception, motor control, learning and mem- signal detection using semiconductor nanowire capaci- ory, and even consciousness, is the final frontier of the tor probes that operate by displacement currents, thus biological sciences. This task requires us to measure and allowing the sensor to be completely encapsulated and ultimately manipulate the dynamic function of large eliminating problems associated with electrochemistry. populations of neurons. Neural degenerative diseases Using our approach, sensors can be integrated and/or and traumatic neurological injuries affect millions of co-fabricated with active microelectronic devices, greatly people worldwide, motivating the development of increasing the feasible number of channels, which can neural prosthetic interfaces to restore sensory or motor be simultaneously recorded. function in affected individuals. Most existing interface systems for such applications are based on a relatively We developed nanowire capacitor devices, and opti- small number of relatively simple electrodes, but many mized fabrication processes to achieve the desired probe applications require higher density arrays, able to resistance and capacitance. Biocompatible perfluoroal- selectively record from or stimulate individual neurons. kane coatings for the 3D probes were developed, opti- Such systems will ultimately require incorporation of mized, and tested for pin-hole free dielectric and saline- active circuitry into packages that can withstand chronic resistant films. We tested the functionality of probes by implantation. Advances in neural sensing and stimula- characterizing photoresponses of isolated retina. We tion interface technology will lead to the development compared capacitive measurements to conventional of hybrid biological-electronic sensor devices for robust, electrode arrays and intrinsic signal imaging techniques functioning neural prosthetic systems, and also provide a in order to validate the new technology and to assess more comprehensive understanding of neural function. the merits relative to competing techniques. We fabri- cated nanowire capacitor detectors on passive electrode In the past several decades, neural activity imaging and arrays and characterized light responses in isolated reti- sensing techniques have been developed to record the na with the new devices using capacitive detection (~60 activity of individual neurons and small networks. An channels). Future projects will develop arrays incorporat- emerging goal is to simultaneously measure the activity ing active readout circuitry (10,000-100,000 channels). of all neurons in the retina or a limited region of brain The sensor array has the sensitivity, spatial resolution, tissue. Advances in sensing technology will allow a more and biocompatibility for basic studies of neural encod- comprehensive understanding of retinal processing and ing and processing, as well as for high-density implant- encoding leading to the development of hybrid biolog- able electro-neural interface systems. Systems based on ical-electronic sensor systems for robust, functioning this technology will be used to explore emerging ideas prosthetic system. We anticipate that such knowledge about how information is encoded by synchronous firing will also serve as the foundation for more sophisticated or covariance within a local neuronal population. machine vision systems that will ultimately emulate the 421

Page 422

perceptual function of biological vision. to drive prosthetic devices for peripheral limb amputees. With support of the LDRD program office we have pursued Both basic science and biomedical applications require these last two applications. electro-neural interface devices to withstand corrosion and induce minimal damage at the electrode/tissue interface. Scientific Approach and Accomplishments Current techniques of neural activity sensing employ multi- During the first two years of the project we made very electrode arrays (MEAs) that are typically made of metal substantial progress on all of the original objectives of the electrodes and utilize direct current measurement. En- project, as documented in previous progress reports and capsulation of MEAs in a protective coating can eliminate project appraisals. We have demonstrated the develop- corrosion and minimize electrochemical charge transfer ment of both micro- and nanoscale 3-D structures (arrays across the electrode-tissue interface, which induce circuit- of silicon or metal pillars) and have manufactured these ry failure and neural damage through irreversible Faradic in sensor arrays that allow connection to the front-end reactions. active electronics used for conventional multi-electrode arrays. We have developed and characterized two classes To overcome the direct current limitations in existing MEA of self-assembled organic coatings for micro-wire sensors. technology, capacitive charge transport between the elec- We have implemented capabilities for electrophysiological trodes and neural tissue might be employed. Capacitive measurements with microelectrodes arrays, using a state sensing could alleviate most of the drawbacks of the Fa- of the art system acquired with LDRD funding. We have radic current stimulation, including corrosion and evolution conducted multi-channel electrophysiological measure- of chemically reactive species. Recent work has advanced ments in retina, combined with simultaneous high-reso- the technology of capacitive stimulation of neurons and lution optical imaging measurements. Experiments have this method will likely prove important for designing future been conducted with both micro- and nanowire sensor electroneural prosthetic systems. Capacitive coupling has arrays demonstrating the feasibility of physiological record- been used for noninvasive measurements in a neuronal ings in retina with LANL devices. culture environment on a semiconductor chip, but the development of this technology has been limited, and Major technological challenges of sensor development present systems do not support implantation and chronic were addressed during the first two years, producing recording. Advances in the design and fabrication of sys- microstructured 3D sensor arrays that have been em- tems for high-density capacitive coupling to neural tissue ployed for electrophysiological measurements (Figure 1). will advance prosthetic applications as well as our knowl- The biggest remaining challenge for the third year was to edge of systems neuroscience. improve the dielectric coating of the micro- and nanowire sensors, and to show that these devices enable capacitive The principal objectives of the original project, as stated in sensing in the absence of galvanic connection to neuronal the proposal include: tissue. Both Huang and Branch worked on this problem, trying to achieve defect free dielectric coatings based on • Develop, design, and fabricate radial semiconductor Halfnium oxide. Such devices could eventually incorporate nanowire capacitors for high density, three-dimension- an organic insulating layer.as demonstrated in previous al implantable electroneural recording devices work. As these devices were fabricated we characterized • Develop prototype implantable devices with self- the resistance across the sensor dielectric coating using assembled, self-healing bio-compatible coatings for impedance measurements (Figure 2) before employing chronic applications. the sensor arrays for electrophysiological measurements (Figure 3). • Study the dynamic function of a neural network by monitoring activity of every neuron in a patch of We undertook an initial demonstration of the foundation retina, using conventional electrode arrays. nanowire technologies for construction of a hybrid sensor array sensors and optical imaging techniques incorporating cultured excitable cells (Figure 4). Because we did not have the required expertise for cultured cell In addition to these central aims we also described other work within the existing project team, we added Rashi Iyer, potential applications of the technology. These include the who in turn was able to recruit a postdoctoral fellow to use of capacitance-based sensors as non-contact pickups participate in this work. We believe that the most feasible for electro-encephalography (EEG); and the possibility of and useful approach will be to establish a neural interface stimulating neurons with capacitive currents.. We also for the peripheral nervous system be to culture muscle discussed the development of hybrid bio-electronic sen- cells (myocytes) in contact with the sensor arrays. The sors for chemical detection and neuromuscular interfaces 422

Page 423

ultimate application would be to acquire control signals would provide both an excellent model system for study of for prosthetic devices from the severed motor neurons of function, as well as a useful biosensor, e.g for detection of peripheral limb amputees. Such implantable interfaces are certain chemical warfare agents. In fact, a patent disclo- actively sought, again principally by NIH(NIBIB) and DARPA. sure was developed for this idea. Because the development of techniques to encourage the The principal bioscience challenges in this work were to regrowth and targeting of myocytes by motor nerve tracts achieve sustainable culture of suitable populations of was beyond the scope of the present project, we pursued nerve and muscles cells, to achieve viable co culture of the use of myocytes that have been transfected with chan- mixed populations of these cell types, to develop electro- nelrhodopsin, allowing their stimulation by light, closely physiological or other strategies for controlling and moni- analogous to our present work in retina. toring activity of these cells, and ultimately to develop integrated systems based on these capabilities. Results During the past year, we developed a three-dimensional Although this portion of the project only began in the last micro pillar sensor array that utilizes capacitive coupling in six months of the ER project, we nevertheless achieved order to sense electrical activity in excised retinal tissue. substantial progress on these objectives. The device incorporates over 3600 capacitive probes orga- We achieved viable primary cultures of both neurons and nized into 60 independent sensor sites (for compatibility myocytes, and observed maturation of the immature cells with existing electronics) spread over an area of 750 µm2. into forms that appear morphologically to be functional, Each sensor site consists of an 8x8 array of Pt micropillars differentiated cells. in order to maximize the capacitive response, intercon- nected by leads on the device. We developed several complementary strategies to allow co-culture of heterogenenous populations of neurons and The device is fully insulated by atomic layer deposition of myocytes. hafnium oxide. Electrochemical impedance spectroscopy was used to study the oxide deposition on the 3D micro We achieved transfection of cells with optogentic mate- pillar sensor array to ensure a pinhole-free dielectric coat- rial that will ultimately allow the activity of these excitable ing. The characteristic impedance magnitudes increased cells to be stimulated and monitored by light. up to 3 orders of magnitude upon oxide deposition and phase measures indicate fully capacitive sensor sites. With the demonstration of capacitive sensing of neuro- physiological responses of retina, and the pilot work on de- The capacitive sensor was interfaced with retinal tissue velopment of cultures of excitable cells we have achieved and upon light stimulation electro-retinograms (ERG) were all of the major technical objectives of the project, as origi- recorded. Our fabrication processes and electrochemical nally proposed. However there are additional steps that impedance measurements demonstrated the usefulness we hope to pursue with follow-on, non-LDRD funding: of such techniques for building high-density 3D arrays that were fully encapsulated with a protective dielectric coat- • Integrate the Hafnium insulated capacitive sensors ing. The work advances the technology towards capacitive with the self-assembled organic coatings previously sensing of neurons with a robust, non-invasive sensing developed. device, providing a base for studying neural encoding and • Build a version of the sensor arrays suitable for integra- processing by networks of neurons and for the future de- tion with commercially available read-out integrated velopment of neural prosthetics. circuits (ROICs) or with custom circuitry to allow sensor During the final year of the project, we requested and re- multiplexing for much higher density data acquisition ceived permission to pursue a new direction in the project: over a serial interface. We anticipate the ability to developing hybrid sensors incorporating capacitive read- record from hundreds of thousands of discrete sensor out and/or stimulation of captive excitable cells (neurons sites. and myocytes). The original motivation for this proposal • Collaborate with others (e.g. Quan Qing at Arizona by the project PI (George) was to develop a electroneural State University or Murat Okandan at Sandia) to de- interface for the perpheral nervous system that would be velop ROICs incorporating front end Field effect tran- suitable for neural control and feedback of a prosthetic sistors ( FETs) for greater sensitivity readout of dense limb. However, in early discussions with Iyer, we realized capacitive sensor arrays. that such a system, allowing the activity of a neuromusclu- • Develop an implantable sensor for the peripheral lar junction formed in vitro to be controlled and monitors, nervous system incorporating captive neurons and 423

Page 424

myocytes for bidirectional interface with motor/sensor nerve trunks. Impact on National Missions The project successfully developed and demonstrated a new technology for sensing of neural responses based on capacitive probes. The approach enables high density sen- sor arrays suitable for long-term studies of neural process- ing as well for chronic implantation for use as interfaces for electroneural prosthetic devices. Bio-hybrid devices based on such devices and incorporating cultured cells should find applications for sensing and monitoring applications 100 as well as for model systems for biological studies. Devel- opment of advanced neural interfaces is a central focus 80 of programs supported by DARPA and NIH/ NIBIB as well as previous programs supported by DOE and LANL LDRD. 60 Development of artificial organ systems is a topic of inter- 40 est to DTRA and NIH. High-density measurement of neural function is a principal focus of the multi-agency BRAIN 20 initiative. Such studies also hold promise for elucidating 0 mechanisms of information processing and encoding by the brain that are increasingly viewed as the basis of a new generation of neuromorphic computing systems that Figure 2. Interface impedance (A) and phase shift (B) vs. will be much smarter and more capable, with a fraction of frequency of a 3D Pt electrode coated with 100-150 cycles of the size, weight, and power requirements of conventional HfO2. computing systems. Figure 1. SEM images of the 60-electrode device with three- dimensional pillar electrodes (Inset A & B) connected by Pt metal traces terminating at bond pads on the outside edge of the device (Inset C). Figure 3. Responses (left axis) of isolated frog retina to light stimulation from 6 individual capacitive sensors with their respective location within the array (A) and a single capacitive sensor response (B). Stimulation (right axis) starts at 8 seconds and remains on for 4 seconds. Note that the retina characteristically responds to both the onset and offset of light stimulus. 424 )seergeD( elgnA esahP 8 10 7 10 6 10 5 10 4 10 3 10 3D Pt Control HfO2 100 cycles HfO2 125 cycles HfO2 150 cycles -1 0 1 2 3 4 5 10 10 10 10 10 10 10 Frequency (Hz) )Ω( ecnadepmI fo edutingaM !” 3D Pt Control HfO2 100 cycles HfO2 125 cycles HfO2 150 cycles -1 0 1 2 3 4 5 10 10 10 10 10 10 10 Frequency (Hz) #” 100 80 60 40 20 0 )seergeD( elgnA esahP 108 107 106 105 104 103 3D Pt Control HfO2 100 cycles HfO2 125 cycles HfO2 150 cycles 10-1 100 101 102 103 104 105 Frequency (Hz) )Ω( ecnadepmI fo edutingaM !” 3D Pt Control HfO2 100 cycles HfO2 125 cycles HfO2 150 cycles 10-1 100 101 102 103 104 105 Frequency (Hz) #“

Page 425

2014. APPLIED SURFACE SCIENCE. 288: 98. Dai, X., S. Dayeh, V. Veeramuthu, A. Larrue, J. Wang, H. Su, and C. Soci. Tailoring the Vapor-Liquid-Solid Growth toward the Self-Assembly of GaAs Nanowire Junctions. 2011. NANO LETTERS. 11 (11): 4947. Dayeh, S. A.. One Dimensional Semiconductor Heterostruc- tures: Potential and Implications for Advanced Technolo- gies. Invited presentation at Nanovation: From Science to Startups. (UC Berkeley, 27 Oct 2012). Dayeh, S. A., A. V. Gin, and S. T. Picraux. Advanced core/ multishell germanium/silicon nanowire heterostructures: Morphology and transport. 2011. Applied Physics Letters. 98 (16): 163112 (3 pp.). Dayeh, S. A., W. Tang, B. M. Nguyen, X. Dai, Y. Liu, Y. Hwang, X. H. Liu, and R. Chen. Nanoscale Heterogeneous Reactions and Interfaces in Ge/Si and for III-V on Si Inte- grated Devices. 2013. SEMICONDUCTORS, DIELECTRICS, AND METALS FOR NANOELECTRONICS 11. 58 (7): 115. Dayeh, S. A., Y. Hwang, J. L. Disterhaupt, J. Yoo, I. H. Camp- Figure 4. Formation of the neuromuscular junction (NMJ). bell, J. S. George, and S. T. Picraux. Nanostructures for (A) Differentiated and un-differentiated C212 (red) skeletal solar energy and probing the brain. Invited presentation myoblasts and NSC34 motoneurons. (B) Fluorescent image at NNSA LDRD 2012 Symposium. (Washington DC, 4 June showing myotubes (green) in differentiation media (DM) (C) 2012). Fluorescent image of neurites (red) sprouting from the cell body (green). (D) Phase contrast images of NMJ (orange) between Dayeh, S. A., Y. T. Hwang, J. L. Disterhaupt, J. Yoo, I. H. C2C12 myotubes (red) and NSC 34 motoneurons (green). Campbell, J. S. George, and S. T. Picraux. Nanostructures for solar energy and probing the brain. Invited presenta- Publications tion at 2012 LDRD Day. (Los Alamos, 23 Oct. 2012). Branch, B. A., Y. Hwang, J. L. Schei, A. M. Dattelbaum, D. Petsev, and J. S. George. Design and fabrication of a 3D Dayeh, S., J. Wang, N. Li, J. Y. Huang, A. Gin, and S. Thomas. pillar electrode array for transduction of retinal responses. Picraux. Growth, defect formation, and morphology con-
2015. ABSTRACTS OF PAPERS OF THE AMERICAN CHEMI- trol of germanium-silicon semiconductor nanowire hetero- CAL SOCIETY. 245: -. structures. 2011. Nano Letters. 11 (10): 4200. Branch, B., J. L. Schei, G. Gupta, A. Dattelbaum, D. Petsev, Dayeh, S., N. MacK, J. Y. Huang, and S. T. Picraux. Advanced and J. S. George. Hafnium oxide encapsulated micropillar core/multishell germanium/silicon nanowire heterostruc- array for transduction of neural responses. IEEE Sensors. tures: The Au-diffusion bottleneck. 2011. Applied Physics Letters. 99 (2): 023102. Branch, B., J. L. Schei, K. Artyushkova, G. Gupta, A. M. Dattelbaum, D. N. Petsev, and J. S. George. 3D Capacitive Dayeh, S., R. Dickerson, and S. Thomas. Picraux. Axial sensor array for detection of neural responses . To appear bandgap engineering in germanium-silicon heterostruc- in Journal of Solid State Science and Technology. tured nanowires. 2011. APPLIED PHYSICS LETTERS. 99 (11):

Branch, B., M. Dubey, A. S. Anderson, K. Artyushkova, J. K. Baldwin, D. Petsev, and A. M. Dattelbaum. Investigating Dayeh, S., X. H. Liu, X. Dai, J. Y. Huang, S. T. Picraux, and C. phosphonate monolayer stability on ALD oxide surfaces. Soci. Rocking chair defect generation in nanowire growth. 2014. APPLIED SURFACE SCIENCE. 288: 98. 2012. APPLIED PHYSICS LETTERS. 101 (5): 053121. Branch, B., M. Dubey, A. S. Anderson, K. Artyushkova, J. K. George, J.. Casting Light on Neural Function: A Subjective Baldwin, D. Petsev, and A. M. Dattelbaum. Investigating History. 2010. In Casting Light on Neural Function: A Sub- phosphonate monolayer stability on ALD oxide surfaces. jective History. , p. 1. 425

Page 426

Gintautas, V., M. Ham, B. Kunsberg, S. Barr, S. Brumby, C. (6): 2251. Rasmussen, J. George, I. Nemenman, L. A. Bettencourt, and Mohite, A. D., D. E. Perea, S. Singh, S. A. Dayeh, I. H. Camp- G. Kenyon. Model Cortical Association Fields Account for bell, S. T. Picraux, and H. Htoon. Highly Efficient Charge the Time Course and Dependence on Target Complexity of Separation and Collection across in Situ Doped Axial VLS- Human Contour Perception. 2011. PLOS COMPUTATIONAL Grown Si Nanowire p-n Junctions. 2012. NANO LETTERS. BIOLOGY. 7 (10): e1002162. 12 (4): 1965. Hwang, Y. T., J. Disterhaupt, J. S. George, S. T. Picraux, and Seo, M. A., S. A. Dayeh, P. C. Upadhya, J. A. Martinez, B. S. S. A. Dayeh. High density capacitive pillar arrays for high Swartzentruber, S. T. Picraux, A. J. Taylor, and R. P. Prasan- fidelity neural sensors. Presented at 2012 MRS Fall Meet- kumar. Understanding ultrafast carrier dynamics in single ing. (Boston, 25-30 Nov. 2012). quasi-one-dimensional Si nanowires. 2012. Applied Physics Hwang, Y. T., J. L. Disterhaupt, J. S. George, S. T. Picraux, Letters. 100 (7): 071104 (5 pp.). and S. A. Dayeh. High density electrically isolated capaci- Seo, M. A., S. A. Dayeh, P. C. Upadhya, S. T. Picraux, J. Mar- tive pillar arrays for high fidelity neural sensors. Presented tinez, B. S. Swartzentruber, A. J. Taylor, and R. P. Prasanku- at Society for Neuroscience Meeting. (New Orleans, 13-17 mar. Polarization anisotropy of transient carrier dynamics Oct. 2012). in single Si nanowires. 2011. In 2011 Conference on Lasers Hwang, Y., B. M. Nguyen, and S. A. Dayeh. Atomic layer de- and Electro-Optics, CLEO 2011 ; 20110501 - 20110506 ; position of platinum with enhanced nucleation and coales- Baltimore, MD, United States. , p. var.pagings. cence by trimethylaluminum pre-pulsing. 2013. APPLIED Seo, M. A., S. A. Dayeh, P. C. Upadhya, S. T. Picraux, J. Mar- PHYSICS LETTERS. 103 (26): -. tinez, B. S. Swartzentruber, A. J. Taylor, and R. P. Prasanku- Hwang, null., null. Dayeh, W. [. Tang, and null. Dai. All-solid mar. Polarization anisotropy of transient carrier dynamics state nickel silicide wafer bonding with wide thermal bud- in single Si nanowires. 2011. In CLEO: 2011 - Laser Science get. 2012. LDRD. to Photonic Applications ; 1-6 May 2011 ; Baltimore, MD, USA. , p. 1. Ji, Z., G. T. Kenyon, S. Brumby, L. M. Bettencourt, and J. S. George. Deep generative model of the where-what path- Tang, W., S. Dayeh, S. Tom. Picraux, J. Y. Huang, and K. Tu. ways in the primate visual cortex. 2011. In 41st Annual Ultrashort channel silicon nanowire transistors with nickel Meeting of the Society-for-Neuroscience ; November 12 silicide source/drain contacts. 2012. Nano Letters. 12 (8): -16, 2011 ; Washington, DC, USA. Vol. 41. 3979. Kar, A., P. Upadhya, S. Dayeh, S. Tom. Picraux, A. Taylor, and R. Prasankumar. Probing Ultrafast Carrier Dynamics in Sili- con Nanowires. 2011. IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS. 17 (4): 889. Li, Xianglong, Jeong-Hyun Cho, Nan Li, Yingying Zhang, D. Williams, S. Dayeh, and S. T. Picraux. Carbon Nanotube- Enhanced Growth of Silicon Nanowires as an Anode for High-Performance Lithium-Ion Batteries. 2012. Advanced Energy Materials. 2 (1): 87. Liu, X. H., H. Zheng, L. Zhong, S. Huan, K. Karki, L. Q. Zhang, Y. Liu, A. Kushima, W. T. Liang, J. W. Wang, J. Cho, E. Ep- stein, S. Dayeh, S. Tom. Picraux, T. Zhu, J. Li, J. Sullivan, J. Cumings, C. Wang, S. Mao, Z. Z. Ye, S. Zhang, and J. Y. Huang. Anisotropic Swelling and Fracture of Silicon Nanow- ires during Lithiation. 2011. NANO LETTERS. 11 (8): 3312. Liu, X. H., L. Q. Zhang, L. Zhong, Y. Liu, H. Zheng, J. W. Wang, J. Cho, S. Dayeh, S. Tom. Picraux, J. Sullivan, S. Mao, Z. Z. Ye, and J. Y. Huang. Ultrafast electrochemical lithiation of individual Si nanowire anodes. 2011. Nano Letters. 11 426

Page 427

Environmental and Biological Sciences Exploratory Research Final Report Restoring Neurological Function with Self-assembled Lipid Bilayers Srinivas Iyer 20110411ER Abstract potential. Multiple disease states result in breakdown of Myelin is a complex tissue that covers nerves and per- myelin, consequently affecting even the most basic mo- mits electrical conduction through our bodies. The role tor functions. Understanding these processes are thus at of myelin is central to all neurological functions, thus the very crux of therapeutic strategies for this important perturbations of this tissue have devastating conse- public health concern. The complexity of myelin struc- quences to human health. These perturbations are typi- ture precludes studies examining the effect of structural cally manifested as “demyelination”, a condition where perturbations on function. Our overall objective is to segments of the nerve are left uncovered which then create assemblies that mimic myelin that will enable leads to a reduction in electrical conduction through systematic studies. the nerve. While conditions such as injury trauma and The main goals of the project included: diabetes result in this condition, it is most prominent in the debilitating disease Multiple Sclerosis. The events Aim 1: Use native myelin proteins to sequentially link surrounding myelin damage and repair has stayed at the several lipid bilayers. forefront of neurochemical research, however insuffi- cient information is available due to the lack of suitable Aim 2: Study size, integrity, fluidity and insulating capac- experimental systems that reproduce the complexity of ity of the assembled structures. the natural tissue. The foremost goal of our effort is to generate an experimental system to investigate the role Aim 3: Mimic known biochemical events surrounding of specific structural features in myelin function. The demyelination and assess their contribution complexity of natural myelin precludes such studies with Scientific Approach and Accomplishments tissue isolates, hence our assembly will enable mecha- nistic interrogations of structural events. The second We have used nanoscale, controlled self-assembly and longer term direction is to establish self-assembly method to create an architecture that closely mimics principles that will guide the ultimate development of structural and functional aspects of the dense line of artificial myelin tissue. natural myelin. We have also performed electrical im- pedance measurements on these structures. This mimic Background and Research Objectives will facilitate studies of the effect of myelin structure on Nature has a huge repertoire of “smart” biomaterials its function and provide insights into hitherto unknown that have evolved to satisfy physiological needs. Perhaps details of this very important tissue. These results will the most fascinating of these is the myelin sheath, an be used to design strategies for treating neurological dis- intricate nanoscale lipid-protein assembly. The easiest orders stemming from myelin damage. In this study, we way to imagine myelin is as a stack of lipid bilayers held examined fluidity and insulating capacity and describe together by proteins. Myelin was defined by Ranvier in the results below. We also addressed aim 3 by using my- 1878 as a sheath that surrounds nerve fibers. Portions of elin basic protein to stitch bilayers; however the results the nerve are left uncovered at points called the nodes were not sufficiently conclusive to determine the effects of Ranvier. It is at these nodes that the nervous impulse of phosphorylation. We are developing a proposal to in the form of an action potential (voltage differential) is the NIH to investigate additional biochemical events as a rejuvenated. The mechanism by which myelin exerts its follow-on project to this LDRD-ER. function is its role as an insulator, preventing the dissipa- Towards our overarching goal of developing tion of ionic gradients that are the basis of the action 427

Page 428

and understanding lipid systems that closely re- (b) Layer-by Layer Approach: In parallel, we sequentially semble myelin sheaths, we used two approaches: link several lipid bilayers in a controlled self-assembly (a) 1-step Approach: We document the synthesis and process using vesicle fusion on conductive surfaces. The characterization of highly ordered and stable phospholipid- layer-by-layer approach enables an alternating structure silica thin films that resemble multi-lamellar architectures with closer resemblance to the dense and Intra-period present in nature such as the myelin sheath. We have lines of myelin. Briefly, lipid vesicles (containing biotin) used a near room temperature chemical vapor deposition were infused on silicon wafers resulting in formation of method to synthesize these robust functional materials first bilayer. Subsequently, streptavidin was added followed (Figure 1). Highly ordered lipid films are exposed to vapors by fusion with another set of vesicles. This procedure was of silica precursor resulting in the formation of nano-struc- performed over and over again resulting in the formation tured hybrid assemblies. This process is simple, scalable of multi lamellar layer. Figure 3 shows the sequential im- and offers advantages such as exclusion of ethanol and no ages of the lipid layers on silicon substrates. As can be seen (or minimal) need for exposure to mineral acids, which are the lipid bilayers recover almost fully in 30-60 minutes. generally required in conventional sol-gel synthetic strate- The diffusion constant estimated for these bilayers is in the gies. The structure of the phospholipid-silica assemblies range of 0.5-1.0 2/s. can be tuned to either lamellar or hexagonal organization Figure 4 shows the impedance spectra of the respective depending on the synthesis conditions. The phospholipid- assemblies. An increase in impedance is observed after silica films exhibit long-term structural stability in air as addition of each subsequent layer. Specifically, the appear- well as when placed in aqueous solutions and maintain ance of a second capacitor is noticeable upon addition of their fluidity under aqueous or humid conditions. This each bilayer. These results clearly indicate that we can cre- platform provides a model for robust implementation ate layer-by-layer model systems that can be used to mimic of phospholipid multilayers and a means toward future and understand lipid multilayers. This is the first report on applications of functional phospholipid supramolecular multi-lamellar assemblies that resemble myelin. assemblies in device integration. As an example, we used POPC lipid to construct our assemblies. Figures 2a and 2b Impact on National Missions show XRD patterns of lipid films before and after expo- sure to TMOS precursors at 37 °C. The patterns reflect the Development of biomimetic models is an example of lamellar structure with a strong (001) diffraction peak at 50 how molecules self-assembled at the nanometer length Å for the control lipid assembly (no silica) and 55 Å for the scale can improve our understanding of complex biologi- hybrid lipid-silica film. Both the control and hybrid films cal systems, a major area of interest of DOE BES. We have exhibit higher order diffraction (002), (003), (004) peaks. created a dynamic model system with which we can begin A TEM micrograph of the surface region of a hybrid POPC- to understand processes critical to myelin sheath forma- silica thin film after 1-hour exposure to TMOS vapor clearly tion and degradation. This exploratory work has laid the indicates a lamellar structure (Figure 2c). No XRD peaks groundwork for an increase in the scientific understanding were observed after calcination, which suggests a collapse of complicated myelin-associated diseases (DHHS interest, of the structure upon removal of the lamellar phospholipid proposal under development for NIH). These efforts tie in assembly. Sequential fluorescence images of the silica- with LANL’s stated mission goals of being a premier materi- coated POPC multi-lamellar assemblies are shown in Figure als laboratory with a focus on developing next-generation 2e. FRAP measurements indicate that phospholipids in biomaterials. The health impacts fits into the Public Health these multi-lamellar ultrathin films are diffusively mobile mission of Bioscience Division and is of interest to certain (D = 0.1 ± 0.04 μm2/s) when placed in aqueous solution. agencies of the DoD especially in the area of treating warf- It is interesting to note that these assemblies did not crack ighters after nerve trauma from battlefield injuries. in aqueous conditions, possibly due to the low synthesis temperature. In conclusion, we are able to easily fabricate tunable robust and fluid hybrid lipid-silica assemblies with various mesostructural features through minimal adjustments to the synthesis conditions. Our laboratories are currently optimizing reaction conditions for successful incorporation of relevant transmembrane peptides and proteins into the hybrid lipid-silica assemblies that resemble myelin. 428

Page 429

Figure 1. Schematic of the CVD process used to prepare silica encapsulated multilamellar lipid assemblies. Spin-coated multilamellar lipid stacks (left) are exposed to vapors of a silica precursor (TEOS or TMOS) and water. Silica forms selectively between the lipid head groups of the lamellar lipid assemblies. Increase in synthesis temperature, use of unsaturated lipid and Figure 3. FRAP images of 1 Bilayer, 2 Bilayers and 5 Bilayers. long exposure times result in the formation of hexagonal phases. After photobleaching, sequential images indicate the recovery of bilayer. Figure 2. XRD, TEM and fluorescence images of POPC-silica assemblies prepared on silicon wafers. (a) XRD pattern before and after exposure to silica precursor vapors at 37°C, and after calcination. (b) XRD patterns of the 4th order diffraction peaks for the samples shown in (a). (c) TEM image of a POPC- silica assembly before calcination. (d) FRAP (luorescence recovery after photobleaching) of a lamellar POPC-NBD-PE silica film (synthesized at 60°C) performed in saturated water vapor. (e) FRAP of a lamellar POPC-NBDE-silica film under water (synthesized at 37°C). The dotted line in Fig. 5a was added to reference the (001) peak for all samples relative to the as- Figure 4. Bode plot of layer-by-layer lipid assemblies showing prepared spin-coated DMPC lipid assembly. increase in impedance with size of assembly. 429

Page 430

Publications Gupta, G., S. Lyer, K. Leasure, N. Virdone, A. M. Dattel- baum, P. B. Atanassov, and G. P. Lopez. Stable and Fluid Multilayer Phospholipid-Silica Thin Films: Mimicking Active Multi-lamellar Biological Assemblies. 2013. ACS NANO. 7 (6): 5300. Gupta, G., w. han, g. lopez, and l. ista. Silicieous nano-bio- materials. To appear in Handbook of nanobiomaterials. By r, . 430

Page 431

Environmental and Biological Sciences Exploratory Research Final Report Understanding the Mechanisms of Biological Transport for Nano-technology Applications Sandrasegaram Gnanakaran 20110412ER Abstract intense scientific debate. This project focuses on trans- The Nuclear Pore Complex (NPC) gates all the traffic port through one particular channel of this kind, the between the cell nucleus and the cytoplasm in eukary- Nuclear Pore Complex (NPC). otic cells, and its transport mechanism is a subject of NPCs are large protein assemblies embedded in the intense scientific debate. NPCs are remarkable biological nuclear envelope that gate all the molecular exchange machines that are directional and selective, but do not between the nucleus and the cytoplasm in all eukaryotic directly consume metabolic energy during the transloca- cells (Figure 1). Transport through the NPC is a center- tion process and seemingly do not undergo large-scale piece of a large number of regulatory processes in the molecular transitions. We seek a mechanistic under- cell, and its overall structure and composition is highly standing of NPC transport and plan to leverage this un- conserved among species – from yeast to humans. The derstanding towards the design of bio-mimetic devices NPC is a remarkable biological machine that combines and sensors. In this project, we have developed several highly optimized and selective transport with speed, computational and theoretical tools that are appropri- versatility, and robustness. It can selectively translocate ate to gain mechanistic understanding into both the thousands of molecules per second in both directions in equilibrium conformations of the system and its dynamic the presence of vast amount of background molecular function. The project resulted in several multi-resolution noise. techniques for a more detailed molecular understanding of the conformational dynamics of the channel. These Transport through the NPC involves interactions of tools have allowed us to develop conceptual models certain transport factors with the natively unfolded that capture the essentials of what is known about the proteins lining the NPC channel. Understanding the transport mechanism and its selectivity, both for the conformational dynamics of these unfolded flexible NPC and for related nano-technological devices. With proteins is a major challenge and requires development the approaches developed in this project, we can now of new hybrid coarse-grained and atomistic simulation show that NPC shares common mechanisms of transport techniques. The research objective is to provide an inte- selectivity and efficiency with many other biological and grated description of how conformational changes of the artificial nano-channels. NPC constituents determine the dynamics of transport through it and, in particular, its selectivity. The concep- Background and Research Objectives tual understanding arising from the study of the NPC can Proper functioning of living cells requires selective mo- be leveraged for design of artificial nano-sorting devices lecular transport into and out of the cell, and between based on the same principles. different cell compartments. Many such channels carry selective transport without an input of metabolic energy Scientific Approach and Accomplishments and without large scale structural transitions from a The centerpiece of the transport mechanism is the layer ‘closed’ to an ‘open’ state; the transport is based on dif- of natively unfolded, unstructured, flexible proteins that fusion only. Understanding selective transport through are attached to the inner surface of the NPC passageway such ‘always open’ channels is a fundamental biological and protrude into its lumen, as well as from nuclear and problem, poses challenging physical questions, and has cytoplasmic faces. These natively unfolded proteins have potentially far-reaching technological applications. The been shown to behave in many aspects as conventional transport mechanisms of many of such channels, and in polymers. The key element producing efficient and selec- particular the selectivity mechanisms, are still subjects of 431

Page 432

tive transport through the NPC is the transient binding of experiments. We developed a theoretical model describ- the transport proteins to specific binding sites on these ing the energetics of the cargo complex’s interactions with polymers. Conformational changes of these polymers have various parts of the NPC that allowed us to calculate the been hypothesized to be involved in the gating and trans- first passage transport times through the channel [4]. In port mechanisms, and several different models have been this model, binding near the cytoplasmic and nucleoplas- proposed [1]. However, many of these models rely on mic openings of the NPC (the “vestibules”) is mediated detailed assumptions about the features of the behavior of by a low density of unfolded proteins. In contrast, the these polymers that are currently unknown. central permeability barrier is comprised of a substantially higher density of the unfolded proteins, which can only be We focused on coarse-grained modeling rooted in polymer overcome with a larger number of favorable interactions. physics and non-equilibrium statistical mechanics to elu- Together, these features are described by a “double-well” cidate the role of these conformational changes. As a first shaped free energy landscape. The conclusion of this work approximation, the models take into account only the most is that the unfolded proteins are arranged in a non-uniform salient features or the polymers and their interactions with fashion such that multivalent interactions predominate in the transport proteins. Such models have an impressive the center of the pore, but not at the periphery. The work record in explaining the properties of polymeric materials, emphasizes several important conclusions not addressed even quite complicated ones. The folded structural pro- by previous models. In particular, transport protein teins comprising the channel and other parts of the device interactions with individual binding sites on the unfolded are be modeled on the coarse grained level as smooth proteins must be relatively weak, significantly weaker than surfaces; their interactions with the chain monomers are the micromolar to nanomolar affinities reported by numer- chosen to represent the required surface interaction prop- ous in vitro studies. This is essential to obtain millisecond erties. The parameters of the coarse grained simulations interaction times and is important for accounting for the are calibrated by comparison with available experimental apparent absence of multivalent interactions in the cyto- data on conformations of unfolded proteins and, where plasmic and nucleoplasmic vestibules. necessary, informed by microscopic atomistic simulations on short time and length scales. We have also provided evidence supporting the hypothesis that the layer of flexible polymers that forms a key compo- In this coarse-grained approach based on polymer phys- nent of biological and artificial channels presents a selec- ics and statistical mechanics, the unfolded proteins are tive potential through which large molecules must diffuse. represented as flexible polymeric chains and the models From this evidence, a recipe for optimizing the choice of can be treated either by analytical methods or by com- the chemical constituents in these devices emerges in the puter simulations. The model has been so far realized in case where the channel serves as a gateway separating the planar geometry where the flexible chains are grafted two finite sized chambers. Using theoretical arguments to a flat surface and are in equilibrium with the solution augmented by Brownian dynamics simulations (Figure of the transport factors. We have done this part in col- 3), we demonstrate that an optimal rate of translocation laboration with Prof. Rob Coalson and Prof. David Jasnow across these gateways may be engineered by systemati- and the graduate student Mr. Michael Opferman at the cally selecting the geometry as well as the synthetic or university of Pittsburgh [2, 3]. In parallel, we have devel- natural components of the polymer layer. The key to this oped simple approximate analytical models that capture optimization lies in tuning the local interactions between the essential properties of the polymer-transport factor all of these constituents and the targeted cargo, striking a mixtures. Results of the models seem to explain the in balance between the transit times and the probability of vitro experiments on the nanomechanic behavior of the crossing the channel. The study reveals four findings sig- filaments with and without transport factors. The model nificant to future artificial molecular sorter designs. First, has also uncovered hitherto unknown collective structural increasing the local binding affinity between the transport transitions in the assemblies of these polymers, induced moieties and the chain segments yields a non-monotonic by the presence of the transport factors [Figure 2]. These variation in the rate of transit across the channel. Second, transitions seem to explain the experimental in vitro data the time to travel across the channel increases as the bind- (hitherto unexplained) and might be important in the gat- ing affinity increases. Third, in accord with gene knockout ing and transport mechanisms of the NPC. experiments, reducing the number of chains within the channel changes the magnitude of the transit rates and Together with the experimental collaborators at Texas times, but preserves the qualitative functionality for large A&M University, we have applied the previously developed cargos. Fourth, one may work within the practical con- diffusion type models of transport to the analysis of trans- straints of synthetic chemistry and still reach a maximal port trajectories observed in the single molecule tracking 432

Page 433

rate of translocation, as prescribed by a simple analytical NPC has also been modulated by phosphorylation, glyco- treatment relating the effective potential to the binding sylation and di-sulfide bond formation. That has prompted strengths of a distribution of sites. us to explore the efforts of such post-translational modi- fications of FG-nups at atomistic scale. In that regard, we We have identified and addressed issues hampering the have been considering a natively unfolded peptide con- development of atomistic level description of about 30 struct with two glycosylations and a disulfide bond. These different natively unfolded proteins known as nucleoporins all-atom MD simulations reveal that the size, secondary or Nups that form the channel. The motivation for these structure and folding of the peptide construct is impacted studies is that an atomistic level description is required to by glycosylation. Also, the disulfide bond between two refine the coarse-grained model to more faithfully reflect unstructured peptides promotes beta strand formation (8). the geometry of the NPC and the sequence of the proteins. Multiple copies of Nups line the NPC, and their collective In summary, mathematical modeling has proven to play an dynamics facilitates transport through NPC. The FG family increasingly important role in the mechanistic understand- of Nups (FG Nups) contains extensive regions of FG motifs ing of the transport through the NPC in particular, because (phenylalanine-glycine) that facilitate the passage of cargo many of the details of in vivo transport are in accessible complexes through the NPCs. However, the mechanism of to direct experimental measurements at this point. In this the interaction of FG Nups with themselves or with cargo project simplified models of diffusion through the NPC complexes is not clear. have provided important insights and the general frame- work for thinking about the transport through the NPC. We considered all-atom molecular dynamics (MD) simula- The modeling of unfolded proteins is challenging both at tions for two types of FG nups, non-cohesive Nsp1 and coarse-grained and all-atom levels and requires new con- cohesive Nup116, in explicit water to study their confor- ceptual mathematical tools as we have implemented. Our mational characteristics and intra-molecular interactions. coarse-grained approaches can capture the emergence of These studies prompted us to reevaluate the applicabil- meso-scale order and function from underlying disorder. ity of commonly used biomolecular force fields towards Importantly, the insights gained from these theoretical/ modeling FG nups. We made extensive comparisons computational studies on NPC can be applied to the design between different water models (Figure 4) and identi- of bio-mimetic systems that can achieve highly regulated fied a better water model to use that properly solvated transport across biological or in vitro membranes. charged residues in FG nups [5]. We find that TIP4P-Ew water model provides hydrodynamic radius in agreement Impact on National Missions with experimental values and resulted in a more extended This project addresses basic questions about the mecha- conformation (Figure 5). Also, we have proposed a net- nisms of intra-cellular transport as well as the techno- work approach that can generally be applied to properly logical and bio-medical applications, in an area where sample large natively unfolded proteins such as FG nups in theoretical modeling is scarce and crucially needed. The all-atom simulations [6]. However, our work on this project research will provide a clear picture of the mechanisms and several of our other projects has suggested a better of intracellular transport. This work supports missions of approach to deal with inaccuracies in the force field and Threat Reduction by understanding the mechanisms of the lack of sampling of FG nup-like proteins, as it remains host-pathogen interactions. The principles of function practically impossible to generate ergodic ensembles for of the NPC and similar biological channels can be used large natively unfolded proteins in MD simulations. The for novel applications such as design of novel vectors for solution lies in using structure based measurements as delivery of materials into the cell nucleus and creation of constraints in simulations, Accordingly, we have proposed artificial bio-mimetic nano-filtering devices. Thus, under- a novel Bayesian framework approach that can use second- standing the functioning of such biological and artificial ary structure and solvation properties to obtain accurate channels poses fundamental biological and physical ques- thermodynamic weights of different conformations by tions and has far-reaching applications in nano-technology quantitatively connecting simulations to infrared spectrum and nano-medicine. of a natively unfolded protein [7]. At LANL, this project has helped to develop the connec- Additionally, MD simulations of multiple copies of FG nups tion between protein biophysics and materials science, show that interactions between FGnups are modulated by paving the way for creation of biosensors for detection of the anchoring distances within NPC. Also, these interac- pathogens and chemical agents. Exquisite selectivity and tions between multiple FG nups promote formation of robustness of biological channels provide inspiration for secondary structure. A manuscript summarizing the above creation of artificial nano-molecular devices based on the findings is under preparation. The dynamics of FG nups in 433

Page 434

same principles. Such artificial devices are cutting-edge in the design of novel molecular sorters and sensors and can be used for single-mismatch DNA detection, protein sort- ing and pathogen detection; functional prototypes of such devices have been built. Figure 2. The morphologies of grafted polymer layers interacting attractively with nanoparticle inclusions, as a function of particle size and the interaction strength, using self-consistent field theory and Langevin dynamics simulations. We find that the addition of nanoparticles causes distinctive changes in the layer morphology. Figure 1. A schematic representation of the NPC structure with the cargo-complex indicated as a kap-beta-bound blue sphere inside the central channel. The NPC structure resembles an hourglass or donut. This channel is sandwiched between the cytoplasmic and nuclear rings. The molecules in different color are important molecules of nucleocytoplasmic transport: Impα, Impβ, Ran, CAS, Ran, RTF/P10, etc. Image from Int Rev Cell Mol Biol 28706: 233. Figure 3. Snapshots from Brownian dynamics simulations of particle transit through a polymer-lined channel. The top image shows a large particle moving through a channel with moderate chain density. The bottom shows a medium-sized particle through a more densely filled channel. The three particles in between illustrate the three different sizes examined, all of which conform to the analytical theory. 434

Page 435

5. Tian, J., A. Zilman, and S. Gnanakaran. Molecular Dynamics Simulation Studies of FG-repeat-containing Nucleoporins. J. Phys. Chem .
6. Sethi, A., J. H. Tian, D. M. Vu, and S. Gnanakaran. Identification of Minimally Interacting Modules in an Intrinsically Disordered Protein. 2012. BIOPHYSICAL JOURNAL. 103 (4): 748.
7. Sethi, A., D. Anunciado, J. Tian, D. M. Vu, and S. Gnana- karan. Deducing conformational variability of intrinsi- Figure 4. Two water models considered in the all-atom MD cally disordered proteins from infrared spectroscopy simulations. with Bayesian statistics. 2013. Chemical Physics. 422:

Publications Gnanakaran, G., J. Tian, and A. Sethi. Conformational study of quaternary epitope region of V1/V2 loop: influence of disulfide bonds and glycosylations. 2012. RETROVI- ROLOGY. 9: -. Jovanovic-Talisman, T., and A. Zilman. Nanobiotechnology: Building a basic nanomachine. 2011. Nature Nanotech- nology. 6: 397. Opferman, M. G., R. D. Coalson, D. Jasnow, and A. Zilman. Morphology of Polymer Brushes Infiltrated by Attrac- tive Nanoinclusions of Various Sizes. 2013. LANGMUIR. 29 (27): 8584. Figure 5. Radius of gyration for Nsp1/Nup116 as a function of Opferman, M., R. Coalson, D. Jasnow, and A. Zilman. Mor- time. (A) and (C) are the systems using TIP3P water model; (B) phological control of grafted polymer films via attrac- and (D) are the systems using TIP4P-Ew water model. Different tion to small nanoparticle inclusions. 2012. PHYSICAL colors in each plot correspond to different simulations with REVIEW E. 86: 031806. different random seeds in each system. Sethi, A., D. Anunciado, J. Tian, D. Vu, and S. Gnanakaran. References Deducing conformational variability of intrinsically

1. Jovanovic-Talisman, T., and A. Zilman. NANOBIOTECH- disordered proteins from infrared spectroscopy with Bayesian statistics. 2013. Chemical Physics. 422: 143. NOLOGY Building a basic nanomachine. 2011. NATURE NANOTECHNOLOGY. 6 (7): 397. Singh, S., A. Junghans, J. H. Tian, M. Dubey, S. Gnanakaran, J. Chlistunoff, and J. Majewski. Polyelectrolyte multilay-
2. Opferman, M. G., R. D. Coalson, D. Jasnow, and A. ers as a platform for pH-responsive lipid bilayers. 2013. Zilman. Morphology of Polymer Brushes Infiltrated by SOFT MATTER. 9 (37): 8938. Attractive Nanoinclusions of Various Sizes. 2013. LANG- MUIR. 29 (27): 8584. Tian, J. H., A. Zilman, and S. Gnanakaran. A Study of Intrin- sically Disordered Proteins from Nuclear Pore Complex.
3. Opferman, M. G., R. D. Coalson, D. Jasnow, and A. Zil- 2013. BIOPHYSICAL JOURNAL. 104 (2): 55A. man. Morphological control of grafted polymer films via attraction to small nanoparticle inclusions. 2012. Tian, J., A. Zilman, and S. Gnanakaran. Molecular Dynamics Simulation Studies of FG-repeat-containing Nucleopo- PHYSICAL REVIEW E. 86 (3): -. rins. J. Phys. Chem..
4. Tu, L. C., G. Fu, A. Zilman, and S. M. Musser. Multiva- Tu, L. C., G. Fu, A. Zilman, and S. M. Musser. Multivalent lent interactions are required for large cargo transport interactions are required for large cargo transport by by nuclear pores: Implications for the spatial nuclear pores: Implications for the spatial orga- organization of the FG-nucleoporins . To appear in nization of the FG-Nucleoporin . To appear in EMBO EMBO Journal. Journal. 435

Page 436

Environmental and Biological Sciences Exploratory Research Final Report Nonlinear Resonances between DNA and THz Radiation Boian Alexandrov 20130765ER Abstract inherent thermal fluctuations at biological temperatures. Recently, it was demonstrated that terahertz radiation The DNA molecule, for example, experiences thermal (THz) can modify cellular gene expression and accelerate motions that induce spontaneous openings and re- stem cells differentiation. The ability to selectively and closings of the double helix known as “DNA breathing remotely activate specific genes and reprogram cells has [8],” or spontaneous “base flipping [9],” or “DNA bubbles an incredible technological and economical potential. To [10],” whose frequencies are also in the THz range. be able to exploit this potential we need to understand Our DNA simulations [11, 12] showed that THz radiation the mechanism behind this complex phenomenon. Our can interact with the natural vibrations of biomolecular mesoscopic modeling demonstrated a novel resonant H-bonds and change the local stability and flexibility of mechanism of THz-DNA interaction. Our simulations the biomolecule, and hence can interfere with biological showed that THz radiation can create unzipping or tran- function. We demonstrated that THz-DNA interac¬tion sient opening (bubbles) in the DNA molecule by inter- can pro¬duce unzipping or enhanced transients open- acting, in a novel nonlinear resonant manner, with the states in the double helix (THz-induced bubbles) [11, natural vibrations of the hydrogen bonds that stabilize 12]. Our numerical data showed that the creation of the double helix. As a first step toward the understand- these new open-states needs a strong assistance by ing of this interaction we built a novel ultra-fast THz-UV the DNA breathing and hence this effect is advanced at pump-probe experimental setup and explored the exis- the DNA vibrational hotspots, such as: gene promot- tence of THz-induced DNA bubbles. ers and protein-DNA binding sites (i.e., at locations that regulate gene expression). Further, led by our simula- Background and Research Objectives tions we demonstrated that THz radiation can change The emergence of reliable and powerful table-top tera- gene expression and cell differentiation [13, 14, 15]. Our hertz (THz) radiation sources has triggered an avalanche results also suggested that THz-induced bubbles depend of applications in security (e.g., airport screening), (bio) on DNA sequence, THz-frequency, time of exposure, and molecular spectroscopy, and medical imaging (cf. [1, 2]). amplitude of the field. As a consequence researchers across the world are be- ginning to explore the effects of THz irradiation on living Scientific Approach and Accomplishments matter. Thus, a range of non-thermal THz-induced bio- The specific goal in this project was, in accordance with logical effects have been observed, including: increased the suggestions and comments of the LANL Strategy DNA production; changes in gene expression; genomic and BIO SAP Teams to our 20130041DR full proposal, to instability; changes in the cell-membrane potential; mi- demonstrate the existence of resonant THz-DNA inter- totic disturbances; acceleration of stem cell differentia- action, or shortly, to demonstrate THz-induced DNA tion, and others (cf. [3, 4, 5]). The main question is: how bubbles. To observe experimentally THz-induced tran- the non-ionizing THz radiation can affect a biomolecule? sient bubbles in DNA, we performed ultrafast THz-UV The structural integrity and flexibility of biomacromol- time-resolved pump-probe spectroscopic experiments. ecules is primarily governed by hydrogen bonds (H- Ultrafast (40 fs) THz pulses, with broadband spectrum bonds), whose natural vibration frequencies are in the (0.1 - 12THz) and strong amplitude of the electric filed ( THz range [6, 7]. Because the H-bonds are weak (only 0.5- 1MV/cm) acted as pump, and, because the absorp- ~few kTs), many biomacromolecules experience rela- tion of UV (at ~265nm) is sensitive to the openings of tively slow conformational motion resulting from the the double helix, variable time-delayed UV pulses were 436

Page 437

used as probe. experimental data shed light on the mechanism of THz- DNA interactions. Our observations can pave the way to The THz-UV pump-probe time-resolved spectroscopic a future mechanism for active remote control of cellular setup was built at the MPA-CINT LUMOS facility. We gained functions (NIH, DOE) and additionally offer a new method multiple spectroscopic data showing THz-induced tran- for THz spectroscopy of biological macromolecules (DoD). sient bubbles in DNA oligomers, selected by DNA dynamics The project is closely aligned with the LANL’s Focus Areas simulations. in Complex Systems, and it is in support of DOE’s research strategies in System and Synthetic Biology. We conducted THz-UV pump-probe experiments with par- ticular DNA oligomers, viz., wtP5: TGAGCACGCAGGGTCTC- CATTTTGAAGCGGGAGG, and its modifications, selected via modeling of the DNA bubble activity. The PAGE purified, and at high concentration, wtP5 oligomers and their modi- fications consecutively were embedded in a temperature controlled 15µm thick water-sandwich holder (Figure 1 inset). This optical (liquid) cell was equipped with specific z-cut crystal quartz windows that are transparent to the THz and UV. The UV and THz pulses were positioned to hit the same spot of the cell. We demonstrated that the THz and UV ultrafast pulses can be measured accurately through 15-30 µm thick water layer, as well as through 15-30 µm thick water-sandwich that contains sodium phosphate physiological buffer (pH Figure 1. The ultrafast THz and UV pulses are successfully 7.4) with various salt concentrations, Figure 1. transmitted through DNA samples in a 15-30 μm water or physiological buffer with added salt. We validated our THz-UV pump-probe setup by measuring the differences in the absorption of UV ultrafast pulses by: i) double stranded DNA, ii) single stranded DNA, and iii) DNA with static bubbles created by insertion (in the wtP5) of 5 nucleotide mismatches (see the inset), Figure 2. Finally, via consecutively changes of the delay between the THz-pump and the UV-probe pulses, we measured the differential transmitted UV probe signal through samples

• with and without DNA, and observed THz-induced DNA bubbles. In Figure 3, we plot the time relaxation of THz- induced bubbles, registered by the change in the delayed UV-absorption. As it can be seen, the demonstrated relax- ation time is ~36 ps. In the inset of the Figure 3, we repre- sent the same process at a full scale. The first bump in the inset is the zero of the UV (the UV laser was switched off). Impact on National Missions Our data and computer simulations strongly suggest that THz radiation interacts (in a novel nonlinear resonant Figure 2. We can tell the difference between single stranded manner) with natural vibrations of biomolecular H-bonds, DNA and double stranded DNA. We can also register static DNA creating new conformational states that can change their bubbles caused by artificially introduced mismatches (e.g., G/T pairing). stability and flexibility. These new open states, in turn, influence gene expression and stem cell differentiation and hence the THz radiation has the potential to be used as a remote regulator of various cellular functions. We observed directly THz-induced bubbles in DNA consis- tent with our modeling. Most importantly the retrieved 437

Page 438

8. GUERON, M., M. KOCHOYAN, and J. L. LEROY. A SIN- GLE-MODE OF DNA BASE-PAIR OPENING DRIVES IMINO PROTON-EXCHANGE. 1987. NATURE. 328 (6125): 89.
9. ROBERTS, R. J.. ON BASE FLIPPING. 1995. CELL. 82 (1):
11. ENGLANDER, S. W., N. R. KALLENBACH, A. J. HEEGER, J. A. KRUMHANSL, and S. LITWIN. NATURE OF THE OPEN STATE IN LONG POLYNUCLEOTIDE DOUBLE HELICES

• POSSIBILITY OF SOLITON EXCITATIONS. 1980. In PRO- CEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA-BIOLOGICAL SCI- ENCES. Vol. 77, 12 Edition, p. 7222.

11. Alexandrov, B. S., V. Gelev, A. R. Bishop, A. Usheva, Figure 3. Observation of time-domain THz induced DNA and K. O. Rasmussen. DNA breathing dynamics in the dynamics, i.e. DNA bubbles induced by the THz-pump. presence of a terahertz field. 2010. PHYSICS LETTERS A. 374 (10): 1214. References
12. Tonouchi, M.. Cutting-edge terahertz technology. 2007.
13. Maniadis, P., B. S. Alexandrov, A. R. Bishop, and K. O. NATURE PHOTONICS. 1 (2): 97. Rasmussen. Feigenbaum cascade of discrete breathers in a model of DNA. 2011. PHYSICAL REVIEW E. 83 (1): -.
14. Brun, M. A., F. Formanek, A. Yasuda, M. Sekine, N. Ando, and Y. Eishii. Terahertz imaging applied to cancer
15. Vafabakhsh, R., and T. Ha. Extreme Bendability of DNA diagnosis. 2010. PHYSICS IN MEDICINE AND BIOLOGY. Less than 100 Base Pairs Long Revealed by Single-Mol- 55 (16): 4615. ecule Cyclization. 2012. SCIENCE. 337 (6098): 1097.
16. Alexandrov, B. S., M. L. Phipps, L. B. Alexandrov, L.
17. Bock, J., Y. Fukuyo, S. Kang, M. L. Phipps, L. B. Alexan- G. Booshehri, A. Erat, J. Zabolotny, C. H. Mielke, H. T. drov, K. O. Rasmussen, A. R. Bishop, E. D. Rosen, J. S. Chen, G. Rodriguez, K. O. Rasmussen, J. S. Martinez, A. Martinez, H. T. Chen, G. Rodriguez, B. S. Alexandrov, R. Bishop, and A. Usheva. Specificity and Heterogene- and A. Usheva. Mammalian Stem Cells Reprogramming ity of Terahertz Radiation Effect on Gene Expression in Response to Terahertz Radiation. 2010. PLOS ONE. 5 in Mouse Mesenchymal Stem Cells. 2013. SCIENTIFIC (12): -. REPORTS. 3: -.
18. Alexandrov, B. S., K. O. Rasmussen, A. R. Bishop, A.
19. Kim, K. T., J. Park, S. J. Jo, S. Jung, O. S. Kwon, G. P. Usheva, L. B. Alexandrov, S. Chong, Y. Dagon, L. G. Gallerano, W. Y. Park, and G. S. Park. High-power fem- Booshehri, C. H. Mielke, M. L. Phipps, J. S. Martinez, tosecond-terahertz pulse induces a wound response in H. T. Chen, and G. Rodriguez. Non-thermal effects of mouse skin. 2013. SCIENTIFIC REPORTS. 3: -. terahertz radiation on gene expression in mouse stem cells. 2011. BIOMEDICAL OPTICS EXPRESS. 2 (9): 2679.
20. Titova, L. V., A. K. Ayesheshim, A. Golubov, R. Rodri- guez-Juarez, R. Woycicki, F. A. Hegmann, and O. Koval- Publications chuk. Intense THz pulses down-regulate genes associ- Alexandrov, B.. DNA Breathing Dynamics, Transcription, ated with skin cancer and psoriasis: a new therapeutic Protein- DNA Binding, and THz Radiation . Invited pre- avenue?. 2013. SCIENTIFIC REPORTS. 3: -. sentation at CINT User Conference, New Mexico, 2013. (Santa Fe, 24-26 September 2013).
21. Fischer, B. M., M. Walther, and P. U. Jepsen. Far-infra- red vibrational modes of DNA components studied by terahertz time-domain spectroscopy. 2002. PHYSICS IN MEDICINE AND BIOLOGY. 47 (21): 3807.
22. URABE, H., and Y. TOMINAGA. LOW-LYING COLLECTIVE MODES OF DNA DOUBLE HELIX BY RAMAN-SPECTROS- COPY. 1982. BIOPOLYMERS. 21 (12): 2477. 438

Page 439

Environmental and Biological Sciences Exploratory Research Final Report Biomarkers for the ‘Signature Wound’ of Iraq and Afghanistan: Traumatic Brain Injury Harshini Mukundan 20130802ER Abstract Background and Research Objectives Traumatic Brain Injury (TBI) afflicts a significant percent- TBI is the signature wound of U.S. soldiers deployed in age of US troops deployed in Iraq and Afghanistan, but is Iraq and Afghanistan, and is broadly defined as any inju- difficult to diagnose using current strategies. The ulti- ry resulting from a blow to the head or penetrative insult mate goal of this project is to develop a rapid biomarker- capable of preventing brain function. While creation of based diagnostic for TBI in blood. However, this cannot mine-resistant ambush protected vehicles has definitely be accomplished until a comprehensive repertoire of saved lives, incidence of TBI associated with rapid head biomarkers secreted during brain injury if established. movement and injury in the proximity of an explosion This requires an integrated biomarker discovery and has greatly increased. The Pentagon estimates that detection approach, directly from human serum and 15,000 troops are afflicted with TBI, whereas the Brain cerebrospinal fluid. The 3-month project reported here Injury Association of America estimates the number at is the first step in that direction wherein we aimed to 360,0001. Part of the reason for this discrepancy is the develop two different methods for the discovery of lack of reliable diagnostics for TBI. TBI is a major con- novel biomarkers of TBI in blood and cerebrospinal cern for not just for deployed troops, but also children fluid, as well as develop assays for two biomarkers on riding bicycles without head protection (~37,000/year), an ultra-sensitive waveguide-based platform that was other accidents and injuries. Even as TBI affects ~ 2 mil- developed at LANL. We were able to evaluate two dif- lion people/year in the U.S alone, there are no reliable ferent methods for biomarker discovery: MALDI mass tests for its diagnosis2. Indeed, the only tests used in spectrometry and 2 dimensional gel electrophoresis in the field are neuropsychological questionnaires such as serum samples. In addition to development of deple- the Automated Neuropsychological Assessment Metrics tion protocols to remove abundant proteins in serum, (ANAM)3. According to the congressional testimony of we were also able to detect spiked TBI biomarkers using Lt. Gen. Dr. E. Schoomaker, results from ANAM are ‘no both methods. However, the results clearly show that better than a coin-flip’, a clear testament to the need for for protein biomarkers, MALDI is much more sensitive better TBI diagnostics4,5. TBI causes a wide spectrum than 2D gel electrophoresis. We also developed assays of pathological outcomes, potentially involving different for two known biomarkers, procalcitonin and cardiolipin, forms of focal and/or diffuse damage, resulting from di- using commercially available antibodies. Procalcitonin rect mechanical impacts to the brain parenchyma. In the was detected using a sandwich immunoassay protocol. human body, TBI induces a complex array of pathological For detection of the amphiphilic cardiolipin that cannot conditions coupled with inflammatory responses. These be measured by conventional immunoassay strategies, responses result in the secretion of several biomarkers we used the novel membrane insertion assay that was that can potentially provide a blood-based reproducible developed by our team. These assays demonstrate our and quantitative diagnostic for TBI. Several potential ability to pull-down and detect biochemically diverse biomarkers have been identified for TBI in blood: lac- categories of biomarkers directly from human serum. tate dehydrogenase, neuron specific enolase, ubiquitin Together, these studies have allowed us to garner signifi- carboxyl terminal esterase-L1, Microtubule-associated cant preliminary data to demonstrate our capabilities to protein-2, Tau protein, NR2 and GluR1 peptides, LRP1, potential sponsors, to further build this project for real- vascular cellular adhesion molecule, IL-6, and cardiolipin. world application. Inspired by the fundamental significance of lipids to the central nervous system, an oxidative lipidomics approach identified biomarkers like cardiolipin, F2-isoprostane, 439

Page 440

thiobutyric acid and malondialdehyde as biomarkers of Scientific Approach and Accomplishments TBI. Despite this, there are no reliable biomarker-based Our team has successfully completed and achieved almost approaches for the diagnosis of TBI. This is because of all proposed objectives within a very short three month many reasons: 1) it is likely that no single biomarker can be interval, as described here. a reliable diagnostic of brain injury in itself, and the use of a combination of biomarkers is necessary; 2) the timeline Documentation and Approvals between the secretion of the biomarker in the cerebro- We have established all requisite approvals for doing this spinal fluid (CSF) following brain injury and its appearance work (IWD, IBC and IRB). We also established contact with in blood has never been studied, 3) most studies have fo- the University of California Brain Bank and requested CSF cused on one biochemical category of biomarkers, namely from 10 individuals (controls and concussion patients). proteins, whereas it is likely that the magic bullet is com- However, this request takes 6-8 weeks to process and pletely varied in composition (sugar or lipid) and finally, unfortunately, we were not able to receive CSF during 4) most of the work has focused on animal models of TBI the course of the reserve proposal. This is the only ob- (especially mice) whereas it is possible that biomarkers in jective we were unable to address during this term. We humans are rather different than those in rodents. performed all method and assay development on serum samples (Randox Laboratories). Research Objectives: For these reasons, the quest for the biomarker(s) that can be the ‘magic bullet’ for diagnosis Best practice methods for sample collection for cardiolipin of TBI remains elusive. The ultimate goal of this proposal and procalcitonin were established in serum. Four different is to identify a comprehensive suite of biomarkers that methods: freeze thaw in liquid nitrogen, filtration, dialysis can provide reliable diagnosis of mild to extreme TBI very and protease inhibitor treatment were tried. Whereas shortly after infection. This reserve proposal is the first the treatments did not affect the stability of procalcitonin step towards achieving that ultimate goal, and the specific positively or adversely, the recovery of cardiolipin was objectives of this reserve are: significantly decreased with filtration, dialysis and protease inhibitor treatment. Thus, rapid freeze thaw in liquid nitro- Core Methodologies for Biomarker Discovery and Detec- gen was determined to be the best method to preserve the tion for TBI: integrity of all biochemical categories of biomarkers. • Establish Institutional Biosafety committee and Institu- Biomarker Discovery by 2D Gel Electrophoresis and Mass tional Review Board approvals to receive human cere- Spectrometry brospinal fluid (CSF) and serum (extension of existing Mass spectrometry and 2D gel electrophoresis methods protocol) form individuals with TBI. We already have a were explored for biomarker discovery in serum. Cardio- functional IBC and IRB for control human serum, and lipin and procalcitonin spiked serum samples were used to can commence that work from the initiation of the demonstrate the feasibility of these methods for biomark- project. er discovery. Because of the abundance of several major proteins (e.g. albumin, -antitrypsin, transferrin, hapto- • Establish best-practice method for sampling and stor- globulin) present in serum, these abundant proteins had age of human CSF and sera to preserve the integrity of to be depleted before any further analysis could be done. three potential biomarkers of TBI, namely cardiolipin, We used an ion-exchange based Abundant Serum Protein pro-calcitonin and microtubule-associated protein. Depletion kit (Norgen Biotek, Ontario, Canada) to deplete • Established methods and limit of detection for pull- these abundant proteins. The sera were first either spiked down and detection of the three biomarkers men- with cardiolipin and procalcitonin at various concentrations tioned above in control CSF and serum using 2D gel and then the spiked samples were processed for abundant electrophoresis, membrane insertion (LANL strategy) protein removal. The samples were then acetone precipi- and mass-spectrometry (DV, 3 months)6,7. tated to concentrate the samples and to remove the eluted buffer. After that, the samples were analyzed by mass Detection of the Biomarkers in Serum and CSF: spectrometry and 2D gel electrophoresis (2DE) methods. • Establish multiplex assays (limit of detection, sensi- For 2DE method (Figure 1), the samples were rehydrated tivity and specificity) for the detection of the three into the appropriate buffer, electro-focused on a 11 cm, biomarker on the waveguide-based biosensor platform pH 3-10 immobilized pH gradient strip (Protean IEF Sys- developed at LANL. tem, BioRad) to separate the molecules according to their • Validate assay performance in buffer vs. complex back- isoelectric point, and then loaded onto 4-16% Tris-HCl ground. SDS-PAGE gels to separate the molecules according to 440

Page 441

their mass size (Criterion Cell, BioRad). The 2DE results with 1% cap-biotinyl-DPPE was injected into this flow are show in Figure 1. By comparing the 2DE images of the cell. After overnight incubation, the flow cell was washed non-depleted human serum gel versus the depleted pro- to yield a supported lipid bilayer (SLB) membrane. The calcitonin spiked human serum gel, we were able to detect membrane was blocked for 1-4 h with 2% BSA/PBS, then the spiked procalcitonin (MW = 12,876 daltons, pI 5.31) rinsed with 0.5% BSA/PBS. Rinsing with 0.5% BSA occurred at a concentration of 443 nM. We could not detect the between all subsequent steps. The flow cell was assembled spiked procalcitonin at lower concentration (87 nM) nor into the plastic holder of that we use to interrogate the could we detect the spiked cardiolipin (23 M) in the serum assay, and the optical methodology was carried out as samples. In addition, we were also able to observe the previously reported . reduction of albumin, antitrypsin, transferrin, and hapto- To analyze for procalcitonin, we utilized a sandwich im- globulin in all the depleted samples. munoassay format on the waveguide-based optical bio- For the mass spectrometry method, the precipitated sam- sensor8. AlexaFluor 647-labeled streptavidin (50 pM) was ple were re-suspended into the appropriate buffer, trypsin added to the flow cell to assess membrane/surface integ- digested to generate small peptide fragments, ran through rity; a strong signal (red trace) indicated that the surface a C18 liquid chromatography column to do a broad separa- was intact. This signal was photobleached, and streptavidin tion of the fragments, and analyzed via a matrix-assisted (10 nM, 5 min) and biotinyl-anti-procalcitonin monoclonal laser desorption/ionization mass (MALDI) spectrometer. antibody (20 nM, 30 min) were added. Fluorescence was With this method we were able to detect the procalcito- again recorded (green trace) to ensure no contamination nin at the lower concentration (87 nM). As with the 2DE with fluorescent species was present. Non-specific binding results, we also observed a reduction in the four abundant was assessed using AlexaFluor 647-labeled anti-procalci- serum proteins in the depleted serum samples when com- tonin polyclonal antibody (AF-pAb, 20 nM, 30 min, purple pared with the non-depleted serum sample. Due to time trace). The non-specific signal was photobleached, and constraints we were unable to test the spiked cardiolipin procalcitonin (1 nM) was incubated at room temperature serum sample with this method. for 4 h. Another aliquot of the AF-pAb was added (30 min), and the signal was recorded (pale blue). We noted a ~20:1 Overall, we have demonstrated some promising results in specific/non-specific binding ratio, indicating successful using these two methods for biomarker discovery in sera. detection of 1 nM procalcitonin, and possible detection of However, improvements can be made to increase the sen- much less (Figure 2). The “corr” (corrected) values indicate sitivity of our methods. It appears that mass spectrometry that the spectra were recorded using a 300 ms integration is more sensitive for the detection of procalcitonin than 2D time rather than the typical 3000 ms integration time due electrophoresis. However, this is a ‘home-brewed’ system to the limited scale of the spectrometeter (4000 counts), that we generated with minimal resources, and we antici- and then corrected by excel. The sensitivity of this assay pate sensitivity enhancement with commercial systems for can be enhanced by longer incubation times and by using the full-fledged effort. Also, 2D is likely to be more efficient a combination of antibodies for the detection (preliminary for proteins with different charge distribution, some of data, not shown here) that exploit the surface avidity ef- which have been suspected to be unregulated during TBI. fect of the binding. Some of the things we can try in the future could be: 1) better depletion of abundant proteins with immune-affini- We also developed a membrane insertion assay for Car- ty columns, 2) better protein separation by using narrower diolipin, which is based on the direct integration of the linear IPG strips (e.g. pH 4-7) or longer length IPG strips molecule to the surface. Again, the sensitivity of this direct and larger SDS PAGE gels, 3) better staining methods (e.g. detection is poor but we suspect this is because the lipidic silver staining), and 4) labeling methods (e.g. iTRAQ) for chain of cardiolipin is truncated in the commercially avail- the mass spectrometry and 5) enhancement of sensitivity able materials, and this integral to membrane insertion. of detection by pre-concentration methods. Cardiolipin is secreted fully in the host, and therefore, we suspect that assay sensitivity will be superior in real-world Detection of the Biomarkers in Serum and CSF samples, and extracted cardiolipin. For all waveguide-based assays, the waveguides and coverslips were freshly cleaned by sonication in chloro- Impact on National Missions form, ethanol, and water (5 min in each solvent), drying Traumatic brain injury is the signature wound of Iraq and under argon, and exposure to UV/ozone. A flow cell was Afghanistan. There are currently no reliable diagnostics constructed using a coverslip and waveguide sandwiching for mild Traumatic Brain Injury, and it remains a very high a 0.5 mm gasket with a flow channel cut into the center. A priority for the Department of Defense, as is evidenced by solution containing sonicated unilamellar vesicles of DOPC 441

Page 442

the congressional testimony of the Surgeon General of the USA in 2012. This work addresses this key gap and has de- veloped key capabilities in this area, which are now being explored for further development using external sponsors. Figure 1. 2DE images of non-depleted human serum (A) and depleted and procalcitonin Fisgpuikreed h1u.m 2anD sEer uimm a(Bg).e s of non-depleted human serum (A) and depleted and procalcitonin spiked human serum (B). 8000 7000 6000 5000 4000 3000 2000 1000 0 550 600 650 700 750 800 850 900 Figure 2. Detection of procalcitonin using the waveguide-based biosensor platform. The specific signal (pale blue) is 20X greater than the non-specific signal associated with the antibody (purple trace). The red line represents binding of AF647 labeled streptavidin to the biotinylated surface, proving membrane integrity. The green line is a measure of the waveguide-associated background. The data are corrected for use of a 10X attenuation filter (see text). 442 )UA( ytisnetnI ecnecseroulF ASAVII160 BG 50 pM AF‐SA corr RGTs BG NSB SB corr Wavelength (nm) Figure 2. Detection of procalcitonin using the waveguide-based biosensor platform. The specific signal (pale blue) is 20X greater than the non-specific signal associated with the antibody (purple trace). The red line represents binding of AF647 labeled streptavidin to the biotinylated surface, proving membrane integrity. The green line is a measure of the waveguide- associated background. The data are corrected for use of a 10X attenuation filter (see text). Publications Anderson, A. S., D. Vu, and H. Mukundan. Biomarkers for Traumatic Brain Injury. To appear in Japan Society of Applied Physics and Materials Research Society, Joint Conference. (Kyoto, Japan, 17-20 Sep. 2013).

Page 443

Environmental and Biological Sciences Postdoctoral Research and Development Continuing Project Determining Physiological Predictors of Climate-driven Forest Mortality Nathan G. Mcdowell 20110756PRD4 Introduction lection for a meta-analysis of forest mortality patterns. We propose to determine which, if any, common This collection is being done via networking with a large physiological responses during heat and drought-driven number of scientists. More than two-thirds of the data mortality exist at a global scale, and thus determine the is now in Dr Adam’s possession. Lastly, Dr Adams is simplest but strongest predictors of tree mortality across working closely with our modelers to learn from them different species of trees. The experiments to determine the key aspects for modeling mortality and to feed them these common predictors will be conducted on trees of information regarding how to model tree death. This is a multiple scales; from seedlings to >100 year old trees. component of a follow on paper that Dr Adams is work- Through the process, we will identify the final steps in ing on. Dr. Adams is well on his way to a very productive mortality and survival, allowing modelers to simplify not career at Los Alamos. only their prediction of stress but of the key aspects of plant physiology that partition plants between survival Future Work and death during drought. The final product(s) will Dr. Adams will continue to collaborate with EES staff, identify the simplest, most common, and statistically post-docs, and students who are working on the Los strongest environmental and physiological predictors of Alamos Survival Mortality (SUMO) project to collect data mortality, and interpretation of their mechanisms. on key physiological responses to drought through mor- tality in pinyon pine and juniper trees. Dr. Adams will Benefit to National Security Missions continue to lead two sampling efforts at SUMO: 1.) to The goal of this research will be to develop a bridge monitor stressed trees for insect attack and characterize between physiology and modeling by identifying the tree defensive capabilities against insects at this experi- simplest level of detail needed to predict tree drought ment; 2.) repeated measurement of inherent hydraulic mortality across multiple forest types. This work directly vulnerability to assess if functional tree water stress supports DOE’s Climate and Environmental Science Divi- relationships change with drought and increased tem- sion goals to understand the role of terrestrial ecosys- perature. Dr. Adams will continue a SUMO supplemental tems in a changing climate and is intended to promote experiment initiated in FY13 aimed at distinguishing the long-term programmatic growth in this arena. Specifical- physiological threshold of survival and mortality us- ly the products of this research will be directly tailored ing transplanted trees in the greenhouse at TA-51. The for incorporation into DOE’s Community Land Model. protocol of this experiment will be to drought stress sets Addressing this key gap for coupled vegetation-climate of trees to hypothesized thresholds of survival, and then models could lead to the rapid development of more re-water these trees, measuring physiological responses accurate global change projections that are urgently and critically, whether the trees ultimately die or survive needed to inform sound climate policy at a global level. when re-watered. This experiment is aimed at elucidat- ing the simplest key thresholds between tree mortality and survival that can be incorporated into mechanistic Progress and predictive models of tree mortality. These data will In the past 12 months Dr Adams has been amazingly inform ongoing modeling efforts by EES staff to address productive. He has collected all of the data for at least DOE’s Climate and Environmental Science Division goals two manuscripts from a field experiment at LANL exam- to understand the role of terrestrial ecosystems in a ining forest mortality, which he is now putting together changing climate. Specifically, data from this experiment into manuscript format. He has also initiated data col- 443

Page 444

will be used to parameterize tree mortality routines in DOE’s Community Land Model. Conclusion Dr. Adams will develop a bridge between physiology and modeling by identifying the simplest level of detail needed to predict tree drought mortality across forest types. This work supports DOE’s Climate and Environmental Science Division goals to understand the role of terrestrial ecosys- tems in a changing climate. Specifically the products of this research will be directly tailored for incorporation into DOE’s Community Land Model. Addressing this key gap for coupled vegetation-climate models could lead to the rapid development of more accurate global change projections that are urgently needed to inform sound climate policy at a global level. Publications Adams, H., and e. al. Empirical and process-based approaches to climate-induced forest mortality models. To appear in Frontiers in Plant Science. Adams, T., and e. al. Nonstructural leaf carbohydrate dynamics of Pinus edulis during drought-induced tree mortality reveal role for carbon metabolism in mortality mechanism. 2013. New Phytologist. : 1142. Breshears, D., H. Adams, and e. al. The critical amplifying role of increasing atmospheric moisture demand on tree mortality and associated regional die-off. . 2013. Frontiers in Plant Science. 4: 266. Franz, T., and e. al. Ecosystem-scale measurements of biomass water using cosmic ray neutrons. . 2013. Geophysical Research Letters. : 3929. Zeppel, M., and e. al. Forest mortality due to drought: latest insights, evidence and unresolved questions on physiological pathways and consequences of tree death. 2013. New Phytologist. : 372. 444

Page 445

Environmental and Biological Sciences Postdoctoral Research and Development Continuing Project Catalytic Mechanism and Inhibition of Metallo-beta-lactamases (MBL), the Ultimate Threat Against Antibiotics Suzanne Z. Fisher 20120776PRD4 Introduction ferent metallo-beta-lactamases, BcII and NDM. These The proposed work will utilize an innovative research seems to express well and having enough protein is the strategy that is LANL-unique for novel antibiotics. We first crucial step for neutron studies. We were also able will combine neutron and X-ray diffraction to obtain high to crystallize one of these enzymes, BcII, at pH 7, which resolution structures and detailed information on how is more physiologically relevant than other reported metallo beta-lactamases (MBLs) work to degrade clini- structures at pH 4.5 and 6. cally used antibiotics. Due to their readiness for horizon- tal transfer and evolution, MBLs entail a serious public For NDM, the enzyme found in deleterious antibiotic-re- health threat, with increasing antibiotic resistance found sistant pathogens, we designed and produced two dou- all over the world. Results will have direct implications ble mutants that is unable to bind the second catalytic for not only understanding how these enzymes work, zinc in the active site. Instead we have populated these but also to enable the design of clinically useful antibi- sites with alkaline earth metals magnesium and cal- otics not susceptible to breakdown by these harmful cium. This allows us to bind antibiotics in the active site, enzymes. but stops hydrolysis and allows us to study the binding mechanism. We are also optimizing conditions for large crystal growth for neutron studies of these variants. Benefit to National Security Missions Metallo betal lactamases comprise a serious human To study the basic binding interactions and to investigate health issue due to the innumerable complications asso- new possible lead compounds, we have co-crystallized ciated with untreatable infections caused by some of the sulfonamide drugs with both BcII and NDM. We are most dangerous bacteria present in clinical settings and working on determining X-ray crystal structure to con- intensive care units. This work, which can only be per- firm binding. In progress also is co-crystal structures formed at LANL’s world-class facilities, will provide the of BcII and NDM in complex with Aztreonam, the only first neutron structures for antibiotic resistance enzymes known antibiotic that is still able to resist hydrolysis by found in multidrug resistant pathogens. This work will these enzymes. advance our basic understanding of enzyme mechanisms and drug binding. Neutron diffraction is underutilized We are ready to progress to the next phase, the in vitro for drug design and is the only technique that can give evolution of NDM to determine which active site muta- unique data on drug binding and degradation. This work tions confers the ability to hydrolyze Aztreonam. This will may lead to novel compounds that will have an enor- deliver an early detection system for real world diag- mous impact on human health and disease. nostic application. We are working on permission and authorization to perform this work in a Biosafety Level 2 We also envision a scenario where detection methods lab. can be developed to find bacteria expressing resistance genes in the environment. This work directly addresses Future Work mission relevance in the fields of basic health research, bioscience, and pathogen detection. For the second year, we are working towards determin- ing X-ray and neutron structures of BcII and/or NDM-1 (wild type and mutants) with different active site metals. Progress We have designed the appropriate expression plasmids Javier was able to obtain high quality clones of two dif- 445

Page 446

and have made enzymes with non-native metals in the active site. This will allow binding of antibiotics, but will prevent hydrolysis allowing us to study the binding interac- tions with the enzyme. We will also pursue complexes with existing antibiotics to investigate binding and the enzyme residues that are involved with catalysis. Neutron crystallography, which will shed light on the H-bond networks in the active site, will facilitate this detailed investigation of drug binding and involvement of water activation. We are pursuing sulfonamide-based drugs as lead com- pounds and are making complexes between these drugs and BcII and/or NDM-1. Sulfonamides are known to bind to other Zn-containing enzymes and we will pursue them here also. Conclusion We will characterize the enzyme NDM-1: (1) Water activa- tion, (2) Substrate binding, (3) In vitro evolution. (1) we will solve the structure of NDM-1 through neutron crystal- lography, which will shed light on the H-bond networks in the active site. (2) we will engineer NDM-1 to enable binding alkali earth metals in the second Zn(II) site. This will eliminate enzyme activity while retaining substrate binding for structural analysis. (3) we will construct NDM- 1 mutant libraries, which will be screened for resistance to monobactams. Mutants will be selected and analyzed to understand the structural traits of enzymes with broad substrate preferences. 446

Page 447

Environmental and Biological Sciences Postdoctoral Research and Development Continuing Project Single Cell Genomics for Better Control of Plant Pathogens Shunsheng Han 20130779PRD1 Introduction mutants will be isolated by both visual observation The proposed research will use the rhizosphere-coloniz- and flow cytometry sorting. ing bacterium Pseudomonas chlororaphis strain 30-84. • Bacterial genome sequencing will be performed in The strain is able to inhibit fungal pathogens and has Dr. Cliff Han’s lab using the cutting edge sequencing become a model for a beneficial commensal bacterium. platform and gel microdroplet culturing technology. This strain demonstrates phenotypic variation resulting Mutational events will be identified by comparing to from spontaneous mutations. The goal of this research is a published wild type P. chlororaphis genome. to identify signature genomic and transcriptomic chang- • To understand the roles of these genomic changes in es in response to environmental signals and investigate bacterial persistence, mutant strains will be com- how these changes benefit the wild type populations pared with the wild type to determine their ability to using a plant associated bacterial as example. control fungal pathogens and form biofilms. Benefit to National Security Missions • We will also perform transcriptomic analysis to identify differentially regulated genes in the mutants The understanding of the relationship between plant compared with the wild type. Selected target genes and bacteria is directly linked to the DOE mission in the will be further investigated using functional genom- area of bioenergy. LANL, supported by DOE, has invested ics tools such as gene deletion and over-expression. significantly in studying plant metabolism and algal biofuel. This work will enrich our portfolio in these areas and bridge the gaps between environmental microbiol- These studies will provide a comprehensive view of bac- ogy and plant study. terial genomic changes/cell differentiation during host association. The results will aid in designing new strate- Progress gies in plant disease control. The postdoc fellow Dongping Wang started on June 10, 2013. For the last 17 days, he has completed many re- Future Work quired training courses including the General employee We will perform the following experiments to fulfill the training. Many other trainings such as biosafety and aims of the proposed research: laser training have been scheduled in July and August. Some experimental materials have arrived at the lab. • To identify mutational events during P. chlororaphis Bacterial strains will be shipped from Texas A&M Univer- colonization, bacterial cells will be inoculated on sity once they are approved by the Institutional Biosafety plant roots. Flow cytometry will be used to deter- Committee at Los Alamos National Laboratory. mine the bacterial size and sort them into different groups. We made an experimental plan for next fiscal year and • Single cell genome will be sequenced using gel mi- will perform the following experiments to fulfill the aims crodroplet culturing in Dr. Cliff Han’s group. Benefi- of the proposed research: cial mutations will be identified by comparing with a published wild type genome sequence. These results • To identify mutational events during P. chlorora- will provide us with a complete set of mutations that phis rhizosphere colonization, bacterial cells will be are involved in host association. inoculated on plant roots. Bacterial spontaneous 447

Page 448

• To study the role of these genomics changes in bacte- rial persistence, beneficial mutants will be compared with the wild type strain in their ability to produce fungal-inhibiting metabolites and form biofilms. • Genomic mutations may lead to changes in gene ex- pression. To understand the beneficial mutation at the transcriptome level, RNA sequencing will be conducted to identify differentially expressed genes between wild type and beneficial mutant strains. Together, these data will give us a comprehensive understanding of cell differentiations in response to environmental signals. • The above analysis will identify candidate genes in- volved in disease control. Selected genes will be inves- tigated using functional genomic approaches. Conclusion We expect to identify mutational events during P. chloro- raphis colonization, to classify and identify the beneficial mutation that are involved in host association. We will also identify the role of these genomic changes in bacterial per- sistence. Understanding of molecular processes involved in bacterial adaptation could lead to better disease manage- ment strategies. Publications Wang, D., C. Han, C. Lo, A. Dichosa, P. Chain, J. M. Yu, L. S. Pierson, and E. A. Pierson. Genomic adaptations of a small colony variant (SCV) in the biofilm of Pseudomonas chloro- raphis 30-84. Genome Biology and Evolution. 448

Page 449

Environmental and Biological Sciences Postdoctoral Research and Development Continuing Project Joint Inversions of Seismic and Gravity Data in Volcanic Areas to Advance Hazards Assessment: A Focus on the Alaskan Subduction Zone and Kilauea, Hawaii Monica Maceira 20130807PRD3 Introduction of the decollement be imaged, constraining the maxi- Three-dimensional passive-source seismic velocity mum magnitude of earthquake this fault can produce?” tomography is a powerful technique that utilizes energy produced by earthquakes to image potentially complex Benefit to National Security Missions subsurface structures, such as magma distributions The proposed research will enhance the capability of the beneath volcanoes or subducting slabs. However, the Laboratory through unique expertise in imaging Earth resolution of these studies are generally limited by the structure. This expertise has applications to Nonpro- natural distribution of earthquakes. While active sources liferation R&D under Nuclear Nonproliferation, where can expand coverage, they are generally only used to accurate Earth models are needed to locate, identify and image the upper crust and are costly. Gravity data, how- determine yield for seismic events of interest such as ever, provide alternate sources of information regarding underground nuclear explosions. High-resolution Earth subsurface structures, and are available at a variety of models are also required for characterization and moni- scales. Due to the inherent relationship between den- toring within several other areas of LANL mission space sity and seismic velocity, gravity and seismic data can including geothermal energy development (Applied be jointly inverted, providing an understanding of Earth Energy Programs Office) carbon sequestration (Fossil structures that is not limited by the distributions of Energy, within the Office of Science Programs), and used seismicity and has stronger constraints on the tempera- fuel disposition and salt repository science (both within ture, compositional, fluid, and magmatic distributions in the office of Civilian Nuclear Programs). a study area. Progress The current distribution of seismic stations in Alaska is This project had just started at the time of this report; focused near the relatively two-dimensional volcanic arc therefore, there is no progress to date. located above the subducting oceanic plate, limiting the ability to resolve its seismic structure. The inclusion of Future Work gravity data in a joint inversion will markedly increase I will simultaneously invert two different data types to the ability to resolve these features in three dimensions. derive the Earth’s 3D velocity structure beneath the I will address questions regarding the sources of volca- Alaska subduction zone and the Kilauea volcano, Hawaii. nism in the Alaska subduction zone. Seismic data will come from local, regional, and national networks and deployments, while gravity data will come Kilauea Volcano presents a unique opportunity to study from US satellite missions as well as in-land measure- one of Earth’s most active volcanoes, in a location that ments. The research paths will be as follows: (1) develop has produced historical tsunamigenic earthquakes. code for efficient and robust multi-parameter inversion, While previous seismic studies have focused on studying (2) gather and pre-process the different data types, and the near-surface beneath the summit caldera, portions (3) perform multi-parameter inversion to produce a vali- of Kilauea’s rift zones, or the seismically active tsunami- dated high resolution high accuracy 3-D velocity image genic decollement fault beneath the volcanic material of the Alaska subduction zone and the Kilauea volcano, and the underlying oceanic crust, the extents of these Hawaii. studies are limited by the natural seismic distributions. By combining gravity and seismic data, I will address The first year, focus will be on code development and questions such as “Can the aseismic and seismic portions 449

Page 450

the Kilauea volcano, Hawaii, where seismic and gravity data are readily available. The expansion of USArray seis- mic stations to Alaska in the coming year will allow seismic models to be significantly improved during the second year of this project. Conclusion The expected results from the research are: I. Develop- ment of the next-generation method for imaging the 3-D structure of the Earth and, II. A high-resolution high-accu- racy 3-D velocity im-age of the crust and upper mantle be- neath the Alaska subduction zone and the Kilauea volcano. I expect to submit two papers to high-impact journals, one for each of the regions of interest, and addressing the scientific questions stated in the abstract. 450

Page 451

Environmental and Biological Sciences Postdoctoral Research and Development Final Report Analysis of Protein Structure-Function Relations in Antibiotic Biosynthesis and Signal Transducing Receptors Louis A. Silks III 20100603PRD2 Abstract standing of complex self-assemblies and the identifica- The vast majority of known antibiotic, chemotherapeu- tion of novel drug leads. tic, anti-viral, anti-inflammatory, analgesic, and immu- Background and Research Objectives nosuppressant agents are natural products. However, only a marginal number of bioactive compounds from The discovery and first direct correlation of antibacterial diverse sources of natural products have been explored properties with a natural isolate from Penicillium chrys- even though they provide an incredible potential for the ogenum in 1928 founded modern medicine. At the time development of new drugs. Natural products or second- of discovery, neither the chemical structure, biochemi- ary metabolites are compounds with diverse bioactivity cal process of activity or the biosynthetic route used by that are isolated from natural sources and are biosyn- the producing fungus was understood. It took over 60 thesized by large multi-functional enzymatic clusters years to elucidate the biosynthesis of L-aminoadipyl-L- in diverse bacteria, fungi and plants and represent the cysteinyl-D-valine (LLD-ACV), the penicillin and cephalo- major source for all drugs and drug leads available to sporin precursor. The basic genetic understanding of this humans. Genetic and enzymatic studies provided insight biosynthetic system has been applied to discover many of isolated enzymatic functions, but have not led to more antibacterial compounds. But reliance on similar strategies for the discovery of novel natural products. biosynthetic mechanisms has resulted in the isolation Bioinformatics and structural biology studies are giving of chemically similar compounds often as -lactams and rise to three-dimensional folding and selective protein- lactones or of other drugs with similar mechanisms of protein recognition processes of key proteins involved in action and cellular targets. This similarity of current anti- the biosynthesis of medically relevant natural products. bacterial compounds contributed to the recent, acceler- This knowledge is the base to understand the spontane- ating emergence of antibiotic resistance and significantly ous self-assembly of multifunctional protein clusters that reduced the availability of effective antibiotic treatment biosynthesize natural products. Once understood, these options. Bacterial infections have once more become the biosynthetic complexes can then be engineered to give third leading cause of death in United States, a major rise to produce new classes of highly specific and more public health threats, and our ability to fight bacterial potent drugs with fewer side effects or novel assembly infections is increasingly problematic. Consequently, an lines can be efficiently identified in nature, their prod- in-depth understanding of the biosynthetic machineries ucts predicted and novel drug leads identified. and pathways that assemble chemical structural features in vivo are critical to the discovery of new biosynthetic To elucidate structures and structural complexes, Dr. clusters that may include not-yet described processing Koglin spent part of his time making stable isotope- steps to produce novel secondary metabolites. enriched proteins for structural studies. Expertise in biochemistry, synthetic chemistry and biophysics enables Biosynthetic processing by Non-Ribosomal Peptide Syn- the design of labeled amino acids and to place them spe- thetase (NRPS) multi-functional enzymatic complexes is cifically into proteins improving the structural resolution accomplished as a series of covalently tethered thioes- of very large biological systems. Proteins of interest are ters similar to fatty acid and polyketide biosynthesis. studied by Nuclear Magnetic Resonance (NMR) spectros- In each chain-elongating condensation, the growing copy, which is the only method to gain understanding of NRP acyl chain participates as an electrophilic partner, the structure and functional dynamic of these proteins. undergoing attacks at the acyl thioester carbonyl while This gained knowledge translates to a detailed under- the monomer, to be incorporated, acts as a nucleophile. 451

Page 452

For NRP chain elongation, the downstream nucleophile The aim of this project is (1) to develop methods that allow is the amine of an aminoacyl-S-peptidyl carrier protein high-resolution structural studies on very large proteins (PCP) with which it forms amide or peptide backbone link- and protein assemblies and (2) further the structural and ages. Carrier domains in general require post-translational functional elucidation of enzymes and assemblies of those modification with a phosphopantetheine (4’-PP) prosthetic biosynthetic systems as a crucial requirement for (3) the group whose terminal thiol function (HS-4’-PP) becomes directed identification of novel drug leads. This research the site of covalent tethering of the growing acyl chains. leads to understand the underlying logic of protein cluster They are often referred to as thiolation or T domains to assembly in general and to reveal the mechanism of natu- emphasize that functionality. NRPS clusters occur either as ral product biosynthesis. distributed assemblies, each catalytic domain on a sepa- The overarching goal is to confront the accelerating devel- rate protein (Type II), or catalytic and carrier domains that opment of antibiotic resistance. Dr. Koglin focuses on the are strung together into one or more functional modules biosynthesis of new natural products with novel chemical within a single polypeptide chain. These multi-modular properties (chemotypes). Those have the potential to be assembly lines are termed Type I. The most common developed into highly effective antibacterial treatments organization of NRP synthetase assembly lines is the Type while being immune to current resistance mechanisms. I pattern, exemplified by aminoadipyl-cysteinyl-valine (ACV) synthetase. The ACV assembly line contains a three- Scientific Approach and Accomplishments module, ten-domain, 480-KDa single subunit polypeptide and represents the first step in penicillin and cephalospo- Dr. Koglin has solved the largest protein structure in solu- rin biosyntheses. In Type I NRP synthetases, a T domain tion by NMR spectroscopy, using his own technologies. is embedded in each module, servicing in cis (domains This record achievement has held for several years. Now, of one protein) adjacent catalytic domains. The chemical Dr. Koglin has applied novel methods of combining selec- logic and enzymatic activity of isolated functions of natural tive isotopic enrichment and NMR spectroscopy at Los product assembly lines are now fairly well deciphered. This Alamos National Laboratory to further push the current logic includes: (1) enzymatic priming of the apo form of a T technology and enabled larger biological macromolecular domain by a phosphopantetheinyl transferase (PPTase) to assemblies to surrender their structures in high resolu- install the active thiol function and the 4’-PP cofactor on a tion. The results of the successfully completed project are carrier protein so that covalent acyl/peptidyl chain growth now applied to build strong collaborations with Harvard can proceed; (2) selection and ATP-dependent activation Medical School and Massachusetts Institute of Technology of monomeric building blocks to allow chain initiation at the Francis Bitter Magnet Laboratory. Scientists of Los catalyzed by adenylation domains (A domains); (3) chain Alamos National Laboratory are a key contributor in efforts elongation that is mechanistically coupled to translocation to study large protein assemblies and membrane protein of the activated substrates leading to amide or peptide structures, dynamics and functions. Dr. Koglin has obtained bond condensation in the C domain; and finally, (4) termi- the first structure of a C-domain by NMR spectroscopy nation of the process which is performed by a thioesterase (Figure 1). This is an important step to establish methods (TE) domain as the most downstream enzymatic function. to study structures and dynamic of very large proteins TE domains may be hydrolases, reductases, or regiospecific and to gain insight into essential biosynthetic processes of macrocycle-forming catalysts and they act to release the bioactive compounds, which directly impacts the ability nascent product from the assembly line. A vast variety of to identify effective antibiotics. So-called condensation (C) tailoring enzymes can modify the peptide chain. By gaining domains perform an essential and generally common step insights into the genetic organization, biosynthetic logic, in natural product biosynthesis. They catalyze the peptide and structural recognition of these assembly lines, we can bond formation in all compounds that are derived from characterize the catalytic processes. non-ribosomal peptides. The high-resolution NMR struc- ture is again the largest structure of its kind and studies Still, very little is known about protein-protein interactions, of enzymatic active protein complexes provide a detailed substrate recognition processes, or the assembly order of understanding of their function and substrate recognition isolated enzymes and modules that define the biosynthesis processes for the first time. Despite a previously published of secondary metabolites. This lack of fundamental knowl- complex enzymatic mechanism, we found that C domains edge makes it extremely difficult to predict the structure are allosteric controlled and their structural dynamic of secondary metabolites produced by large mega-enzyme limits space within the active site and the peptide bond is clusters, such as antibiotics, solely based on their genetic formed by chance between two activated amino acid resi- encoding. The current development strategy needs to be dues. Since the peptide bond has a lower energy compared re-imagined to identify new and effective drug candidates. to thioester activated amino acid residues, this process 452

Page 453

is not reversible. This finding is absolutely crucial for the II NRPS activation-and-carrier di-domain, this assembly biotechnological design of systems synthesizing novel line contains a short type I NRPS consistent with a sec- engineered drug leads, since no complex catalytic cascade ond activation domain, one condensation domain for the must be reconstructed for peptide bond biosynthetic synthesis of a cytosolic di-peptide, a thioesterase domain, steps. In addition this knowledge and methods enable (1) and a C-terminal membrane-inserted export protein. An to block the biosynthesis of bacterial and fungal toxins, (2) associated cytoplasmic glycosyl-transferase is responsible to understand a family of proteins in the development of for decorating and bridging the short peptide during the antibiotic resistance, (3) and to understand the drug niacin precursor biosynthesis. After hydrolysis off the assembly and its binding to what is called GPCR. This could give rise due to the thioesterase and presumed N-C bond forma- to alternative treatment of type II diabetes. tion at the anomeric center of the first sugar moiety, the C-terminal membrane transport domain immediately Dr. Koglin further applied the gained knowledge of struc- exports the precursor. The precursor is never released into tures and protein complexes to identify and initiated the cytoplasm and its processing is continued within the the characterization of a complete new class of natural periplasmic space by a second glycosyl-transferase to form products with antibacterial properties and developed a a disaccharide and a copper-amine-oxidase responsible for new approach to determine the cellular target of a novel converting an amine into an aldehyde function. Both are compound or any given drug. Each of the identified novel membrane-bound and located in the periplasmic space, natural products demonstrates confirmed broad-spectrum outside the cytoplasm. As a consequence, the functional antibacterial activity and neither have been described organization of this assembly line spans the cytoplasm in the literature or deposited in any database. These membrane and the biosynthesis of the identified com- glycopeptides are produced by Non-Ribosomal Peptide pounds is subdivided. While precursors of the matured Synthetase (NRPS) assembly lines from divergent species. products are synthesized inside the cell, the final assembly However, they show structural similarity and possess a of the bioactive secondary metabolites is located outside novel common chemotype. Preliminary antibacterial ef- the cell. Since we could not identify an encoded resistance ficacy results and promising toxicology studies, validated mechanism in the genome of the producing strain, this externally, suggest that we have identified a new class of suggests that the bacterial toxicity profile of this second- antibiotic compounds that act on an entirely new bacterial ary metabolite requires that the compound be not to be target. We further validated antibacterial efficacy, includ- released inside the cell. We assume that the host itself is ing against multi-drug resistant Staphylococcus aureus and susceptible to its own metabolite based on the organiza- Clostridium difficile. We initiated cytotoxic profiling on hu- tion of the biosynthetic pathway that assembles cytoplas- man cell-lines and small animals. The studied compounds mic precursors in the periplasm. This also suggests that the show very low toxicity and single doses of up to 80mg/kg evolution of antibiotic resistance against this compound body weight are tolerated. This offers a wide therapeutic may not be easily accomplished. We discovered the first window. We elucidated the cellular target and currently and never before described split or fragmented biosynthe- establish the mechanism of action. As it turned out, these sis of a secondary metabolite. novel antibacterial drug candidates inhibit the assembly of the bacterial ribosome on mRNA and effectively prevent Based on our initial results, we identified a second example the biosynthesis of proteins in bacteria. of a similar biosynthetic complex that incorporates a mem- brane-transport domain and extra-cellular maturation of a Initially, we have discovered a unique non-ribosomal secondary metabolite. This second example is also involv- peptide synthetase (NRPS) cluster for the biosynthesis of ing a second glycosyl-transferase and an oxidative enzyme an unusual, novel secondary metabolite in the sequenced system to generate an aldehyde function. Compared to the genome of a hyper-thermophile anaerobic bacterial strain. identified first example, isolated from a hyper-thermophilic We purified the predicted secondary metabolite from the we found this second example in the genome of an aerobic cultured producer strain and identified all the enzymatic cold lake bacterium. The living environments or ecological functions of the proteins associated with this novel NRPS niches of both producer strains could not be more differ- assembly line by in vitro assays. For this first example, we ent. Recently, we identified and purified additional three have identified the chemical structure as a glycopeptide and novel natural products, beside the two described and identified its biosynthetic pathway (Figure 2). This examples. We elucidated their chemical structures by NMR type of secondary metabolite is highly unusual. We have spectroscopy and continue to develop purification meth- further linked the novel and unique enzymatic func- ods to produce pure material for antibacterial efficacy and tions of the biosynthetic pathway to the structure of the cytotoxic assays. isolated compound. In addition to a freestanding type 453

Page 454

Impact on National Missions Publications This project supports the DOE mission in biological sci- Bowers, A. A., M. G. Acker, A. Koglin, and C. T. Walsh. Ma- ences and biosecurity by enhancing our understanding of nipulation of thiocillin variants by prepeptide gene replacement: structure, conformation, and activity of how cellular processes work in tandem to form natural heterocycle substitution mutants. 2010. Journal of the products of high biomedical value. We build capabilities for American Chemical Society. 132 (21): 7519. every agency that has a health mission, including HHS, DoD and DHS. Frueh, D. P., A. Leed, H. Arthanari, A. Koglin, C. T. Walsh, and G. Wagner. Time-shared HSQC-NOESY for accurate distance constraints measured at high-field in (15) N-(13)C-ILV methyl labeled proteins. 2009. Journal of Biomolecular NMR. 45 (3): 311. Frueh, D. P., H. Arthanari, A. Koglin, C. T. Walsh, and G. Wagner. A double TROSY hNCAnH experiment for ef- ficient assignment of large and challenging proteins. 2009. Journal of the American Chemical Society. 131 (36): 12880. Frueh, D. P., S. Nichols, S. Mishra, H. Arthanari, A. Koglin, C. T. Walsh, and G. Wagner. Transient Domain Interac- tions in Non-Ribosomal Peptide Synthetases . 2011. BIOPHYSICAL JOURNAL. 100 (3): 17. Frueh, D. P., S. Nichols, S. Mishra, H. Arthanari, A. Koglin, Figure 1. High-resolution structure of a condensation domain. C. T. Walsh, and G. Wagner. Transient Domain Interac- The left image represents the structural bundle of the NMR tions in Non-Ribosomal Peptide Synthetases. 2011. structure calculation selected by lowest energy values. The RMSD Biophysical Journal. 100 (3): 17. is 1.2Angstrom and the structure is shown from N-terminus in blue to C-terminus colored red. At this resolution, dynamic and Koglin, A., V. Doetsch, and F. Bernhard. Molecular engi- functional studies reveal the selective interaction and substrate neering aspects for the production of new and modi- recognition processes. The right panel shows the result of fied biosurfactants. 2010. Advances in Experimental substrate recognition and binding shown in blue. Single amino Medicine and Biology. 672: 158. acid residues responsible for the allosteric controlled structural dynamic causing the ‘grinding’ movement of the upper and lower half of the enzyme (red arrows in left panel) is shown in red. Figure 2. Biosynthetic assemblies of novel natural products. The graphic presentation of first discovered split assembly line (labeled ‘1’) that produces novel antibacterial compound reveals its uniqueness. While precursors are assembled in the cytoplasm, the final processing and maturation is located in the periplasm. Number ‘2’ shows the graphic presentation of the biosynthetic cluster of the second identified and characterized natural prod- uct. Though the biosynthesis is entirely different, the produced compounds are chemically very similar as shown in the two 13C-HSQC NMR spectra. The table represents a summary of the antibacterial efficacy studies for both compounds. With increased purity, the efficacy improved dramatically including against hu- man pathogenic bacteria. 454

Page 455

Environmental and Biological Sciences Postdoctoral Research and Development Final Report Quantitative Modeling of Cellular Noise Michael E. Wall 20100604PRD2 Abstract putational realms and to define an integrated research We analyzed single-cell experimental data on various strategy in which these tools can be systematically com- natural and synthetic gene regulatory systems. In bacte- bined. ria we analyzed flow cytometry data to predict the be- Analyzing Single-Cell Fluctuations. A system of ordinary havior of genetic parts for synthetic biology applications differential equations, known as the master equation, such as biofuel production. In yeast we analyzed single is often used to describe the variability measured in molecule fluorescence in situ hybridization (smFISH) biological systems at the level of single cells. Unfortu- data to predict the response of cells to osmotic stress. nately, the master equation is difficult to solve for most In mammalian cell cultures we developed methods for systems. Prior to this project, Dr. Munsky invented a measuring single-cell levels of small RNAs and used fast way of solving the master equation. His technique, them to analyze the response of these cells to pathogen known as the Finite State Projection (FSP) algorithm, exposure. We gained important new insights into each enables computational researchers to ignore unimport- of these systems and showed that distributions of single- ant contributions to the master equation. By improving cell behavior for diverse cell types can be quantitatively the efficiency and accuracy of stochastic analyses, his predicted with a very high level of accuracy. This work method set the stage for a large impact to the field. A led to publication of a book chapter, an article in Nature sub-goal of this project was to further improve these Biotechnology, and two articles in Science on the iden- tools for more efficient analysis of stochasticity in bio- tification of predictive models for gene regulation. The logical processes. methods have been implemented in computer software that the community can use for similar studies. The Integrating Experimental and Computational Analyses. results of this project will help to transform biology by Another goal of this project was to developed approach- enabling researchers to add an entirely new dimension es to integrate computational FSP analyses with quanti- to understanding predicting and controlling complex cel- tative single-cell experimental analyses. Combining dis- lular behavior. crete stochastic analysis with single-cell/single-molecule measurements can reveal features of gene regulation Background and Research Objectives that are hidden by bulk assays or deterministic analyses Even genetically identical cells in identical environments (Figure 1). To study single-cell responses, today’s gold exhibit wildly different phenotypical behaviors due to standards are flow cytometry, which enables researchers cellular fluctuations known as gene expression ‘noise’. to quantify protein expression in thousands of cells per Previously, such noise was considered a nuisance that second, and single-molecule fluorescent in situ hybrid- compromised cellular responses, complicated modeling, ization (smFISH), which enables direct counting of mRNA and made predictive understanding all but impossible. transcripts in individual cells. At the start of this project Many past studies focused on how cellular processes re- LANL had existing expertise in flow cytometry, led by Dr. move or exploit noise to a cell’s advantage. From these, Munsky’s cosponsor (Babetta Marrone, B-DO). Expertise it has become clear that different cellular mechanisms in smFISH has been developed at LANL during the course affect these cellular fluctuations in different ways. As of this project through the recruitment of D. Shepherd, a consequence, it is now clear that cellular fluctuations a postdoc co-advised by Drs. Munsky and Werner. Both may actually contain valuable information about under- flow cytometry and smFISH enables building distribu- lying cellular mechanisms. Finding and exploiting this tions of cellular protein or RNA content, respectively information requires a strong integration of single-cell/ (Figure. 1C) for different experimental conditions or at single-molecule measurements with discrete stochastic different times during a system response. In turn, these analyses. The overall goal of this project has been to de- distributions can be computationally described by the velop these tools in both the experimental and the com- master equation and analyzed using the tools devel- 455

Page 456

oped by Dr. Munsky (see above). For relatively simple gene Gene Regulation (with Gregor Neuert (MIT/Vanderbily), Rui regulatory systems, temporally resolved mRNA or protein Zhen Tan (MIT), Leonid Teytelman (MIT), Alex van Oude- distributions have proved sufficient to constrain the model naarden (MIT/Hubrecht) and Mustafa Khammash (UCSB/ space to achieve understanding. Dr. Munsky reasoned that ETH)). Single molecule fluorescence in situ hybridization the systematic integration of single cell experiments and (smFISH) allows researchers to count the number of mRNA stochastic analyses would enable rapid quantitative hy- molecules in individual cells. When combined with Dr. pothesis testing and the ability to identify system mecha- Munsky’s Finite State Projection method, these data help nism and/or parameters. The second sub-goal of this proj- discriminate between vast numbers of quantitative model ect was to combine data and stochastic modeling to enable hypotheses. Dr. Munsky developed an iterative approach this rapid hypothesis testing. to (i) extract more quantitative information from single- cell experimental measurements and (ii) guide the design Developing Predictive Models for Complex Biological of subsequent experiments. He and his collaborators have Systems. In the course of this research, Dr. Munsky recog- identified and validated a stochastic model to accurately nized that and given complex biological behavior could be predict the time-varying behavior of three shock-response described at many different scales and at many different genes in yeast under the influence of multiple genetic mu- levels of abstraction. As a result, modeling of biologi- tations. This integration of experimental and computation- cal systems commonly faces a tradeoff between simple al methods represents a powerful new approach that can models that bear little resemblance to reality and detailed be applied to any gene, pathway, or organism. Dr Munsky mechanistic models that are rife with unknown param- and collaborators have published this work in two recent eters. Although both approaches can improve conceptual high-impact articles, both in Science: [Neuert, Munsky, et understanding, neither is ideal to make quantitatively ac- al, Systematic Identification of Signal-Activated Stochastic curate predictions. Dr. Munsky then set a goal to develop a Gene Regulation, Science, 339:6119, 584-587, 2013], and new approach to consider many different complexity levels [Munsky, Neuert, et al. Using Gene Expression Noise to Un- and use uncertainty quantification methods to weigh trad- derstand Gene Regulation, Science, 336(6078), 183–187, eoffs between differing model complexities in the context 2012]. Dr. Munsky and his experimentalist colleague Dr. of available data. By determining when models are too Gregor Neuert (now assistant professor at Vanderbult Uni- simple, when they are too complex, and then they are just versity) share first authorship on both papers. right, it would become possible to acquire useful quantita- tive predictions. 3) Identification of Transcriptional Regulation Dynamics During Host-Pathogen Interactions (with Elizabeth Hong- Scientific Approach and Accomplishments Geller (B7-LANL), Nan Li (B7-LANL), James Werner (MPA- Dr. Munsky has made great progress on four tasks related CINT-LANL), Douglas Shepherd). As part of the project to this project: “Illuminating the Dark Matter of the Genome: Small RNAs as Novel Targets for Bioterrorism,” Dr. Munsky has been

1. Identification of Gene Regulatory Networks Based Upon helping to apply his techniques to discover and character- Stochastic Modeling and Flow Cytometry Measurements ize small, non-coding RNAs in biothreat agents. An overall (with Chunbo Lou (MIT), Christopher Voigt (MIT)). Single- goal is to characterize the dynamics by which small non- cell measurements using high throughput flow cytometry coding RNA affects gene regulation during host-pathogen provide a wealth of information, which can reveal impor- interactions. Dr Munsky and colleagues seek to improve tant aspects of genetic regulation. Dr Munsky used these experimental approaches to count RNA copy numbers in measurements to identify the mechanisms and parameters single cells and computational/theoretical approaches to of gene regulatory constructs, which may otherwise be dif- analyze these data. In this project, Dr. Munsky and Dr. Wer- ficult to measure. In this project, Dr. Munsky worked with ner co-sponsored a postdoc (Dr. Douglas Shepherd), who collaborators Chunbo Lou and Christopher Voigt to identify started in May 2011. Preliminary work on this project was predictive models that would guide the design of synthetic published at [Shepherd, et al., New tools for discovering biological systems for eventual use in biofuel applications. the role sRNA plays in cellular regulation, Proc. SPIE 8228, Their results were published in a high impact journal: 2012]. This work led to experimental and computational [Ribozyme-based “insulator parts” buffer synthetic circuits results that hint that spatial fluctuations provide a wealth from genetic context, Nature Biotechnology, 30:11, 1137- of information about biological processes. As part of this 1142, 2012]. work, Dr. Munsky and colleagues discovered a new tech- nique to improve the single-to-noise ratio of individually Dr. Munsky has begun explorations in similar collabora- labeled RNA molecules. This technique enables the reli- tions with LANL researchers to quantify cell-to-cell variabil- able quantification of much smaller RNA molecules than ity to identify predictive models that can guide the design were previously detectable. The results of this work are of algae strains and harvesting technologies for biofuel ap- under review in the Annals of Chemistry, [D. Shepherd, et plications in low quality water. al, Counting small RNA in pathogenic bacteria. In revision for Analytical Chemistry, 2013].
2. Systematic Identification of Signal-Activated Stochastic 456

Page 457

4. Coarse graining strategies for transient analysis of sto- chastic gene regulation. At the 2011 q-bio Summer School A B (co-organized by Dr. Munsky), Dr. Munsky initiated collabo- ration with University of Pittsburgh researchers to develop ro efficient computational tools for stochastic processes. The k tc On a F resulting methods have been published in the proceed- g o ings of the 51st IEEE Conference on Decision and Control n a F [J. Tapia, J. Faeder, B. Munsky, Adaptive Coarse-Graining for Transient and Quasi-Equilibrium Analyses of Stochas- tic Gene Regulation, Proc. of the 51st IEEE Conference on Variabilikty vers/ugs parameters. C Off Decision and Control, 5361-5366, Maui, HI, Dec. 2012]. Dr. Munsky recently released stochastic analysis software as part of a review chapter that he published in 2012: [B. Munsky, Modeling Cellular Variability, in Quantitative Biol- ogy From Molecular to Cellular Systems, pp. 234-266, M. Wall, Ed., Taylor & Francis Group, New York, NY, 2012.]. Figure 1. Single-molecule measurements. (A) Using smFISH, Impact on National Missions specific mRNAs appear as bright diffraction-limited spots, whose numbers and 3D locations can be quantified. (B) Measurements The work in this project addressed the problem of model- of the variances and full distributions are much more informative ing stochastic systems in general and sought to validate than average expression levels. Induction from one bulk expres- methods using bacterial, yeast and mammalian gene sion level to another (i.e., from one dashed white line to the next) regulation as test beds. The results have advanced several to another can be accomplished by different mechanisms (red technical problems that involve modeling of stochastic or purple arrows). By keeping track of the corresponding change behavior, which crosscuts the needs of many agencies and in variance, one can identify the hidden mechanism of induction missions including DOE (biothreat reduction and using (change in color). (C) These three distinctly different distribu- microbes for biofuel production), NIH (design of therapeu- tions all have the same mean but correspond to different classes tics for bacterial or viral infections), and broad problems of gene regulation—by measuring the shape of the distribution in chem-bio, materials, chemistry, bioscience, health re- with smFISH, we can learn more about the underlying system, search, and bioenergy research. whereas qPCR would return only the less informative mean. (Im- Dr. Munsky’s enthusiasm for the analysis of single-cell ages in panel A adapted from [Munsky, Science, 2012; panels B-C behavior has helped LANL to build a first-in-class team to adapted from [Shepherd, SPIE, 2012]). integrate computational and experimental investigations of single-cell gene regulation. The team has developed the References experience and experimental apparatus to take these in- 1. Neuert, G., B. Munsky, R. Z. Tan, L. Teytelman, M. vestigations to the next level. Khammash, and A. van Oudenaarden. Systematic iden- tification of signal-activated stochastic gene regulation. For the FY14 DR pre-proposal cycle, Dr. Munsky’s stochas- 2013. Science. 339: 584. tic modeling methods developed in this project are now also currently under consideration for use in several ap- 2. Munsky, B., G. Neuert, and A. van Oudenaarden. Using plications: (1) cyber security; (2) pathogenic bacterial resis- Gene Expression Noise to Understand Gene Regula- tance; (3) immune response; and (4) algal biofuels. tion. 2012. Science. 336 (6078): 183. Dr. Munsky’s work as the main organizer of the q-bio Sum- 3. Munsky, B.. Modeling Cellular Variability. 2012. In mer School during the course of this project has helped Quantitative Biology From Molecular to Cellular Sys- LANL gain recognition in the global quantitative biology tems. By Wall, M. E.. p. 233. Boca Raton, FL: Chapman arena. Dr. Munsky, William Hlavacek (T6), Lev Tsimring and Hall. (UC San Diego), and Jeff Hasty (UC San Diego) applied for an NIG R25 grant related to the summer school. This grant 4. Tapia, J., J. Faeder, and B. Munsky. Adaptive Coarse- received a perfect score in review, and is currently pend- Graining for Transient and Quasi-Equilibrium Analyses ing funding of up to $1M over five years through the New of Stochastic Gene Regulation. 2012. In 51st IEEE Con- Mexico Consortium. ference on Decision and Control. (Maui, HI, December 10-13, 2012). , p. 5361. Maui, HI: IEEE. Upon its conclusion, this project will lead to the conversion of Dr. Munsky to full time technical staff member in CCS-3. 5. Shepherd, D., N. Li, E. Hong-Geller, B. Munsky, and J. This project also led to the recruitment of a postdoctoral Werner. New tools for discovering the role sRNA plays researcher, Dr. Douglas Shepherd (MPA-CINT), who is co- in cellular regulation. 2012. (San Francisco, California, sponsored by Dr. Munsky. Jan. 2012). Vol. 8228p. . : . 457

Page 458

6. Shepherd, D., N. Li, S. Micheva-Viteva, B. Munsky, E. Shepherd, D., N. Li, E. Hong-Geller, J. Werner, and B. Mun- Hong-Geller, and J. Werner. Counting small RNA in sky. Spatiotemporal measurements and modeling of pathogenic bacteria. To appear in Analytical Chemistry. genetic expression . Presented at The Sixth Annual q- Bio Conference. (Santa Fe, NY, 8-12, Aug. 2012).
7. Lou, C., B. Stanton, Y. Chen, B. Munsky, and C. A. Voigt. Ribozyme-based insulator parts buffer synthetic cir- Shepherd, D., N. Li, S. Micheva-Viteva, B. Munsky, E. Hong- cuits from genetic context. 2012. Nature Biotechnol- Geller, and J. Werner. Counting small RNA in pathogen- ogy. 30 (11): 1137. ic bacteria. To appear in Analytical Chemistry. Publications Tapia, J., J. Faeder, and B. Munsky. Adaptive Coarse-Grain- ing for Transient and Quasi-Equilibrium Analyses of Lou, C., B. Stanton, Y. Chen, B. Munsky, and C. A. Voigt. Ri- Stochastic Gene Regulation. 2012. In 51st IEEE Confer- bozyme-based insulator parts buffer synthetic circuits ence on Decision and Control. (Maui, HI, 10-13, Dec. from genetic context. 2012. Nature Biotechnology. 30 2012). , p. 5361. Maui, HI: IEEE. (11): 1137. Munsky, B.. Modeling Cellular Variability. 2012. In Quan- titative Biology From Molecular to Cellular Systems. By Wall, M. E.. , p. 233. Boca Raton, FL: Chapman and Hall. Munsky, B.. Integrated experimental and computational identification of stochastic gene regulation . Presented at Gordon Research Conference in Stochastic Physics in Biology. (Ventura, CA, 23-25 Jan. 2011). Munsky, B., G. Neuert, and A. van Oudenaarden. Using Gene Expression Noise to Understand Gene Regula- tion. 2012. Science. 336 (6078): 183. Munsky, B., S. Lampoudi, D. Thorsley, and A. Walczak. Conference Report. Presented at The 3rd workshop on stochasticity in biochemical reaction networks . (Banff, Canada, 12-16, Sept. 2011). Munsky, B., and M. Khammash. Identification from sto- chastic cell-to-cell variation: A genetic switch case study. 2010. IET Systems Biology. 4: 356. Nemenman, I., S. Gnanakaran, W. Hlavacek, Y. Jiang, B. Munsky, M. Wall, and J. Faeder. The Fifth Annual q-bio Conference on Cellular Information Processing. 2012. Physical Biology. 9: 050201. Neuert, G., B. Munsky, R. Tan, L. Teytleman, M. Khammash, and A. van Oudenaarden. Systematic Identification of Signal-Activated Stochastic Gene Regulation. 2013. Sci- ence. 339 (6119): 584. Shepherd, D., N. Li, E. Hong-Geller, B. Munsky, and J. Wer- ner. New tools for discovering the role sRNA plays in cellular regulation. 2012. In Single Molecule Spectros- copy and Superresolution Imaging. (San Francisco, Cali- fornia, Jan. 2012). Vol. 8228, p. . : . Shepherd, D., N. Li, E. Hong-Geller, B. Munsky, and J. Wer- ner. Spatiotemporal Measurements and Modeling of Genetic Expression. Presented at Stochastic Physics in Biology. (Ventura, CA, 16-19, Jan. 2013). 458

Page 459

Environmental and Biological Sciences Postdoctoral Research and Development Final Report Modeling the Surface Mass Balance and Freshwater Runoff from the Greenland Ice Sheet in a Changing Climate Matthew W. Hecht 20100631PRD4 Abstract Glaciers and Ice Caps: Contribution to sea level rise, also This Final Report summarizes the very successful work of produced a press conference and wide coverage in the a Director’s Postdoctoral Fellow on the surface mass bal- media, including an interview by the BBC. ance and runoff from various glacial systems, including Dr. Mernild continues to establish himself as an inter- the Greenland Ice Sheet. national expert on state of the World’s glaciers and ice sheets. He has been asked to serve as a contributing Background and Research Objectives author on the Sea Level Rise chapter of the upcoming This project has focused on the hydrological cycle over Fifth Assessment Report of the International Panel on Greenland, in particular on the balance between cold- Climate Change. This is remarkable recognition for so season accumulation of snowpack and warm season young a scientist. runoff. The stated aim of the proposal was to predict local and regional changes in Greenland’s surface mass Impact on National Missions balance and runoff in a changing climate based on com- LANL contributes to the US effort in climate modeling puter modeling of the snowpack, with a dual approach through the Community Earth System Model, in part- of modeling the entire Greenland Ice Sheet while also nership with the NSF. The DOE has a particularly strong applying more detailed study to two particularly well focus in high latitude climate, in part due to LANL’s observed regions, the Kangerlussuaq and Sermilik Fjord expertise in sea ice, and in the dynamical modeling of drainage areas. ice sheets. There had been no particular expertise in the Dr. Mernild took good advantage of the opportunity DOE before Dr. Mernild’s arrival in the study of year-to- provided by the Director’s Fellowship. He not only met year variations and trends in snowpack and runoff. This the objectives of our Proposal, but also found opportuni- work has had a substantial impact on the national effort ties to make excellent use of a variety of observational to monitor and understand trends in climate at the high sources, with a tremendously deep set of collaborators latitudes. from the Laboratory and around the world, and has References furthermore applied his analysis well beyond the bound- aries of Greenland. 1. Mernild, , and B. Hasholt. Climate-driven fluctua- tions in freshwater to Sermilik Fjord, East Greenland, Scientific Approach and Accomplishments during the last 4000 years. 2012. The Holocene. 22: One measure of accomplishment is the record of pub- 155–164. lications. In 2010, Dr. Mernild had six peer-reviewed 2. Hanna, . Discerning the influence of North Atlantic publications, five of which were first-authored. In 2011 atmospheric and oceanic forcing effects on 1900– he had another three first-authored publications. As 2010 Greenland summer climate and melt. 2012. the LDRD-supported work has come to full fruition, Dr. International Journal of Climatology. Mernild has seen nine manuscripts appear in refereed journals already in 2012, and listed in the reference 3. Mernild, , and N. T. Knudsen. Multi-decadal marine section, with another six in review. He has also had and land-terminating glacier retreat in Ammassalik several invited presentations over the past year, includ- region, Southeast Greenland. 2012. The Cryosphere. ing most notably at the annual meeting of the European 6: 625–639. Geophysical Union in Vienna. His EGU presentation, on 459

Page 460

4. Mernild, S. H., and G. E. Liston. Greenland freshwater E. Liston. Climate-driven fluctuations in freshwater to runoff. Part II: Distribution and trends, 1960–2010. \{S\}ermilik \{F\}jord, \{E\}ast \{G\}reenland. 2011. The Holo-
5. Journal of Climate. cene. : in press. Mernild, S. H., N. T. Knudsen, W. H. Lipscomb, J. C. Yde, J.
6. Liston, G. E., and S. H. Mernild. Greenland freshwater K. Malmros, B. H. Jakobsen, and B. Hasholt. Record runoff. Part I: A runoff routing model for glaciated and mass loss from Greenland’s best-observed local glacier. non-glaciated landscapes (HydroFlow). 2012. Journal
7. The Cryosphere Discussions. 5: 461. of Climate. Mernild, S. H., N. T. Knudsen, W. H. Lipscomb, J. C. Yde, J.
8. Hinzman, , and S. H. Mernild. Trajectory of the Arctic as K. Malmros, B. Hasholt, and B. H. Jakobsen. Increas- an Integrated System. 2012. Ecological Applications. ing mass loss from Greenland’s Mittivakkat Gletscher.
9. The Cryosphere. 5: 341.
10. Pedersen, , and C. S. Andersen. Sediment accumulation in a fjord in front of the Mittivakkat glacier delta, SE Mernild, S. H., N. T. Knudsen, and E. Hanna. Mittivakkat \{G\} Greenland. 2012. The Holocene. letscher, \{SE\} \{G\}reenland. 2011. In The Arctic Report Card, Greenland. Edited by J. E. Box et al.. NOAA Re-
11. Hanna, , and K. Steffen. Recent warming in Greenland port. in a long-term instrumental (1881–2012) climatic Mernild, S. H., T. L. Mote, and G. E. Liston. \{Greenland ice context. Part 1: Evaluation of surface air temperature sheet surface melt extent and trends: 1960-2010\}. records. 2012. Environmental Research Letters. 7. \{201. \{JOURNAL OF GLACIOLOGY\}. \{57\}: \{621.
12. Mernild, , and M. van den Broeke. Simulated internal Mernild, S. H., and G. E. Liston. Greenland freshwa- storage build-up, release, and runoff from Greenland ter runoff. \{P\}art \{II\}: \{D\}istribution and trends, Ice Sheet at Kangerlussuaq, West Greenland. 2012. 1960–2010. 2011. J. Climate. : submitted. Arctic, Antarctic, and Alpine Research. 44: 83.
13. Fausto, R. S., S.H. Mernild, B. Hasholt and A.P. Ahlstrom, and N.T. Knudsen. Modelling suspended sediment concentration and transport for the hydrological year 2004—2006. 2012. Arctic, Antarc- tic and Alpine Research. 44: 306.
14. Mernild, S. H., N. T. Knudsen, W. H. Lipscomb, J. C. Yde, J. K. Malmros, B. Hasholt, and B. H. Jakobsen. Increas- ing mass loss from Greenland’s Mittivakkat Gletscher.
15. The Cryosphere. 5: 341.
16. Mernild, S. H., T. L. Mote, and G. E. Liston. Greenland ice sheet surface melt extent and trends: 1960-2010.
17. JOURNAL OF GLACIOLOGY. 57: 621.
18. Mernild, S. H., G. E. Liston, C. A. Hiemstra, J. H. Chris- tensen, M. Stendel, and B. Hasholt. Surface mass- balance and runoff modeling using \{HIRAM4\} \{RCM\} at \{K\}angerlussuaq (\{S\}ondre \{S\}trom- fjord), \{W\}est \{G\}reenland, 1950—2080. 2011. Journal of Climate. 24: 609. Publications Mernild, S. H., G. E. Liston, C. A. Hiemstra, J. H. Chris- tensen, M. Stendel, and B. Hasholt. \{Surface Mass Balance and Runoff Modeling Using HIRHAM4 RCM at Kangerlussuaq (Sondre Stromfjord), West Greenland, 1950-2080\}. \{201. \{JOURNAL OF CLIMATE\}. \{24\}: \{609. Mernild, S. H., M.-S. Seidenkrantz, P. Chylek, and and G. 460

Page 461

Environmental and Biological Sciences Postdoctoral Research and Development Final Report Novel Laboratory and Field Observations to Model Climate Warming by Aged Absorbing Aerosols Manvendra K. Dubey 20110751PRD3 Abstract co’s Las Conchas wildfire in 2011 also contribute to ab- Climate forcing from absorbing aerosols such as black sorption, and together lead to increased warming of the carbon (BC), a.k.a. “soot”, emitted by the combustion atmosphere. Further measurements are required, but it of fossil fuels and wildfires are the most uncertain ele- appears that with climate change leading to the pres- ments in terms of our ability to assess climate change, ence of more and larger wildfires that this effect could and currently BC is thought to be the second most compound global warming, and should not be over- important contributor to warming behind CO2. State- looked in climate models. Additionally, single scatter of-the-art techniques were applied to measure BC in albedo from fires correlates with combustion efficiency, the field and the laboratory for improved treatments in and this relationship can be applied within models to climate models. Aerosol optical and physical properties constrain the radiative forcing from BB emissions. were measured in the ambient atmosphere in addition BC Absorption Enhancement: Field measurements of BC to the use of laboratory studies as a more controlled in the United Kingdom indicate the first ambient mea- setting to study aerosol processing. This research is surements of an appreciable absorption enhancement focused on the coupling of measurements made with (~30%) due to coatings on BC. This has been predicted a single particle BC-specific instrument that detects BC based on laboratory work and modeling studies, but has mass via laser-induced incandescence and photoacoustic not been confirmed in the field yet. Our measurements spectrometers that measure the absorption and scatter- indicate that BC can have increased absorption due to ing due to particles, in addition to using various other coatings, and it should be noted that our numbers are supportive measurements to characterize the aerosol a lower bound for this measurement so it is possible and gas-phase precursors. From the empirical measure- that the enhancement could be closer to the predicted ments, simplified parameters are elucidated for applica- factor of 2, which would have a large significance on the tion in climate models with the overall goal of reducing amount of warming due to BC. The methodology and the uncertainty in the climactic impacts of aerosols, cur- key results are highlighted here. rently one of the largest uncertainties in global models today. BC Removal: A method was developed for the online removal of BC from mixed aerosol samples. Labora- Background and Research Objectives tory studies with a BC surrogate, Aquadag, indicate Gaps in our knowledge of emissions of absorbing that laser-induced vaporization can be used to remove aerosols like black carbon from fires and Aeolian dust 90% on average of the original BC mass present. This that absorb sunlight to warm the climate are the largest technique could be applied to probe the optical proper- source of uncertainty in climate model projections of ties of brown carbon from both laboratory and ambient climate change. We identified opportunities and used samples, which would be of great interest to both the LANL state-of-the-art aerosol instruments in field and aerosol and climate community in general since these laboratory studies to understand how absorbing aero- other absorbing aerosols are poorly characterized and sols transformations influence their ability to absorb and could be significant contributors to the warming of scatter sunlight. the atmosphere, especially for locations with biomass burning. Other applications include removing BC to Wildfires: BC is not the only significant absorbing spe- determine whether or not it has a significant role in both cies in wildfire emissions. Tar balls, found to be present cloud and ice nucleation, which will also be very impor- at 10x the amount of BC by number during New Mexi- 461

Page 462

tant in determining both BC lifetime in the atmosphere in terms of morphology and classified into four categories and also its role in cloud processing and indirect effects on and are shown in Figure 1 below: ~50% are embedded radiative forcing. (heavily coated), ~34% are partly coated, ~12% have inclu- sions, ~4% are bare. Tar balls were divided into two differ- Improved Absorption Measurements: Direct online aero- ent classes, one being more oxidized than the other. Since sol absorption measurements using photoacoustic spec- the mixing of BC particles with other material affects their troscopy are considered noisy in comparison to more tradi- optical, chemical and physical properties inclusion of these tional filter measurements. However, they are not subject observations should improve climate model performances. to the biases that reduce accuracy of the filter measure- ments. From analyzing the aerosol absorption measure- Single scattering albedo (ω) from BB aerosols produced ments at the stationary DOE ARM site at SGP, upgraded from ninety-nine experimental single source burns at the lasers and an external acoustic filter improve the data Missoula Fire Sciences Laboratory during FLAME-IV were quality significantly, by factors of 1.3-5.8, depending on the compared with New Mexico’s two largest wildfires, Las wavelength of the measurement. Another technique was Conchas in 2011 and the Whitewater-Baldy Complex in also explored that utilizes direct measurements, but from 2012. Laboratory measured ω ranges from ~0.2 to 1, with two different instruments. From this work, there will be an the variation depending strongly on combustion conditions IOP next year at the SGP site to determine the best tech- rather than fuel type. The flaming phase that precedes the nique for measuring aerosol absorption in the field. smoldering phase has a lower ω due to a higher mass frac- tion of BC that increases as combustion efficiency decreas- Transported BC: During the BEACHON field campaign in es. During the later smoldering phase, ω remains relatively 2011 that was located in a remote forest in Colorado, BC constant ~0.95 due to the largely scattering organics that dominantly originated from transported pollution from cit- are more dominant during this phase. ω is parameter- ies 20-50 miles away, Colorado Springs and Denver. More ized as a function of modified combustion efficiency using than half of the BC was thickly coated, consistent with the laboratory data, which the field observations support, aged BC, and was usually <5% of the total aerosol mass making it possible for use for evaluating the climate im- except during pollution events when it was ~10%. BC cor- pacts of BB aerosols within radiative transfer models. related with organic aerosols during the pollution periods, and there was also increased aerosol absorption indicating Absorption Enhancement Observed in the Field the presence of either external absorbing species, such as Currently, absorption from BC is thought to be the second brown carbon, and/or enhanced absorption due to the BC largest contribution to global warming after carbon diox- mixing state. ide, but with high uncertainty. Laboratory studies on the optical properties of BC and modeling studies indicate that Absorbing Dusts and Hematite: In addition to the major BC could have up to a two-fold increase in absorption due findings discussed here, we also performed laboratory to coatings on BC, which is not currently accounted for in experiments on size-selected absorbing dusts from differ- most models. However, this kind of increase in absorption ent sites across the globe and on hematite, a dominant due to coatings on BC has yet to be measured in the field. absorbing component in dusts. We measured the optical As a matter of fact, the only study out currently indicates a properties and determined a calibration for measuring very small increase in the absorption and considers coat- hematite using laser-induced incandescence. We hope to ing BC to be a null effect. However, these are measure- continue this work and use the data to develop an optical ments from only one location, and the BC there was not model for dust that could be tested in the field. well-aged nor did it have thick coatings present, therefore, supporting the published findings. Scientific Approach and Accomplishments BC and Tar Balls in Wildfire Emissions During the ClearfLo campaign in the UK, we denuded the Biomass burning (BB) is one of the largest sources of car- BC up to 250 degrees C and measured the optical proper- bonaceous aerosols in the atmosphere, significantly affect- ties in order to determine whether or not there was an ing earth’s radiation budget and climate. Tar balls, abun- appreciable absorption enhancement due to the semi-vol- dant in biomass burning smoke, absorb sunlight and have atile coatings found on BC. For two locations in the UK, we highly variable optical properties, typically not accounted found ~30% absorption enhancement (see Figure 2) due for in climate models. Individual biomass burning particles to coatings on BC as determined by focusing on the longer from the Las Conchas fire (New Mexico, 2011) were ana- wavelengths, 781 nm, where BC dominates the absorption lyzed using electron microscopy. The relative abundance of spectrum. The absorption enhancement is also correlated tar balls (80%) was found to be 10 times greater than soot with coating thickness and photochemical age, which sup- particles (8%). Soot, a.k.a. black carbon (BC) was quantified ports theoretical predictions and laboratory simulations. 462

Page 463

These results indicate that internally mixed BC can have Aerosol absorption and scattering from the three-wave- increased absorption signals and that more studies need to length photoacoustic soot spectrometer (PASS3; Droplet be conducted to determine how often this is present in the Measurement Technologies) has been located in the DOE atmosphere. AOS at SGP since 2009. Approximately six months of measurements from January 2013 to July 2013 at 405 nm, Removal of BC for Isolated Studies on Other Absorbing 532 nm, and 781 nm with an upgraded instrument were Aerosols analyzed. The uncertainty within the absorption measure- The ability to separate black carbon (BC) from organic ment was improved by factors of 5.8, 3.8, and 1.3, respec- carbon (OC) in an online aerosol sample would enable the tively, for the three wavelengths with upgraded lasers and physical processes and optical properties to be studied by an external acoustic filter. carbon class. A novel method was developed to remove BC from an online aerosol sample with the soot surrogate Direct absorption measurements from the PASS-3 are Aquadag. This method can be applied to ambient and compared with two collocated filter measurements from source aerosol samples to isolate the roles that different the Continuous Light Absorption Photometer (CLAP, NOAA/ carbon classes have on climate, e.g. atmospheric warming, GMD) at 467 nm, 528 nm, and 652 nm, and the Particle cloud formation and lifetime. Removal of BC from an am- Soot Absorption Photometer (PSAP; Radiance Research) bient aerosol sample would allow for the isolated study of at 467 nm, 530 nm, and 660 nm in Figure 4. The CLAP other absorbing aerosols such as dust and brown carbon. appears to have less bias and better corrections for filter issues than the PSAP, and compares well with the direct Two Single-Particle Soot Photometers (SP2s) were used in PASS absorption measurements (Figure 4). Scattering series, along with a Three-Wavelength Photoacoustic Soot measurements from the PASS-3 are compared with an Spectrometer (PASS-3) for measuring optical properties. integrating nephelometer (TSI) at 450 nm, 550 nm, and Polydisperse and monodisperse particles were sampled 700 nm. On average, the absorption and scattering signals from 100 nm to 800 nm in diameter. For all monodisperse indicate a generally clean location. The average absorp- samples, the average percent remaining by mass and tion angstrom exponent at this site was determined to be number of the original size selected are 11.1% ± 18.6% 1.03, indicating minimal presence of absorbing species and 12.7% ± 21.9%, respectively. Figure 3 below shows other than black carbon. The average single scatter albedo an example of the particle mass size distribution before across all wavelengths and measurements is 0.8, showing a and after the removal of 500 nm diameter particles. The dominantly scattering aerosol at the site. Angstrom expo- results are applicable to other instruments employing LII, nents are used to interpolate between the three different e.g. the Soot Particle Aerosol Mass Spectrometer (SP- wavelength ranges, which are less than 20%. AMS), and could also potentially be used for the study of new and/or altered carbonaceous particles in a controlled During the ClearfLo campaign in 2012 we had both the environment. extinction and scattering measurements, and attempted to perform resolving aerosol absorption via subtraction. Scat- Improved Absorption Measurements at an ARM Site tering and extinction from the two instruments compare Direct aerosol absorption measurements using photo- well, however, the measured and calculated absorption acoustic spectroscopy are considered “noisy.” However, in- have larger discrepancies and uncertainties as can be seen direct measurements that have better precision use filters in the Figure below. This method will require careful cali- and as a result are subject to biases that significantly re- bration, collocation and any corrections for measurements duce their accuracy. Another proposed method has been at different wavelengths since it requires the subtraction of to measure extinction and scattering and to calculate the two large numbers (extinction and scattering) to determine absorption from the subtraction of the two measurements a relatively small number (absorption). being made by two different instruments. There has been much debate within the last year about which technique Transported BC in a Remote Forest is best suited for measuring aerosol absorption, especially BC and its optical properties were measured during the in relatively clean areas with low signals. From our data BEACHON campaign in 2011. The SP2 and 3 PASS instru- and extensive experience with absorption measurements ments were deployed with complementary aerosol equip- it was determined that the DOE would fund an Intensive ment to measure size-distributions and denude (heat the Operating Period (IOP) that LANL would lead next year at aerosol to remove semi-volatile components) the aerosol the DOE Atmospheric Radiation Measurement’s Aerosol as were used during the sampling of the Las Conchas fire. Observing System (AOS) at the Southern Great Plains (SGP) We tried to conduct denuding studies to look for absorp- Central Facility in Ponca City, OK to try to answer this ques- tion enhancements due to coatings on BC, but the concen- tion. trations at this remote site were too small to achieve this. 463

Page 464

BEACHON-RoMBAS was determined to be a “semi-clean” Gorkowski, K., M. K. Dubey, B. A. Flowers, A. C. Aiken, B. environment impacted by pollution transported from Z. Klein, C. Mazzoleni, N. Sharma, and S. China. Aero- nearby cities (e.g., Colorado Springs, Denver). Average BC sol Optical Properties of Smoke from the Las Conchas Wildfire, Los Alamos, NM. 2011. In American Geophys- mass concentrations were ~ 50 ng m-3 with peak periods ical Union Annual Meeting. (San Francisco, 5-9 Dec. exceeding 500 ng m-3 due to transported pollution events. 2011). , p. A52A. San Francisco: AGU, EOS. The BC mass concentration was <5% of the total PM1 mass on average during the campaign and is closer to ~10% dur- Mohr, C., F. D. Lopez-Hilfiker, P. Zotter, A. S. Prevot, S. C. ing pollution events. Herndon, L. R. Williams, J. P. Franklin, M. S. Zahniser, D. R. Worsnop, W. B. Knighton, A. C. Aiken, K. J. Gorkows- Most of the BC particles (55 ± 10%) were thickly coated, ki, M. K. Dubey, and J. A. Thornton. Contribution of which is consistent to the sources being from transported/ nitrated aromatics to wood burning brown carbon light aged pollution. AAE’s indicated aerosol absorption was absorption in Detling, UK during winter time. 2013. En- dominantly from BC, indicating minimal presence of brown vironmental Science and Technology. 47 (12): 6316. carbon or other absorbing organic species. This was also Ulbrich, I. M., M. R. Canagaratna, M. J. Cubison, Q. Zhang, consistent with the mass absorption coefficients (MACs) N. L. Ng, A. C. Aiken, and J. L. Jimenez. Three-dimen- that we measured at the site that were similar to measure- sional Factorization of Size-resolved Organic Aerosol ments of pure BC. However, during periods of identified Mass Spectra from Mexico City. 2012. Atmospheric pollution plumes, there were elevated MACs, indicating Measurement Techniques. 5: 195. possibly some transported absorbing organics unlike the organics present in the forest background. BC correlated with organics in the plumes, but none of the other mea- sured aerosol chemical species (e.g., ammonium, chloride, nitrate, sulfate). It was also determined that BC was not significant in terms of aerosol hygroscopicity, the initial wa- ter uptake necessary for a particle to form a cloud droplet, at this site during the campaign (Ezra Levin, CSU, personal communication). Participation in Large Interdisciplinary Field Campaigns FLAME-IV: Fire Lab at Missoula Experiment-IV. Organized by Colorado State University, Carnegie Mellon University and the University of Montana, funded by NSF. Publications Aiken, A. C., B. A. Flowers, and M. K. Dubey. : Absorbing Aerosols: Field and Laboratory Studies of Black Carbon and Dust. 2011. In American Geophysical Union Annual Meeting. (San Francisco, 5-9 Dec. 2011). , p. : A53A. San Francisco: AGU, EOS. China, S., C. Mazzoleni, K. Gorkowski, A. C. Aiken, and M. K. Dubey. Morphology and Mixing State of Individual Freshly Emitted Wildfire Carbonaceous Particles. 2013. Nature Communications. 4: 3122. Docherty, K. S., A. C. Aiken, J. A. Huffman, I. M. Ulbrich, P. F. DeCarlo, D. Sueper, and D. R. Worsnop. The 2005 Study of Organic Aerosols at Riverside (SOAR-1): Instrumen- tal Intercomparisons and Fine Particle composition. 2011. Atmospheric Chemistry and Physics. 11: 12387. Duplissy, J., P. F. DeCarlo, J. Dommen, M. R. Alfarra, A. C. Aiken, and J. L. Jimenez. Relating Hygroscopicity and Composition of Organic Aerosol Particulate Matter. 2011. Atmospheric Chemistry and Physics. 11: 1155. 464

Page 465

Information Science and Technology

Page 466

Information Science and Technology Directed Research Continuing Project Optimization Principles for Co-design Applied to Molecular Dynamics Stephan J. Eidenbenz 20110710DR Introduction lations on future novel computer architectures, and Our goal is to develop a framework for hardware- provide a basis for more decisive influence on hardware software codesign as a formally-posed optimization design. For our molecular dynamics application domain, problem. While the optimization framework will be ap- we will reach regimes that enable us to study a wealth plicable to multiple problem domains, for the target ap- of poorly understood phenomena of critical importance plication we use molecular dynamics (MD), an exemplar to the lab mission, including nuclear weapons, energy for the need for computational scaling, and archetypal of systems, and armor, to name but a few. the obstacles thereof. We view codesign as search and selection from a vast space of hardware and software Progress designs that map to performance metrics. The objec- Along the two major tasks of building the Accelerated tive function that we aim to optimize contains, as main Molecular Dynamics (AMD) simulator AMDSim and de- components, run time (or computational rate), problem veloping a more detailed Profile-Parallelize-Predict (PPP) size, simulated time duration, energy use, and hardware approach, we have made progress as follows. cost. The framework will manifest itself in the form of (i) optimization software addressing hardware and We have developed the AMDSim code suite, so far software design space enumeration, and performance consisting of TADSim to simulate temperature acceler- prediction methods at multiple scales, (ii) documented ated dynamics simulations, and ParRepSim to simulate best practice insights, and (iii) theorems on upper and parallel replica dynamics simulations. The codes have lower bounds on theoretically achievable performance been thoroughly tested and validated against the real for some algorithmic and hardware architecture com- code using a metallic surface diffusion system. The binations. On the application side, our goal is to en- validation process took longer than we had anticipated able the atomistic simulation tools that can probe the due to a few software errors but also due to the phys- physics of processes such as ductile spall failure under ics of the studied system, which turned out to be more shock conditions and the evolution of radiation damage. complex than we expected. We developed a model for Achieving this requires two orders of magnitude increase the escape rates (e.g., a diffusive jump of an atom on the in simulated time over current state-of-the-art petascale surface) that includes an additional temperature depen- computing, and cannot be realized without this code- dence to account for anharmonicity. We also found it sign approach, as direct implementation on an exascale was important to include a temperature dependence in platform would not achieve it. These two problems are the number of steepest-descent iterations required to of great intrinsic interest to the lab, addressing the re- detect transitions. With this more refined model of the sponse of materials in extreme conditions and enabling physics built into the TADSim code, a very satisfactory the design of more effective and safe fission power set of validation results was obtained. After the valida- plants, respectively. tion period, we have started large parameter scans, which constitute the first step in a process: first getting Benefit to National Security Missions a global view of the multidimensional landscape, sec- Successful implementation of our optimization frame- ondly using optimization to find local extrema in regions work will provide the unique capability of codesigning of interest. A new hurdle we are now facing is that the hardware and software on a sound semi-formal basis. due to the large number of parameters that we test It will enable effective implementation of physics simu- in our fully factorial design, a huge number (millions) 466

Page 467

of simulated simulations must be performed. This leads ASCR Workshop on Modeling and Simulation of Exascale to significant HPC cycle requirements as well as creating Systems and Applications are currently under review; and unusual challenges in effectively using the job queuing two more papers on AMDSim and PPP are in preparation. systems. We have been successful in our request for ad- ditional needed computing resources. Future Work The goals and objectives for the optimization framework The PPP component has progressed most significantly in are structured along each major task as follows: the automatic parallelization of serial programs via a multi- step tool chain. The tool chain itself consists of several dis-

1. Complete the accelerated molecular dynamics simula- crete components, some freely available open-source soft- tion simulator (sic) to include all algorithmic variations ware (most significantly the LLVM compiler infrastructure), for the temperature-accelerated dynamics (TAD) meth- and some developed by this project. In brief: a program in od, the parallel replica method (ParRep) and combina- any language that LLVM or GCC can accept (C, C++, Fortran, tions of both. Variations include speculative execution etc.) is transformed to the LLVM intermediate representa- of transition branches, alternate variations of checking tion, then transformed to introduce and maximize parallel- for transitions, multiple high-temperature settings, etc. ism, essentially by respecting exactly the data dependen- Our goal is to expand our current performance predic- cies, and otherwise ignoring the control dependencies that tion method (AMDSim) to quickly adapt to algorithmic are often largely artifacts of the sequential programming changes in terms of the ability to model them, which language. In this form the program exhibits a theoretically then can be fed into a set of optimization rules. maximal degree of parallelism. Subsequent analysis and
2. Complete the full optimization cycle for the hardware transformation schedules the instructions with respect oriented PPP (Profile, Parallelize, Predict) approach, to a processor budget, memory access constraints, data which aims to find shortcuts to optimal designs by transfer costs, etc., in other words, optimally or near-opti- exposing the inherent parallelism in a problem with mally on an arbitrary hypothetical computer architecture. near-optimum accuracy. A key challenge will be to find Typically the goal is to minimize makespan—the time for all a balance between maximizing parallelism without computations to complete, but other metrics are possible, losing performance to data movement. This task will such as total energy consumption. In this way we can build on the promising work of FY12 that included the answer such questions as “Given an architectural specifica- development of a the ASET tool (for Architecture Space tion, what is fastest a given program could possibly run on Exploration Tool). it?” or “Given a program, what is the fastest it could run with current semiconductor technology but with unlimited 3. Complete implementation of a suite of optimization budget?” In the context of the whole project, this gives us rule sets, including genetic algorithms, simulated an- the automated performance prediction needed in the co- nealing, and other metaheuristics. design optimization loop. Independently, we believe that
3. Integrate the optimization tools from (3) with the per- the PPP capability has the potential to be a game-changer formance and algorithm-space exploration tools from in automated program parallelization. (1) and (2) to run experiments at large scale. Our goal is to find novel architecture and software combina- To validate our results we are currently targeting real tions that show promising performance in a simulated extant hardware (Intel x86) on which to run the auto- environment. parallelized code. This has proven to be rather more complicated than originally expected, mainly because the original nominal target of the transformed code was Conclusion custom or simulated hardware. Nonetheless, we are on A successful codesign optimization framework will fun- track to achieve this. A potential next step (next FY) would damentally change the way scientific applications are be to target a more exotic extant architecture such as the constructed: HW/SW codesign fundamentals will become Intel MIC, which would be particularly appropriate because an essential part of software engineering. The optimiza- a) we have access to such hardware; and, b) a future ver- tion framework will have manual and automated aspects. sion of the MIC may provide the bulk of the computational Similar to compiler optimization techniques used in clas- power of LANL’s next capability machine. sical compilers for the automated aspects, it will achieve peak performance when the human designer allows the Our work has resulted in a publication in the 2012 Winter optimizer to test a large set of potential solutions in a high- Simulation Conference; three position papers for the DOE/ throughput optimization loop. 467

Page 468

Publications Santhi, N., M. K. Davis, and S. Eidenbenz. Graphical models Optimization Principles for Hardware/Software Co-Design and a scalable methodology for exascale performance with Applications in Molecular Dynamics. 2012. In prediction. 2013. LA-UR 13-24435. Poster at Supercomputing conference SC’11. (Seattle, Williams, S., and S. Eidenbenz. Extensible design for multi- WA, 12-18 Nov. 2011). , p. 1. Seattle, WA: ACM. agent serial processes. 2013. LA-UR 13-27222 . Delengov, V., Y. Li, J. Thompson, and L. Kroc. Hardware- Software Codesign. 2012. Claremont Graduate Univer- sity Mathematics Clinic - Final Report. Eidenbenz, S., A. Voter, K. Davis, S. Mniszewski, D. Perez, N. Santhi, S. Thulasidasan, H. Djidjev, L. Gurvits, and C. Junghans. Optimization Principles for Codesign applied to Molecular Dynamics: Design Space Exploration, Performance Prediction, and Optimization Strategies. 2012. In DOE ASCR/ASC Exascale Research Conference 2012. (Portland, NY, 16-18, April, 2012). , p. n.a.. Port- land, OR: DOE. Eidenbenz, S., A. Voter, K. Davis, S. Mniszewski, D. Perez, N. Santhi, S. Thulasidasan, H. Djidjev, L. Gurvits, and L. Kroc. Optimization Principles for Hardware/Software Co-Design with Applications in Molecular Dynamics. 2011. Los Alamos National Laboratory Unlimited Re- lease. Eidenbenz, S., K. Davis, A. Voter, S. Mniszewski, H. Djidjev, L. Gurvits, C. Junghans, D. Perez, N. Santhi, and S. Thu- lasidasan. Optimization Principles for Computational Co-Design with Applications to Molecular Dynamics . 2012. Poster at LANL Materials Capability Review. Eidenbenz, S., N. Santhi, S. Mniszewski, and M. K. Davis. Performance modeling at Multiple Interacting Scales: hierarchical generalized instructions sets to move through different abstraction levels using stochastic substitution. 2013. LA-UR 13-24444. Gurvits, L.. Simulating one Markov Process based on a Black Box for another: Rigorous Proofs. 2012. Draft ver- sion. Kroc, L., N. Santhi, L. Kroc, V. Delengov, Y. Li, J. Thompson, and S. Eidenbenz. Optimization Principles for Arith- metic Functions in Hardware-Software Co-Design

• Extended Abstract. 2012. In The Winter Simulation Conference. (Berlin, Germany, 9-12 December 2012). , p. 337. Berlin, Germany: SCS. Mniszewski, S.. TADSim: Discrete Event Simulation of Tem- perature Accelerated Dynamics Performance. 2011. Los Alamos National Laboratory Unlimited Release Report. Mniszewski, S., S. Eidenbenz, A. F. Voter, D. Perez, and C. Junghans. Parametrized discrete-event-based applica- tion simulators for performance modeling. 2013. LA- UR 13-24386. 468

Page 469

Information Science and Technology Directed Research Continuing Project CoCoMANS: Computational Co-design for Multi-scale Applications in the Natural Sciences Dana A. Knoll 20110737DR Introduction explosives and armor, and climate/energy effects. We will forge a qualitatively new predictive-science capability exploiting evolving high-performance com- Progress puter architectures for multiple national-security-critical We have had two external reviews of the CoCoMANS application areas—including materials, plasmas, and project in FY-13. The first was the standard external climate—by simultaneously evolving the four corners review required by the LDRD office at roughly the mid- of science, methods, software, and hardware in an point of the project. This occurred at LANL on October integrated computational co-design process. We will 11, 2012. There were four external committee members develop new “applications-based”, self-consistent, two- (LLNL, MIT, Univ. of Utah, and Univ. of New Hampshire) way, scale-bridging methods that have broad applicabil- and three internal committee members (XCP (2) and ity to the targeted science, will map well to emerging HPC). The second external review occurred on May 22, heterogeneous computing models (while concurrently 2013 as part of the LANL Computational Physics and guiding hardware and software to maturity), and provide Applied Math (CPAM) capability review. This commit- the algorithmic acceleration necessary to probe new tee had seven external members. In both reviews the scientific challenges at unprecedented scales. Outcomes CoCoMANS project was rated overall as outstanding / will 1) represent a paradigm shift in the pursuit of world- excellent. Both committees see the project as making class science at LANL by bringing a wide spectrum of good progress, having the potential for significant long- pertinent expertise in the computer and natural sciences term impact, and being in a leadership role. in to play; 2) deliver a documented computational co- design knowledge-base, built upon an evolving software The plasma application sub-area has made excellent infrastructure; 3) foster broad, long-term research col- progress in FY-13. A fundamental manuscript on the laborations with appropriate hardware partners; and 4) High-Order Low-Order (HO-LO) approach to electrostatic deepen understanding of, and our experience-base with, problems has been accepted for publication in the SIAM consistent, two-way, scale-bridging algorithms. J. of Sci. Comput. The fundamental HO-LO electromag- netic algorithm has been developed and tested in proto- type software. Two high profile paradigm shift physics Benefit to National Security Missions problems have been defined and a paradigm shift report The need for computational co-design and multiscale has been written. We have made good progress on mini- solution algorithms resides in almost every mission at apps. Initial implementation of the HO-LO algorithm LANL that involves computational science. Nowhere is on multi-node many-core hardware has been success- the need greater than in the NNSA ASC program. We ful and has clearly validated our “partial replication” have carefully chosen two of our application areas, ma- implementation strategy. We have clearly demonstrated terials science and plasma physics, as these areas repre- significant flops for minimal node-to-node communica- sent current and future multi-scale challenges to ASC. All tion. Initial HO solver implementation on GPUs has been three chosen application areas are relevant to the Office successful and has clearly validated our “physics code of Science mission. Most importantly, the research abstraction” concept allowing for efficient implementa- proposed here is a required investment to prepare for a tion on multiple hardware types. We have submitted a LANL leadership role in the DoE Exascale Initiative. Mis- manuscript on aspects of this success to the Supercom- sion applications for computing at this scale are legion, puting 2013 (SC13) conference in Denver CO (November and include nuclear weapons, fusion and fission energy, 469

Page 470

2013). Additionally, CoCoMANS plasma application talks must be made in order to implement the mini-apps on the where given at the APS Division of Plasma Physics meet- CCS-7 Darwin cluster. With the mini-apps implemented on ing (Nov 2012) and the SIAM Computational Science and Darwin, we must collect enough profiling data to isolate Engineering (CS&E) meeting (February 2013). communication bottlenecks, determine the level of asyn- chronous computation, and to help understand where and The ocean application sub-area has made excellent prog- how to overlap communication with computation. ress in FY-13. The HO-LO algorithmic approach to the free- surface ocean model has been solidified through prototype With this profiling information in hand we can perform, testing. We will proceed with an implicit LO solver and an and document, a co-design iteration. This will result in explicit HO solver implemented utilizing the Jacobian-Free some level of refinement of the scale-bridging algorithms Newton-Krylov method and nonlinear elimination. We to account for the feedback from the initial mini-app work. have defined a first level paradigm shift problem and vet- At this point we will be able to add a level of detail to our ted it with LANL ocean modeling experts. We have begun compact-app plan in each application area. At this point to write a paradigm shift document for the ocean applica- we will also be able to define and develop specific software tion. We have a well define prototype / mini-app plan kernels, in support of each compact-app, which will be which will provide the required testing prior to compact used in specific hardware vendor interactions. application development. There has been initial progress on mini-app development and implementation on multi- Conclusion node many-core hardware. We have executed initial on- The technical impact of this project will be a roadmap lead- node many-core parallel scaling studies and profiling stud- ing to a quantum leap forward in computational science ies. An initial HO-LO algorithm manuscript is in the review for a variety of applications on modern architectures. No process. We have given talks on the ocean application at general scale-bridging algorithm of the type we propose Frontiers in Computational Physics (Dec 2012), the SIAM has gone through a computational co-design activity. As a Computational Science and Engineering (CS&E) meeting result of demonstrating a paradigm shift in specific applica- (February 2013), and the Copper Mountain Conference on tion areas, we will define a computational co-design blue Multigrid Methods (April, 2013). print for system-scale simulation with fine-scale physics fi- delity for several applications. Furthermore, we will obtain During FY-13 we have had one PhD student (Jeff Willert, a quantifiable understanding of the potential of various Applied math, NCSU) successfully defend his dissertation architectures to achieve such simulations resulting from and accept a post doc position in T-3. We have converted a our detailed interaction with vendors. post masters student (Josh Payne, MS Nuc E, MIT) to a staff position in CCS-7. Finally, we began supporting a new PhD Publications student (Chris Leibs, Applied Math, UC Boulder). Chen, G., L. Chacon, C. A. Leibs, D. A. Knoll, and W. Taitano. Fluid preconditioning for Newton-Krylov-based, fully Future Work implicit, electrostatic particle-in-cell simulations. To ap- In each application area, a large faction of initial algorithm pear in J. Comput. Phys.. development is done, significant prototyping software Knoll, D. A.. Computational co-design of multi-scale-ap- has been developed, and a first iteration idea of mapping plications in the natural sciences: CoCoMANS. 2013. these algorithms onto emerging architectures is in place. LANL Computational physics and applied math external Our primary task for FY14 will be the development and capability review. implementation of mini-apps onto the Darwin cluster and gathering some profiling data from these mini-apps. Of Knoll, D. A., L. Chacon, W. Daughton, and J. Payne. Plasma course, this primary task will be executed with significant paradigm shift document for the CoCoMANS project. communication across the two application areas. The goal 2013. LA-UR-13-22126. of this task will be to obtain the profiling data required to Newman, C., and D. A. Knoll. Physics-based precondition- execute the first and second co-design iterations for each ers for ocean simulation. To appear in SIAM J. of Scien- of the application areas. tific Computing. To execute on our primary task we will need to execute Payne, J., D. A. Knoll, A. McPherson, W. Taitano, L. Chacon, on a number of sub-tasks in order. First, we must build G. Chen, and S. Pakin. Computational co-design of a adequate documentation of our prototype software to multiscale plasma application: a process and initial aid in the planning and development of the mini-apps. results. Presented at International Parallel and Distrib- uted Processing Symposium. (Phoenix AZ, May 19-23, Next, some choices of language and programming models 470

Page 471

2013). Taitano, W., D. A. Knoll, L. Chacon, and G. Chen. Develop- ment of a consistent and stable fully implicit moment method for a Vlasov-Ampere particle-in-cell system. To appear in SIAM J. of Scientific Computing. Willert, J., C. T. Kelley, D. A. Knoll, and H. Park. Hybrid deterministic / monte carlo neutronics. To appear in SIAM J. of Scientific Computing. 471

Page 472

Information Science and Technology Directed Research Continuing Project Hierarchical Sparse Models for Robust Analysis of Video Data Steven P. Brumby 20120103DR Introduction Benefit to National Security Missions Vision is one of the hardest and most intriguing prob- The theoretical framework, algorithms and codes pro- lems in artificial intelligence and computer science. posed in this project will provide a solid basis for new Advances in the field have hinged on three fundamental information science and technology capability to sup- challenges: (1) how to create mathematical models of port several of LANL’s core mission areas. Robust, fast, natural video sequences that can generate and interpret automated understanding of video has the potential to the visual scene; (2) how to learn image feature repre- revolutionize analysis of large-scale video data of ex- sentations in such models across many spatial and tem- periments and simulations, global-scale environmental poral scales and for many object categories; and, (3) how science, proliferation detection, and safety and security to exploit the sheer size and richness of large-scale video applications (perimeter security, workplace safety moni- datasets now becoming available. These three problems toring, remote hazardous inspections). This capability not only need to be solved together but are now within will also significantly impact DOE/NNSA and our USG reach as a result of new theoretical, experimental and partners by enabling: autonomous vehicles using pas- computational advances. Our project will create novel sive video sensors, automated threat detection, forensic video analysis models and algorithms that we expect will imagery and video analysis tools, persistent surveillance approach human analyst performance and will surpass video tools for warfighter support, forest fire monitoring, human speed. border security, remote inspections of infrastructure, disaster response, and robotic space exploration. We Out team brings together world-class researchers in will present our results at leading conferences, publish statistical modeling, computer vision, and high perfor- in high impact scientific journals, seek opportunities to mance computing. We will pursue new research direc- brief potential sponsors, and develop white papers and tions for three complementary areas of information proposals for demonstrations and follow-on projects. science and technology: (1) develop semi-supervised learning algorithms to build multi-category deep proba- Progress bilistic networks and study the performance of these Project is proceeding well, with progress on the core models and algorithms on real-world data-sets with ap- technical goals. plication to object detection, classification, and change/ anomaly detection; (2) extend the theoretical framework As planned, as of mid-FY13 we have now integrated of sparse representations and compressive sensing to the advanced sparse representation algorithms (non- learn hierarchical sparse representations that describe convex basis pursuit denoising algorithms) into the high video data of natural scenes at many spatial/tempo- performance codes developed in FY13 (PANN), produc- ral scales and many levels of object complexity; and, ing a new code we call VAST (Video Analysis and Search (3) exploit many-core CPU and massively parallel GPU Technology). We were the highest-ranked Institutional (graphical processing unit) computing technology for Computing proposal for 2013, and have been awarded efficient training and fast execution of our models and 1.5M core-hours/years for 2013 and 2014, which will algorithms, and explore size, weight and power (SWaP) cover all the key work and demonstrations planned. constraints on algorithm and software design for next- The team has now produced a total of 22 publications: 1 generation smart sensors. published Journal publication (IEEE Trans. Signal Pro- cessing), 14 refereed/published conference proceedings 472

Page 473

papers (including high-impact ICASSP and IJCNN), 2 ref- are also supporting one of the student teams at this year’s ereed published abstracts, and 5 unpublished conference LANL Engineering Institute’s Dynamics Summer School, papers (SIAM, DOE CoDA, invitation-only Snowbird Learn- which provides us with the opportunity to work with video ing Workshop). data collected by a small remotely-piloted helicopter. The team has identified and is collecting a new partially- Future Work labeled video dataset to support our core research, the As per our original project proposal, FY13 work will focus Twitter Vine social media video dataset of public data, cur- on the following tasks: rently being posted worldwide at a rate of ~1M videos/day. We now have over 1 petapixel of video from this source,

1. Semi-supervised multi-class learning - we will extend which has generated significant interest amongst sponsors our models of sparse dictionary learning to greater and visitors. Social media video is an exciting new type numbers of object categories (i.e., ~1000 common of data, with many real-world applications ranging from object categories), using unsupervised learning to epidemiology to facility security to intelligence, and the describe the unlabeled video frame backgrounds. Our volume of social media video is expected to explode with goal is to compare our new algorithms to the state-of- the arrival of wearable networked video cameras (e.g., the-art using the annual public ImageNet Large-Scale Google Glass). This data complements existing holdings of Visual Recognition Challenge (ILSVRC) as our bench- satellite imagery, aerial video, ground video, astronomical mark. imagery, and social media (still) imagery that we continue
2. Non-linear low dimensional + sparse models - we will to work on. extend our models of low rank + sparse frame decom- positions to incorporate sensor platform motion and We are working with LANL Technology Transfer (TT) Divi- the non-linear video frame transformations induced by sion to support Intellectual Property (IP) protection and our multi-scale, multi-object learned dictionaries. This licensing of tools developed under our LDRD-DR. TT has work will contribute to our ILSVRC benchmarking, and contracted external counsel to support three patent filings to analysis of video sequences. based on our work.
3. MPI/GPU code development - we will implement the This year we organized and co-sponsored a CNLS Work- above novel algorithms for our supercomputing plat- shop on Statistical Image Analysis in conjunction with our forms, including LANL’s new 200+ GPU “Moonlight” mid-project review. Feedback from our mid-project review Institutional Computing resource. was very encouraging. Conclusion The capability we are developing under this LDRD-DR has We will develop a neuroscience-inspired computer vision have brought in a number of DoD and USG Phase I dem- system based on a hierarchical sparse generative model onstration projects and members of our team are support- executing faster-than-real-time processing of natural video. ing new projects won from DARPA (DARPA/MTO UPSIDE This system will learn robust, multi-scale features for reli- program) and NNSA+DTRA, with total funding exceeding able data triage, and learn to reconstruct the relevant parts 2M$. Several proposals for additional follow-on projects of the signal from a representation that serves as a basis are in preparation, with Program Development funding for classification, segmentation, and change detection. from GS-PO. We have also started developing a collabora- This capability will support applications in proliferation tion with New York University (NYU) Polytechnic Center for detection, warfighter support, environmental science, and Urban Science and Progress (CUSP), which grants us access large-scale visualization of simulations. Our theoretical to urban remote sensing video and social media datasets framework and high performance computing algorithms that directly support our core technical goals and position could also be applied to forensics, genomics, materials us for follow-on projects. science, spectral analysis, and audio and text data-sets including web text. With regards to personnel changes, Post-Doc Zhengping Ji left LANL for a position at Samsung Research in Pasadena, CA. Graduate Student Dylan Paiton was accepted into the Publications UC Berkeley Vision Science Ph.D. program as the student of Bachega, L., J. Theiler, and C. Bouman. Evaluating and im- Prof. Bruno Olshausen. Post-Masters student Scott Halv- proving local hyperspectral anomaly detectors. 2011. In 2011 IEEE Applied Imagery Pattern Recognition erson from D-3 has joined the team, and we are currently Workshop: Imaging for Decision Making, AIPR 2011 ; interviewing potential Post-Docs to replace Zhengping. We 20111011 - 20111013 ; Washington, DC, United States. 473

Page 474

, p. var.pagings. (Baltimore, MD, 29 April- 3 May 2013). , p. 87500K. Baltimore: SPIE. Brumby, S.. Image fusion for remote sensing using fast, large-scale neuroscience models. 2011. In Multisen- Ji, Z., W. Huang, and S. Brumby. Learning sparse represen- sor, Multisource Information Fusion: Architectures, tation via a nonlinear shrinkage encoder and a linear Algorithms, and Applications 2011 ; 27-28 April 2011 ; sparse decoder. 2012. In 2012 Annual International Orlando, FL, USA. Vol. 8064, p. 806402 (8 pp.). Joint Conference on Neural Networks, IJCNN 2012, Part of the 2012 IEEE World Congress on Computational Brumby, S. P., M. I. Ham, and G. T. Kenyon. Semi-supervised Intelligence, WCCI 2012 ; 20120610 - 20120615 ; Bris- learning of high-level representations of natural video bane, QLD, Australia. , p. var.pagings. sequences. Presented at Computational and Systems Neuroscience (COSYNE) . (Salt Lake City, UT, 23-26 Feb Ji, Z., and S. P. Brumby. Integrating Bottom-up and Top- 2012). down Visual Attention for Object Segmentation. Pre- sented at Vision Sciences Society (VSS) 2012 Annual Chartrand, R.. Nonconvex Splitting for Regularized Low- Meeting. (Naples, FL, 11-16 May 2012). Rank plus Sparse Decomposition. Invited presentation at Joint Mathematics Meetings. (Boston, MA, 4-7 Jan. Moody, D. I., S. P. Brumby, J. C. Rowland, and C. Gango- 2012). dagamage. Learning sparse discriminative representa- tions for land cover classification in the Arctic. 2012. Chartrand, R.. Nonconvex Splitting for Regularized Low- In SPIE Optics + Photonics Annual Meeting 2012. (San Rank + Sparse Decomposition. Invited presentation at Diego, CA, 11-16 Aug. 2012). , p. 85140Q. San Diego: Compressive Sensing Workshop: Leveraging Sparsity at SPIE. UCLA & Beyond. (Los Angeles, CA, 6-8 March 2012). Moody, D., S. Brumby, K. Myers, and N. Pawley. Sparse clas- Chartrand, R., and B. Wohlberg. A nonconvex ADMM algo- sification of rf transients using chirplets and learned rithm for group sparsity with sparse groups . 2013. In dictionaries. 2011. In 45th Asilomar Conference on IEEE International Conference on Acoustics, Speech, Signals, Systems and Computers, ASILOMAR 2011 and Signal Processing . (Vancouver, BC, 26-31 May ; 20111106 - 20111109 ; Pacific Grove, CA, United 2013). , p. ?. Vancouver: IEEE. States. , p. 1888. Chartrand, null.. Image processing and reconstruction. Paiton, D. M., S. P. Brumby, G. T. Kenyon, G. J. Kunde, K. 2012. LDRD. D. Peterson, M. I. Ham, P. F. Schultz, and J. S. George. Combining Multiple Visual Processing Streams for Galbraith, A. E., R. Chartrand, B. Wohlberg, and S. P. Locating and Classifying Objects. Presented at IEEE Brumby. Simulating vision through time: Hierarchical, Southwest Symposium on Image Analysis and Interpre- sparse models of visual cortex for motion imagery. tation (SSIAI) 2012. (Santa Fe, NM, April 22-24, 2012). 2012. In IEEE Applied Imagery and Pattern Recognition (AIPR) 2012. (Washington, DC, 23-25 Oct, 2012). , p. 1. Rodriguez, P., and B. Wohlberg. Iteratively Reweighted Washington, DC: IEEE. Least Squares for L1 Problems: Boring But Still Effec- tive. Presented at 2012 SIAM Conference on Imaging Gunawardena, N., J. Heit, G. Lederman, A. Galbraith, D. Science. (Philadelphia, PA, May 20-22 2012). Mascarenas, and S. P. Brumby. Remote placement of magnetically coupled ultrasonic sensors for structural Rodriguez, P., and B. Wohlberg. Fast principal component health monitoring . Invited presentation at Interna- pursuit via alternating minimization . Invited presenta- tional Modal Analysis Conference. (Orlando, FL, 3-6 tion at IEEE International Conference on Image Pro- February, 2014). cessing . (Melbourne, AU, 15-18 Sept, 2013). Huang, W., Z. Ji, S. Brumby, G. Kenyon, and L. A. Betten- Rodriguez, P., and B. Wohlberg. A comparison of the com- court. Development of invariant feature maps via a putational performance of iteratively reweighted least computational model of simple and complex cells. squares and alternating minimization algorithms for 2012. In 2012 Annual International Joint Conference l1 inverse problems . 2012. In IEEE International Con- on Neural Networks, IJCNN 2012, Part of the 2012 IEEE ference on Image Processing. (Orlando, FL, 30 Sept- 3 World Congress on Computational Intelligence, WCCI Oct, 2012). , p. 3069. Orlando: IEEE. 2012 ; 20120610 - 20120615 ; Brisbane, QLD, Australia. , p. var.pagings. Theiler, J., B. Wohlberg, and R. Chartrand. Local Robust Principal Components for Nonlinear Datasets,. Pre- Ji, Z., J. Theiler, R. Chartrand, G. Kenyon, and S. P. Brumby. sented at 2012 SIAM Conference on Imaging Science. SIFT-based sparse coding for large-scale visual recog- (Philadelphia, PA, May 20-22 2012). nition. 2013. In SPIE Defense, Security, and Sensing. 474

Page 475

Theiler, J., G. Cao, L. Bachega, and C. Bouman. Sparse Ma- trix Transform for Hyperspectral Image Processing. 2011. IEEE JOURNAL OF SELECTED TOPICS IN SIGNAL PROCESSING. 5 (3): 424. Theiler, J., and B. Wohlberg. Detection of spectrally sparse anomalies in hyperspectral imagery. 2012. In 2012 IEEE Southwest Symposium on Image Analysis and Interpre- tation, SSIAI 2012 ; 20120422 - 20120424 ; Santa Fe, NM, United States. , p. 117. Theiler, J., and B. Wohlberg. Regression framework for background estimation in remote sensing imagery. To appear in Institute of Electrical and Electronics Engi- neers (IEEE) Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing (WHIS- PERS) . (Gainesville, FL, 25-28 June 2013). Theiler, J., and B. Wohlberg. Detection of unknown gas- phase chemical plumes in hyperspectral imagery. 2013. In SPIE Defense, Security, and Sensing. (Baltimore, MD, 29 April- 3 May 2013). , p. 874315. Baltimore: SPIE. Venkatakrishnan, S. V., C. A. Bouman, and B. Wohlberg. Plug-and-play priors for model based reconstruction. 2013. Purdue University, ECE Technical Reports. Venkatakrishnan, S. V., C. A. Bouman, and B. Wohlberg. Plug-and-play priors for model based reconstruction. To appear in IEEE Global Conference on Signal and In- formation Processing . (Austin, TX, 3-5 Dec. 2013). Wohlberg, B.. Inpainting by Joint Optimization of Linear Combinations of Exemplars. 2011. IEEE Signal Process- ing Letters. 18 (1): 75. Wohlberg, B., R. Chartrand, and J. Theiler. A Robust Lo- cal Linear Data Decomposition. Presented at Learning Workshop. (Snowbird, UT, April 3-6, 2012). Wohlberg, B., R. Chartrand, and J. Theiler. Local Principal Component Pursuit for Nonlinear Datasets. Presented at IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP) (2012). (Kyoto, Japan, March 25-30, 2012). 475

Page 476

Information Science and Technology Directed Research Continuing Project Empowering the Expert: Machine Learning with User Intelligence Reid B. Porter 20130013DR Introduction fective tools can be built. Scientists and intelligence analysts are very good at un- derstanding and exploiting the specialized datasets with Benefit to National Security Missions which they work; the problem is that there is too much This project empowers scientists and defense analysts to of this data for them to look at. Data processing tools exploit specialized data in less time, and with greater ac- help to clean up, filter and identify the most relevant curacy, both by automating tedious and repetitive tasks subsets of data, but recent developments in machine and by enabling them to focus on validation and higher- learning have indicated that there is a grand opportunity level objectives. This general capability will be valuable to dramatically enhance the efficiency and accuracy of to a large number of government agencies and missions. this process, by involving the user in a more interactive This project will also develop specific tools for (at least) dialog. two LANL applications. Traditional machine learning employs a fixed set of Nuclear Material Forensics labels, supplied up-front by the user, to optimize data Experts draw upon a vast understanding of the nuclear processing tools before they are applied to a larger fuel cycle to quantify spectroscopy images, but they dataset or data archive. This approach has proven suc- spend weeks each year manually correcting tool esti- cessful both in theory (rigorous proofs) and in practice mates of particle and grain boundaries. This project’s (commercial applications). Machine learning with user image quantification advances will be valuable to DOE, intelligence pushes the state of the art in two directions. DHS, and DOD. From up-front learning to interactive learning Cyber Security Traditional machine learning tools stop when the human Analysts draw upon a deep understanding of the com- steps in. This means the expert is still the bottleneck in puter network as well as attacker behavior to determine data exploitation: spending too much time in tedious if anomalous network traffic requires further investiga- and repetitive post-processing tasks and not enough tion. However there is simply too many anomalies for time on validation and research. Machine learning with analysts to perform even the most cursory triage. This user intelligence starts when the human steps in, and project’s cyber-security tools will be valuable to DOE, tools are a force multiplier, where the user’s intelligence DHS and DOD, in particular for the protection of the is the force to be multiplied, not an expense to be mini- NNSA complex itself. mized. Progress From label learning to relational learning Within the first year we developed and submitted a Machine learning with user intelligence formalizes the comprehensive review article on ‘Interactive Machine interactive dialog with relational objects, stated in terms Learning.’ This article provides a unified view of this of motifs, clusters, alignments, matches and other gen- emerging field and provides a road-map for both our eralizations of standard labels. By providing additional project, and the larger community. Our team was invited interface “bandwidth” between the domain expert and to present this article at one of the first workshops on the data processing tool (and thus creating a better Interactive Machine Learning that was held in February expression of the dialog in terms of graphical models), 2013. The roadmap articulates the two main technical higher quality associations are produced, and more ef- 476

Page 477

thrusts in our project. tions between Belief Propagation and Perfect Match- ings. We call the first thrust the ‘Training Dialog’, and it moves us from up-front learning to interactive learning (as de- A number of proposals have been developed based on our scribed in the project description). Progress included: work. These include: • We developed a number of experiments to quantify • ‘Millions of Spies, Millions of Eyes: Active Crowdsourc- the benefit of having the ‘user-in-the-loop’ compared ing for Nonproliferation’, Christy Ruggiero, Reid Porter. to traditional batch learning. This proposal builds on work with our collaborators at • We are also developing an Expectation Maximization UCSC [10] and it proposes to develop a crowd-sourced based framework for interactive anomaly detection interactive learning system for nonproliferation and that can adapt to different types of user input. The arms control (submitted to DOE). approach starts with non-interactive machine learn- • ‘Maximizing Relevance Feedback for Intelligence Mes- ing tools (e.g. anomaly or change detection tools) and sage Traffic Analysis’, David Izraelevitz, Reid Porter, then incorporates increasingly complex user feedback submitted to DHS. This proposal builds on our ad- (e.g. labels for one class, labels for multiple classes vances in the ‘Training Dialog’ and proposes to inte- etc.). grate, test and deploy these advances to DHS tools and • We developed a number of new machine learning applications. methods that expand the range of tools and potential applications that we can use for interactive learning. Future Work • We also collaborated with the Cyber-Security team at Task: Develop methods for structured output prediction. Los Alamos, to develop a prototype software interface Goal: Efficient and general purpose learning algorithms to for triage of network traffic anomalies. This prototype support increased user interaction. will enable us to test our methods on real network traffic data. Task: Develop interactive learning strategies for Graphical Models. We call the second thrust the ‘Training Vocabulary, and Goal: Algorithms that are robust to missing data and/or it moves us from label learning to relational learning (as limited training examples. described in the project description). Progress included: Task: Identify and develop efficient strategies for combin- • We developed new algorithms that learn from user ing inference and learning. examples to ‘merge’ segments of an image into larger Goal: A better understanding of the accuracy and robust- (more meaningful) regions. These algorithms exploit ness of interactive learning with graphical models. ‘connectivity’ properties and provide a foundation for developing tools for image quantification applications Task: Implement initial tools for image quantification and in material science, bio-medical imaging and remote cyber-security applications. sensing. Goal: Software that domain experts can use on practical datasets and an experimental platform for the project. • We developed a test and evaluation framework for ‘learning to merge’ and also developed a software Task: Develop synthetic and benchmark experimental prototype for demonstrating our tools. We also started frameworks for interactive learning. to prototype a web interface for this application with Goal: An ability to evaluate the accuracy and utility of tools the objective of hosting a crowd-sourced experiment. for practical applications at various levels of interaction. This experiment will provide a more extensive test and evaluation framework for ‘learning to merge’ and Conclusion will help increase the profile of our project within the larger research community. Our goal is to provide a general framework for data pro- cessing tools that directly targets the post-processing • The ‘merge’ problem is an example of a much larger bottleneck. We emphasize graphical models in this devel- (and fundamental) problem in machine learning opment, as these provide a balance of rigor and flexibil- known as structured output prediction, and this year ity on the learning front and domain adaptability on the our team has advanced the theoretical underpinnings interactivity front. An additional goal is to develop new for this problem. These advances make new connec- 477

Page 478

tools, optimized for specific applications in nuclear mate- rial forensics and cyber-security, where there is a critical need to make data exploitation more accurate and more efficient. Publications Chertkov, M., A. Gelfand, and J. Shin. Loop calculus and bootstrap-belief propagation for perfect matchings on arbitrary graphs. arXiv:1306.1267. Gelfand, A., J. Shin, and M. Chertkov. Belief propagation for linear programming. CoRR, arXiv:1305.4130. Harvey, N.. User driven sampling strategies in visual data exploitation. To appear in Visualization and Data Analy- sis, SPIE, 2014.. (San Franciso, 5-6 February, 2014). Loxley, P.. Learning dictionary statistics from natural im- ages. CoRR. abs/1304.0023. Oyen, D., A. Niculescu-Mizil, R. Ostroff, A. Stewart, and V. P. Clark. Reliable differential dependency network analy- sis. Presented at International Conference on Machine Learning. (Beijing, China, 21-26 June). Oyen, D., and T. Lane. Bayesian discovery of multiple Bayesian networks via transfer learning. To appear in IEEE International Conference on Data Mining. (Dallas, Texas, 7-10 Dec). Porter, R. B., S. Lundquist, and C. Ruggiero. Learning to merge: a new tool for interactive mapping. Porter, R., J. Theiler, and D. Hush. Interactive machine learning in data exploitation. To appear in IEEE Com- puters in Science and Engineering. Shin, J., A. Gelfand, and M. Chertkov. A graphical trans- formation for belief propagation: maximum weight matchings and odd-sized cycles. CoRR, abs/1306.1167. Theiler, J.. Spatio-spectral anomalous change detection in hyperspectral imagery. To appear in 1st IEEE Global Signal and Information Processing Conference. (Austin, Texas, 2-3 Dec). Theiler, J.. Matched pair machine learning. To appear in Technometrics. Vaish, R., S. Lundquist, R. Porter, and J. Davis. Crowd-sourc- ing bounding box annotations for computer vision. Presented at HCOM 2013: Human computation. (Palm Springs, California, 6-9 Nov). Wolpert, D., and D. Rajnarayan. Using machine learning to improve stochastic optimization. 2013. In Twenty- Seventh AAAI Conference on Artificial Intelligence. (Bellevue, Washington, 14-15 July). , p. 146. Bellevue, Washington: AAAI. 478

Page 479

Information Science and Technology Directed Research Final Report Multi-perspective Network-scale Modeling and Detection for Cyber Systems Michael E. Fisk 20110093DR Abstract are as important as time-series observed at single points Our work was at the intersection of “big data” or data- in the graph. intensive computing, statistics, graph modeling, and Scientific Approach and Accomplishments graph algorithms. We established a new regime of cyber security detection and defense that uses whole-network Our accomplishments were in two major areas which are traffic modeling and detection. In contrast, previous described in the following sections on subgraph mod- detection techniques were largely myopic and unable to eling and detection and on whole-graph modeling to detect the most subtle of contemporary attacks. In our detect and cluster malware execution. approach, cyberspace activity from multiple observation Model-Based Anomalous Subgraph Detection perspectives is represented as a graph with edges (and For this work, we have developed a family of temporal nodes) described as a time-series measurements. We graph models. The building blocks for those models then build data-driven models of the localized dynamics are models for individual edges and nodes. To model in these graphs and use those models to detect anoma- activity time-series for each edge in a graph, we first lous change in the dynamics. In particular we focus on developed two-state hidden Markov model that learns change that is consistent with attacker behavior. We two different distributions of activity representing active use similar techniques to examine the dynamic execu- and inactive (or high and low) levels of activity com- tion of malicious software and to detect and cluster monly seen in these edges. These models are then used malicious software based on a combination of dynamic to measure the likelihood of new events with respect to models and static features of the software. Finally, we previous activity. Fitting these models to data is com- developed predictive models that examine the resilience putationally intensive. For example, using our data- of command and control overlay networks (including intensive computing framework, we modeled a month’s botnets) to different strategies for detecting and disrupt- worth of traffic on the LANL network (550,000 edges ing those communications networks. modeled). That computation required 1.5 CPU-years, but was completed in just over 2 days on a 256-core cluster. Background and Research Objectives In subsequent work, we discovered that for some impor- The field of computer intrusion detection is dominated tant real-world data sets, new edge formation is much by detection decisions that can be made form single more statistically significant than increased usage of pre- observation points of activity from a single actor and/ viously existing edges. We developed and implemented or against a single target. Further, most previous work models for the formation of new edges in the graph. We is focused on signature-based techniques. Our objec- modeled the creation of new edges as a process that tives were to fuse different observational perspectives is dependent on the edge creation behavior of the two together in a single analytical context, to model the nodes participating in that edge. normal dynamics of these behaviors, to enable model- based detection of anomalous change that is consistent We have examined multiple methods for identifying with malicious activity. subgraphs which contain coincident anomalies. First, we developed a detection approach that enumerates every This work leveraged previous work at Los Alamos subgraph that contains a fixed shape (such as a 3-hop National Laboratory on computer misuse and anoma- line or a star) and computes a joint likelihood of the lous change detection and to expand that work to the time-series data observed across that entire graphlet. context of graphs where relationships between entities 479

Page 480

By simple combinatorics, there are immense numbers of such a core of a large graph, after removing which the these graphlets in real-world graphs, but we have de- remaining graph has only small strongly connected com- veloped a hybrid parallel (both SMP and MPI) library for ponents; thus only a small number of nodes and messages enumerating these graphlets and computing likelihood need to be monitored. We worked on the Twitter dataset (or other arbitrary computation) on each. Using the LANL with more than 40 million nodes and 1.2 billion follow- network data set again, we enumerated and measured the ing relationships. We have modeled how malware spread anomalousness of 300 million graphlets in 5 seconds on in online social networks is affected by the structure and a 48-core computer. We have also scaled this anomalous activity patterns of social network users. graphlet detection work up to 1015 graphlets in less than Finally, we have also combined Bayesian networks and 2 minutes of computation. We validated this techniques game theory to develop a comprehensive framework for against well-understood data sets in LANL’s own network evaluating consequences of DDoS attacks and effectiveness data and demonstrated that actual intrusions were among of multi-layer defense mechanisms. We have also develop- the most anomalous events during those time periods. ing graph similarity-based metrics and software to study We have also modeled the frequency at which graphlets evolution of Email communications. We also examined occur in a given location in a graph in order to determine social networks, the spread of information in those net- when anomalous structures exist and utilized additional works, and how to detect malicious use of those networks. criteria to reduce the search space to events that are con- We developed graph-theoretic tools to identify suspicious sistent with certain malicious behaviors. For example, we nodes that can potentially pass C&C (command and con- developed the notion of telescoping subgraphs in which trol) information to a large number of individual bots using the duration of one edge encompasses the duration of online social networks. subsequent edges in the subgraph. Finally, we performed Malware Detection and Clustering some initial work to model attacker utility functions such We developed a novel approach to malware detection that we can more generally restrict the search to anoma- based on the analysis of dynamic instruction traces from lies that represent rational decision making by an adver- a large number of executable files. The base method can sary. be trivially extended to an online procedure, provided the Much of our work was on network flow data, but we also engineering hurdle of obtaining a sample of the instruc- developed models and associated detection results that tions executed by a given process can be satisfactorily incorporate observations of web requests, anomalous addressed. An execution trace collected from a running processes on nodes, and network traffic between nodes. process is summarized by a state transition probability We also developed techniques to register e-mail events matrix, P, like that used to represent the probabilistic on computer communication graphs by mapping users to evolution of a Markov Chain. P is estimated by combining computers based on e-mail authentication and data and counts on all possible consecutive pairs of instructions with where e-mail is sent. a prior model posing that the rows of P are independent Dirichlet vectors. We have also evaluated how well peer-to-peer botnet communications graphs can be detected when sensor One binary classification approach that we developed is coverage limits visibility to subsets of the communications to model from estimated P matrices for a large number of graph and strategies that bot-herders could deploy to im- known malicious and known benign software samples us- prove the resilience and robustness of peer-to-peer based ing nonparametric logistic regression on the elements of P. botnet structures and possible countermeasures by the The elastic net technique is used to perform logistic regres- defender as a response. We have expanded this work, by sion with careful selection of useful variables from among allowing bots to use spoofing packets to hide their identi- the many thousands of predictors represented by the ties. The number of spoofing source IP addresses per bot elements of P. We illustrate the utility of the method on a is generated from high-variations distributions, so the sum sample of malware, and compare the results favorably to over all observed bots converges slowly. Such behavior other leading malware detection schemes (both signature allows a bot-herder to obfuscate the size of a botnet. As and classification based). We also pursued a more classi- a final part of our focus on botnets, we explored the idea cally statistical approach, using Dirichlet models to perform of graph-based detection bots hidden in Twitter-like online the classification. The false positive/true positive trade-off social networks. Our key idea is to look for the core of large showed slightly better performance than the previous graphs, which remain relatively stable over its evolution. approach. In addition, these methods provide confidence Botnets are more likely to stay in the periphery of such intervals around the probability of maliciousness, provid- large graphs. We developed efficient algorithms to find ing analysts and automated methods with a quantification 480

Page 481

of the uncertainty of the classification result. A top-tier Publications statistics journal article on this work is in under review. Anderson, B., C. Storlie, and T. Lane. Multiple Kernel Learn- ing Clustering with an Application to Malware. 2012. We have also pursued machine learning techniques on In Data Mining (ICDM), 2012 IEEE 12th International this same data. In this approach we use multiple graph Conference on. , p. 804. IEEE. kernels to provide distance metrics between pairs of Anderson, B., C. Storlie, and T. Lane. Improving malware Markov model graphs and use machine learning algo- classification: bridging the static/dynamic gap. 2012. In rithms to build classifiers. We have developed a similar Proceedings of the 5th ACM workshop on Security and approach for anomalous change detection that identifies artificial intelligence. , p. 3. ACM. when a program’s execution changes more than it usually does. We have developed new machine learning tech- Anderson, B., D. Quist, J. Neil, C. Storlie, and T. Lane. niques, motivated by two goals. The first goal is to improve Graph-Based Malware Detection Using Dynamic Analy- the classification of malware versus benign software. To sis. 2011. Journal in Computer Virology. 7 (4): 247. this end, we have developed multi-view kernel methods, Anderson, B., D. Quist, and T. Lane. Detecting Code Injec- which combine various features of a given executable file. tion Attacks in Internet Explorer. 2011. In 35th Annual This approach is novel in the literature, and a paper has IEEE International Computer Software and Applications been submitted for review. The results are very promising, Conference STPSA workshop, 2011. (Saarbruecken, achieving far better performance than current state-of-the- 6-10 June 2011). , p. 90. Saarbrucken: IEEE. art signature methods, as well as other statistical methods. Arackaparambil, C., and G. Yan. Wiki-watchdog: Anomaly Second, we have developed methods for relating classes of detection in wikipedia through a distributional lens. malware to each other, with the ultimate goal of identify- 2011. In Web Intelligence and Intelligent Agent Tech- ing families of malware and establishing a phylogenetic nology (WI-IAT), 2011 IEEE/WIC/ACM International tree. Initial findings include good clustering of different Conference on. (Lyon, France, 22-27 August 2011). Vol. types of malware. In addition, the malware which did 1, p. 257. New York: IEEE. not fall cleanly into a cluster, but was fairly close to two Djidjev, H. N., and M. Onus. Scalable and Accurate Graph different clusters actually informed us on the content of Clustering and Community Structure Detection. 2013. those two clusters. It was discovered that in fact those two IEEE Trans. Parallel Distrib. Syst.. 24 (5): 1022. clusters tended to have some features which were com- mon between them, and the malware which lay between Djidjev, H., G. Sandine, C. Storlie, and S. Vander Wiel. the clusters were exemplified by these features. This is a Graph Based Statistical Analysis of Network Traffic. promising aspect of this research, as it provides promise 2011. In Ninth Workshop on Mining and Learning with that research into phylogenetic trees is indeed tractable. Graphs. (San Diego, CA, 21-22 Aug. 2011). , p. 1. San Diego, CA: IEEE. Impact on National Missions Djidjev, H., and G. Sandine. Graph-Based Traffic Analysis for This project will positively impact cyber security missions Network Intrusion Detection. 2011. TSC Science High- executed by both LANL’s internal security program, essen- lights, LA-UR 10-06320. tial to our DP mission, and Global Security programs, with secondary benefits on our commercial CRADAs (we are Djidjev, H., and G. Sandine. Graph-Based Traffic Analysis for Network Intrusion Detection. 2010. In 2010 Annual not funding any CRADAs, but technology developed under Computer Security Applications Conference (ACSAC). this LDRD may become background intellectual property (Orlando, F, 5-9 December 2010). , p. 1. Orlando, FL: that results in funding of subsequent work by an existing ACM. CRADA partner). Our advances have put us on a path to catch-up with and keep pace with contemporary, undetect- Eidenbenz, S., M. Marathe, and A. Sen. Editorial for Com- able adversaries, and build a base for the future. Moving puter Networks special issue on ‘’Towards a Science of forward, we are also pursing ways to apply these dynamic Cyber Security’’. 2013. Computer Networks. 57 (10): graph modeling capabilities to other strategic analytical 2119. problems including biosurveillance and video analysis. Fi- Jin, D., D. Nicol, and G. Yan. An Event Buffer Flooding At- nally the data-intensive demands of this problem have led tack against DNP3-Controlled SCADA Systems. 2011. In to insight on data-intensive computing, use of LANL’s HPC Proceedings of the 2011 Winter Simulation Conference platforms for data-intensive computing, and data manage- (WSC’11). (Phoenix, AZ, 5-9 December, 2011). , p. 1. ment methodologies and abstraction layers for diverse, Phoenix, AZ: IEEE. high-volume data. 481

Page 482

Kakumanu, S., S. Eidenbenz, and R. Sivakumar. Lattice rout- peer-to-peer botnets. 2011. Computer Networks. 55 ing: A 4D routing scheme for multiradio multichannel (8): 1941. ad hoc networks. 2011. Ad Hoc Networks. 9 (1): 95. Yan, G., G. Chen, S. Eidenbenz, and N. Li. Malware propaga- Kent, , and L. M. Liebrock. Differentiating User Authentica- tion in online social networks: Nature, dynamics, and tion Graphs. 2013. In Security and Privacy Workshops defense implications. 2011. In 6th International Sym- (SPW), 2013 IEEE. , p. 72. IEEE. posium on Information, Computer and Communica- tions Security, ASIACCS 2011 ; 20110322 - 20110324 ; Liu, L., X. Zhang, G. Yan, and S. Chen. Chrome extensions: Hong Kong, China. , p. 196. Threat analysis and countermeasures. 2012. Network and Distributed System Security Symposium (NDSS). : Yan, G., S. Chen, and S. Eidenbenz. RatBot: Anti-Enumera-

1. tion Peer-to-Peer Botnets. 2011. Information Security. 1 (1): 135. Neil, J. C., B. D. Uphoff, C. B. Storlie, and C. L. Hash. To- wards Improved Detection of Attackers in Computer Zeng, Yuanyuan, Guanhua Yan, S. Eidenbenz, and K. Shin. Networks: New Edges, Fast Updating, and Host Agents. Measuring the Effectiveness of Infrastructure-level
2. In 1st International Symposium on Resilient Cy- Detection of Large-scale Botnets. 2011. In 2011 IEEE ber Systems. Nineteenth IEEE International Workshop on Quality of Service (IWQoS 2011) ; 6-7 June 2011 ; San Jose, CA, Neil, J., C. B. Storlie, C. Hash, Brugh, Alexander, and M. USA. , p. 1. Fisk. Scan Statistics for the Online Detection of Locally Anomalous Subgraphs. 2013. Technometrics. Zhang, R., J. Sun, G. Yan, and S. Zhang. Fine-grained private matching for proximity-based mobile social network- Nguyen, N. P., G. Yan, M. T. Thai, and S. Eidenbenz. Con- ing. 2012. In INFOCOM, 2012 Proceedings IEEE. (Or- tainment of Misinformation Spread in Online Social lando, FL, Mrach 25 - 30). , p. 1969. Orlando, FL: IEEE. Networks. 2012. WebSci ‘12 Proceedings of the 3rd Annual ACM Web Science Conference . : 213. Pelechrinis, K., G. Yan, S. Eidenbenz, and S. Krishnamurthy. Detection of Selfish Manipulation of Carrier Sensing in 802.11 Networks. 2011. Mobile Computing, IEEE Trans- actions . : 1. Rahman, M., G. Yan, H. Madhyastha, M. Faloutsos, S. Ei- denbenz, and M. Fisk. iDispatcher: A unified platform for secure planet-scale information dissemination.
3. Peer-to-Peer Networking and Applications. 6 (1):

Rahman, S., G. Yan, M. Faloutsos, H. Madhyashta, S. Eiden- benz, and M. Fisk. iDispatcher: A Unified Platform for Secure Planet-Scale Information Dissemination. 2011. Los Alamos Unlimited Release LA-UR 11-02386. Yan, G.. Peri-Watchdog: Hunting for Hidden Botnets in the Periphery of Online Social Networks. 2011. Los Alamos Unlimited Release. Yan, G.. Peri-Watchdog: Hunting for Hidden Botnets in The Periphery of Online Social Networks. 2012. Computer Networks. : 1. Yan, G., C. Arackaparambil, S. Bratus, and A. Caglayan. On- Tuning the Knobs of Distribution-based Methods for Detecting VoIP Covert Channels. 2012. In Proceedings of Hawaii International Conference on System Sciences (HICSS-45). (Hawaii, 4-9 January, 2011). , p. 2431 . Ha- waii: IEEE. Yan, G., D. Ha, and S. Eidenbenz. AntBot: Anti-pollution 482

Page 483

Information Science and Technology Directed Research Final Report Correlations and Dynamics in Information Science Robert E. Ecke 20110434DR Abstract chanics techniques to understand information science Information, and its processing, transport, and vulner- questions in the large systems limit. ability, are matters of national security. The understand- Within these themes we applied information process- ing and management of information systems in com- ing tools to address the application areas of HIV and SIV puter and communication networks, biological systems transmission, cyberphysical security, electrical power and social dynamics play a vital role in multiple national systems (smart grids), computer vision, cybersecurity, security challenges. With this project, the Center for transport networks, and epidemic spreading. Nonlinear Studies (CNLS) targeted fundamental science challenges that are central to long-term strategies for Scientific Approach and Accomplishments improving information management. The science efforts Bioinformatics which focus on understanding the effects of “correla- We began a study of the genetic mutations of the Hepa- tions and dynamics” in information systems explored the titis C virus (HCV). The genetic diversity can be analyzed effects of random influences (stochasticity), the con- together with a model of random mutations to predict sequences and prospects of coordinating the action of time since infection. Preliminary results show that the pairs or higher numbers of units or ‘nodes’, the dynamics different mutation mechanisms in HCV will require mod- of timing and coordinating signals, and the accuracy of eling using multiple overlapping random processes [1]. network models. We studied the spread and evolution of viruses. We Background and Research Objectives traced the spread and dynamics of the Latvian HIV-1 epi- The science of this project focused on understanding demic using phylogenetic tools, finding an unexpected the effects of “correlations and dynamics” in informa- decrease in viral divergence due to the existence of a tion systems. This project targeted fundamental science fast and slow rate of spread occurring simultaneously challenges that are central to long-term strategies in in this epidemic. We re-analyzed a previously described information processing systems. HIV-1 transmission chain in light of a phylogenetic model and found, contrary to previous thinking, that it might Research was organized around three basic themes in not be possible to extract a unique transmission history information science: [2]. This work has implications for court cases that rely Bioinformatics: Sensing and processing of information, on sequencing data to support transmission claims. bioinformatics, and bio-inspired methods of speeding Most notably embodied by Google Flu Trends, real-time searches and fast processing of information. estimates of the impact of disease based upon Internet Complex Networks: Modeling and analysis of complex data such as search queries and Twitter messages are systems, applied mathematics analysis of tractable emerging. These approaches use linear models of a models, methods to develop real-world models, descrip- single data source to measure a small number of dis- tion of epidemics, and the development of fast, robust eases (influenza and dengue fever). Our current work methods for real-time cyber security analysis. extended this research direction by using non-linear models that include interaction features to measure a Algorithms, Optimization, & Control: Inference and variety of diseases across the globe, using multiple publi- learning with stochastic information, and statistical me- cally available data sources (Twitter messages, Wikipedia 479

Page 484

access logs, and Wikipedia edit logs). This work also seeks to approximate an arbitrary collection of binary random to provide short-term forecasts or identify fundamental variables (possibly from sample data), and applied the limitations of the data streams that make such forecasting algorithm to modeling voting [8]. impossible. We considered various important problems in communica- We studied methods to improve and extend simulations of tions, e.g., decoding of Low-Density Parity-Check codes, bio-chemical networks such as signal transduction net- detection in two-dimensional intersymbol interference works. A better understanding of the dynamics of signal- channels (magnetic and optical storage, cellular networks), ing pathways would have a significant impact on cancer etc., and modeled them as inference problems over research. Unfortunately, typical systems of interest are graphs. too large to model using ordinary differential equations We used our information science capabilities to improve and are currently only amenable to simulation that is often materials modeling. In a major advance, we developed a slow. We developed an approximate model reduction new algorithm to simulate fermions coupled to a classical method that takes advantage of the modular character field. By using the technique of “automatic differentia- of most biochemical networks. The technique harnesses tion” as the basis of a numerical method, our algorithm short bursts of simulation to approximate knowledge of decreases the cost of a calculation significantly, enabling the (unknown) underlying differential equations [3]. This simulations of unprecedented system sizes, and revealing can then be used to significantly speed up simulations and previously unknown quantum phases. We obtained coarse- even fed back into standard numerical analysis tool-kits to grained dynamics of a quenched ferromagnet, generalizing do stability or bifurcation analysis of the system. conformal field theory results on a two-dimensional Ising Algorithms, Optimization, and Control model to higher dimensional, anisotropic systems. We developed image processing algorithms to learn ob- In the area of energy distribution, we developed an algo- jects from images of natural scenes. A major achievement rithm for determining the optimal placement and timing of was made by showing how you can learn a sparse (small battery exchange and charging stations (for plugin hybrid data) representation of images to create an algorithm that electric vehicles – PHEV) [9]. This project produced a novel accurately categorizes objects. This algorithm performs gradient descent algorithm for optimizing the placement well on for recognizing noisy handwritten numerals and of the stations that relied on continuous time optimization beats existing algorithms on community standard image formulations. One of the more interesting aspects of the labeling tests. One of the main approaches [4], [5] is a approach was the first-ever ability to model demand for deep generative model. The models are generative in that PHEV charging as a function of the stations already built. they postulate possible images that fit the data, and deep In other words, we can model situations where technology in that they fuse information at multiple semantic levels. availability increases technology adoption (i.e., the avail- The models also incorporate sparse coding, processing ability of stations increases the adoption of PHEVs). We information by representing it with a sparse set of coef- developed a heuristic procedure for optimizing the place- ficients. The hierarchical sparse coding approach, which ment and sizing of flexible energy storage in a grid that allowed information to flow both up and down the seman- includes renewable energy sources, and studied a model of tic hierarchy, won the best poster paper award at the 2011 stochastic temporal fluctuations in a power grid. IEEE International Joint Conference on Neural Networks. More recently, ideas from continuous optimization led to Networks the development of surrogate functions that allow the We analyzed data integrity attacks for smart grid applica- computational learning tasks to be performed much more tions [10]. Real power injections at loads and generators, efficiently. This work has been applied to classification of and real power flows on selected lines in a transmission images [5], [6], identification of objects within images [5], network are monitored and transmitted to the system and identification of transient astronomical objects. operator. These are used in state estimation algorithms to make dispatch, re-balance and other energy management In general information science-related efforts, we analyzed system decisions. Coordinated cyber attacks on power and generalized the Belief Propagation (BP) approach meter readings can be designed to be undetectable by any to computing the permanent (a quantity much like the bad data detection algorithm. These unobservable attacks determinant, but harder to calculate) of a non-negative present a serious threat to grid operations. In [11] we matrix [7]. Known bounds and conjectures were verified in presented an efficient algorithm to find all unobservable experiments, and some new theoretical relations, bounds attacks, involving the compromise of exactly two power and conjectures were proposed. We developed a simple injection meters and an arbitrary number of power meters greedy algorithm for learning the best planar Ising model 484

Page 485

on lines. We characterized such attacks and studied coun- project helped bring 16 postdocs and 15 students to LANL termeasures. This work continued with a better algorithm to work in the area of Information Science and Technology. for detection and the study of consequence of any attack. Two mission areas that will potential use output from this We applied [12] the result from our study on attacks to project are energy security with the Smart Grid efforts and power systems to develop a multi criteria decision mak- bio-security with the quantitative biology, epidemiology ing algorithm based on observability, cost importance and and synthetic vision aspects of the IS&T work. security. We began working on issues related to state estimation References of power systems. Power system operators rely critically 1. Giorgi, E. E., and T. Bhattacharya. A note on two-sam- on state estimation for verification, fault detection and ple tests for comparing intra-individual genetic se- localization, and redispatch under contingency operations. quence diversity between populations. 2012. Biomet- We explored techniques to accelerate state estimation by rics. 68 (4): 1323. computing state estimates at a small subset of buses using limited measurements from the power subsystem of inter- 2. Giorgi, E. E., B. T. Korber, A. S. Perelson, and T. Bhat- est. These could be operator selected “important” buses, tacharya. Modeling sequence evolution in HIV-1 infec- which connect to “important” lines with significant real tion with recombination. 2013. Journal of Theoretical power transfer. Our techniques are inspired by uncertainty Biology. 329: 82. quantification methods. 3. Lemons, N., B. Hu, and W. Hlavacek. Hierarchical We developed several algorithms and techniques for graphs for rule-based modeling of biochemical sys- power systems. We developed a technique to measure the tems. 2011. BMC Bioinformatics. 12 (1). vulnerability of power networks to generator failure that 4. Huang, W., Z. Ji, S. P. Brumby, G. Kenyon, and L. M. Bet- uses instanton techniques to estimate probabilities of un- tencourt. Development of invariant feature maps via likely events. This allows more accurate “N-1 contingency a computational model of simple and complex cells. failure” analysis than the with the existing techniques. 2012. p. 1. With a student we developed mathematically rigorous mapping from first-principles models of power generators 5. Ji, Z., S. Brumby, G. Kenyon, and L. A. Bettencourt. to a coupled oscillator model. For that model, provably Integrating Bottom-up and Top-down Visual Attention correct closed-form synchronization conditions were found for Object Segmentation. 2012. Journal of Vision. 12 that reveal the connection between synchronization in (9): 926. power networks and the network topology and parameters [13]. 6. Ji, Z., W. Huang, and S. P. Brumby. Learning sparse representation via a nonlinear shrinkage encoder and a For random intersection graphs, which are models of social linear sparse decoder. 2012. p. 1. networks, we found conditions for model parameters that determine when the graph is mostly connected (has 7. Chertkov, M., and A. B. Yedidia. Computing the Perma- a “giant component”) or not. These proofs are important nent with Belief Propagation. 2011. arXiv:1108. for estimating, e.g. the extent of the spread of diseases in network models of social interactions [14], [15]. 8. Johnson, J., P. Netrapalli, and M. Chertkov. Learning Planar Ising Models. 2010. arXiv:1011. Impact on National Missions 9. Dvijotham, K., S. Backhaus, and M. Chertkov. Oper- This project supported and enhanced the fundamental ations-Based Planning for Placement and Sizing of science of Information Science and Technology, an area of Energy Storage in a Grid With a High Penetration of strategic interest to the laboratory. The CNLS provided a Renewables. 2011. arXiv:1107. basic science-oriented postdoc program in which the CNLS shared postdocs with staff involved in IS&T research in T, 10. Giani, A., and R. Bent. Addressing smart grid cyber CCS, P and D-divisions. security. 2013. p. 44:1–44:4. \{ACM\}. By underpinning the fundamental part of the ‘S’ (Science) 11. Giani, A., R. Bent, M. Hinrichs, M. McQueen, and K. in IS&T, the project helped the laboratory fulfill its security Poolla. Metrics for assessment of smart grid data integ- mission, growing the area of Information Science. These rity attacks. 2012. p. 1. efforts also increased LANL’s presence in the academic world and helped with recruitment and retention. This 12. Linda, O., M. Manic, A. Giani, and M. McQueen. Multi- 485

Page 486

criteria based staging of Optimal PMU Placement using deficiency virus (SIV) sequences but fail to contain very Fuzzy Weighted Average. 2013. p. 1. early evolution and eventual fixation of epitope escape mutations during SIV infection. 2011. Journal of Virol- 13. Dörfler, F., M. Chertkov, and F. Bullo. Synchronization ogy. 85 (8): 3746. in complex oscillator networks and smart grids. 2013. Chandrasekaran, V., M. Chertkov, D. Gamarnik, D. Shah, Proceedings of the National Academy of Sciences. 110 and J. Shin. Counting Independent Sets Using the (6): 2005. Bethe Approximation. 2011. SIAM Journal on Discrete Mathematics. 25 (2). 14. Bradonjić, M., A. Hagberg, N. Hengartner, N. Lemons, and A. Percus. The phase transition in inhomogeneous Chernyak, V. Y., and M. Chertkov. Planar graphical models random intersection graphs. 2013. arXiv:1301. which are easy. 2010. Journal of Statistical Mechanics: Theory and Experiment. 2010 (11). 15. Bradonjić, M., A. Hagberg, N. Hengartner, and A. Per- cus. Component Evolution in General Random Inter- Chertkov, M., I. Kolokolov, and V. Lebedev. Tail-constraining section Graphs. 2010. In Lecture Notes in Computer stochastic linear–quadratic control: a large deviation Science. Edited by Kumar, R., and D. Sivakumar. p. 36. and statistical physics approach. 2012. Journal of Sta- tistical Mechanics: Theory and Experiment. 2012 (08). Publications Chertkov, M., and A. B. Yedidia. Computing the Permanent Backhaus, S., and M. Chertkov. Getting a grip on the elec- with Belief Propagation. 2011. arXiv:1108. trical grid. 2013. Physics Today. 66 (5): 42. Crespi, V., G. Cybenko, and A. Giani. Engineering Statistical Bienstock, D., M. Chertkov, and S. Harnett. Chance Con- Behaviors for Attacking and Defending Covert Chan- strained Optimal Power Flow: Risk-Aware Network nels. 2013. IEEE Journal of Selected Topics in Signal Control under Uncertainty. 2012. arXiv:1209. Processing. 7 (1): 124. Boeras, D., P. Hraber, M. Hurlston, T. Evans-Strickfaden, T. Duclut, C., S. Backhaus, and M. Chertkov. Hysteresis, Phase Bhattacharya, E. Giorgi, J. Mulenga, E. Karita, B. Korber, Transitions and Dangerous Transients in Electrical Pow- S. Allen, C. Hart, C. Derdeyn, and E. Hunter. Role of er Distribution Systems. 2012. arXiv:1212. donor genital tract HIV-1 diversity in the transmission bottleneck. 2011. Proceedings of the National Acad- Dvijotham, K., S. Backhaus, and M. Chertkov. Operations- emy of Sciences. 108 (46): e1156. Based Planning for Placement and Sizing of Energy Storage in a Grid With a High Penetration of Renew- Bonachela, J., M. Raghib, and S. Levin. Dynamic model of ables. 2011. arXiv:1107. flexible phytoplankton nutrient uptake. 2011. Proceed- ings of the National Academy of Sciences. 108 (51): Dörfler, F., M. Chertkov, and F. Bullo. Synchronization in 20633. complex oscillator networks and smart grids. 2013. Proceedings of the National Academy of Sciences. 110 Bradonjić, M., A. Hagberg, N. Hengartner, N. Lemons, and (6): 2005. A. Percus. The phase transition in inhomogeneous ran- dom intersection graphs. 2013. arXiv:1301. Erdos, P., D. Gerbner, N. Lemons, D. Mubayi, C. Palmer, and B. Patkos. Two-Part Set Systems. 2012. Electronic Jour- Bradonjić, M., A. Hagberg, N. Hengartner, and A. Percus. nal of Combinatorics. 19 (1). Component Evolution in General Random Intersection Graphs. 2010. In Lecture Notes in Computer Science. Gerbner, D., N. Lemons, C. Palmer, B. Patkós, and V. Szécsi. Edited by Kumar, R., and D. Sivakumar. , p. 36. Almost Intersecting Families of Sets. 2012. SIAM Jour- nal on Discrete Mathematics. 26 (4): 1657. Bradonjić, M., and L. Lazos. Graph-based criteria for spec- trum-aware clustering in cognitive radio networks. Giani, A., E. Bitar, M. Garcia, M. McQueen, P. Khargonekar, 2012. Ad Hoc Networks. 10 (1): 75. and K. Poolla. Smart Grid Data Integrity Attacks. 2013. IEEE Transactions on Smart Grid. 4 (3): 1244. Brandão, F. S., M. Christandl, and J. Yard. Erratum to: Faith- ful Squashed Entanglement. 2012. Communications in Giani, A., E. Bitar, M. Garcia, M. McQueen, P. Khargonekar, Mathematical Physics. 316 (1): 287. and K. Poolla. Smart grid data integrity attacks: charac- terizations and countermeasures. 2011. In 2011 IEEE Cale, E. M., P. Hraber, E. E. Giorgi, W. Fischer, T. Bhattacha- International Conference on Smart Grid Communica- rya, T. Leitner, W. W. Yeh, C. Gleasner, L. D. Green, C. S. tions (SmartGridComm). , p. 232. Han, B. Korber, and N. L. Letvin. Epitope-specific CD8+ T lymphocytes cross-recognize mutant simian immuno- Giani, A., R. Bent, M. Hinrichs, M. McQueen, and K. Poolla. 486

Page 487

Metrics for assessment of smart grid data integrity sentation via a nonlinear shrinkage encoder and a lin- attacks. 2012. In 2012 IEEE Power and Energy Society ear sparse decoder. 2012. , p. 1. General Meeting. , p. 1. Johnson, J., P. Netrapalli, and M. Chertkov. Learning Planar Giani, A., and R. Bent. Addressing smart grid cyber security. Ising Models. 2010. arXiv:1011. 2013. In Proceedings of the Eighth Annual Cyber Secu- rity and Information Intelligence Research Workshop. , Johnson, M., and A. Gutfraind. Evader interdiction and col- p. 44:1–44:4. \{ACM\}. lateral damage. 2012. , p. 86. Gintautas, V., B. Kunsberg, M. Ham, S. Barr, S. Zucker, S. Keszegh, B., N. Lemons, and D. Pálvölgyi. Online and Brumby, L. A. Bettencourt, and G. Kenyon. An im- quasi-online colorings of wedges and intervals. 2012. proved model for contour completion in V1 using arXiv:1207. learned feature correlation statistics. 2010. Journal of Kudekar, S., J. Johnson, and M. Chertkov. Improved Linear Vision. 10 (7): 1162. Programming Decoding using Frustrated Cycles. 2011. Gintautas, V., M. Ham, B. Kunsberg, S. Barr, S. Brumby, C. arXiv:1105. Rasmussen, J. George, I. Nemenman, L. A. Bettencourt, Kudekar, S., J. Johnson, and M. Chertkov. Linear Program- and G. Kenyon. Model Cortical Association Fields Ac- ming based Detectors for Two-Dimensional Intersym- count for the Time Course and Dependence on Target bol Interference Channels. 2011. arXiv:1102. Complexity of Human Contour Perception. 2011. PLoS Computational Biology. 7 (10). Kudekar, S., and N. Macris. Decay of Correlations for Sparse Graph Error Correcting Codes. 2011. SIAM Journal on Giorgi, E. E., B. T. Korber, A. S. Perelson, and T. Bhattacha- Discrete Mathematics. 25 (2). rya. Modeling sequence evolution in HIV-1 infection with recombination. 2013. Journal of Theoretical Biol- Lemons, N., B. Hu, and W. Hlavacek. Hierarchical graphs ogy. 329: 82. for rule-based modeling of biochemical systems. 2011. BMC Bioinformatics. 12 (1). Giorgi, E. E., and T. Bhattacharya. A note on two-sample tests for comparing intra-individual genetic sequence Li, H., M. Stoddard, S. Wang, L. Blair, E. Giorgi, E. Parrish, diversity between populations. 2012. Biometrics. 68 G. Learn, P. Hraber, P. Goepfert, M. Saag, T. Denny, B. (4): 1323. Haynes, B. Hahn, R. Ribeiro, A. Perelson, B. Korber, T. Bhattacharya, and G. Shaw. Elucidation of Hepatitis C Gutfraind, A.. Optimizing Topological Cascade Resilience Virus Transmission and Early Diversification by Single Based on the Structure of Terrorist Networks. 2010. Genome Sequencing. 2012. PLoS Pathogens. 8 (8). PLoS ONE. 5 (11). Linda, O., M. Manic, A. Giani, and M. McQueen. Multi- Gutfraind, A., A. Hagberg, D. Izraelevitz, and F. Pan. Inter- criteria based staging of Optimal PMU Placement using diction of a Markovian Evader. 2011. In Proceedings of Fuzzy Weighted Average. 2013. In 2013 IEEE Interna- the 12th INFORMS Computing Society Conference on tional Symposium on Industrial Electronics (ISIE). , p. 1. OR, Computing, and Homeland Defense. , p. 3. Liu, X. B., M. Wang, J. Meng, E. Ben-Naim, and Z. Y. Guo. Huang, W., Z. Ji, S. P. Brumby, G. Kenyon, and L. M. Bet- Minimum Entransy Dissipation Principle for Optimiza- tencourt. Development of invariant feature maps via tion of Transport Networks. 2010. International Journal a computational model of simple and complex cells. of Nonlinear Sciences and Numerical Simulation. 11 2012. , p. 1. (2): 113. Ji, Z., J. Theiler, R. Chartrand, G. Kenyon, and S. Brumby. Loxley, P. N., L. M. Bettencourt, and G. T. Kenyon. Ultra-fast Decoupling sparse coding of SIFT descriptors for large- detection of salient contours through horizontal con- scale visual recognition. 2013. Vol. 8750, p. 87500k. nections in the primary visual cortex. 2011. EPL (Euro- Ji, Z., M. Luciw, J. Weng, and S. Zeng. Incremental Online physics Letters). 93 (6). Object Learning in a Vehicular Radar-Vision Fusion Loxley, P. N., and L. M. Bettencourt. Visually-salient contour Framework. 2011. Intelligent Transportation Systems, detection using a V1 neural model with horizontal con- IEEE Transactions on. 12 (2): 402. nections. 2011. arXiv:1103. Ji, Z., S. Brumby, G. Kenyon, and L. A. Bettencourt. Integrat- Parrish, N., F. Gao, H. Li, E. Giorgi, H. Barbian, E. Parrish, ing Bottom-up and Top-down Visual Attention for Ob- L. Zajic, S. Iyer, J. Decker, A. Kumar, B. Hora, A. Berg, ject Segmentation. 2012. Journal of Vision. 12 (9): 926. F. Cai, J. Hopper, T. Denny, H. Ding, C. Ochsenbauer, J. Ji, Z., W. Huang, and S. P. Brumby. Learning sparse repre- Kappes, R. Galimidi, A. West, P. Bjorkman, C. Wilen, R. 487

Page 488

Doms, M. O’Brien, N. Bhardwaj, P. Borrow, B. Haynes, M. Muldoon, J. Theiler, B. Korber, G. Shaw, and B. Hahn. Phenotypic properties of transmitted founder HIV-1. 2013. Proceedings of the National Academy of Sciences. 110 (17): 6626. Samaniego, H., G. Sérandour, and B. Milne. Analyzing Tay- lor’s Scaling Law: qualitative differences of social and territorial behavior on colonization/extinction dynam- ics. 2012. Population Ecology. 54 (1): 213. Silva, A., G. Jander, H. Samaniego, J. S. Ramsey, and C. Figueroa. Insecticide Resistance Mechanisms in the Green Peach Aphid Myzus persicae (Hemiptera: Aphi- didae) I: A Transcriptomic Survey. 2012. PLoS ONE. 7 (6). Turitsyn, K., M. Chertkov, and M. Vucelja. Irreversible Monte Carlo algorithms for efficient sampling. 2011. Physica D: Nonlinear Phenomena. 240 (4–5): 410. Whitney, J., P. Hraber, C. Luedemann, E. Giorgi, M. Daniels, T. Bhattacharya, S. Rao, J. Mascola, G. Nabel, B. Korber, and N. Letvin. Genital Tract Sequestration of SIV fol- lowing Acute Infection. 2011. PLoS Pathogens. 7 (2). Wolpert, D., M. Harré, E. Olbrich, N. Bertschinger, and J. Jost. Hysteresis effects of changing the parameters of noncooperative games. 2012. Physical Review E. 85 (3). Wozniak, P., D. I. Moody, Z. Ji, S. P. Brumby, H. Brink, J. Rich- ards, and J. S. Bloom. Automated Variability Selection in Time-domain Imaging Surveys Using Sparse Repre- sentations with Learned Dictionaries. 2013. In Ameri- can Astronomical Society Meeting Abstracts. Vol. 221. Yusim, K., R. Dilan, E. Borducchi, K. Stanley, E. Giorgi, W. Fischer, J. Theiler, J. Marcotrigiano, B. Korber, and D. Barouch. Hepatitis C Genotype 1 Mosaic Vaccines Are Immunogenic in Mice and Induce Stronger T-Cell Responses than Natural Strains. 2013. Clinical and Vac- cine Immunology : CVI. 20 (2): 302. Zdeborová, L., and F. Krzakala. Quiet Planting in the Locked Constraint Satisfaction Problems. 2011. SIAM Journal on Discrete Mathematics. 25 (2). 488

Page 489

Information Science and Technology Early Career Research and Development Continuing Project Gating and Emission Enhancement of Diamond Field-emitter Arrays Heather L. Andrews 20120522ECR Introduction ture. The next set of samples are in preparation and will High-quality beams of electrons are critical to a wide be delivered shortly. The pulsed power supply necessary array of ubiquitous devices. These devices range from to drive the cathodes up to their operational limit has microwave systems, electronic displays, and vacuum been recommissioned and is being integrated into the microelectronics to research equipment such as electron experimental chamber. Finally, we have made significant microscopes, terahertz sources, and free-electron lasers. progress testing the design of the cathode holder and anode. Each of these parts are custom, and critical to the The quality of an electron beam can only degrade once experiment. They must withstand a very high electric the beam leaves its source, the cathode, so it is criti- field in order to test true DFEA performance. While they cal to develop cathodes that produce inherently high- are designed to withstand the anticipated electric field, quality beams. The broad goal of this research project is they must be tested because minor surface imperfec- to further the development of a novel cathode called a tions or even material defects can cause their behavior diamond field-emitter array (DFEA) whose fundamental to deviate from that predicted by modeling. element produces nearly the highest quality electron beam allowed by the laws of physics. The specific objec- Future Work tive for this project is to prove the relevance of DFEAs to The main goals for the next fiscal year include testing high-power free-electron lasers (FELs) by demonstrating the limits of DIamond Field-Emitter Array (DFEA) per- enhancement and control of emission and determining formance, and investigating whether laser light can be the limits of DFEA performance. used to turn off and on DFEA emission. To accomplish the first goal we will progressively test larger and denser Benefit to National Security Missions arrays to see how high current density can be achieved. Cathodes that produce inherently high-quality electron beams are of key importance to many focus areas of We will continue a small amount of modeling work to global security including high-power free-electron lasers, ensure the experimental set-up is sufficient to withstand directed energy sources and terahertz sources. The the high-current density we expect to produce. Finally, development of DFEAs could have a dramatic impact on we will investigate whether laser light can turn on and the accelerator segment of LANL operations and assure off the current emitted by DFEAs. We will do this by shin- LANL a place at the forefront of free-electron laser and ing a laser on the front, side, and/or back surface of the high-power microwave source development. DFEA, while the voltage used to produce emission is very low. If we observe emission in one of these configura- tions, we will make more thorough investigation to begin Progress to determine the limits and conditions under which laser Over the past year this project has accomplished a num- light gates DFEA cathodes. ber of critical goals. First, the test chamber was fully assembled and pumped down to a very good vacuum. Conclusion This is critical because one of the questions we need to address is Diamond Field-Emission Array (DFEA) perfor- This research is expected to produce experimental data mance under different vacuum levels. Next we received describing the limits of operation of DFEAs. This data will our first DFEA samples from our collaborators at Vander- inform the scope of potential applications of DFEAs as bilt University. These will be tested in the very near fu- robust, high-quality cathodes. In the event DFEAs are 489

Page 490

as widely applicable as they promise, this novel technol- ogy has the potential to greatly improve existing vacuum micro-electronic devices as well as allow new devices to be created. 490

Page 491

Information Science and Technology Exploratory Research Continuing Project Uncertainty Quantification for Networks Earl C. Lawrence 20120123ER Introduction operations, situational awareness after a disaster, cost Dependence on network-delivered commodities, such as estimation, and climate impact in future networks; influ- information and energy, places infrastructure networks ences between interconnected networks; and targeted in a prominent and growing role in national and global data collection for uncertainty reduction. security. Because they have often grown incrementally and without comprehensive plans, these networks are Progress complex, vulnerable, and incompletely known or de- This past fiscal year focused on two research direc- scribed. Nevertheless, policy analysts need to make deci- tions. First, we have worked on expanding the network sions about infrastructure resiliency and criticality. This state under consideration. The first year focused on research seeks to improve the decision-making process network topology. This past year, we have worked on by providing methods for network uncertainty quantifi- adding other network quantities like loads and genera- cation (UQ): predicting unknown components and esti- tions. These additions assume results from the topology mating uncertainty in these predictions. Specifically, our estimation that gives a list of possible topologies with goal is topology, attribute, and flow inference in partially associated probabilities. For a given topology, we could observed infrastructure networks using science-based use the power flow equations to see how settings of the simulation tools. Statistical methods exist for graph unknown generations and loads produce lines flows and exploration and for incorporating complex simulators, how those simulated line flows compare to measured but these two areas are separate and combining them data. Unfortunately, this is a very computationally in- requires advances in statistical methodology. These tech- tensive process. Thus, for each topology with significant niques will leverage the scientific knowledge built into a probability, we build an approximation to the relation- simulator for use with statistical methods for character- ship between generations/loads and the line flows that izing the space of graphs (topology) and attributes that correspond to measured data. This approximation describe networks. allows us to quickly find the distribution of generations/ loads that match observed data. Combined with earlier Benefit to National Security Missions work, this will give us a fast way to estimate the general This research will impact LANL programs in Global state of power grid based on a data measured at a few Security (PADGS) and Computer, Computational, and points in the network. This work is currently being writ- Statistics (CCS) Sciences and the missions of the En- ten into a paper, with plans to submit to IEEE Transac- ergy Security Institute and the Information Science and tions on Power Systems by the end of the summer. Technology Institute. PADGS will benefit from validated, science-based methodologies that estimate unobserved The second direction we have pursued is to use the network features with known confidence. Many PADGS topology uncertainty quantification techniques from problems in homeland security and intelligence are cur- the first year in a network control application: optimal rently intractable without the proposed capability. The power flow (OPF). In the OPF problem, a grid operator core information science is also applicable to a wide va- determines the settings for generation capabilities based riety of network domains including cyber, supply chains, on the current state of the network (including demand, natural gas, water, and oil, as well as social and biological generation, and topology). When the state is precisely networks. Possible applications include criticality, impact known, the generation settings can be determined analysis, state estimation, event detection, effects-based using a fairly simple optimization routine. When the 491

Page 492

estimated state is incorrect, the solution could lead to line FORMS Annual Meeting 2013. (Minneapolis, 6-9 Oct. overages and network failure, so incorporating uncertainty 2013). is important. Our work this year has been to expand the Garcia, M., A. Giani, and K. Poolla. Partial state estimation optimization framework to include the output of the topol- for Electricity Grids. To appear in 52nd IEEE Conference ogy estimation, a list of possible topologies with associated on Decision and Control. (Florence, Italy, 10-13 Dec. probabilities. This new chance-constrained OPF frame- 2013). work can be used to determine generation settings when the topology is not completely known, while minimizing Lawrence, E., S. Vander Wiel, and R. Bent. Model bank the probability of line overloads. This work is also current- state estimation for power grids using importance ly being written for IEEE Transactions on Power Systems sampling. 2013. Technometrics. 55 (4): ?. with plans to submit by the end of summer. Wiel, S. Vander, E. Lawrence, and R. Bent. Uncertainty quantification for networks with power distribution ap- In addition, a summary of the entire project so far was plications. Invited presentation at Conference on Data given in an invited talk at the 2013 Quality and Productivity Analysis. (Santa Fe NM, 29 Feb - 2 Mar 2012). Research Conference in the session Data-Driven Decisions in Smart City Infrastructure. Wiel, S. Vander, E. Lawrence, and R. Bent. Uncertainty quantification for power networks. Presented at Op- timization and Control for Smart Grids. (Santa Fe NM, Future Work 21-25 May 2012). We will shift our model development to use data from phasor measurement units (PMUs). A PMU is an example Wiel, S. Vander, R. Bent, E. Casleton, and E. Lawrence. of the type of data collection device that so-called “smart Identification of Downed Lines in Power Grids. Applied grids” will use. They measure voltage and current informa- Stochastic Models in Business and Industry. tion with high precision. This should position our work for Wiel, S. Vander, R. Bent, E. Lawrence, and E. Casleton. Un- use with newly developing technology. certainty quantification for networks with power distri- bution applications. Invited presentation at Quality and We will continue to expand space of possibly network Productivity Research Conference 2013. (Niskayuna, states, particularly for topology. We would like to move NY, 5-7 June 2013). beyond the model bank formulation of year one and con- sider more general spaces of possible network topologies. We will continue to expand the optimal power flow for- mulation to better account for sources of uncertainty. This will keep pace with our other work so that we can continue to demonstrate an end-to-end process for uncertainty quantification: take measurements, estimate uncertainty, and take uncertainty-aware actions in the grid. Conclusion The result of this research will be methodology for estimat- ing, with uncertainty, the unobserved parts of infrastruc- ture networks, such as electric power. This methodology will be a tool for better understanding how these networks perform, where and why they are vulnerable, and how to improve them. Some possible applications include critical- ity assessment of domestic and foreign networks, disaster recovery, and improving reliability and efficiency for the power grid. Publications Bent, R., D. Bienstock, E. Chertkov, E. Lawrence, and S. Vander Wiel. Chance constrained optimal power flow for topology uncertainty. Invited presentation at IN- 492

Page 493

Information Science and Technology Exploratory Research Continuing Project Phylogenetic Classification and Analysis of Computer Malware Guanhua Yan 20120443ER Introduction efforts is the realization that cyber-threat reduction is Computer malware have plagued the Internet for more among the core missions of LANL. This project seeks to than two decades and are now responsible for the connect that capability with our leadership in the appli- majority of malicious activities in the cyber space. The cation of phylogenetics to biological (viral) systems and two key challenges in defending against a malware at- develop a novel capability within cyber security science. tack are finding who originated this attack so that law enforcement can get involved (malware attribution), Progress and identifying which family the malware belongs to so In the past year, we finished these tasks under the aus- that effective mitigation techniques can be quickly de- pices of this project. (1) We developed a principled solu- ployed (malware classification). Addressing these issues tion to finding consensus in data clustering by multiple demands in-depth analysis of malware code, which is experts. Our project requires labeled data as ground currently done by either labor-intensive manual reverse truth to study malware evolution. As manually label- engineering or error-prone signature-based schemes. ing a large number of malware samples is difficult, we rely on the classification results by multiple AV software Against this backdrop, we propose methods to analyze to group malware samples. We formulate the problem malware programs efficiently and accurately using ideas rigorously and show that finding an optimal solution from biological phylogenetics.We will develop malware is computationally intractable. We further proposed a phylogeny to organize previously observed malware graph theory-based method to find sub-optimal solu- instances into a structure that facilitates classification tions. (2) We developed a framework for automated of new ones captured from the wild (e.g., email attach- malware classification. This framework can combine ments). Moreover, malware phylogeny that accurately various types of malware features, some of which may characterizes the evolution of malware instances sheds have missing values, and use the Neyman-Pearson light on the originators of malware attacks.This project criterion to search optimal parameters of cost-sensitive also concerns data-intensive phylogenetic inference, as classifiers. (3) We also developed a new transductive it will be conducted on a large locally collected malware malware classification method to spread label informa- repository with millions of malware instances. tion from labeled samples to unlabeled ones. (4) We have run phylogenetic software based on maximum- Benefit to National Security Missions likelihood principles on the features we have collected This projects meets two FY12 priorities of Information, from a few malware families to show their evolution. We Science and Technology. If we are successful in develop- are currently interpreting the results and evaluating how ing phylogenetic classification tools, we contribute to the accurately the malware phylogeny reflects the realism. Inference/Prediction priority by demonstrating statistical In the past year, we have three papers accepted for pub- inference methods for the history of malware programs lication directly from the results of this project. even though their evolution involves human intelligence. Even if we are ultimately unsuccessful in developing such Future Work tools, we advance the data intensive computing priority The popularity of online social media networks has by developing efficient algorithms to extract representa- inspired researchers in diverse fields to mine events of tive features from a large malware database that con- interest (e.g., stock prices, epidemic disease, and earth- tains millions of malware programs. Underlying these quakes). These efforts, however, are mainly performed 493

Page 494

in an offline fashion. Few efforts have been focused on Measurement and modeling of computer systems (SIG- the computational challenges involved in real-time event METRICS’13). (Pittsburgh, PA, 17-21 June 2013). Vol. mining from streaming data sources. In this project, we 41, 1 Edition, p. 347. New York: ACM. will tackle two fundamental problems. (1) How to choose Kong, D., and G. Yan. Discriminant malware distance learn- features that can be used to predict the events of interest? ing on structural information for automated malware classification. 2013. In Proceedings of the 19th ACM In previous works, feature selection from streaming data SIGKDD international conference on Knowledge dis- sources is mainly done manually by domain experts. Such covery and data mining (KDD’13). (Chicago, 11-14 Aug. an approach does not scale well when we deal with large 2013). , p. 1357. New York: ACM. volumes of data. Based on the methods we have already Kong, D., and G. Yan. Transductive malware classification: developed, we will extend them to find features that are ir- find your lineage from your neighbors. 2013. LA-UR 13- relevant to the events of interest. With a high-dimensional 22771, Los Alamos National Laboratory.. feature space, the combinatorial optimization problem may be computationally intractable. One possible solution Liu, L., X. Zhang, G. Yan, and S. Chen. Chrome Extensions: is ranking features based on the distributional differences Threat Analysis and Countermeasures. 2012. In The during the periods with and without events of interest and 19th Annual Network & Distributed System Security then applying a greedy approach to select relevant fea- Symposium (NDSS’2012). (San Diego, California, 5-8 tures. If we have a large number of features, this method Feb. 2012). , p. 1. Reston, VA, USA: Internet Society. may still be computationally expensive. An alternative Yan, G.. Finding common ground among experts’ opinions: method may be leveraging the structural information conflict-free malware clustering for malware lineage revealed by clusters of features to accelerate the process analysis. 2013. LA-UR 13-22751, Los Alamos National of feature selection. (2) Given a set of features, how to Laboratory.. monitor events of interest? We will develop change-point detection algorithms that monitor the changes in the Yan, G., N. Brown, D. Kong, and M. Anghel. Towards a distributions of features (or combinations of features) that Comprehensive View of Feature Selection for Malware are indicative of occurrences of events of interest. Classification. 2012. LA-UR 12-25308, Los Alamos Na- tional Laboratory. There are a broad spectrum of applications that can Yan, G., N. Brown, and D. Kong. Exploring discriminatory benefit from this project, including but not limited to features for automated malware classication. 2013. In bio-surveillence and logistics. The deliverables from this Proceedings of the 10th Conference on Detection of project will be effective and efficient algorithms for real- Intrusions and Malware & Vulnerability Assessment time monitoring of events of interest from streaming data (DIMVA’13). (Berlin, Germany, 17-19 July 2013). Vol. sources such as Twitter tweets. 7967, p. 41. Berlin Heidelberg: Springer. Conclusion The end results of this project include (1) Fundamental understanding of evolution of computer malware based on a locally collected malware repository; (2) Classification of malware instances into different malware families using unsupervised machine learning techniques; (3) Malware phylogeny for each malware family that characterizes evolution of malware variants in the same malware family. Malware phylogeny will be used to find close ancestors when a new malware instance is spotted, and information about these ancestors will be used to both understand the origin of the new malware attack and estimate its behavior and potential damages. Publications Kong, D., and G. Yan. Poster: Discriminant Malware Dis- tance Learning on Structural Information for Auto- mated Malware Classication. 2013. In Proceedings of the ACM SIGMETRICS/international conference on 494

Page 495

Information Science and Technology Exploratory Research Continuing Project Combinatorial Approaches To Graphical Models: Theory and Applications Michael Chertkov 20120469ER Introduction ties to support DOE mission in the areas of information In recent years, graphical models have come to play technology for very large-scale inference, estimation and an important role in modern approaches to computer optimization. science and machine learning and have found many applications in various fields of science and engineer- Benefit to National Security Missions ing. In short, a graphical model is a statistical model for Our project will provide new powerful tools in learn- a collection of (possibly very many) random variables in ing of predictive statistical models from observations. which correlations among the variables are succinctly Applications for these algorithmic and theoretical IT described by a sparse set of interactions or constraints know how” tools are many, ranging from bio- and among small subsets of variables. The conditional in- genetic networks, detection of anomalies in infrastruc- dependence relations (Markov properties) satisfied by ture networks, inverse problems in seismology, reservoir these variables are then summarized by a graph in which exploration and discovery in oil fields, to probabilistic vertices represent random variables and edges denote co-design based on limited measurements. Our project interactions among the variables. Although graphi- primarily addresses Information Science & Technology cal models provide a powerful and versatile modeling Grand Challenge by providing data-driven algorithmic framework, they suffer from fundamental challenges in tool for Intelligent Data Acquisition, Management, And that inference in general graphical models is computa- Analysis” (especially in what concerns Knowledge extrac- tionally intractable. This has motivated an extensive and tion, Automated discovery, Active learning, Anomaly on-going research effort to develop computationally- detection, Data mining and Semi-supervised learning) tractable, heuristic methods for approximate inference and Computational Co-design” (in what concerns inte- in graphical models. gration of domain expertise, e.g. in statistical modeling of observational data, and novel machine learning tech- We propose to pioneer advanced methods for graphical niques the proposal develops. Consequently, proposed models derived from combinatorial methods of statisti- activities will help LANL to respond to incoming National cal physics. More specifically we intend to build on trac- Initiatives, DOE, DOD, DTRA, DARPA and DHS related to table methods developed for special classes of graphical monitoring, detection, control and analysis of large data models, namely Ising and dimer models defined on sets. graphs of simple topology (planar and bounded-genus graphs). While inference is intractable for the Ising and Progress dimer models on general graphs, it becomes tractable We developed a general purpose Belief Propagation for planar graphs being equivalent to calculation of a (BP) algorithm for Linear Programming. BP is a popular, matrix determinant or Pfaffian. These methods were distributed heuristic for performing MAP computa- originally developed to obtain estimates of lattice mod- tions in Graphical Models. BP can be interpreted, from els in statistical physics and are little-known and under- a variational perspective, as minimizing the Bethe Free utilized in the machine learning community. Energy (BFE). BP can also be used to solve a special class of Linear Programming (LP) problems. For this class of We believe that this research agenda will advance the problems, MAP inference can be stated as an integer LP field of graphical models in new directions and enhance with an LP relaxation that coincides with minimization LANL’s presence and reputation in the domain of graphi- of the BFE at zero temperature”. We generalized these cal models as well as cultivate fundamental capabili- 495

Page 496

prior results and establish a tight characterization of the LP inal model via the BFE of an alternative PM problem. We problems that can be formulated as an equivalent LP relax- then study the zero-temperature version of this Bootstrap- ation of MAP inference. Moreover, we suggest an efficient, BP formula for approximately solving the MWPM prob- iterative annealing BP algorithm for solving this broader lem. We did so by leveraging the Bootstrap-BP formula to class of LP problems. We demonstrated the algorithm’s construct a sequence of MWPM problems, where each performance on a set of weighted matching problems by new problem in the sequence is formed by contracting using it as a cutting plane method to solve a sequence of odd-sized cycles (or blossoms) from the previous problem. LPs tightened by adding blossom’’ inequalities. This Bootstrap-and-Contract procedure converges reliably and generates an empirically tight upper bound for the We have developed a Graphical Transformation for Belief MWPM. We concluded this study by discussing the rela- Propagation: Maximum Weight Matchings and Odd-Sized tionship between our iterative procedure and the famous Cycles approach. Max-product `belief propagation’ (BP) is Blossom Algorithm of Edmonds ‘65 and demonstrate the a popular distributed heuristic for finding the Maximum A performance of the Bootstrap-and-Contract approach on a Posteriori (MAP) assignment in a joint probability distribu- variety of weighted PM problems. tion represented by a Graphical Model (GM). It was recent- ly shown that BP converges to the correct MAP assignment Future Work for a class of loopy GMs with the following common fea- We expect to continue our work on learning graphical ture: the Linear Programming (LP) relaxation to the MAP models with the help of statistical physics models to de- problem is tight (has no integrality gap). Unfortunately, velop computationally-tractable, heuristic methods for ap- tightness of the LP relaxation does not, in general, guaran- proximate inference. The focus next year will be two fold, tee convergence and correctness of the BP algorithm. first continuation of the work on inference algorithms and ideas, and second incorporating these ideas into learning, The failure of BP in such cases motivates reverse engineer- both of factor functions and of graphical relations. ing a solution -- namely, given a tight LP, can we design a `good’ BP algorithm. We have designed a BP algorithm On the first task, empowered by our recent progress in the for the Maximum Weight Matching (MWM) problem over field of counting perfect matchings over bipartite graphs, general graphs. We prove that the algorithm converges to and developing fractal Belief Propagation (BP) and other the correct optimum if the respective LP relaxation, which related counting and reconstruction ideas for permanents, may include inequalities associated with non-intersecting we plan to attack a more challenging problem of general odd-sized cycles, is tight. The most significant part of our matchings (not only perfect and defined beyond bipartite approach is the introduction of a novel graph transforma- graphs). Maximum-A-Posteriori versions/formulations tion designed to force convergence of BP. Our theoreti- of the problems are known as allowing computationally cal result suggests an efficient BP-based heuristic for the sound solution, however these known algorithms are not MWM problem, which consists of making sequential, of BP type. We will attempt to find the respective BP algo- cutting plane’’, modifications to the underlying GM. Our rithms for MAP and then extend it to computing marginal experiments showed that this heuristic performs as well probabilities and solving the counting problem. This re- as traditional cutting-plane algorithms using LP solvers on search will also benefit from recently introduced Cumulant MWM problems. Expansion method for counting and related Loop Calculus approach developed at LANL. We also analyzed Loop Calculus and Bootstrap-Belief Propagation for Perfect Matchings on Arbitrary Graphs. Second, we will continue to work on learning, advancing In this sub-project we have discussed computation of the learning of planar graphical models with the aforemen- Partition Function (PF) and the Minimum Weight Perfect tioned cumulant expansion method and related General- Matching (MWPM) on arbitrary, non-bipartite graphs. ized BP approach. The main idea here is to develop an iterative process which first ignore part of correlations, We present two novel problem formulations - one for reconstructing the best it can on a relatively small sub- computing the PF of a Perfect Matching (PM) and one for graph and then extend the structure sequentially to better finding MWPMs - that build upon the inter-related Bethe fit the data. We will also explore very recent and promising Free Energy, Belief Propagation (BP), Loop Calculus (LC), ideas related to inference and learning of mixed problems Integer Linear Programming (ILP) and Linear Programming defined over both discrete and continuous variables. These (LP) frameworks. First, we described an extension of the problems emerge naturally LC framework to the PM problem. The resulting formulas, coined (fractional) Bootstrap-BP, express the PF of the orig- in the context of optimization and control of infrastruc- 496

Page 497

tures and are more challenging than either of the pure problems (discrete or continuous). Conclusion Our research goals can be broadly summarized as follows: • Develop new methods for learning graphical models aimed at exploiting tractable computations on planar and bounded-genus graphs. • Develop new combinatorial analysis of inference in graphical models and of some popular approximate inference methods such as belief propagation and its generalizations, providing improved convergence and error analysis of these methods • Iinvent entirely new approaches to approximate infer- ence for intractable graphs that leverage efficient com- putations for planar graphs (e.g., on planar subgraphs or planar covers of a non-planar graph) and demon- strate improvements over existing methods. Publications Chertkov, M., A. Gelfand, and J. Shin. Loop Calculus and Bootstrap-Belief Propagation for Perfect Matchings on Arbitrary Graphs. To appear in The Journal of Physics: Conference Series (JPCS). Chertkov, M., and A. B. Yedidia. Approximating the Perma- nent with Fractional Belief Propagation. 2013. JOUR- NAL OF MACHINE LEARNING RESEARCH. 14: 2029. Chertkov, M., and V. Chernyak. Statistical Inference in Structured Graphical Models: Gauge Transformations, Belief Propagation & Beyond. 2012. this is a big review — still work in progress. Gelfand, A., J. Shin, and M. Chertkov. Belief Propagation for Linear Programming. To appear in ISIT 2013. (Istanbul, Turkey, July 2013). Shin, J., A. Gelfand, and M. Chertkov. A Graphical Trans- formation for Belief Propagation: Maximum Weight Matchings and Odd-Sized Cycles. To appear in NIPS 2013. (Lake Tahoe, CA, December 2013). 497

Page 498

Information Science and Technology Exploratory Research Continuing Project Combining Information Theory and Game Theory David H. Wolpert 20120708ER Introduction Benefit to National Security Missions Game theory is the formalism for analyzing how inter- This project will lay the foundation for quantitative LANL acting, utility-maximizing agents (humans, organizations, tools for the design of systems that gather and analyze firms) behave. Recent work has shown how to formulate intelligence in a broad range of scenarios, and how best game theory in terms of Bayes nets. Other recent work to devote resources to impeding intelligence-gathering has shown how to analyze information flow across Bayes of potential adversaries. Those tools will complement nets. In this project I will combine these two break- LANL’s strengths in measurement technology. As the throughs to analyze information “flow” among a set of work progresses, I will seek opportunities to contribute interacting, agents and the effects of that flow on agent to LANL missions in Warfighter support and Global Secu- behavior. rity instrumentation. This work will build capabilities for D division, ISR, T division, CCS division, and ACS. Game theory analyzes the joint behavior of interacting agents when each agent predicts the behavior of the Progress other agents, aware that those other agents are predict- My first task is to construct a novel game theory repre- ing their behavior. Any modification to the information sentation, based on multi-agent influence diagrams, and flowing among such mutually predicting agents will establish the equivalence of this novel representation affect their behavior. Conversely, any change in the be- with the representations of conventional game theory. I havior of an agent will affect the information flowing to have now completed this task, and am in the process of the other agents. To understand this interplay between writing up the results for submission. information flow and behavior, I integrate information theory into game theory. I will start by analyzing issues My second task is to complete a research project show- such as: ing how to formalize such unawareness using the extension of multi-agent influence diagrams mentioned • Often reducing noise in the observations of an agent above. I have now completed this task, and am in the will increase their expected utility. Sometimes it will process of writing up the results for submission. decrease the agent’s expected utility. The reason is My third task was to complete the organizing of the that in certain circumstances, if you reduce informa- joint LANL / Santa Fe Institute workshop on Combin- tion flow to yourself - and your opponent knows ing Information Theory and Game Theory and run that that you now have less information - they will workshop. I did this successfully. change their behavior accordingly, and that may end up benefiting you more than the loss of information Future Work hurts you. The first research issue is to characterize Conventional game theory uses two representations of when it will benefit a player increase increase the games: “normal form”, and “extensive form.” Neither noise in their observations. representation is amenable to Shannon information • A related issue is understanding when an agent ben- theory. Accordingly, one of the central features of my efits if the noise level is reduced in the observations application of Shannon information theory to study of the other agents. (In general, I may either be hurt games is by using a different representation, based on or helped by a change in your noise level, indepen- multi-agent influence diagrams. To relate this new work dently of whether it helps or hurts you.) to the results in conventional game theory, I must estab- 498

Page 499

lish the equivalence of this novel representation with the representations of conventional game theory. This is the first task for the coming year, which will result in a paper submission. My second task is to complete the organizing of the joint LANL / Santa Fe Institute workshop on Combining Informa- tion Theory and Game Theory and run that workshop, and then publish a book of proceedings / classic papers. Conventional game theory has considered two types of limitations on information: incomplete and imperfect in- formation. However in many scenarios the limitation is due to inherent unawareness by a player of the form of the un- derlying game. For example, it may be that a player is not even aware of all the potential moves of their opponents, but then gains some information concerning those moves. The third task will be to complete a research project first showing how to formalize such unawareness using the ex- tension of multi-agent influence diagrams mentioned and then analyzing the information theoretic aspects of such unawareness. This will be completed in FY14. Conclusion This project will create a new formalism for describing and analyzing how the interaction of multiple agents depends on and affects the information flow governing their inter- action. This formalism will lay the groundwork for a range of applications including the design of large distributed systems for intelligence gathering and analysis. Publications Wolpert, D., M. Harre, E. Olbrich, N. Bertschinger, and J. Jost. Hysteresis effects of changing the parameters of noncooperative games. 2012. PHYSICAL REVIEW E. 85 (3, 2): 036102. Wolpert, D., and D. Leslie. Information Theory and Ob- servational Limitations in Decision Making. 2012. B E JOURNAL OF THEORETICAL ECONOMICS. 12 (1): 5. 499

Page 500

Information Science and Technology Exploratory Research Continuing Project Software/Hardware Mapping for Data Locality Optimization Hristo N. Djidjev 20130252ER Introduction order to maximize data reuse and locality and increase We consider the problem of mapping an existing se- the amount of parallelism. The success of this project quential code onto a model of a multi-core high-per- will advance the ability of LANL and DOE in solving the formance computing (HPC) system so that the memory co-design problem and building competence for achiev- access/com-munication cost is minimized. This can be ing exascale capabilities. This will lead to computers with achieved by achieving high data access locality, one of improved performance and price parameters, improving the most important properties of efficient HPC codes. the existing simulation capabilities. It will positively af- For instance, modern computer processors such as fect multiple application areas that rely on high perfor- Graphics Processor Unit (GPU) and Cell have limited mance computing and computational simulation for amount of local memory and the time of data transfer their data analysis and prediction. between that memory and the global memory is huge compared to the cost of computation or the cost of local Progress access. Hence, it is important to have a method that Since the co-PI Leonid Gurvits left the lab in December maps the operations of a software code onto a hardware 2012, we had to look for collaborators with experience model in a way that maximizes locality and data reuse, in polyhedral optimization and high-performance com- i.e., such that as large as possible amount of computa- puting, which was a difficult task given the small number tions are performed on the data in the local memory of researchers working in both areas. We established before that data is replaced with new data. Unfortu- cooperation with the group of Prof. Sanjay Rajopad- nately, the corresponding mapping optimization problem hye from Colorado State University and created a joint is very difficult to solve efficiently as it is NP-hard. Hence, research plan with him very closely tied to the ER. Also, we will look at a subclass of all software codes, namely one of his students, Vamsi Tandrapati, came in June 2013 nested-loops, and at a subclass of all transformations, for 3 months at LANL as a summer intern working on the namely affine transformations. Previous research has project. Also, we are expecting Ramakrishna Upadrasta, shown that those subclasses capture the most relevant who is very experienced in the above topics, to come as from practical point of view types of codes (as loops a postdoc in July 2013 for two years to also work on the are usually responsible for most of the total computa- project. With him and Vamsi on the team, the staffing tion time) and, on the other hand, result in optimization problems have been successfully resolved. problems of manageable computational complexity. Our goal is to address the algorithmic challenges and to The work so far has concentrated on developing the design mapping algorithms that are both accurate and high-level optimization framework based on the polyhe- scalable and to test their performance on a prototype dral method. One of the important observations made mapping tool. is that the computational complexity of the optimization procedure can be significantly reduced, compared with Benefit to National Security Missions the procedure originally described in the proposal, by Co-design has been recognized as one of the high- eliminating the need to search the space of all hardware priority areas of research for Information Science and parameters for an optimal solution. Previously, for each Technology at LANL. This project focuses on a key step set of hardware parameters (each point in the search of the software/hardware co-design process – mapping space), one had to run a problem for optimally mapping the software representation onto the hardware model in the software representation onto the corresponding 500

Page 501

hardware. The new approach directly finds the sets of the Djidjev, Hristo N.. Efficient Shortest Path Computations on optimal hardware parameters by including them as vari- Multi-GPU Platforms. 2013. Los Alamos National Labo- ables in the mapping optimization problem. We are trying ratory, LA-UR-13-26585. the new framework on simple algorithms with the purpose Djidjev, Hristo N., Guillaume Chapuis, Sunil Thulasidasan, of computing the parameters of that optimization prob- and Rumen Andonov. On Solving the Shortest Path lem, using as input Jacobi-1D, Gauss-Siedel-1D and Smith- Problem for Planar Graphs using Graphics Proces- Watermann algorithms and as a software tools the Pluto sors. 2012. Los Alamos National Laboratory, LA- polyhedral method tool. UR-12-25700. Djidjev, Hristo N., Rumen Andonov, Sanjay Rajopadhye, and Future Work Vamsi Tandrapati. Co-Design Optimization of Jacobi During the second year, we plan to complete the algorithm Kernel for GPU Architectures. 2013. Los Alamos Na- for codesign for a single algorithm and test it on a GPU- tional Laboratory, LA-UR-13-28110. type architecture. We will start by choosing a codelet (a computationally intensive code snippet) from PolyBench Djidjev, Hristo N., Sunil Thulasidasan, Rumen Andonov, or a “polyhedral” fragment from some code of interest. We Guillaume Chapuis, and Dominique Lavenier. Ef- will transform into Alpha polyhedral representation and ficient Multi-GPU Computation of All-Pairs Shortest Paths. 2013. Los Alamos National Laboratory, LA- pick a shape using the Pluto tool. Then we will choose a UR-13-28111. tile-parallelization strategy including schedule of tiles, allo- cation of tiles to processors, and inter-tile communication based on Pluto or similar tool. This will yield all parameters needed to formulate the constraints of the optimization problem. We will then try to solve that optimization prob- lem for different meaningful objective functions. The next steps will be to generalize that framework for the case of an algorithm with multiple codelets. Conclusion We will formulate several versions of the optimization problem discussed above and develop efficient algorithms for their solutions. We will combine our new algorithms in a proof-of-principle tool and will test it on optimizing nested loops for GPU type architectures. The loops will be chosen both from simple linear algebra and numerical methods codes during the development phase, as well as on advanced codes used at LANL (e.g., in molecular dynamics simulations) for which GPU type implementa- tions exist. The goal will be to compare the quality of our solution against hand-optimized GPU implementations in order to validate the results. Publications Aleksandrov, L., H. Djidjev, A. Maheshwari, and J. R. Sack. An Approximation Algorithm for Computing Shortest Paths in Weighted 3-d Domains. 2013. DISCRETE & COMPUTATIONAL GEOMETRY. 50 (1): 124. Djidjev, H., and M. Onus. Using graph partitioning for ef- ficient network modularity optimization. 2013. GRAPH PARTITIONING AND GRAPH CLUSTERING. 588: 103. Djidjev, Hristo N.. Parallel seed-based approach to protein structure similarity detection. 2013. Los Alamos Na- tional Laboratory, LA-UR-13-27448. 501

Page 502

Information Science and Technology Exploratory Research Continuing Project Contextual Learning and Recognition Alexei N. Skurikhin 20130265ER Introduction ment of breakthrough mathematics and computer sci- Event and object recognition requires analysis of massive ence required to extract knowledge from massive quan- quantities of data to detect and identify static or dy- tities of data. Current approaches to event and object namic patterns of interest. One challenge for automated recognition in imagery are overwhelmed by the number object recognition, and machine learning in general, of images, the sizes of images (e.g., high-resolution arises from the fact that in spite of huge and growing satellite image can be up to 10^9 pixels and larger), and quantities of data, complex event recognition often re- the diversity of visual features associated with objects mains ill-posed. The object recognition ambiguity can be of interest. Current approaches are also facing a need addressed by exploiting the fact that events rarely occur to deal with increasing use of signature suppression, in isolation; they tend to co-occur and co-vary, and this camouflage, concealment, and deception. Context aware correlation structure, known as context, provides impor- image data mining will provide an ability to focus human tant information for the disambiguation of the event. analysts on “hot spots” in such data. However, while being extremely useful for disambigua- tion of events and objects, such context-aware knowl- The project is also related to the Threat Reduction mis- edge discovery typically leads to a significant increase in sion. If successful, our methods will increase the produc- the number of features to be analyzed. Modeling contex- tivity of human analysts, optimize the use of high-value tual information and scaling machine learning methods sensors (which typically have a narrow field of view) for for context aware object recognition to large problems monitoring of wide areas, and improve the timeliness are among the most challenging issues in developing the and accuracy of decision making. next generation of information extraction approaches. Progress We aim to develop novel efficient machine learning and Contextual learning is posed as a structured machine inference methods for context aware probabilistic recog- learning problem, which is about learning knowledge nition of complex events and objects based on Markov from data with internal structure in the form of one and Conditional Random Fields. Development of scalable or more relations between co-occurring, co-located or context learning and inference methods for probabilistic co-varying objects, which are also known as inter-linked object recognition will have implications in many do- objects. Rather than recognizing each object indepen- mains, such as computer vision, image interpretation, dently, classification of such interlinked objects is done bioinformatics, text analytics, and web search. Further- collectively by performing simultaneous classification of more, the problem of efficient learning and inference several objects, which act as context for each other. in general structure random fields is closely related to During the first year of the project, we made advance- several important combinatorial optimization problems. ments in two areas. The first is the investigation and Methods developed in this project will therefore facili- development of descriptors to characterize contextual tate progress in combinatorial optimization, probabilistic cues. We focused on the computer vision area and graphical modeling, and related areas. developed algorithms to estimate similarities of image elements (such as image pixel patches) by considering Benefit to National Security Missions longer range interactions between the elements across This project contributes to addressing the Information multiple spatial scales, using descriptors based on edge Science and Technology Grand Challenge - the develop- orientation histograms, phase congruency, differen- 502

Page 503

tial invariants, and scale invariant interest points. These will investigate message scheduling schemes used during descriptors are used to characterize patches and their inference, as well as approaches to approximate inference relationships (e.g., similarities), while patches themselves in intractable graphs via ensembles of tractable sub-mod- constitute complex objects which are modeled as flexible els. constellations of parts augmented with inter-part contex- tual relations. This provides a context modeling framework For validation, we will utilize datasets used in the computer with statistical summary of the whole image, spatial layout vision community, as well as seek for applications outside properties within local image regions surrounding pixel of computer vision area. patches, as well as reduces original image data size and thus learning and inference computational complexity Conclusion due to the use of lower number of pixel patches instead Our goal is to develop efficient machine learning and infer- of huge number of individual pixels constituting original ence algorithms for probabilistic modeling of contextual image. information and context-aware object recognition based on Markov and Conditional Random Fields (M&CRFs). This The second is an optimization approach for performing will broadly entail investigating and developing algorithms approximate probabilistic inference in contextual models for (1) multi-scale contextual modeling and (2) scaling an of object recognition. The context framework based on approximate probabilistic inference and M&CRF learning graphical probabilistic models appears quite promising to large datasets. While developing these algorithms will but faces a challenge of computational intractability of have implications in many problem domains, we will focus exact inference if underlying probabilistic graphical model on computer vision, on a challenging problem of object contains cycles, which is the case in many real world cases, recognition in an unconstrained environment, for validat- e.g. in computer vision where an image is modeled as grid ing our algorithms. structured pairwise Markov random field (MRF) and/or Conditional random field (CRF). To address the problem, Publications we prototyped and investigated approximate inference Burr, T., and A. N. Skurikhin. Approximate inference options based on pseudo-likelihood (PL) approach. Instead of in data models for image analysis. 2013. presented at calculating the probability of individual variables at sites in the Statistical Image Analysis Workshop, Santa Fe, NM, MRF/CRF given all other variables at different sites, the PL LAUR 13-21603. approach approximates the joint probability of the vari- ables at nodes in MRF/CRF as a product of individual condi- Burr, T., and A. Skurikhin. Pseudo-likelihood inference for tional probabilities over each variable, where the condi- Gaussian Markov random fields. 2013. Statistics Re- search Letters. 2 (3): 63. tioning is on the neighbors of the variable. This property makes it a very efficient parameter learning method with Burr, T., and A. Skurikhin. Selecting summary statistics in some loss of accuracy, for it avoids exactly calculating the approximate Bayesian computation for calibrating sto- likelihood partition function. The PL-based approximation chastic models. 2013. BioMed Research International. has been shown to produce consistent estimates in the 2013: 10 pages. large lattice limit. We investigated inference based on PL in “flat’ and hierarchical CRF models. Skurikhin, A. N., S. R. Garrity, N. G. McDowell, and D. M. Cai. Automated tree crown detection and size estima- tion using multi-scale analysis of high resolution satel- Finally, when the data and/or parameter dimension is ex- lite imagery. 2013. Remote Sensing Letters. 4 (5): 465. tremely large, particularly whenever the likelihood involves summing probabilities over many unobserved states such as genealogies in biology, and, for example, in applications in astronomy and cosmology, and corresponding models do not have an analytical likelihood function, we devel- oped and investigated an approach to choose summary statistics for approximate Bayesian computation to enable inference without explicit likelihood. Future Work We plan to investigate and develop algorithms for ap- proximate probabilistic inference to reduce computational complexity of the learning and recognition processes. We 503

Page 504

Information Science and Technology Exploratory Research Continuing Project Physics-based Data Models and Architecture-optimized Backends for a Portable Data-parallel Computation Library Christopher M. Sewell 20130418ER Introduction This proposal is intended as a significant first step to- Portability is a major challenge for software developers wards allowing scientists to easily map their simulations due to the increasing heterogeneity of high-performance onto a Thrust-based data model and efficiently run their computing hardware. To address this issue, we have algorithms on emerging parallel technologies. recently implemented several standard algorithms using NVIDIA’s Thrust library. Thrust provides basic vector data Benefit to National Security Missions types and implementations of a number of data-parallel The deliverables and the experience resulting from primitives (such as scans, reductions, stream compac- the proposed work will build laboratory capabilities in tions, etc.) on these vectors in backends that target programming models within the larger exascale com- OpenMP and CUDA. While we believe that the data-par- puting co-design missions. The physics targeted by our allel model embodied by Thrust is the right approach to front-end work has been chosen to support the type of achieve both portability and performance across a wide simulations critical to the DOE Nuclear Energy Programs range of architectures, Thrust suffers from several major and other DOE agencies, while the architectures tar- limitations that prevent its widespread use. geted by our back-end work are being chosen to support the specific hardware that will be likely to be used in the First, Thrust has an extremely simplistic data model, laboratory’s next generation of supercomputers. The de- consisting only of one-dimensional vectors, which makes liverables of this project, including interface reconstruc- it difficult for a programmer to use with more complex tion and advection operators for both structured and un- data types commonly used in physics simulations, such structured grids that can achieve good performance on as multidimensional and unstructured meshes. several next-generation supercomputing architectures, could be used to interface with compressible multi-ma- Secondly, Thrust’s OpenMP backend is not very efficient terial hydro codes used extensively in the ASC program on many platforms, especially vector architectures, and in X Division. This work should also be complemen- and devices which do not use CUDA or OpenMP are tary to other work on cross-platform high-performance not supported. Thus, we propose to devise an abstract computing at the laboratory. For example, the phys- data model on top of Thrust which is co-designed with ics code resulting from this work could potentially be developers of a multimaterial hydrodynamics simula- compiled by the compiler being developed by the Scout tion. Furthermore, we intend to expand and improve the project, with this work providing a useful application and backends for Thrust to better target implementations of data abstraction for Scout, and the low-level compiler the data-parallel primitives to emerging technologies. optimizations made by Scout providing an opportunity for further performance enhancement for this work. While we will target vendor technologies based on lab computing priorities, we expect to target the Intel MIC Progress architecture and at least one additional platform, such We have implemented 1D, 2D, and 3D mesh data struc- as an IBM Power System or the ARM Architecture. Co- tures that allow physicists to easily map a structured design will be integrated across all levels, from algorithm grid to the 1D vectors used by Thrust. The design of the development to data model design to backend imple- mesh data structures provides an interface familiar to mentations to (on a limited scale) hardware vendor computational physicists while introducing them to the engagement. concept of the data parallel programming model. This 504

Page 505

enables them to take advantage of the available parallel- project status update e-mails. ism while still thinking in terms of familiar structures such as cells, vertices, and edges. This is the result of several We believe that we are on track to meeting our two Year iterations of tight and productive cooperation between the 1 milestones by the end of the fiscal year. Specifically, computational physicist and the computer scientists of the we expect to complete our implementations of interface team. We have implemented volume fraction initializa- reconstruction in 2D and in 3D using the structured mesh tion using all three of these mesh data structures, and are data structures that we have developed, and to be able to currently implementing interface reconstruction for the 2D run our algorithms efficiently on a single Xeon Phi (MIC) case. coprocessor and on multiple Xeon Phi coprocessors. We have purchased a small cluster of Intel Xeon Phi (MIC) Future Work nodes, which have been successfully installed within our Second-year Physics Co-Design Milestone: Complete im- group’s Darwin cluster, and we have extensively studied plementation of volume fraction initialization and interface this architecture (including attending several MIC program- reconstruction on unstructured meshes. ming workshops). We have run our PISTON isosurface Second-year Backend Co-Design Milestone: Decide on operator, using the existing OpenMP backend, directly on additional target architecture(s), acquire access to neces- an Intel Xeon Phi coprocessor (i.e., in “native” MIC mode), sary hardware, and learn how to use and understand the with good scaling results with the number of threads. architecture(s). We then modified the OpenMP backend for selected Conclusion data-parallel primitives in order to improve vectorization We expect to deliver a fluid flow simulation of multiple ma- performance on the Xeon Phi. We have also started imple- terials that seamlessly maps its mesh data structures onto menting backends for a small number of Thrust primitives a new Thrust data model, enabling it to run key algorithms using the “offload” mode (in which some processing is per- and operators with high parallel efficiency on multiple formed on the host and some on the coprocessor). In the emerging supercomputing technologies. At a minimum, near future, we plan to improve vectorization and support this would include visualization and analysis operators that offload mode for additional Thrust primitives. compute information such as interface boundaries, tak- ing advantage of cutting-edge parallel architectures such Furthermore, we have extended the Thrust library to sup- as Intel MIC. The data model developed should be ap- port concurrency in distributed memory environments plicable to a variety of other simulations, and the experi- across multiple nodes. This enables the application devel- ence gained from optimizing for the targeted architectures oper to write data-parallel algorithms while viewing the should be applicable to a variety of additional platforms. data as single, long vectors, essentially without needing to explicitly take into consideration whether the values are Publications actually distributed across nodes. Our distributed wrapper for Thrust handles the communication in the backend us- Sewell, C., L. Lo, and J. Ahrens. Portable data-parallel visu- ing MPI, while still using the standard Thrust library to take alization and analysis in distributed memory environments. advantage of available on-node parallelism. We have writ- To appear in IEEE Symposium on Large-Scale Data Analysis ten distributed implementations of several key data-paral- and Visualization. (Atlanta, Georgia, 13-14 Oct. 2013). lel primitives, including scan, scatter/gather, sort, reduce, and upper/lower bound. In the near future, we expect to run algorithms using this distributed implementation on multiple Xeon Phi (MIC) coprocessors in our cluster. We have submitted a paper about our distributed Thrust design to the Large Data Analysis and Visualization Sym- posium at the IEEE Visualization Conference. Marianne Francois gave an invited talk about the underlying physics for the interface reconstruction method entitled “Advances and Challenges in Modeling Interfacial Flows” at the March meeting of the American Physical Society. Infrastructure for project management has also been established, includ- ing a git repository on a shared network space and periodic 505

Page 506

Information Science and Technology Exploratory Research Continuing Project Co-Design of Burst Buffer Hardware and Data Analysis/Visualization Software for Large-scale Simulations Jonathan L. Woodring 20130457ER Introduction Benefit to National Security Missions Burst buffers are new concepts for storage architectures The primary technical goal of this co-design effort is to that will enable larger supercomputers. A burst buffer provide specification documentation for burst buffers, can be thought of as a cache that sits between the high large-scale simulations, and visualization and analysis performance supercomputer and its file system. Its workflows. Technical goals that must be achieved in- main purpose is to support high-speed checkpoint and clude building a software framework that enables post, restart for fault tolerance enabling larger supercomput- in situ, in transit, and hybrid visualization and analysis ers. The supercomputing communities are uncertain as processing, performance analysis of this software frame- to the burst buffers specifications and what they should work and building a prototype of a simulation working in be capable of performing. One extreme is that the burst this framework. buffer is no more than an extension of the file system acting as a buffer to the slower, larger, parallel file sys- The impact of this work will influence the design of tem. Another extreme for the burst buffer configuration future supercomputers: both hardware and software. In is a robust data intensive parallel supercomputer. addition, this work will influence computational science applications (across many disciplines) for large-scale The high potential of burst buffers starts with the fault design and analysis of results. tolerance enabling the next jump in scale of super- computers. This potential is extended by this proposal: The A-relevant categories are sciences and programs researching a burst buffer’s ability to visualize, analyze, that have a large simulation and/or supercomputing triage, and transform large-scale data while it is in transit component to their research and programmatic needs. (as the data are moving from compute to storage) or The results of this research will impact how future some combination thereof. This extended possible uses supercomputers, simulations, and analysis workflows for the burst buffer will enable larger scale supercomput- are designed to work with burst buffer enabled super- ers and decrease time to results for users of supercom- computing. To be able to run on the next-generation puters. supercomputers, the science programs will need com- putational and large-scale data assistance to utilize the We propose to use the Parallel Ocean Program (POP) hardware (such as burst buffers) for the simulations and and possibly other simulation codes, the Darwin Cluster, other computational components to scale up in size and Burst Buffer prototype nodes to configure simula- and complexity. Our research will be able to inform the tion-to-analysis data workflows. We will use combina- simulation and computational scientists, analysts, and tions of different modes of analysis and burst buffer system designers the best practices for using burst buf- hardware configurations, such as in situ, in transit, and fers in visualization and analysis, in addition to check- post-processing visualization and analysis. We will ex- point/restart. plore the hardware and software configuration space for these proposed burst buffer nodes ranging from simple Progress cache appliances to a full-fledged data intensive system. The tasks for last year were: Understanding what a burst buffer should be and what it

1. Install and Configure Burst Buffer Hardware is capable of, is a current barrier to designing and speci-
2. Prototype of POP as a Benchmark fying the next generation of supercomputers.
3. Generate Plots and Ratios of Data Production and 506

Page 507

Ingestion Rates of each Component in the System Architec- integrated into CESM (Community Earth Systems Model, ture while Running POP a combined climate model used in the IPCC, Intergovern- mental Panel on Climate), due to our acceleration of the We have successfully completed each of these tasks and MOC (Meridional Overturning Current), which will greatly created or finishing up the end of year deliverables for accelerate the CESM/COSIM analysis workflow. each task. We are currently finishing the benchmark report for POP, The burst buffer prototype hardware has been successfully POP analysis, and the prototype burst buffer itself. This deployed in the darwin research cluster. It is a small com- report will be useful for us (as baseline numbers for future pute cluster consisting of compute nodes and solid state research in the second and third years), but for other disk technology. In addition, we have set up a “scratch” researchers and engineers trying to evaluate burst buffer parallel file system to model the typical parallel file systems hardware and HPC environment configurations. used in HPC computing environments. The scratch paral- lel file system and burst buffers will be used to compare The base file system and hardware IO performance tests against in situ vs. in transit vs. post-processing analysis have been completed. We are currently in the phase workloads, coexisting with checkpoint/restart. such that we are tuning parameters and doing a further parameter sweeps to compare the IO performance im- The burst buffer prototype is flexible in that we can explore provements through tuning the IO system and different different hardware and software configurations, for experi- configurations. This parameter experimentation includes menting with checkpoint/restart workloads and analysis tuning the file system layers and hardware for optimization configurations. We can reconfigure the burst buffers to act comparison and adds potential avenues for future experi- like traditional IO nodes (dumb buffers) compared to our ments. desired usage of co-analysis and visualization offloaded from the compute, coexisting with compute heavy work- Additionally, we have benchmarked the MOC analysis op- loads and checkpoint/restart fault tolerance. Our experi- erations for in transit mode and will be instrumenting POP ments will be how we can optimize the time to results, itself with an in situ MOC calculation in the near future while also providing fault tolerance services. for direct comparison. After the parallel IO routines are finished this month, we will benchmark the checkpoint/re- The current software stack under evaluation includes GPFS start capabilities on the burst buffer and baseline systems. with file placement option (new beta software that was This will include running the Lagrangian tracer analysis used in the IBM Watson system), JBOD (just a bunch of code and performance tests on in situ vs. in transit vs. disks), PLFS, HPFS (Hadoop), existing POP analytic soft- post-processing tracers. ware, new POP analysis code, and software derived from ParaView/VTK. We will also be exploring other portions The expected rough draft of the report is due in July and of the software stack in the next year, such as schedul- the final benchmark report will be completed in August. ing analysis with checkpoint/restart and how to perform These benchmark experimental numbers will give us migration of code to data (rather that data to code). the grounding to look for improvements in the analysis, computation, and scientific workflows in the next year and POP and a test input deck for simulation and analysis work- following year. loads has been successfully deployed on darwin and on the burst buffers as our test experiments. We are currently in Future Work the final stages of integration of parallel file system rou- Explore dynamic scheduling architecture for the burst tines into POP and POP analysis code (this includes MPI-IO buffer and PLFS file acceleration) for final benchmarking, listed in Currently in HPC systems, resource provision is based the next milestone. This will be useful to compare parallel around the batch queue model. If we assume that the writes to existing serial Fortran (POSIX) writes, and N-1 and burst buffers are part of the resource provisioning, this N-N workloads. means in traditional HPC queues that burst buffers will be under utilized. For example, a simulation will only use a Additionally, our current POP analysis work may have burst buffer temporarily, leaving it idle the rest of the time. near future external impact, as it will be leveraged by the To exploit the idle cycles, we will explore an IaaS model in COSIM/CESM climate community. One of our benchmark an HPC environment. Other researchers have attempted analysis routines, developed for our experiements, is this in the past, but with unsatisfactory results due to not currently under evaluation by the COSIM team. It may be fitting an HPC workload. We will configure our prototype 507

Page 508

with HPC environments in mind, prioritizing checkpoint/re- Mitchell, C., and D. Bonnie. In-Transit Processing: Data starts with an analysis (multi-user workload). This is unique Analysis using Burst Buffers. High Performance Parallel outside of existing IaaS models and previous research. I/O. Edited by Prabhat, , and Q. Koziol. Explore a programming model for applications, to move compute to data under distributed parallelism model (SPMD) Currently, the models for moving computation to data or data to computation only exist in the loosely-coupled cloud computing environment (see Hadoop and map-reduce). In scientific computing, most researchers use the SPMD (single program, multiple data) programming paradigm, that is most personified by MPI and PGAS programming models. We will look to extend the SPMD scientific programming model to account for data locality, which will be important for burst buffer analysis. If we assume that HPC data will be scattered across a burst buffer file system, especially without a global namespace, analysis applications need the capability to move computations to data, because data will not be able to be moved to computations. This pro- gramming model will allow scientific SPMD analysis adapt to “scattered data.” Explore advanced compiler technology Advances in compilers (LLVM) will support scheduling, moving and executing analysis for automatic in situ, in transit, and post-processing placement for reduced time to results. Conclusion The primary technical goal of this co-design effort is to provide specifications for burst buffers and analysis work- flows. Technical goals that must be achieved include building a software framework that enables post, in situ, in transit, and hybrid visualization and analysis process- ing, performance analysis of this software framework and building a prototype of Parallel Ocean Program working in this framework. The impact of this work will influence the design of future supercomputers and simulations: both hardware and soft- ware. In addition, this work will influence computational science applications (across many disciplines) for large- scale design and analysis of results. Publications Mitchell, C., D. Bonnie, and R. Knight. Co- Design of Burst Buffer Hardware and Data Analysis/Visualization Soft- ware for Large-Scale Simulations: Burst Buffer Initial Test Report, August 2013. 2013. LA-UR-13-27974, Los Alamos National Laboratory. 508

Page 509

Information Science and Technology Exploratory Research Continuing Project Validating Near-ideal Data Throughput Using a Rate-compatible Protocol Scott H. Robinson 20130504ER Introduction This project will demonstrate, both in the lab and under Benefit to National Security Missions real world conditions, an advanced information coding This project demonstrates technology that can signifi- technique, combined with a specialized transmission cantly increase the data throughput capacity of exist- protocol, which will maintain near-ideal data through- ing wireless (RF) communication links. Any mission or put performance regardless of the noise present in a agency concerned with the efficient and reliable delivery wireless (Radio Frequency or RF) communication chan- of digital information in the presence of dynamically nel. Working together, the coding technique and the changing interference could benefit from our work. transmission protocol dynamically tunes the throughput NNSA is responsible for satellite instruments to meet na- rate of each data packet to adjust to rapidly changing tional treaty monitoring requirements. This includes the interference in the transmission channel. This is done delivery of sensor data to the ground during ionospheric automatically with a minimum amount of feedback and disruptions due to a high altitude nuclear event. Our no specialized equipment to measure the quality of the technology would ensure maximum data delivery under channel. The result is a data transfer rate that approach- these dynamically changing conditions. Additionally, es the theoretical maximum set by the Shannon limit. NNSA remote sensing missions also have the need to ef- ficiently transmit data across a RF link and could benefit Our technique maximizes data throughput capac- from our technology. ity across a wide range of channel conditions; adjusts dynamically to changing conditions in the transmission Agencies including the DOD, NASA and numerous Intelli- channel without a priori knowledge; minimizes the gence Agencies rely heavily on wireless communications amount of back channel communication required from to perform their missions. Often these RF links must op- the receiver; can be implemented using common com- erate when the condition of the communication channel munication hardware; and is compatible with existing is poor due to natural interference, weak transmission modulation and channel bandwidths. Previous work signals or intentional jamming. Our coding algorithm included a detailed description of how data packets are and transmission protocol can significantly improve the constructed; how data packets are decoded; how hand- performance of these links. shaking is done between transmitter and receiver; and a computer simulation comparing expected performance In the world of commerce, the annual growth in wire- of our technique to the performance of other existing less data traffic has been predicted at 74% over the next techniques over a wide range of channel quality. The re- couple years. Much of this growth is driven by expanded sults of this simulation showed significant improvement use of smartphones. Due to the mobility of these over the other coding methods. devices, channel conditions constantly change due to their distance from a base station; intervening buildings Our work will implement the coding algorithm and the or terrain; and noise present in the environment. Our transmission protocol in actual hardware and validate technology can improve the performance under all these its data throughput performance under a wide range of conditions while increasing the data carrying capacity of interfering noise. the RF spectrum. 509

Page 510

Progress pleted. The encoder can handle permutation pattern sizes Work this year has focused on three main areas: construct- up to 65336 bits in length which is important to provide ing a model of the information encoding technique and flexibility during future testing. The encoder uses block transmission protocol; implementing the design in logic memory structures in the FPGA for storing message bits compatible with hardware; and purchasing and gaining and parity check bits allowing it to support a variety of experience with the hardware and tools that will be used puncturing schemes. An initial VHDL design for the decod- for building the prototype. er has also been completed. This design uses a technique that allows the size of the decoder matrix and bit-widths to A variety of Low Density Parity Check (LDPC) codes were be adjusted to analyze performance trade-offs. A moder- researched to identify the best form of a LDPC code that ate sized version of the decoder was then synthesized provides both good rate-compatible code performance into hardware so FPGA logic resources could be assessed and is well-matched to the prototype hardware. Through including logic gate and block memory usage. Work contin- this research the Weight-3-Repeat-Accumulate (W3RA) ues to refine the design of the decoder to optimize the use code was selected. These codes have a simple encoder of logic resources to allow larger versions to be efficiently structure while still maintaining a lower bit-error-rate floor realized. compared to other standard Repeat-Accumulate LDPC codes. Two methods to convert the W3RA code into a rate- Two wireless development kits have been purchased from compatible code were then explored. These techniques Texas Instruments (part number CC1101DK) to provide the are check node splitting and check node puncturing. After frontend radio component for the prototype. These units evaluation, check node puncturing was selected due to the use the unlicensed instrumentation frequency band of simplicity of its implementation. 433 MHz. Initially, these radios will be connected together using a coax cable. Actual over-the-air transmission will be High level simulation models of the encoder and decoder done during year two of the project and after permission have been fully developed and tested. This work has in- to transmit is obtained from the laboratory. These radios cluded the creation of a unique algorithm that constructs will be connected to a FPGA Development board from a parity check matrix in a way that avoids a problem with Xilinx, the ML605. The FPGA on the ML605 will support the decoding known as small Tanner graph cycles. If a check W3RA encoder, decoder and transmission protocol while matrix has small Tanner graph cycles then the amount of the radio is used to physically transmit and receive the redundant information in the code will be low and it will information. not operate well in the presence of noise. This algorithm is important to this effort so that the developed W3RA Development is currently in progress to connect the vari- code will exhibit good decoding performance across a wide ous segments of the hardware design and create commu- range of transmission channel conditions. Work continues nication between the CC1101DK, the ML605 and a host on construction of a full end-to-end model that will use a PC. Once completed, the host PC will be used to configure Monte Carlo simulation to compare the bit-error-rate of the FPGA on the ML605 with the encoder, decoder and the W3RA code when using different amounts of punctur- protocol design; adjust transmission parameters; send and ing that provides the rate-compatible performance re- receive messages; and monitor performance statistics. quired. Future Work Using the model of the encoder and decoder, logic has The goal during the first year of the effort has been to been developed to implement the design in a Field Pro- implement the information coding algorithm and transmis- grammable Gate Array (FPGA). This required the high level sion protocol in a prototype wireless device and perform simulation model to be converted to a low level design initial data throughput testing in the lab. using fixed point arithmetic and lookup tables to simplify The goal of the second year of the effort will be to demon- hardware implementation. These designs are written in strate the ability of the design to maximize data through- VHDL, a standard hardware description language used put across a dynamically changing noisy channel. Initially for FPGA design. Throughout the development process, work will be done in the lab to optimize the performance simulation of the low level fixed point design and the VHDi of the coding algorithm, the transmission protocol and its design are performed and results are compared to the implementation in hardware. Additionally, tools will be original design to ensure correct implementation. developed to monitor, in real time, the throughput rate of A VHDL design of the W3RA code encoder has been com- the data being transferred by the devices and the signal to 510

Page 511

noise ratio of the communication channel. The prototype devices will then be taken outside and tested wirelessly in an operational environment. Major tasks for the second year of the effort are: • Establish an experimental setup in the lab between two prototype devices communicating with each other using a coax cable. Use lab equipment to inject noise into the communication channel and monitor perfor- mance. • Optimize performance of the information coding algo- rithm and transmission protocol in the hardware. • Document the data throughput performance of the prototype devices across a wide range of simulated noise using the lab experimental setup. • Move to wireless operations outside and document the data throughput performance of the prototype devices under operational conditions. These tests will include varying the power of the transmitter and moni- toring performance while both devices are in motion. Conclusion At the completion of this project, a demonstration proto- type will exist with sufficient quantitative test results to validate the data throughput performance of the informa- tion coding algorithm and transmission protocol across a wide range of channel conditions. These tests will be done both in the lab and over the air in an operational environ- ment. These results will provide a solid foundation for future operational programs to evaluate the benefits and computational complexities of our technology. 511

Page 512

Information Science and Technology Exploratory Research Continuing Project Sparse, Distributed, and Robust Network Control Marian Anghel 20130558ER Introduction Benefit to National Security Missions The goal of this project is the design of distributed con- Recent workshops organized by the Department of trollers for complex interconnected systems. Although Energy (DOE) in partnership with Office of Electricity emerging network control applications include large Delivery and Energy Reliability (OE) were focused on the mechanical structures, distributed sensing, and interact- challenges of grid modernization efforts (March 2009) ing mobile robot agents, the focus of this project is the and on the computational needs for the next generation design of distributed controllers for large scale power electric grid (April 2011). The proposed research directly grid systems. supports these program development directions. This project aims to address two critical weaknesses of More specifically, the report on “Advanced Control power grid systems. First, the current power grid has Methods” conducted by the National Energy Technology not yet fully embraced the most recent advances in Laboratory for the DOE in 2007 has identified the need distributed sensors, embedded computing systems, and for the development of distributed intelligent agents advanced communication networks. Second, the power that respond rapidly at the local level to unburden cen- grid faces the simultaneous convergence of several tralized control systems. Our project directly addresses emergent difficulties (intermittent renewable energy this problem by developing a distributed control meth- sources, frequent transmission congestions, relentless odology which is robust with respect to system distur- market pressure) which the existing centralized control bances. infrastructure will not be able to solve. Furthermore, the Smart Grid Research and Develop- In response to these challenges, this project seeks to ment Program of the DOE/OE, identifies in its multi-year construct an algorithmic synthesis of nonlinear and dis- program plan the development of advanced control tributed control techniques for large networked power methods. As the smart grid evolves, this program notes systems. More specifically, we propose a decentralized that more complex automated control systems will be approach which exploits separability and decomposition necessary to maintain optimum operation of the grid. of the corresponding control problem into nonlinear The control algorithms that will be investigated in this sub-problems which can be solved locally and efficiently. project directly address the fundamental conceptual This project explores new ideas from compressed sens- challenges posed by such automated control systems. ing, convex optimization, and algebraic geometry, that Finally, the April 2011 DOE workshop, “Computational have the potential to radically change the ability to for- Needs for the Next Generation Electric Grid Proceed- mulate the stability and control of the future Smart Grid ings,” highlights a class of mathematical and computa- (and in general large networked dynamical systems). tional problems relevant for potential power systems research. It identifies, in particular, the difficulties cur- Success of this project will position LANL at the forefront rently encountered by direct transient stability analysis of this research and enable engineering control systems methods. This project directly addresses this problem to be developed that can operate at the global network and our preliminary results have already offered sig- scale. nificant breakthrough results based on novel algebraic control techniques. 512

Page 513

Progress tive of minimizing the effect that the neglected interaction We have built the software infrastructure needed to dynamics between subsystems have on the subsystem drift simulate the dynamics of classical power systems. We dynamics. Hence, this decomposition will facilitate the have also built the software framework needed to cast the construction of a composite Lyapunov function. power systems equations into a polynomial form that Sum of Squares (SOS) methods can use for system analysis and We have applied the combination of decomposition and controller design. The SOS method is a recently developed SOS Lyapunov techniques to a 7 generator 26 bus power algebraic framework for analysis of nonlinear systems, system model. We remark that this system is too big to be which uses the SOS decomposition of polynomials as a analyzed directly using SOS techniques. computational relaxation to polynomial non-negativity conditions. The algorithmic search for non-negative SOS The analysis has been described in a paper titled “Stability polynomial enables the algorithmic construction of Lyapu- Analysis of Power Systems using Network Decomposition nov functions for nonlinear dynamical systems. and Local Gain Analysis,” that was submitted and accepted to the 2013 IREP Symposium-Bulk Power System Dynamics Moreover, we have implemented a decomposition tech- and Control. nique that decomposes the power system into a number of smaller, weakly interacting subsystems. This algorithm We have also significantly revised two papers that were was proposed by our collaborators at Oxford University submitted for publication before the start of this project and employs graph theoretic methods that takes into and that were under review when this project started. The account both the strength of interactions and the energy papers are highly relevant for this project. The first one flows between the generators in different subsystems. titled “Algorithmic Construction of Lyapunov Functions for This algorithm generalizes early decomposition approaches Power System Stability Analysis,” has been accepted for that were based only on the magnitude of the elements of publication in IEEE TRANSACTIONS ON CIRCUITS AND SYS- the system matrix - the strength of the interaction terms TEMS and is already published online. This paper describes between the system generators. It was shown that this is a SOS based algorithm that is used for computing Lyapu- not a good approach, as magnitude alone does not always nov functions for each individual subsystem that results provide a good indication of coupling strength between after the system decomposition algorithm is applied. subsystems. The proposed algorithm takes both energy flow and dynamics into account whereas many other A second paper titled “Spontaneous synchrony in power- methods simply account for the system topology. grid networks,” has been accepted and published by Na- ture Physics. This paper derives a condition under which After we have performed a system decomposition into the desired synchronous state of a power grid is stable, a number of dynamic subsystems specified a priori, we and use this condition to identify tunable parameters of have computed Lyapunov functions for each subsystem the generators that are determinants of spontaneous and used them to estimate the region of attraction (ROA) synchronization. We intend to use this finding in the con- pertaining to the equilibrium point of each isolated subsys- troller design phase of our project. tem. We have used sparsity constraints to force a simple expression for these functions - minimize the number of Future Work monomials in our polynomial Lyapunov functions - and we In the first year of the project we have implemented a have also computed sparsity free Lyapunov functions for decomposition technique that seeks to decompose the which no sparsity constraints were imposed. We started to system into a number of smaller, weakly interacting subsys- analyze how well these Lyapunov functions approximate tems. This algorithm employs graph theoretic methods the region of attraction of the stable operating point and that takes into account both the strength of interactions how sparsity constraints affects the quality of these ap- and the energy flows between the generators in different proximations. subsystems. Finally, after we have identified Lyapunov functions for We have also computed independent Lyapunov functions each isolated subsystem, we started to aggregate together for each subsystem using a Sum of Squares (SOS) algorithm the subsystem Lyapunov functions in order to compute a that we have recently introduced. We have used sparsity composite Lyapunov function for the entire system. We constraints to force a simple expression for these polyno- note that using a good decomposition algorithm is critical mial Lyapunov functions. for this approach to succeed. Indeed, the decomposition algorithm seeks a partition of the system with the objec- We have also computed estimates for the region of at- 513

Page 514

traction (ROA) of each subsystem when its interactions ty Analysis of Power Systems using Network Decompo- with the rest of the system are neglected. These estimates sition and Local Gain Analysis. To appear in Bulk Power provide a bound on the largest disturbance (angle and System Dynamics and Control Symposium.. (Crete, Greece, 25-30 Aug. 2013). frequency deviations ) that the subsystem can withstand without loosing its stability. We started to analyze the trad- Motter, A. E., S. A. Myers, M. Anghel, and T. Nishikawa. eoff between highly sparse (few monomial terms) polyno- Spontaneous synchrony in power-grid networks. 2013. mial Lyapunov function and the more complex functions NATURE PHYSICS. 9 (3): 191. that can be computed when no sparsity constraints are used in the SOS programming framework. Furthermore, we started to implement SOS techniques to aggregate together these subsystem Lyapunov functions in order to compute a composite Lyapunov function for the entire system that takes into account the interactions between subsystems. In the second year we will couple the decomposition ap- proach with the SOS analysis framework to analyze power grid systems of increasing size. We will apply recursive decomposition techniques to larger power grid systems and we will compute the input-to-output stability proper- ties of each stable subsystem using the SOS framework. By patching together the input-to-output response of each subsystem we will derive stability bounds for the large- scale complex system. We will improve the stability of each subsystem by designing local state and output feed- back controllers and estimate their disturbance rejection performance. Conclusion This project will provide for the first time a methodology for the algorithmic construction of Lyapunov functions and for the design of distributed, nonlinear controllers for power grid systems. The algorithm will use sparsity constraints to control the complexity of the controllers and a hierarchical decompo- sition methodology in order to extend modern nonlinear algebraic control techniques to large scale distributed systems. The successful completion of this project will generate a renaissance of direct analysis methods based on Lyapunov techniques and will open new research directions in the distributed control of power grid systems. Publications Anghel, M., F. Milano, and A. Papachristodoulou. Algo- rithmic Construction of Lyapunov Functions for Power System Stability Analysis. 2013. IEEE Transactions on Circuits and Systems I: Regular Papers. 60 (9): 2533. Anghel, M., J. Anderson, and A. Papachristodoulou. Stabili- 514

Page 515

Information Science and Technology Exploratory Research Final Report Rare Category Detection James P. Theiler 20110126ER Abstract and proliferation detection for national security. This We investigated a new kind of anomaly detection, called project will develop theory, algorithms, and validation Rare Category Detection, with the aim of more efficient- experiments for a new approach called rare category ly finding and more accurately identifying anomalies that detection (RCD) that promises substantial improvements are meaningful. We looked at this issue from a num- in our ability to detect meaningful anomalies. ber of different angles, developing theoretical insights, Scientific Approach and Accomplishments practical implementations, and looking into a variety of potential applications, ranging from medical imagery to Rare Category Detection generally involves two stages: satellite state of health. stage 1 identifies the initial regions of interest in the data (typically via more traditional anomaly detection) and Background and Research Objectives stage 2 modifies and/or updates these regions after each A seemingly impossible task in data mining is that of user transaction. extracting meaningful rare-event data from massive data In the first year, we developed a new kind of anomaly sets when we don’t know what is meaningful until we detection where instead of locating data from sparsely see it. On the one hand, any credible solution, whatever populated regions (as in standard anomaly detection) its form, must incorporate a human expert to determine we locate data from regions with high concentration what is meaningful and what is not. On the other hand, (so they are meaningful) but low probability mass (so human experts can directly examine only an infinitesi- they are still rare). We have discovered that the key to mal fraction of the data, thus creating what appears to doing this efficiently is to synthesize a background data be a fundamentally intractable problem. To confront set and then solve a 2-class pattern recognition problem this intractability we must make prior assumptions, and that discriminates between the observed data and the hope that they capture some relevant characteristics of background. The utility of this method depends on how meaningful data. The most common prior assumption we synthesize the background. Instead of adopting a is that meaningful rare-event samples are unlike the conventional approach that would synthesize the back- rest of the data, that they are anomalous. And indeed, ground from scratch using an artificial model, we have anomaly detection is an important component of the so- begun the development and testing of methods that lution, though standard anomaly detection may by itself produce a background by applying a special mapping to be inadequate. For example, in an astronomical image the observed data. This has numerous advantages for archive, we can identify the most anomalous 0.1% of the implementation, performance, and robustness. objects in the sky survey using standard anomaly detec- tion, but only 1% of these anomalies (and only 0.001% In the first year, we also made progress on stage 2 with of the full dataset) might actually be meaningful in the the development of “hemi-supervised” learning. This sense that they correspond to potential new discoveries approach is effective in cases where the ground truth by the astronomer. The goal is to identify these meaning- provided by the user is both one-sided and limited. In ful anomalies as efficiently as possible. Similar examples particular the user is only required to identify samples are can be found in the areas of network monitoring and that are in a category (without needing to counterbal- file system forensics for cybersecurity, and nuclear foren- ance those with samples that are “not” in a category). sics, scenario extraction, structural health monitoring, We have developed hemi-supervised methods that are satellite health monitoring for the space nuclear detona- effective for a wide range of situations that extend from tion detection mission, remote sensing image analysis, 515

Page 516

cases where the data can be described by simple Gaussian An important step in the design of an optimal detector is probability distributions to cases where the data distribu- to obtain a robust estimate of the broadband noise. This tion is arbitrarily complex. Furthermore, since the ground is a surprisingly difficult problem when the received signal truth provided by the user is very limited we have begun to contains a significant number of narrowband components, incorporate methods of “transfer learning” and “statistical which is typical in a remote sensing environment. One of relational learning” into our framework. Transfer learning our second-year accomplishments was to develop a new is concerned with the storage of knowledge gained while method for estimating the broadband noise, and to prove solving one problem and then applying it to a different that this method provides superior performance in these but related problem. This will allow us incorporate outside environments. knowledge efficiently and effectively without burdening We also developed a new method for distinguishing be- the user. Statistical relational learning is concerned with tween a proliferate signal and narrowband clutter. Many the discovery of relationships between different compo- types of proliferate activities generate a harmonic series; nents of the data, and will be used to help us make con- i.e., a time series that consists an oscillatory signal plus nections between the small amount of ground truth data its harmonics. However, a remotely sensed signal usually provided by the user and the other parts of the data. contains a large amount of sinusoidal clutter; i.e., a large A specific effort during the first year was the development number of arbitrarily spaced sinusoids. The difference of a new empirical error-minimizing algorithm for nearest- between the proliferate harmonic series and the clutter neighbor classifiers that reduces the number of exemplars sinusoids is that the clutter sinusoids are not harmonically (and hence the memory storage, and exploitation time) aligned. Thus we can discriminate between the two by and obtains competitive performance to the conventional detecting the presence or absence of harmonic alignment. k-nearest-neighbor classifiers, which have significantly The usual solution to this problem is to identify all the greater computational complexity. In addition, by directly sinusoidal components in the signal, and then to see how minimizing error, our approach enables nearest-neighbor many of them are harmonically aligned. We have devel- classifiers to be used for cost sensitive pattern recognition oped a simpler method that detects the presence of a problems that arise in anomaly and rare category detec- harmonic signal more directly without having to identify all tion. the sinusoidal components of the signal. This eliminates an intermediate step and improves detection performance. The second year pursued the further development of hemi-supervised detection methods for nonprolifera- The signal detection problem is one of a class of statistical tion problems. In a typical nonproliferation problem, the problems called “composite hypothesis testing;” what’s proliferate activity produces a signal that is mostly known composite is that the signal of interest can take on multiple ahead of time, and the challenge is to detect this signal in forms (for instance, signal strength is typically not known a deployed environment where it is buried in significant a priori, and one therefore wants to detect signals over a amounts of broadband noise and narrowband clutter that range of signal strengths). Although the simple hypoth- are not known ahead of time. One aspect that makes this esis testing problem admits a well-known optimal solu- problem difficult is that the noise and clutter are not con- tion – the likelihood ratio test – the composite hypothesis stant; they change from one deployment to the next, and problem is more ambiguous. The standard solution is the they change from one time period to the next for a specific generalized likelihood ratio test (GLRT), which involves deployment. Existing detection methods assume a fixed replacing unknown “nuisance” parameters with their maxi- model for the noise and clutter, and provide mixed results: mum likelihood estimates, and then following that up with they provide good performance when the model is ap- a likelihood ratio test. Recently, a further generalization of proximately correct, and poor performance when it is not. the GLRT has been proposed which combines (or “fuses”) On the other hand, the hemi-supervised detection meth- what are known as “clairvoyant solutions” (these are the ods developed in this project can adapt to the changes in likelihood ratio solutions that would be obtained in the noise and clutter, and can therefore provide good perfor- hypothetical situation that the nuisance parameters were mance in all cases. The hemi-supervised approach requires known). We investigated these new clairvoyant fusion that the detector be designed “on the fly” while the data algorithms, and identified both efficient formulations and is being collected in the field. As part of this effort, we suc- theoretical difficulties. cessfully developed several components of the hemi-su- In the third year, we focused on an application in Space pervised detection algorithm, and worked on making these Nuclear Detonation Detection (SNDD), and in particular, components efficient so that they can be used to design investigated an archival dataset for the Space and Atmo- and implement the detector “on the fly”. spheric Burst Reporting System Validation Experiment 516

Page 517

(SAVE). This instrument was launched in November 2007, Publications and began experiencing a series of unexpected events, Kwon, H., X. Hu, J. Theiler, and A. Zare. Editorial: Algorithms starting in January of 2008, and culminating in a load bus for Multispectral and Hyperspectral Image Analysis. glitch and instrument shut-down on March 12, 2008. Our 2013. Journal of Electrical and Computer Engineering. interest was in identifying and characterizing anomalous 2013: 908906. behavior during the lifetime of the satellite. In the process Mascarenas, D., A. Cattaneo, J. Theiler, and C. Farrar. Com- of examining these data, we developed a further appre- pressed sensing techniques for detecting damage in ciation for the importance of feature selection, and some structures. 2013. Structural Health Monitoring. 12: new tools and ideas. One new notion is the idea of “subtle 325. anomalies” – these are data samples that are unusual in the context of a multidimensional distribution, but are not Porter, R., D. Hush, and B. G. Zimmer. Error Minimizing far from normal in the context of the individual features. Algorithms for Nearest Neighbor Classifiers. 2011. In Proceedings of the International Society for Optics and Thus, if the temperature at a sensor is unusually high, then Photonics. (San Francisco, January 23-27, 2011). , p. 1. that is an anomaly; but if the temperature in one sensor New York: SPIE. is just slightly higher than average, while the voltage at an adjacent sensor is slightly lower than average, and if Prasad, L., J. Theiler, M. Fair, and S. Swaminarayan. Feature those sensors usually track each other, then that may be a Extraction, Anomaly, and Change Detection on World- subtle anomaly. These anomalies are interesting because View 2 Imagery by Hierarchical Image Segmentation- a they won’t be found by simple “envelope based” anomaly study. 2012. In Annual Conference on Algorithms and detectors based on individual sensors. Technologies for Multispectral, Hyperspectral, and Ul- traspectral Imagery XVIII ; 20120423 - 20120427 ; Balti- The SNDD problem also forced us to confront the essen- more, MD. Vol. 8390, p. 83901U. tially non-Gaussian character of these data, and to ad- Stearns, S. D., and D. R. Hush. Digital Signal Processing with dress a distinction between outliers and anomalies. A few Examples in MATLAB. 2011. extreme outliers, due for instance to data drop-outs, might not themselves be very meaningful as anomalies, but can Theiler, J.. Confusion and clairvoyance: some remarks on distort the covariance estimators that are central to the the composite hypothesis testing problem. 2012. In anomaly detection process. The Catch-22 is that covari- Annual Conference on Algorithms and Technologies for ance estimators are required for outlier detection, but the Multispectral, Hyperspectral, and Ultraspectral Imag- outliers need to be identified in order to obtain robust ery XVIII ; 20120423 - 20120427 ; Baltimore, MD. Vol. covariance estimates. Traditional algorithms approach 8390, p. 839003. this problem iteratively. But we can exploit the funda- Theiler, J.. The incredible shrinking covariance estimator. mentally heterogeneous nature of the features (some are 2012. In Automatic Target Recognition XXII ; 23-24 temperatures, some are voltages, some are on/off, etc.) April 2012 ; Baltimore, MD, USA. Vol. 8391, p. 83910P to our advantage, and identify outliers using individual (12 pp.). envelope-based detectors. Removing those outliers, we then employ multivariate covariance-based methods for Theiler, J.. Formulation for min-max clairvoyant fusion more meaningful anomaly detection. based on monotonic transformation of statistics. 2012. Optical Engineering. 51: 111714. Impact on National Missions Theiler, J., and B. Wohlberg. Local coregistration adjust- This project is directly linked to the core data analysis ment for anomalous change detection. 2012. IEEE component of national security missions in cybersecurity Transactions on Geoscience and Remote Sensing. 50 and nonproliferation. It helps build a unique capabil- (8): 3107. ity in machine learning that directly supports the future program development plans described by the LANL IS&T Zimmer, B. G., D. Hush, and R. Porter. Ordered Hypothesis Center. Furthermore it provides capability that can be used Machines. 2011. Journal of Mathematical Imaging and Vision. : 1. to leverage program development in several key areas including: file system forensics and network monitoring for Zuluaga, M., D. Hush, E. F. Leyton, M. Hernandez Hoyos, cybersecurity (DOE/NSA/CIA), satellite health monitoring and M. Orkisz. Learning from Only Positive and Unla- (DOE/NNSA) and proliferation detection and monitoring beled Data to Detect Lesions in Vascular CT Images ; (DOE/NNSA). pt.3. 2011. In Medical Image Computing and Comput- er-Assisted Intervention (MICCAI 2011). ( 14th Interna- tional Conference ; 18-22 Sept. 2011 ). , p. 9. 517

Page 518

Information Science and Technology Exploratory Research Final Report Exact Renormalization Method for Frustrated Systems and Optimization Cristian D. Batista 20110181ER Abstract tions because small perturbations produce big changes We suggest a renormalization framework to study com- in the response of the system. Frustrated systems are plex phenomena in interacting and frustrated systems, also at the root of Optimization: Certain instances of with additional applications in combinatorial optimiza- combinatorial optimization problems can be associated tion. Such phenomena are characterized by the competi- with frustrated systems whose ground states encode the tion and potential coexistence of various states of mat- solution to the problem. ter. Unveiling the organizing principles behind complex Under this project we developed a code based on a behavior in matter is a central problem in physics and novel algorithm for solving models that are relevant for other sciences. Our prime goal was to devise an exact describing such materials. We obtained all of the exact and efficient computational method for understanding ground states under control and unraveled novel states emergent behavior in physical systems with competing of matter that are relevant for certain classes of frustrat- interactions. We applied recent and advanced results ed materials. We also collaborated with our experimen- in information theory, mathematics, and quantum talists from MPA-CMMS as some of the materials can be many-body physics, to construct a scale-renormalization understood with our code. algorithm that is exact under certain assumptions. Next, we used this algorithm to predict the low temperature Scientific Approach and Accomplishments physical properties of relevant models that remain Code Development: We developed a code to efficiently unsolved. In addition, since combinatorial optimization solve a family of XXZ spin Hamiltonians on a ladder problems can be reformulated in the physics language by (1D) by using a new algorithm that was suggested in associating a cost function with the energy of a physi- our proposal. The algorithm is an exact renormalization cal system, we studied our method to solve spin glasses method that can be used to obtain all the ground states that encode the solution to these problems and devised of the Hamiltonian. In particular, our code demonstrated a quantum algorithm that provides provable quantum to be very efficient for the problem under consideration: speedups for these problems. the memory requirements and CPU time scale with the square of the lattice size. This represents an exponential Background and Research Objectives speedup with respect to other exact methods for the One of the most important challenges in materials sci- same problem, such as exact diagonalization (e.g., Lanc- ence is to discover states of matter with novel responses zos). Remarkably, the numerical results played a crucial and functionalities. The presence of strong competing role in the determination of analytic solutions for this interactions is the basic ingredient for achieving this Hamiltonian family. Also, our algorithm was later gener- goal. However, this characteristic also increases the alized in order to solve other classes of highly frustrated complexity of these materials and challenges our ability spin models in D>1 dimensions (see below). of predicting their properties. By running our code we were able to identify the exact Frustrated systems give rise to rich and exotic phe- ground state degeneracy as a function of the system or nomena that can lead to novel physical responses and lattice size. This information was crucial for guiding the functionalities. The variety of exotic states is due to the analytical efforts that we describe below. The results of highly degenerate low-energy spectrum that results from our combined approach (numerical plus analytical) led us the frustrated nature of the interactions. This property to the discovery of a new state of matter that consists of makes frustrated systems highly desirable for applica- 518

Page 519

an exact condensation of anyons [1]. Anyons are particles to solve the optimization problem more efficiently, Figure with fractional statistics, i.e., their statistics interpolates 2. between bosons and fermions. Although anyons explain The family of Hamiltonians considered in the project some aspects of the fractional Hall effect in two dimen- satisfies certain interesting properties. One such property sions, they are not common in physics. In particular, there is that every ground state is also a ground state of local are very few one-dimensional models that contain anyonic operators (projectors) that act trivially on it. We introduced solutions. Our method allowed us to identify the first exact the so called “spectral gap amplification problem” (GAP) solution of a very simple spin Hamiltonian that corre- for this family and showed that a quadratic amplification of sponds to a condensate of anyons (generalization to frac- the gap is possible and optimal. This allows for new quan- tional statistics of the Bose-Einstein condensate). Although tum speedups for optimization and for the preparation of a the notion of fractional statistics has been discussed in dif- large class of relevant quantum states (projected entangled ferent contexts, we are not aware of any exact solution of pairs or PEPS). Our results were accepted for a presenta- a spin model that corresponds to an anyonic condensation. tion at QIP, the most important conference in quantum Therefore, we believe that our exact anyonic solutions are information, and published in SIAM J. Comp, a high-impact the first example of anyonic statistics in quantum magnets. journal in computer science - Ref. [3]. We also computed relevant correlation functions that are necessary for characterizing these new states of matter. The main difficulty for studying interacting systems is the The results of this work were published in Phys. Rev. Lett. exponential increase of the number of states with the sys- -Ref. [1], Figure 1. tem size. Renormalization approaches avoid this problem by retaining only the state-variables that play a significant In a second step of the project, we generalized our algo- role in the low-temperature properties. We remark that rithm and developed a code for obtaining the entire set of computing ground-state properties of frustrated models ground states of two-dimensional (D=2) Hamiltonians that cannot be done via convex optimization: The cost function fulfill the hypothesis of applicability of our algorithm [2]. to minimize has a complex landscape with multiple mini- Such Hamiltonians do not need to be of the XXZ type. The ma, each associated with the corresponding ground state. tree structure of our novel renormalization algorithm is Our method goes beyond convex optimization, and allows depicted in Figures 1 and 2. for attacking combinatorial optimization problems where By developing this code, we were able to demonstrate the cost function is non-convex. that it is possible to compute relevant correlation func- Applications to real materials: We also applied our algo- tions for the characterization of the zero temperature state rithm to the computation of field-induced magnetization of matter that is described by the Hamiltonians under in a highly frustrated compound Ca3Co2O6 that consists consideration. In particular, we discovered a very unusual of a triangular lattice of ferromagnetic Co chains [4]. The quantum critical point, which corresponds to the onset inter-chain coupling is antiferromagnetic leading to an of antiferromagnetism-induced by an external magnetic enormous number of ground states. We applied our algo- field. Our results demonstrated that the novel numerical rithm to the generation of all the possible ground states method developed under this project is extremely power- and subsequent computation of the magnetization vs. ful, efficient and exact. These very unusual characteristics field curve starting from each of them. The output of this open the possibility of investigating new states of matter in calculation was used to fit the experiments on compounds highly frustrated materials. As we explain by the end of this of the same family that were done at the National High document, during the last year of the project we started Magnetic Field Laboratory of Los Alamos [5]. to consider materials problems, which are describable by Hamiltonians that fulfill the assumptions for applicability of Our renormalization method allowed us to: a) Connect our method. seemingly unrelated physical phenomena; b) Identify the general symmetry principles behind complex phase dia- Optimization: In optimization, the ground state of the grams; Unveil new states of matter; c) Obtain exact solu- Hamiltonian encodes the solution to the problem. Fast tions with complex ordering. algorithms to prepare such ground states are thus desired. Typically, the time complexity of known algorithms for this Impact on National Missions problem depends strongly on the inverse of the spectral gaps to the next excited eigenstate; the smaller the gap, We developed a novel scientific approach that created the larger the complexity. It is important then to devise fundamentally new capabilities for bridging two LANL other Hamiltonians that have the same ground states but a grand challenges: “Information, Science and Technol- much larger spectral gap. These Hamiltonians can be used ogy” (IS&T) and “Materials: Discovery Science to Strategic 519

Page 520

Applications”(MDSSA). The research that we produced is a direct response to FY11 IS&T Grand Challenge priorities: development of methods for inference and prediction of large-scale complex systems and design of algorithms to efficiently ex-tract information from massive amounts of data. Our research is also a direct response to FY11 priori- ties of the Materials Grand Challenge and its three central themes: prediction and control, extreme environments, and emergent phenomena. Frustration is a challenging problem, identified in DOE’s ‘Directing Matter and Energy: Five Challenges for Science and the Imagination’, that is ex- citing because of both, the scientific consequences of new states of matter and the technologically relevant multi- functional responses. We received follow-up funds from the Air Force (AFOSR) for a grant on quantum methods for aerospace problems, where some of the techniques developed in this ER pro- posal (e.g., spectral gap amplification) will play a crucial role. Figure 2. Representation of the network contraction for each matrix element for the same system of Fig. 1. References

1. Batista, C. D., and R. D. Somma. Condensation of Any- ons in Frustrated Quantum Magnets. 2012. PHYSICAL REVIEW LETTERS. 109 (22): -.
2. Feiguin, A., R. D. Somma, and C. D. Batista. Exact real- space renormalization method and applications. 2013. Physical Review B. 88: 075145.
3. Somma, R. D., and S. Boixo. Spectral Gap Amplification.
4. SIAM Journal on Computing. 42 (2): 593–610.
5. Kamiya, Y., and C. D. Batista. Formation of Magnetic Figure 1. Representation of the tree-tensor network contrac- Microphases in Ca3Co2O6. 2012. Physical Review Let- tion for a system with p spins (blue circles). In this example, the ters. 109: 067204. sets vl refer to pairs of nearest-neighbor spins and each wl is a single spin. The tensors Tl live in the vertices of the tree and the 5. Kim, J., Y. Kamiya, E. D. Mun, M. Jaime, N. Harrison, contraction indices are in the edges. An arrow means a sum over J. D. Thompson, T. H. Yi, Y. S. Oh, S. W. Cheong, C. D. the corresponding index. Batista, and V. S. Zapf. Multiferroicity with coexisting isotropic and anisotropic spins in Ca3Co\{2-x\}MnxO6. Physical Review Letters . Publications Altarawneh, M. M., G. W. Chern, N. Harrison, C. D. Batista, A. Uchida, M. Jaime, D. G. Rickel, S. A. Crooker, C. H. Mielke, J. B. Betts, J. F. Mitchell, and M. J. R. Hoch. Cascade of Magnetic Field Induced Spin Transitions in LaCoO3. 2012. PHYSICAL REVIEW LETTERS. 109 (3): -. Batista, C. D., and R. D. Somma. Condensation of Anyons in Frustrated Quantum Magnets. 2012. PHYSICAL REVIEW LETTERS. 109 (22): -. 520

Page 521

Delgado, F., C. D. Batista, and J. Fernandez-Rossier. Local probe of fractional edge states of S=1 Heisenberg spin chains. To appear in Physical Review Letters. Feiguin, A., R. D. Somma, and C. D. Batista. An exact real- space renormalization method and applications. 2013. Physical Review B. 88 (7): 075145. Kamiya, Y., and C. D. Batista. Formation of Magnetic Mi- crophases in Ca3Co2O6. 2012. Physical Review Letters. 109: 067204. Kamiya, Y., and C. D. Batista. Magnetic vortex crystals in frustrated Mott insulator . Physical Review X. Kim, J., E. D. Mun, Y. Kamiya, M. Jaime, N. Harrison, J. Thompson, H. T. Yi, Y. S. Oh, S. W. Cheong, C. D. Batista, and V. Zapf. Multiferroicity with coexisting isotropic and anisotropic spins in Ca3Co\{2-x\}MnxO6. Physical Review Letters. Koutroulakis, G., T. Zhou, C. D. Batista, Y. Kamiya, J. D. Thompson, S. E. Brown, and H. D. Zhou. Magnetic Phases of the Spin-1/2 Triangular-Lattice Antiferromag- net Ba3CoSb2O9 . Physical Review Letters. Lin, S., and C. D. Batista. Orbital magnetism induced by heat currents in Mott insulators. Physical Review Let- ters. Mun, E. D., G. Chern, V. Pardo, F. Rivadulla, V. Zapf, and C. D. Batista. Magnetic Field Induced Transition in the Multiferroic Spinel CdV2O4. Physical Review Letters. Muniz, R., Y. Kato, and C. D. Batista. Generalized spin-wave theory: application to the bilinear-biquadratic model. Physical Review B. Nishida, Y., Y. Kato, and C. D. Batista. Efimov effect in quan- tum magnets. 2013. NATURE PHYSICS. 9 (2): 93. Somma, R. D., and S. Boixo. Spectral Gap Amplification. 2013. SIAM Journal on Computing. 42 (2): 593. Wierschem, K., Y. Kato, Y. Nishida, C. D. Batista, and P. Sen- gupta. Magnetic and nematic orderings in spin-1 anti- ferromagnets with single-ion anisotropy. 2012. PHYSI- CAL REVIEW B. 86 (20): -. Zhang, Z. F., K. Wierschem, I. Yap, Y. Kato, C. D. Batista, and P. Sengupta. Phase diagram and magnetic excitations of anisotropic spin-one magnets. 2013. PHYSICAL RE- VIEW B. 87 (17): -. 521

Page 522

Information Science and Technology Exploratory Research Final Report Computational Modeling of Topo-taxis: Directing the Motion of Bacteria and Cells with Microfabricated Topologies Charles Reichhardt 20110189ER Abstract types of non-equilibrium states. Are these transitions This project focused on computational modeling of like a transition from a solid to liquid? Finley can we nonequilibrium phenomena for particles undergoing impose some kind of boundary conditions or a substrate motion on patterned substrates. Particular emphasis to control the transition from the different states or con- is on particles undergoing motion for different types of trol the motion of the microscopic degrees of freedom in dynamics such as externally driven or self driven as well these systems. as understanding the effects of a substrate or a combina- Although the non equilibrium system can include very tion of a substrate and bulk driving on directed motion, complex systems even understanding some of the sim- aggregation, and functionality. These systems include plistic and looking for general principles could provide externally driven particles such colloids, vortex and sky- powerful tools for utilizing properties of non-equlibrum rmions as well as self-particles such as bacteria and ar- systems for applications. Some of the simplistic non-equ- tificial swimmers. We showed that these substrates can librium systems with relatively few parameters are dis- lead to directed motion is the a funnel like geometry is crete particle bases systems. Examples include colloids used as well as varies types of novel dynamics. We find or solid particles in solutions, electron crystals, vortices that a strong substrate can arrest aggregation; however, in superconductors, magnetic nano particles and so on. under an applied drive the system with a substrate can In these system an external force can be applied by the aggregate much faster than systems without a substrate. addition of electric field, magnetic field, or by having This opens a new route for controlling pattern formation fluid flow through the systems. Another class of particle in aggregating materials. Additionally, we showed that based non-equlibrium systems are self-driven particles aggregation can lead to enhanced diffusion or mobility or active matter. An example would be that of swim- in cases where there is an asymmetry in the substrate, ming bacteria, where the particles can be considered a which can be useful for particle separation. Our work self-driven elongated particles that swim according to also showed that skyrmions are very promising for pos- some type of run and tumble dynamics and interact with sible applications of novel memory devices. other bacteria or obstacles. Other examples could be for pedestrian flow, vehicular flow, and crawling cells. More Background and Research Objectives recently there have been several advances in artificial Equilibrium systems are systems that are not driven and swimmers such as self-driven colloids. These system examples include solids, liquid and gases. As external an exhibit a wealth of nonequlibrium type behaviors parameters such as presser, temperature or volume including pattern formation, chaotic type flows, and changes these system can exhibit transition such as from time dependent behaviors. Far less is known about what a liquid to a solid or a gas to a liquid. For solids there can happens where these types of nonequilibrium systems also be transitions from one type of crystalline solid to are placed on some kind of topographic substrates. Such another. Over the past 100 years withe developments in a substrates cane be create id my a micro fabricated sur- Statistical thermodynamics we now have considerable faces, nanostructures and by optical means. The goal of understanding of transitions in these systems and how the project was to understand how substrates with vari- to characterize transitions in them. Another class of ous typologies can change non-equilibrium behaviors, be systems is that of not in equlibrium or driven systems. A useful for apparitions of these systems, and understand simple example of this is a continuously stirred liquids, if universal principles apply to broad class of externally which show a turbulent like motion. The open questions and self driven particle systems. in this class of systems is there transitions from different 522

Page 523

Scientific Approach and Accomplishments substrate is effectively wetted by a portion of the particles Our approach has been to use a combination of large while the remaining particles form a single clump. (3) For scale simulations and theory to model different aspects strong substrates, the substrate dominates the behavior of particles externally driven and self particle assemblies. and there is little or no aggregation. We next address the We first examined the effects of self-driven aggregating or question of what happens under an applied drive. We flocking particles, and found that if the substrate is asym- found that even systems with strong substrates can exhibit metric, a novel ratchet effect occurs in which the aggregate dynamically induced aggregation when driven, and that exhibits a net dc motion even when there is no applied dc the shape of the aggregation can be controlled and can drive. We also showed that the size of the particles in the take the form of stripe or labyrinth patterns. Applications aggregate can lead to motion in different directions, so for this system include a new dynamic method to create that a new type of particle separation can be achieved. We new kind of patterns. If such methods could be used on have also found that if the self driven particles are placed the nanoscale that having the ability to control the pat- in between movable barriers that are very strong self-ag- terns would give us a way to tailor materials properties. gregation effect can occur in what we call an active matter Additionally there are examples of particles that can be Casimir which can be hundred of time stronger than for toxic however when they aggregate they are actually less non active matter thermal particles. We also show that the toxic so understanding to how to control aggregation on a for certain cases increasing the mobility of the individual substrate is also has applications in medical fields. self-driven particles will actually lead do a decreases mobil- Over the three year time from of this project we have pub- ity of the interacting particles of as a whole do to what is lished over 25 papers. This includes publications in high called an active matter jamming effect. This is particularity impact journals such as Nature Materials, PNAS, Physical to understand when trying to use active matter systems Review Letters, Applied Physics Letters, Physical Review B, to increase diffusion of mixing in certain systems where Physical Review E, Soft Matter, and New Journal of phys- collective effects may work against certain desired effects. ics. Additionally we have 5 more publications currently We have also been examining the kinds of patterns that under review or to be submitted shortly. Additionally this can occur for externally driven particles on random and work resulted in over 30 invited colloquial, and seminars at periodic substrates. Recently a new kind of magnetic de- international conferences, universities and other national fect called a skyrmion was found in certain chiral magnets laboratories. The work partially supported two postdocs. however any piratical use of such objectives means we One of them Jeff Drocco has now started as a permanent have to understand how they move and how to control the staff member at LLNL in the bio-threats division. The other motion of such objects. We first developed a microscopic is still at LANL. This work was also used to partially sup- model for the motion of a single skyrrmion, which was var- port two graduate students who will be getting there PhD’s ied in both continuum and molecular dynamics types simu- from Notre Dame in 2014. lations. We than found that by adding a random substrates that a pinning effect occurs leading to a critical value of the Impact on National Missions externally applied force being need to get the particles to Particle separation. There are many examples of processes move. We also found that if the particles are placed in a relevant to the mission where one would want to separate circular nanodiks that a novel rectifier should be possible out different components of a mixture on the nanoscale. and that they are indeed useful for applications. We have Our work suggests that the addition of active swimmers or also been examine driving colloidal particles. active aggregating particles can be used to create a ratchet The last problem we address is that of particles with short effect in which aggregates of one size move in one direc- range repulsion and longer range attraction. This interac- tion while aggregates of another size move in the opposite tion was modeled with two Bessel functions and compared direction, leading to enhanced sorting. This could be used to systems with only repulsion. Previous work showed for a wealth of application for the lab mission of bio threat that this model forms multiple clump and stripe phases; detection, bio fuels, new mineral extractions techniques, however, we demonstrated that these are not the true filtration, and environmental work. ground states, and that instead the system forms a single Toxicology. Certain nanoparticles can have very toxic ef- clump under long time thermal motion. The next question fects; however, the toxicity can be altered if the particles we addressed was to understand what happens to aggre- aggregate, and in certain cases the aggregates are nontoxic gation in the presence of quenched disorder. We model even though their component particles are toxic. This sug- both random and periodic substrates and identify three gested that controlled aggregation could be used to reduce regimes. (1) For weak substrates, the system completely the toxicity of many types of materials that are byproducts phase separates. (2) For intermediate substrates, the 523

Page 524

of missions relevant to the Nation. Our results suggest tems. 2012. PHYSICAL REVIEW E. 86 (2, 1): 021406. that the aggregation can be enhanced by moving nanopar- Lin, S. Z., C. Reichhardt, C. D. Batista, and A. Saxena. Par- ticles over a substrate with a strong applied drive. ticle model for skyrmions in metallic chiral magnets: Dynamics, pinning, and creep (vol 87, 214419, 2013). Novel non-volatile memory. Magnetic domain walls show 2013. PHYSICAL REVIEW B. 88 (1): -. considerable promise for applications in non volatile memory and novel logic devices, however there are sev- Lin, S. Z., C. Reichhardt, C. D. Batista, and A. Saxena. Par- eral problems with these system in that domain walls are ticle model for skyrmions in metallic chiral magnets: generally strongly pinned and hard to move, which means Dynamics, pinning, and creep. 2013. PHYSICAL REVIEW very high currents are needed. Recently a new type mag- B. 87 (21): -. netic object called a skryrmion was found which is weakly Lin, S. Z., C. Reichhardt, C. D. Batista, and A. Saxena. Driven pinned. We have shown how to create and destroy skry- Skyrmions and Dynamical Transitions in Chiral Mag- moins in these systems and how to control there motion nets. 2013. PHYSICAL REVIEW LETTERS. 110 (20): -. with external fields and substrates. Lin, S. Z., C. Reichhardt, and A. Saxena. Manipulation of Publications skyrmions in nanodisks with a current pulse and sky- Chern, G. W., C. Reichhardt, and C. J. O. Reichhardt. Frus- rmion rectifier. 2013. APPLIED PHYSICS LETTERS. 102 trated colloidal ordering and fully packed loops in ar- (22): -. rays of optical traps. 2013. PHYSICAL REVIEW E. 87 (6): -. Lin, S. Z., and C. Reichhardt. Stabilizing fractional vortices in multiband superconductors with periodic pinning ar- Drocco, J. A., C. J. O. Reichhardt, C. Reichhardt, and A. R. rays. 2013. PHYSICAL REVIEW B. 87 (10): -. Bishop. Static and dynamic phases for magnetic vortex matter with attractive and repulsive interactions. 2013. McDermott, D., J. Amelang, L. M. Lopatina, C. J. O. Reich- JOURNAL OF PHYSICS-CONDENSED MATTER. 25 (34): -. hardt, and C. Reichhardt. Domain and stripe formation between hexagonal and square ordered fillings of col- Drocco, J. A., C. J. Reichhardt, and C. Reichhardt. Character- loidal particles on periodic pinning substrates. 2013. izing plastic depinning dynamics with the fluctuation SOFT MATTER. 9 (18): 4607. theorem. 2011. European Physical Journal E. 34 (10): 117 (7 pp.). Ray, D., C. J. O. Reichhardt, B. Janko, and C. Reichhardt. Strongly Enhanced Pinning of Magnetic Vortices in Drocco, J., C. J. Olson Reichhardt, and C. Reichhardt. Bidi- Type-II Superconductors by Conformal Crystal Arrays. rectional sorting of flocking particles in the presence of 2013. PHYSICAL REVIEW LETTERS. 110 (26): -. asymmetric barriers. 2012. Physical Review E - Statisti- cal, Nonlinear, and Soft Matter Physics. 85 (5): 056102. Regev, I., and C. Reichhardt. Rheology and shear band sup- pression in particle and chain mixtures. 2013. PHYSI- Drocco, J., C. Reichhardt, C. J. O. Reichhardt, and A. R. CAL REVIEW E. 87 (2): -. Bishop. Statics and dynamics of wetting-dewetting transitions for particles with attractive interactions on Reichhardt, C. J. O., E. Groopman, Z. Nussinov, and C. periodic substrates. 2012. OPTICAL TRAPPING AND OP- Reichhardt. Jamming in systems with quenched disor- TICAL MICROMANIPULATION IX. 8458: -. der. 2012. PHYSICAL REVIEW E. 86 (6): -. Drocco, J., C. Reichhardt, and C. Reichhardt. Bidirectional Reichhardt, C. J. Olson, J. Drocco, T. Mai, M. B. Wan, and C. Sorting of Self-Propelled Microswimmers in the Pres- Reichhardt. Active matter on asymmetric substrates. ence of Asymmetric Barriers. 2013. BIOPHYSICAL 2011. In Optical Trapping and Optical Micromanipula- JOURNAL. 104 (2): 496A. tion VIII ; 21-25 Aug. ( 2011 ; San Diego, CA, USA). Vol. 8097, p. 80970A (13 pp.). Drocco, J., L. M. Lopatina, C. Reichhardt, and C. J. O. Reich- hardt. Dynamics of Self-Driven and Flocking Particles Reichhardt, C. J. Olson, and C. Reichhardt. Vortex dynamics on Periodic Arrays. 2012. OPTICAL TRAPPING AND OP- and symmetry locking on quasiperiodic and periodic TICAL MICROMANIPULATION IX. 8458: -. substrates. 2012. Physica C: Superconductivity and its Applications. 479: 45. Libal, A., B. M. Csiki, C. J. O. Reichhardt, and C. Reichhardt. Colloidal lattice shearing and rupturing with a driven Reichhardt, C. J., Liba&acute, and C. Reichhardt. Multi-step line of particles. 2013. PHYSICAL REVIEW E. 87 (2): -. ordering in kagome and square artificial spin ice. 2012. New Journal of Physics. 14 (2): 025006 (15 pp.). Libal, A., C. Reichhardt, and C. J. Reichhardt. Hysteresis and return-point memory in colloidal artificial spin ice sys- Reichhardt, C. J., and C. Reichhardt. Vortex dynamics and 524

Page 525

symmetry locking on quasiperiodic and periodic sub- Zhou, C., C. Reichhardt, M. Graf, J. Su, A. Balatsky, and I. strates. 2012. PHYSICA C-SUPERCONDUCTIVITY AND Beyerlein. Comment on “Giant Plasticity of a Quan- ITS APPLICATIONS. 479: 45. tum Crystal”. 2013. Physical Review Letters . (111): 119601. Reichhardt, C., J. Drocco, C. J. Reichhardt, and A. R. Bishop. The effect of pinning on vortex states with attractive and repulsive interactions. 2012. PHYSICA C-SUPER- CONDUCTIVITY AND ITS APPLICATIONS. 479: 15. Reichhardt, C., Z. Nussinov, and C. J. O. Reichhardt. Con- necting Jamming and Depinning Transitions. 2013. SOLID STATE PHYSICS, VOL 57. 1512: 7. Reichhardt, C., and C. J. Olson Reichhardt. Statics and dynamics of Yukawa cluster crystals on ordered sub- strates. 2012. Physical Review E (Statistical, Nonlinear, and Soft Matter Physics). 85 (5): 051401 (13 pp.). Reichhardt, C., and C. J. Reichhardt. Dynamics and Direc- tional Locking of Colloids on Quasicrystalline Sub- strates. 2011. In Optical Trapping and Optical Micro- manipulation VIII ; 21-25 Aug. ( 2011 ; San Diego, CA, USA). Vol. 8097, p. 80970V (11 pp.). Reichhardt, C., and C. J. Reichhardt. Structural transitions and dynamical regimes for directional locking of vorti- ces and colloids driven over periodic substrates. 2012. Journal of Physics: Condensed Matter. 24 (22): 225702 (8 pp.). Reichhardt, C., and C. J. Reichhardt. Dynamically induced locking and unlocking transitions in driven layered sys- tems with quenched disorder. 2011. Physical Review B (Condensed Matter and Materials Physics). 84 (17): 174208 (16 pp.). Reichhardt, C., and C. O. Reichhardt. Dynamical freezing of active matter. 2011. PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA. 108 (48): 19099. Reichhardt, C., and C. O. Reichhardt. Dynamical freezing of active matter. 2011. PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA. 108 (48): 19099. Reichhardt, C., and C. O. Reichhardt. DEFECT DYNAMICS Breaking up in a curved plane. 2012. NATURE MATERI- ALS. 11 (11): 912. Zhou, C. Z., J. J. Su, M. J. Graf, C. Reichhardt, A. V. Balatsky, and I. J. Beyerlein. Plastic response of dislocation glide in solid helium under dc strain-rate loading. 2013. PHYSICAL REVIEW B. 88 (2): -. Zhou, C. Z., J. J. Su, M. J. Graf, C. Reichhardt, A. V. Balatsky, and I. J. Beyerlein. Dislocation-induced anomalous soft- ening of solid helium. 2012. PHILOSOPHICAL MAGA- ZINE LETTERS. 92 (11): 608. 525

Page 526

Information Science and Technology Exploratory Research Final Report Algorithmic Co-design: Paradigms for Unstructured Problems on Accelerated Architectures Sunil Thulasidasan 20110195ER Abstract ciated matrices and showed how such problems can be This goal of this research effort was to develop algorith- elegantly formulated for GPU architectures [2]. Further mic techniques to efficiently utilize the massive com- graph-based research included analysis and implementa- puting power afforded by today’s parallel processors tion of fast parallel algorithms for protein matching prob- for solving problems that lack obvious data parallelism. lems in computational biology [6], efficient algorithms Such problems are often characterized by irregular mem- for modularity-based graph clustering [7] and analysis of ory accesses and have traditionally been better suited dynamics of influence on large-scale social networks [8] for CPU architectures that are optimized for sequential On the software front, we developed and open sourced and branch-heavy code. a library for easily prototyping simulations of large-scale The irregularity of these problems is a natural outcome complex networks on distributed memory parallel ar- of graph-based representations; graphs are mathemati- chitectures. The library, named, SimX [3,15] is a general- cal objects that capture relationships between entities purpose simulation library for rapidly developing parallel and many important scientific problems can be elegantly discrete-event simulation applications. formulated in terms of graph-exploration and optimiza- Our publication portfolio includes conferences and tion. However, efficient graph representations, which journals in the areas of algorithms, high performance are pointer-dense and unstructured, do not translate to computing, simulation tools and methodologies and data efficient computations on parallel processors and many analytics. As part of this project, we also established optimal graph algorithms do not lend themselves to easy successful collaborations with other researchers in the parallelization. Nevertheless, it is becoming increasingly above fields, both domestically (MIT) and internationally clear that massive on-chip parallelism will continue to (France, India and Turkey). be the architectural trend in the foreseeable future. The potential performance benefits in extracting parallelism Background and Research Objectives on unstructured problems are thus too compelling to While GPUs and other stream processors were originally ignore. developed for numerically intensive media applications, The contributions in this project have had a common advances in performances and programmability have theme: namely providing algorithmic techniques and resulted in these processors being successfully deployed software implementations to exploit the massive par- for numerous computationally demanding scientific allelism for graph based problems using both shared tasks. The impressive arithmetic power and massive and distributed memory designs. Specifically, for graph parallelism of today’s graphics processors is an outcome based problems, we attacked the classical shortest- of the highly parallel and numerically intensive nature of path problem for planar graphs [3,5], and showed how computer graphics. Not surprisingly, the general pur- reformulating the graph problem in terms of matrices pose computing tasks that have been able to leverage and exploiting the resulting structured nature of matrix the performance benefits of these processors exhibit computations resulted in greatly improved speed-ups many computational similarities to graphics applications, for path problems on certain kinds of real world graphs, namely 1) high data parallelism, 2) high compute inten- such as those found in road networks. In addition, we sity, and 3) data locality which results in structured (or also attacked the coverage problems in sensor network coalesced) memory accesses. graphs by exploiting the spectral properties of the asso- 526

Page 527

However, many problems do not readily exhibit the above network that is based on a reduction of the modularity op- properties. An important example of this is problems deal- timization problem to the minimum weighted cut problem, ing with graph-based data structures, which are found in a We give an experimental evaluation of an implementation variety of scientific domains. Graphs, which are essentially based on that approach on graphs from the 10th DIMACS a representation of relationships between objects, find Implementation Challenge Test bed [9]. applications in such diverse domains as molecular biol- Another graph-based problem we tackled was the cov- ogy epidemiology, transportation, social network analysis erage problem in sensor networks where we studied a and data mining, to name a few. Efficient graph represen- distributed algorithm for coverage in mobile sensor when tations, (which store as little redundant information as pos- very little information is available regarding terrain topol- sible) result in highly unstructured data; this makes them ogy. We use virtual forces based on potential fields to less than ideal candidates for streaming processor manipu- achieve coverage and obstacle avoidance. Additionally, lations. Nevertheless, the massive parallelism afforded by a distributed dynamic relocation algorithm – triggered today’s graphics processors is too compelling to ignore. by event occurrences – is also simulated which achieves Current GPUs support hundreds of cores per chip and even convergence to a formation that surrounds the event of future CPUs will be many cores. To harness this computing interest, in the absence of prior information regarding power for computing power for unstructured problems, we formation geometry. Control laws are formulated based need a shift in algorithmic thinking since approaches that on potential fields; convergence to equilibrium, graph are theoretically non-optimal, but afford a greater degree connectivity properties and sensitivity to parameters are of parallelism might actually be better suited to GPU-type simulated and analyzed using the spectral properties of the architectures. matrices associated with the sensor network graphs. [2] As This has been confirmed by our research thus far into effi- part of this project, we also looked at graphs induced by cient computations for shortest-paths problems for certain large-scale real-world social networks. Specifically, we look kinds of graphs. By reformulating the graph problem in at several graph-theory based heuristics for understanding terms of matrix operations, and using matrix theoretic the dynamics of influence and celebrity notion on Twitter properties, we have been able to exploit the structured [8]. nature of matrix computations that are much more ame- On the software front, we were able to open source a nable to fine grained parallelization of GPUs. One of our library for easily prototyping simulations on distributed research tasks focused on shortest-path (SP) problems in memory parallel architectures. The library, named, SimX planar graphs; the SP problem is a fundamental one in [15] is a general-purpose simulation library for rapidly computer science, while planar graphs (graphs that can be developing parallel discrete-event simulation applications. embedded in a plane such that no two edges cross each SimX provides the simulation developer with core func- other) are found in diverse areas such as transportation, tionality such as event queuing, entity management, time robotics, and chip design. The problem is to find paths of advancement, domain partitioning, synchronization and minimum cost between pairs of nodes in edge weighted- message passing. Parallel simulations are implemented graphs, where the cost of a path p is defined as the sum of using distributed memory and message passing. SimX has the cost of all edges that makeup p. The distance between been written to be flexible enough to suit a variety of ap- two nodes v and w is defined as the minimum cost of a plication domains that use discrete-event simulations for path between v and w. modeling, and has been designed to allow applications to Finding similarities between protein structures is a crucial scale from single processor workstations to high-perfor- task in molecular biology. Many tools exist for finding an mance clusters. optimal alignment between two proteins. These tools, however, only find one alignment, even when multiple Scientific Approach and Accomplishments similar regions exist. We propose a new parallel heuristic- Path problems on graphs: The algorithm we designed for based approach to structural similarity detection between the All-Pair Shortest Path problem (APSP) for planar graphs proteins that discovers multiple pairs of similar regions is based on a modification to the FW (FW) algorithm [10], using graph theoretic techniques. We prove that returned exploiting the property of planarity. The complexity of our alignments have root-mean-square deviation lower than a algorithm with respect to the number of nodes is close to given threshold. Computational complexity is addressed by quadratic (O(n2)), while its structure is regular enough to taking advantage of both fine and coarse-grain parallelism. allow for an efficient parallel implementation, enabling us to exploit GPUs. Our algorithm decomposes the graph In our work on modularity based graph clustering, we into p parts, solves the APSP problem for the sub-graph propose an approach for finding the communities of a 527

Page 528

induced by each part (in parallel, if the graph is too large graphs (32-k nodes). For the largest instance, the cost to fit into memory and more than one GPU is available), adjacency matrix requires about 4.2 GB of RAM. Instances and then uses that information to compute the distances larger than about 38,000 vertices would require more RAM between pairs. The details of the algorithm are given in [3]. than currently available on the GPU. One way around the memory size problem is to exploit the spatially constrained The above algorithm was implemented on a GPU using nature of paths in real-world graphs that will allow us to CUDA, nVIDIA’s parallel programming framework for the consider only a relatively small subset of the original graph GPU. Modern GPUs are efficient at manipulating struc- [1]. We are also currently extending this work to tackle tured data like matrices, and their highly parallel architec- larger graph instances (up to one million nodes) by using ture (a GPU trades the complicated cache and control logic LANL’s new multi-GPU hybrid super computers. in a CPU for a large number–often hundreds–of arithmetic units) makes them ideally suited for processing large blocks Protein Similarity: Finding similarities between protein of data simultaneously (referred to as the SIMD–Single structures are a crucial task in molecular biology. Many Instruction Multiple Data–paradigm). tools exist for finding an optimal alignment between two proteins. These tools, however, only find one alignment, In order to comply with the GPU paradigm, we define a even when multiple similar regions exist. We propose a computational kernel that implements Floyd-Warshall’s new parallel heuristic-based approach [6] to structural sim- algorithm and computes the APSP over a sub-matrix of the ilarity detection between proteins that discovers multiple initial matrix. We then define computation grids, where pairs of similar regions. We prove that returned alignments blocks correspond to sub-matrices that can be computed have root-mean-square deviations (RMSD) lower than a simultaneously. Phase 1 of the algorithm consists of given threshold. Computational complexity is addressed by computing APSP in self-dependent (in terms of data de- taking advantage of both fine and coarse-grain parallelism. pendencies) diagonal sub-matrices, using a block-parallel version of the Floyd-Warshall algorithm. Computations Our approach reduces the optimal alignment problem to for these diagonal sub-matrices can be run in parallel in the problem of finding a maximum clique in a graph called different GPU blocks. Phase 2 of the algorithm consists of an alignment graph (Figure 3), where each edge corre- computing APSP on a sub-graph of the initial graph solely sponds to a matching of similar internal distances (up to comprised of boundary vertices. For this purpose, we a user-defined threshold T). Since the latter problem is implemented a GPU version of the blocked APSP algorithm an NP-hard problem, we relax this condition and accept described in [12], a variant of the Floyd-Warshall algorithm cliques such that edges correspond to matching of similar where computations are divided into groups that can be internal distances up to 2T. For this relaxed problem we easily be mapped to GPU blocks. Phase 3 of the algorithm propose a polynomial algorithm and its efficient parallel consists of computing APSP in the remaining non-diagonal implementation comparing two protein structures that sub-matrices, using the same parallel Floyd-Warshall algo- guarantees to return alignments with RMSD less than a rithm. Since Phases 1 and 3 use the same data patterns we given threshold value, if such alignments exist. illustrate their data dependencies in Figure 1. In order to test our parallel implementation, we compare In order to test the efficiency of the algorithm, we compare run times obtained with various numbers of threads on a two implementations of the partitioned APSP algorithm: single large instance. The input alignment graph for this (1) a single core CPU implementation of the partitioned all- instance contains 15024 vertices and about 9.5 million pairs shortest-path algorithm (referred to as CPU version), edges. Computations were run using a varying number of and (2) a single GPU implementation of the partitioned cores of an Intel(R) Xeon (R) CPU E5645 @ 2.40GHz. Figure all-pairs shortest-path algorithm (later referred to as GPU 4 shows run times and speedups with respect to the num- version). ber of CPU cores. The gain in terms of speedup becomes less significant beyond 12 cores. Note that similar results - The CPU version runs on an Intel(R) Xeon(R) CPU X5675 at both in terms of length and RMSD scores - can be obtained 3.07GHz. The GPU version runs on an nVidia Tesla m2090 in less than 30 seconds with a sparser alignment graph. consisting of 512 cores at 1.3 GHz. The benchmark consists of random cost adjacency matrices, representing planar Graph Modularity: For our graph modularity work [7], graphs with sizes ranging from 1024 vertices to 32,768 ver- we designed an approach for finding the communities of tices. These graphs were generated using the LEDA graph a network that is based on a reduction of the modular- generator [11]. Figure 2 shows the GPU versus CPU speed- ity optimization problem to the minimum weighted cut up comparison for progressively larger graphs, with the problem, and we give an experimental evaluation of an GPU being up to 14 times faster than the GPU for larger implementation based on that approach on graphs from 528

Page 529

the 10th DIMACS Implementation Challenge Test bed [9]. with LANL and DOE missions. For example, the distributed Specifically, we describe a reduction from the problem of memory GPGPU computing algorithms for path problems finding a partition of the nodes of a graph G that maximiz- in graphs is relatively new territory and represents a signifi- es the modularity to the problem of finding a partition that cant opportunity for LANL to establish itself in this area. minimizes the weight of the cut in a complete graph on the Other specific examples of impact on national missions same node set as G, and weights dependent on a random include: (a) Advances in parallel graph clustering and com- graph model associated with G. Modifying existing codes munity detection algorithms will find numerous uses for for graph partitioning can then solve the resulting mini- analyzing signaling pathways and metabolic channels in mum cut problem. biological networks (Grand Challenge: Complex Biological We have tested our algorithm against Ovelgonne and Systems), and network analysis in epidemiology (Bio-secu- Geyer-Schulz’s algorithm [14], which was ranked at the rity Science) (b) Community detection algorithms can also top in the DIMACS Challenge with respect to its accuracy. be readily applied to detecting and characterizing terror- Figure 5 compares the performance of our algorithm with ist networks and channels (Grand Challenge: Sensing and that algorithm. For each experiment, the table shows the Measurement for Global Security) (c) Graph algorithms average running time and modularity of the partition for have also found numerous uses in LANLs efforts in infra- each of the algorithms. The results show modularity of structure analysis (NISAC). The results of our effort can be the clusters that our algorithm finds is 7% less on average, used to vastly improve the scale and speed of current mod- but our algorithm is 48 times faster on average. For one els and simulations. (d) LANL’s efforts in data-intensive and instance (kron-g500-logn16), our algorithm is 390 times exascale computing (Grand Challenge: IS&T) can directly faster. leverage our work, since one of the major computational tasks in this domain will be inferring semantic relationships SimX Parallel Simulation Library: The SimX library for inside huge data sets that will necessarily involve massive developing parallel, discrete-event simulations in Python amounts of unstructured computation. [3,15] of large-scale complex systems are written in C++ and Python. SimX APIs are exposed to Python, enabling rapid development and prototyping of a parallel simulation entirely in Python. SimX provides the simulation modeler with the core functionality needed in a parallel simulation library such as event queuing, process management, time advancement, domain partitioning, synchronization and message passing. Arbitrary Python objects can be sent and received between different simulation processes without having to write explicit serialization code. Further, the fast C++ core and distributed memory design of SimX makes the parallelization transparent to the user and allows appli- cations to scale from multi-core user workstations to high Figure 1. Block-level data parallelism and data dependencies in performance clusters. the adjacency matrix for phase 1 and phase 3 of the algorithm. SimX is free software and has been open sourced under Sub-matrices for which computations are required are shown in red. Arrows indicate data dependencies. the GNU LGPL license. It is publicly available at http:// github.com/sim-x. It has already been used in developing an agent-based simulation model of a monetary reserve system validated against real world data [13]. More SimX based applications are currently being developed at LANL and elsewhere. Impact on National Missions LANL is uniquely poised to make important contributions to the emerging field of heterogeneous computing. Our Figure 2. Run-times (left) with respect to input matrix size on aim in this proposal was to take a significant step in that di- the CPU and GPU versions. Figure on right shows first four data rection by exploring new algorithmic methods to enable us points zoomed in. to unleash the potential of heterogeneous and accelerated computing on a lager variety of problems that are aligned 529

Page 530

Planar Graphs Using Graphics Processors. 2012. LANL Technical Report LA-UR-12-25700. 4. Thulasidasan, S., L. Kroc, and S. Eidenbenz. Developing Parallel Discrete Event Simulations in Python: First Re- sults and User Experiences with the SimX Library. 2012. LANL Technical Report, LA-UR-12-26739. Submitted to 6th International ICST Conference on Simulation Tools and Techniques (SimuTools 2013). 5. Djidjev, H., S. Thulasidasan, G. Chapuis, and R. An- Figure 3. Example of an alignment graph used here to compare donov. Accelerating Graph Algorithms Using Graphics the structures of two proteins. The presence of an edge between Processors: Shortest Paths for Planar Graphs. 2013. vertex (1, 1) and vertex (3, 2) means that the distance between ADTSC Science Highlights 2013: Information Science atoms 1 and 2 of protein 1 is similar to the distance between and Technology Section. atoms 1 and 3 of protein 2. 6. Chapuis, G., M. Le Boudic-Jamin, R. Andonov, H. Dji- djev, and D. Lavenier. Parallel Seed-based Approach to Protein Structure Similarity Detection. (Warsaw, Poland, 8-11 Sep 2013). Figure 4. Dependence of the run times and speedups on the num- 7. Djidjev, H., and M. Onus. Using Graph Partitioning ber of CPU cores. for Efficient Network Modularity Optimization. 2013. Graph Partitioning and Graph Clustering: Contempo- rary Mathematics. : 103. 8. Srinivasan, M. S., S. Srinivasa, and S. Thulasidasan. Exploring Celebrity Dynamics on Twitter. (New Delhi, India, 17-19 Oct 2013). 9. Tenth Dimacs Implementation Challenge. 2013. http:// www.cc.gatech.edu/dimacs10/. 10. Cormen, T. H.. Introduction to Algorithms. 1990. 11. Melhorn, K. etal.. LEDA: A Platform for Combinatorial and Geometric Computing. 1999. 12. Venkatraman, G. etal.. A Blocked All-Pair Shortest- Paths Algorithm. 2003. Journal of Experimental Algo- Figure 5. Comparison of our algorithm with Ovelgonne and rithmics. : 2.2. Geyer-Schulz’s algorithm. 13. Williams, S., and S. Eidenbenz. Themis-1: An Agent References Based Model of a Modern Monetary Reserve System. Presented at Agent-Directed Simulation Symposium.

1. Thulasidasan, S.. Heuristic Acceleration of Routing in (San Diego, 7-10 April 2013). Transportation simulations using GPUs. 2011. (Barcelo- na, Spain, 21-5 March 2011). p. 90. Brussels, Belgium:
2. Ovelgonne, M., and A. Geyer-Schulz. An Ensemble ICST . Learning Strategy for Graph Clustering. To appear in Contemporary Mathematics 2010.
3. Thulasidasan, S.. A Simulation Based Analysis of Dis- tributed Coverage and Dynamic Relocation for Mobile
4. Thulasidasan, S.. simx 0.1. 2013. https://pypi.python. Sensor Networks Using Potential Fields. 2012. Los org/pypi/simx/0.1. Alamos Technical Report LA-UR-12-25669. Publications
5. Djidjev, H., G. Chapuis, S. Thulasidasan, and R. An- Chapuis, G., M. Le Boudic-Jamin, R. Andonov, H. Djidjev, donov. On Solving the Shortest Path Problems for 530

Page 531

and D. Lavenier. Parallel Seed-based Approach to Pro- tein Structure Similarity Detection. Presented at Tenth International Conference on Parallel Processing and Applied Mathematics (PPAM 2013). (Warsaw, Poland, 8-11 Sep 2013). Djidjev, H., G. Chapuis, S. Thulasidasan, and R. Andonov. On Solving the Shortest Path Problems for Planar Graphs Using Graphics Processors. 2012. LANL Techni- cal Report LA-UR-12-25700. Djidjev, H., S. Thulasidasan, G. Chapuis, and R. Andonov. Accelerating Graph Algorithms Using Graphics Proces- sors: Shortest Paths for Planar Graphs. 2013. ADTSC Science Highlights 2013: Information Science and Tech- nology Section. Djidjev, H., and C. Sommer. Approximate distance queries for weighted polyhedral surfaces. 2011. In Algorithms

• ESA 2011. (19th Annual European Symposium ; 5-9 Sept. 2011 ; ). , p. 579. Djidjev, H., and M. Onus. Using Graph Partitioning for Ef- ficient Network Modularity Optimization. 2013. Graph Partitioning and Graph Clustering: Contemporary Mathematics. : 103. Srinivasan, M. S., S. Srinivasa, and S. Thulasidasan. Explor- ing Celebrity Dynamics on Twitter. Presented at 5th IBM Collaborative Academia Research Exchange Con- ference (I-CARE 2013). (New Delhi, India, 17-19 Oct 2013). Thulasidasan, S.. Heuristic Acceleration of Routing in Trans- portation simulations using GPUs. 2011. In 4th Interna- tional Conference on Simulation Tools and Techniques (SimuTools 2011). (Barcelona, Spain, 21-5 March 2011). , p. 90. Brussels, Belgium: ICST . Thulasidasan, S.. The Graph Laplacian and the Dynamics of Complex Networks. 2012. Talk given at University of Washington, Department of Applied Mathematics [LA- UR-12-22065]. Thulasidasan, S.. A Simulation Based Analysis of Distrib- uted Coverage and Dynamic Relocation for Mobile Sen- sor Networks Using Potential Fields. 2012. Los Alamos Technical Report LA-UR-12-25669. Thulasidasan, S., L. Kroc, and S. Eidenbenz. Developing Par- allel Discrete Event Simulations in Python: First Results and User Experiences with the SimX Library. 2012. LANL Technical Report, LA-UR-12-26739. Submitted to 6th International ICST Conference on Simulation Tools and Techniques (SimuTools 2013). 531

Page 532

Information Science and Technology Exploratory Research Final Report Extreme Quantum Simulation: Co-design from Desktop to Exascale Nicolas Bock 20110230ER Abstract of these algorithms to take full advantage of the next We present a novel approach to linear scaling matrix- generation of high performance computing platforms: matrix multiplication, the Sparse Approximate Matrix A thousandfold increase in computational power trans- Multiply (SpAMM), a fast algorithm for matrices with de- lates into a thousandfold increase in system size while cay that achieves a linear computational complexity with only into a tenfold increase in system size using conven- respect to matrix size. We find that the error achieved tional methods. with SpAMM is lower than that of a conventional dense One of the first linear scaling approaches was Yang’s method (SGEMM) for small tolerances. In serial, SpAMM “divide and conquer” (D&C) [1], an ad hoc scheme for outperforms SGEMM with a cross-over already for small patching together density matrices corresponding to matrices and in parallel, we find near ideal scaling and fictitious quantum sub-domains. The appeal of D&C is load balancing on both shared and distributed memory that it is perfectly parallel, as troublesome off-diagonal systems, including modern accelerator architectures information is ignored. However, it was found early that such as Intel’s Xeon Phi. ignoring this off-diagonal information leads to uncon- trolled errors and D&C was abandoned. More recently Background and Research Objectives Wang et al. rediscovered D&C, however, this version of Continued trends in computer architectures are trans- D&C is likewise unable to converge the total energy with forming quantum simulations into an increasingly increasing fragment size [2]. powerful, rigorous, predictive, and interpretive tool for chemistry, biology, and materials science. However, algo- Over the past 12 years a rigorous alternative to linear rithms have not kept pace with the increasingly complex scaling has been pioneered at LANL, demonstrating the hardware. On the one hand, conventional algorithms ability to control the error in ground state, and static for electronic structure are able to take full advantage and dynamic response calculations [3-13]. Our early of large parallel platforms through standard software work to parallelize SpMM focused on spatial locality of libraries, but are throttled by a computational complex- the underlying physical system to maximize diagonal ity that scales with system size cubed, i.e. a thousandfold dominance, but did not address the spatial and tempo- increase in computational power translates only into a ral locality of work and data of the abstract convolution tenfold increase in the system size that can be computed space of the multiply. Currently, a majority of work in in a given time. On the other hand, novel linear scal- electronic structure theory employs some version of ing algorithms, which exploit quantum locality of non- SpMM, although no scalable algorithm for the general metallic quantum systems to achieve a sparse matrix problem exists. representation and linear scaling through methods such as the Sparse Matrix-Multiply (SpMM) lead to irregular Scientific Approach and Accomplishments work and data and do not lend themselves to Cartesian The dominating obstacle towards a scalable high-per- domains or standard libraries. In addition, increasingly formance SpMM is the irregular nature of data access, complex memory hierarchies and use of heterogeneous which cannot be predicted a priori. Costly latency and architectures through accelerator technologies such as bandwidth barriers between memory hierarchies can GPUs only exacerbate this shortcoming on the path to quickly render an implementation I/O bound, and make the exa-scale. Despite the challenges facing linear scaling data and load distribution challenging. A so-called algorithms, the near exponential growth of computa- cache-oblivious approach promises an elegant solution: tional power underscores the tremendous potential 532

Page 533

By ordering the abstract convolution along a space filling chemical solvers [31-33], which will lead to a quantum curve the spatial and temporal locality of data access is chemical solvers eco system based on a common frame- enhanced, which naturally leads to an algorithm that opti- work and data structures, allowing for a significant reduc- mally overlaps I/O with computation, with potentially very tion in code size and complexity of linear scaling quantum competitive performance [14, 15]. We developed a serial, chemistry programs. In addition, the recursive approach recursive version with encouraging results [16]. However, employed by SpAMM allows for overdecomposition we found that convolution ordering effects applied to providing significant opportunities for parallelism. We are sparse matrices decrease with increasing matrix sparsity currently preparing a publication detailing our parallel due to the increasingly irregular memory access and result- SpAMM implementation on shared and distributed mem- ing lack of temporal locality and the inability of the CPU’s ory systems using advanced middle ware solutions to data hardware prefetch unit to hide memory access, ultimately distribution and load balancing. We find near ideal parallel rendering this cache-oblivious approach infeasible. scaling of SpAMM on shared memory architectures, includ- ing accelerator systems such as Intel’s Xeon Phi, Figure 4, Keeping recursion, but dropping the ordering of the con- and are currently investigating scaling of SpAMM to thou- volution, we refined our approach and introduced a new sands of nodes on a distributed computer platform, Figure Sparse Approximate Matrix-Multiply (SpAMM) algorithm 5. So far, results on scalability and load balancing are very which multiplies potentially full matrices with decay and promising, strongly indicating that the generalized N-Body recursively eliminates insignificant contributions in the framework with its potential for task overdecomposition product space, leading to complexity reduction in the presents a scalable, high-performance path towards large sparse product space [17], as opposed to truncation in the scale linear scaling electronic structure calculations. vector space and applying standard sparse matrix algebra (SpMM) [4]. The recursive approach is similar to Strassen’s Impact on National Missions algorithm for fast dense matrix multiplication [18], and This research will deploy complimentary, LANL unique also to methods that improve locality across memory hier- capabilities to establish a strategic thrust with broad Mis- archies (cache-oblivious approaches), pioneered by Wise sion Relevance and a Program Development impact that is and coworkers [19, 20]. SpAMM belongs to a broad class translational, providing coherence between basic research of generalized N-Body solvers, which are “fast” algorithms and engineering specifics. This work complements the that achieve reduced complexity through hierarchical developing SciDAC program Exascale for Energy, as well approximations, which can be found in high-performance as DOE’s Basic Research Needs: Catalysis for Energy, the implementations across a broad range of disciplines. Re- BES grand challenge the ability to design and to imple- lated solvers include the database join operation [21-23], ment new catalytic processes by design. More broadly, this gravitational force summation in astrophysics simulations project develops basic capabilities with broad laboratory [24, 25], the fast Gauß transform in machine learning [26, mission relevance, including DARPA’s Quantum Effects in 27], and visualization applications [28, 29]. Biological Environments (QuBE) and LANL’s Matter-Radia- tion Interactions in Extremes (MaRIE). This project will also Despite the significant challenges due to irregular data directly broadly impact the DOE and NNSA’s push toward access inherent in sparse solvers, it is possible to write exascale computing, as well as BES and OBER missions in high performance implementations of SpAMM. Our Energy and DHS Biothreat missions. serial, single-precision implementation [30] employs a hand coded assembly kernel operating on 4 x 4 matrices and achieves around 70% performance compared to the 20 SGEMM function of a vendor tuned library such as MKL or ACML, Figure 1. In combination with a linear tree to avoid explicit recursion and potentially costly branching in the 15 code, we find an early cross-over to linear scaling for water clusters, and better performance than SGEMM between 50 10 and 90 water molecules, Figure 2. Eliminating insignificant contributions in product space achieves a smaller error compared to SGEMM for small SpAMM tolerances, which 5 we attribute to differences in execution order; hierarchical vs. row-column, Figure 3. 0 Matrix-matrix multiplication is now aligned with the N- 16 50 100 150 200 250 300 Body programming model employed by other quantum 533 ]s/polfG[ ecnamrofrep MMEGS SpAMM kernel (τ = 0) on Intel SpAMM kernel (τ = 0) on AMD MKL on AMD ACML on AMD MKL on Intel ACML on Intel Matrix size n

Page 534

Figure 1. SpAMM kernel performance compared to library SGEMM. The horizontal lines indicate the SpAMM kernel perfor- mance, operating on fixed-size 4x4 matrices. The performance of SGEMM improves with increasing matrix size and levels off for larger matrices. The SpAMM kernel performs is about 70% of the large matrix performance of SGEMM. 40 35 30 25 20 15 10 5 0 0 1000 2000 3000 4000 5000 6000 7000 8000 534 selcycagiG SpAMM, 4x4, 2.0e-8 SpAMM, 4x4, 2.0e-7 SpAMM, 16x16, 1.0e-7 SpAMM, 16x16, 1.0e-6 SpAMM, MKL, 1.0e-7 SpAMM, ACML, 1.0e-7 18 16 14 12 10 8 6 0 2000 Matrix size n s/polfG Figure 2. SpAMM serial performance on quantum chemical matrics of water clusters. The thick red line in the inset indicates SGEMM performance. SpAMM exhibits better performance than SGEMM already for small water clusters, and an early onset of linear scaling behavior. 0.0001 1e-05 1e-06 1e-07 0 1000 2000 3000 4000 5000 6000 7000 xam||�|| rorre 100 10 1 0.1 4 8 16 32 64 128 240 SGEMM, MKL SpAMM, 4x4, 0.0 SpAMM, 4x4, 2.0e-8 SpAMM, 4x4, 2.0e-7 SpAMM, 16x16, 1.0e-7 SpAMM, 16x16, 1.0e-6 Matrix size n Figure 3. Error comparison of SpAMM and dense SGEMM. The thick red line indicates the SGEMM error with matrix size. Notice that SpAMM yields an error lower than SGEMM for small toler- ances. ]s[ emitllaw SpAMM 4096x16 on Xeon Phi SpAMM ideal

threads

Figure 4. Parallel scaling of SpAMM on Intel Xeon Phi. Note that this version of the accelerator has 60 physical cores and can run up to 4 threads per core. We find continuing performance gains beyond 60 threads, however, parallel scaling slows from ideal. 1000 100 10 1 1 2 4 8 16 32 64 128 256 ]s[ emitllaw SpAMM 8192x256 on mustang SpAMM ideal

nodes (1 PE per node)

Figure 5. Parallel scaling of SpAMM on distributed memory com- puter cluster. We find near ideal parallel scaling up to 256 nodes. References

1. Yang, W.. Direct Calculation of Electron Density in Density-Functional Theory. 1991. Phys. Rev. Lett.. 66:
3. Wang, Lin-Wang, Zhengji Zhao, and Juan Meza. Linear- scaling three-dimensional fragment method for large- scale electronic structure calculations. 2008. Phys. Rev. B. 77: 165113.
4. Challacombe, M., N. Bock, Chee Kwan Gan, Graeme Henkelman and Karoly Nemeth, Anders Niklasson, Anders Odell, Eric Schwegler, C.J. Tymczak, and Valery Weber. FreeON, a suite of linear scaling quan- tum chemistry programs, http://www.freeon.org/.

Page 535

4. Challacombe, M.. A general parallel sparse-blocked M. N. Niklasson, and Matt Challacombe. Cache matrix multiply for linear scaling. 2000. Comput. Phys. oblivious storage and access heuristics for blocked ma- Commun.. 128: 93. trix-matrix multiplication. 2008. arXiv:0808.1108.
5. Németh, K., and M. Challacombe. The quasi-indepen- 17. Challacombe, Matt, and Nicolas Bock. Fast Multiplica- dent curvilinear coordinate approximation for geom- tion of Matrices with Decay. 2010. arXiv:1011.3534. etry optimization. 2004. J. Chem. Phys.. 121:
6. Strassen, Volker. Gaussian elimination is not optimal.
8. Numer. Math.. 13: 354.
9. Challacombe, M.. A simplified density matrix minimiza-
10. Frens, Jeremy D, and David S. Wise. Auto-Blocking tion for linear scaling self-consistent field Matrix-Multiplication or Tracking BLAS3 Performance theory. 1999. J. Chem. Phys.. 110: 2332. from Source Code. 1997.
11. Weber, V., A. M. N. Niklasson, and M. Challacombe.
12. Wise, David S., Jeremy D. Frens, Yuhong Gu, and Higher-order response in O(N) by perturbed projec- Gregory A. Alexander. Language support for Morton- tion. 2005. J. Chem. Phys.. 123: 044106. order matrices. 2001.
13. Niklasson, A., C. J. Tymcak, and M. Challacombe. Time-
14. Schneider, Donovan A., and David J. DeWitt. Tradeoffs reversible ab initio molecular dynamics. 2007. J. Chem. in processing complex join queries via hashing in multi- Phys.. 126: 144103. processor database machines. 1990.
15. Weber, V., and M. Challacombe. Parallel algorithm for
16. Mishra, Priti, and Margaret H. Eich. Join processing in the computation of the Hartree-Fock exchange relational databases. 1992. ACM Comput. Surv.. 24: 63. matrix: Gas phase and periodic parallel ONX. 2006. J. Chem. Phys.. 125: 104110.
17. Chen, Shimin, Anastassia Ailamaki, Phillip B. Gibbons, and Todd C. Mowry. Improving hash join perfor-
18. Challacombe, M., and E. Schwegler. Linear scaling mance through prefetching. 2007. ACM Trans. Data- computation of the Fock matrix. 1997. J. Chem. Phys.. base Syst.. : 32. 106: 5526.
19. Warren, Michael S., and John K. Salmon. A portable
20. Niklasson, A. M. N., C. J. Tymczak, and M. Challa- parallel particle program. 1995. Computer Physics combe. Trace Re-Setting Density Matrix Purification in Communications. 87: 266. O(N) Self-Consistent Field Theory. 2003. J. Chem. Phys.. 118: 8611.
21. Warren, Michael S., and John K. Salmon. A parallel hashed Oct-Tree N-body algorithm. 1993. Proceedings
22. Gan, C. K., C. J. Tymczak, and M. Challacombe. Linear of the 1993 ACM/IEEE conference on Supercomputing. scaling computation of the Fock matrix. VII. Parallel : 12. computation of the Coulomb matrix. 2004. J. Chem. Phys.. 121: 6608.
23. Gray, Alexander G., and Andrew W. Moore. N-Body Problems in Statistical Learning. 2001.
24. Schwegler, E., M. Challacombe, and M. Head-Gordon. Linear scaling computation of the Fock matrix. II. Rigor-
25. Greengard, Leslie, and John Strain. The fast Gauss ous bounds on exchange integrals and incre- transform. 1991. SIAM Journal on Scientific and Statis- mental Fock build. 1997. J. Chem. Phys.. 106: 9708. tical Computing. 12: 79.
26. Bader, M., and C. Zenger. A Cache Oblivious Algorithm
27. Choi, Myung Geol, Eunjung Ju, Jung-Woo Chang, and for Matrix Multiplication Based on Peano’s Space Fill- Jehee Lee and Young J. Kim. Linkless octree using ing Curve. 2006. Lecture Notes in Computer Science. multi-level perfect hashing. 2009. Computer Graphics 3911: 1042. Forum. 28: 1773.
28. Bader, M., S. Franz, and A. Heinecke. Hardware-Orient-
29. Lewinger, Thomas, Vinicius Mello, Adelailson Peixoto, ed Implementation of Cache Oblivious Matrix Opera- Sinesio Pesco, and Helio Lopes. Fast generation tions Based on Space-Filling Curves. 2008. Lec- of pointerless octree duals. 2010. Computer Graphics ture Notes in Computer Science. 4967: 628. Forum. 29: 1661.
30. Bock, N., Emanuel H. Rubensson, Paweł Sałek, Anders 535

Page 536

30. Bock, N., and M. Challacombe. An Optimized Sparse Approximate Matrix Multiply for Matrices with Decay. 2013. SIAM J. Sci. Comput. 35 (1): C72.
31. Challacombe, Matt, Eric Schwegler, and Jan Almlöf. Fast assembly of the Coulomb matrix: A quantum chemical tree code. 1996. J. Chem. Phys.. 104: 4685.
32. Challacombe, Matt. Linear scaling computation of the Fock matrix. V. hierarchical cubature for nu- merical integration of the exchange-correlation matrix.
33. J. Chem. Phys.. 113: 10037.
34. Challacombe, Matt, Eric Schwegler, and Jan Almlöf. Chemistry: Review of current trends. 1996. Publications Bock, N., and M. Challacombe. An Optimized Sparse Ap- proximate Matrix Multiply for Matrices with Decay.
35. arXiv:cs.NA. 1203.1692: 1. Bock, N., and M. Challacombe. An Optimized Sparse Ap- proximate Matrix Multiply for Matrices with Decay.
36. SIAM Journal on Scientific Computing. 35 (1): C72. Challacombe, M., and N. Bock. Fast Multiplication of Matri- ces with Decay. 2010. arXiv:cs.DS. 1011.3534: 1. 536

Page 537

Information Science and Technology Exploratory Research Final Report Smart Grid Control and Optimization Michael Chertkov 20130766ER Abstract We combined (i.e. “reconstructed”) our prior complex Over the three years preceding these efforts, a bigger systems approach with control theory to incorporate (smart grid) DR team made significant progress towards our recently developed modeling of a feeder with large developing an entirely new field of science consisting of number of inductive motors for designing controls to a combination of Complex System Theory, Control Theo- maintain local grid stability. ry, and Computer Science and Optimization Theory. As is Our past work on simulation-driven optimization meth- often the case, research generates new questions as well ods for transmission expansion planning assumed that as answers. This one year reserve ER addressed ques- generator outputs are constant and well controlled, tions in Grid Science that were opened by that work. We however, grid expansion that includes time-intermittent developed new approaches to modeling dynamical phe- renewable generation breaks this assumption making nomena in power grids, such as electro-magnetic waves optimal grid expansion planning far more difficult. We in transmission and dynamics of voltage drop in distribu- combined this optimization approach with a complex tion. We also extend classic optimization and planning systems approach that will bring probabilistic measures paradigms in power system with an innovative complex of electrical grid failure into the simulation-driven op- system approach, allowing a computationally advanta- timization methods to create an entirely new network geous way of dealing with uncertainties and fluctuations planning tool that balances probabilistic measure of risk of renewable energy resources. against the costs of grid expansion. Background and Research Objectives Impact on National Missions This one year reserve project has focused on extend- This project leverages LANL/DOE capabilities in informa- ing our prior work on developing new optimization and tion theory, infrastructure analysis, and complex net- control algorithm for operations and planning of modern works for research in a national priority, the smart grid. power grids on both transmission and distribution levels. The team’s combined theoretical and applied expertise The new research objectives of the project included – (a) constitutes a forward-looking DOE capability in analysis focusing on dynamic phenomena in power grids, specifi- of energy systems. Development of Information Science cally on analysis of electro-mechanical waves in trans- and Technology tools for smart grids contribute to en- mission systems and voltage dynamics in distribution ergy security, and predictive science more broadly. Our systems; and (b) building the complex system approach, research on optimization of renewable placement and e.g. in modeling uncertainties in power grids, into more the expansion of current laboratory infrastructure analy- traditional optimization approaches. sis tools to include intermittent renewables generation targets problems on the intersection of energy, climate, Scientific Approach and Accomplishments and infrastructure. We extended our past work on modeling and simula- Our smart grid effort enables LANL/DOE to expand its tion of electro-mechanical and voltage dynamics in infrastructure planning and analysis programs, initially distribution circuits, focusing on theoretical and intuitive focused mainly on infrastructure security and sponsored understanding of the highly nonlinear and hysteretic by the Department of Homeland Security, to a broader phenomena that occur when voltage dynamics interact set of programs in energy infrastructure planning and with large numbers of induction motors—phenomena efficiency analysis with the Department of Energy as pri- that are becoming serious issues for local grid stability. 537

Page 538

mary sponsor. The project thereby positions LANL closer to the center of mass in energy research, a national and DOE priority area. This project has led to programmatic funding from a range of federal agencies — DTRA, NSF, and DOE/OE. Publications Backhaus, S., and M. Chertkov. Getting a Grip on the Elec- trical Grid. 2013. Physics Today. : 10. Barrows, C., S. Blumsack, and R. Bent. Using Network Metrics to Achieve Computationally Efficient Optimal Transmission Switching. To appear in Proceedings of the 46th Hawaii International Conference on System Sciences (HICSS 2013). (Grand Wailea, Hawaii, Jan 2013). Bienstock, D., M. Chertkov, and S. Harnett. Chance Con- strained Optimal Power Flow: Risk-Aware Network Control Under Uncertainty. SIAM Review. Bienstock, D., R. Bent, and S. Backhaus. Synchronization- Aware and Algorithm-Efficient Chance Constrained Optimal Power Flow. Presented at 2013 IREP Sympo- sium-Bulk Power System Dynamics and Control. (Crete, Greece, 25-30 Aug, 2013). Duclut, C., S. Backhaus, and M. Chertkov. Phase Transitions and Dangerous Transients in Power Distribution Sys- tems. 2013. Physical Review E. 87: 062802. Dvijotam, K., S. Backhaus, and M. Chertkov. Storage Sizing and Placement through Operational and Uncertainty- Aware Simulations. To appear in HICSS 2014. (Big Is- land, Hawaii, Jan 6-9, 2014). dorfler, F., F. Bullo, and M. Chertkov. Synchronization in Complex Oscillator Networks and Smart Grids. 2013. PNAS. 110 (6): 2005. 538

Page 539

Information Science and Technology Exploratory Research Final Report Development of a Novel Algorithm to Generate User-defined Databases of Genomic Signatures Patrick S. Chain 20130786ER Abstract gorithm to identify nucleotide signatures specific and We have implemented a novel algorithm that identifies unique to any organism or set of organisms. Although nucleotide signatures – the unique genome – indepen- this algorithm conceptually has many uses, we will apply dent of functional annotation that is currently applicable this method of generating ultra-specific databases to to bacteria and viruses at any taxonomic rank. We have the problem of rapid and accurate organism identifica- named it the GOTTCHA workflow for Genome Origins tion based on sequence data. The database we initially Through Taxonomic CHAllenge. Through many test cases, propose constructing will contain all organism-specific we have demonstrated its accuracy to be far superior to (or clade-specific) signatures found anywhere in any all other published tools available at the time of analysis. genome, regardless of any predicted functions that may Specifically, we have examined in detail the performance be associated with a given signature. The signatures can of GOTTCHA versus other available tools on 24 synthetic then be used to rapidly, accurately, and sensitively clas- metagenomes of varying complexity and read distribu- sify any sample based on sequence data and using open- tion, 4 mock metagenomes whose sequences were source or proprietary tools. The proposed algorithm will derived from actual sequencers and two real metage- be flexible and extensible, allowing end-users to gener- nomes. In addition we examined GOTTCHA’s ability to ate specifically tailored databases for a multitude of successfully profile 2924 novel bacteria across all taxo- purposes. Use of this database can be implemented to nomic ranks. fortify current or future biosurveillance efforts, improve diagnostics in the public health or forensics fields, as Background and Research Objectives well as providing and essential reference for fundamen- Read-based sequence identification is highly difficult, tal science research in the fields of metagenomics and very computationally intensive given today’s sequencer pathogen diagnostics. throughput capabilities, and particularly difficult when Scientific Approach and Accomplishments data is obtained from complex (metagenomic) samples (e.g. pathogens present during infection, in blood, The GOTTCHA workflow consists of six tools: two pre- sputum, feces, or organisms found in soil, water, etc.). processing tools, the GOTTCHA generator, a read split- While there are very rapid tools that provide a snapshot trimmer, the GOTTCHA profiler, and the GOTTCHA filter, of possible taxonomic representation, there are no tools which will be folded into the profiler at a later date. We that rapidly examine all the data to identify even very used the Perl language to quickly implement the algo- rare members of natural or perturbed systems, primar- rithm, though greater speed and memory efficiencies ily due to RAM and data throughput limitations as a will be realized when this is ported to a compiled lan- result of extreme data redundancy. The database and guage. The two pre-processing tools written to prepare tools proposed here will minimize the impact of data data for the GOTTCHA generator include mkSpeciesTree. redundancy and strengthen our ability to detect and pl and taxIndexer.pl. The first generates a single, univer- respond to biological outbreaks in a rapid, reliable, and sal taxonomic tree file using the GenBank files available tailored manner, by maximizing the specificity of biologi- to the user as well as a file containing all the genomic cal knowledge one can obtain from a sample, as well as vitals of the bacteria to be analyzed (name, length, minimizing the time to acquire this knowledge, and to sequence type, TAXID, last update, etc.). The taxIndexer. do so with as much sensitivity as possible. pl tool uses the taxonomic tree file and the genomic vitals files to generate a single XML file for input into the This proposal will implement and deploy a novel al- GOTTCHA generator. It is at this level that the desired 539

Page 540

taxonomic levels are specified. The GOTTCHA generator This work ties directly to two LANL missions, the reduc- uses these data to compute the signatures, which cur- tion of global threats and solving other emerging national rently takes approximately 12 days on a 2TB RAM 80-core security and energy challenges. The suite of tools and al- server. Incrementally updating the database is a possibility gorithms have been directly applied to the area of human given the intermediate files generated, but is not currently pathogen (bacterial or viral) detection from within complex implemented. samples, primarily clinical. In addition, we propose to uti- lize these tools to identify emerging pathogens of impor- Upon generation of the desired GOTTCHA database(s), tance to agriculture and food chain supply safety, as well as three corresponding database statistics files are generated, to understanding the interactions and impacts of various organized according to the individual contigs, organisms bacterial communities on algal biofuel production strains or the database as a whole. Once these databases are when grown in production pond environments. generated, they are used as references to which metage- nomic datasets are aligned and profiled. The read set is first trimmed with our split-trimming tool written in the D programming language, then mapped to the GOTTCHA database(s) with a read mapper. We assessed GOTTCHA’s performance at the species level with the Burrows Wheel- er Aligner (BWA) version 0.6.2-r126-tpx since it provides two important features: it allows local alignment and randomly assigns multiply-mapped reads. The GOTTCHA profiler then parses the resulting alignment file, assigns data to relevant taxonomies, and generates a preliminary result. The GOTTCHA filter is then applied on top of this to remove excess false positives due to read stacking, se- quencing errors, and polymorphisms. Essential parameters of the filter were extracted from GOTTCHA’s performance on 24 synthetic metagenomes ranging from 250k reads (25 organisms) to 1M reads (100 organisms), to 300M reads (>400 organisms). These were then applied to four mock metagenomes and two real metagenomes – one air filter and one clinical fecal sample – and GOTTCHA was shown to work well even for RNA samples. Our preliminary results of the clinical sample even showed that a linear relation- ship exists between recovered hits and the concentration of pathogen in the sample, though further testing is neces- sary to solidify this relationship under various conditions. In order to assess GOTTCHA’s performance on novel organ- isms, we attempted to classify almost 3000 novel draft genomes (not present in the GOTTCHA database) to their parent taxonomy. This included 1923 novel strains, 791 novel species, 175 novel genera, 26 novel families, 4 novel orders and 5 novel classes. GOTTCHA was able to identify many contaminated genome projects and even misnamed organisms, emphasizing GOTTCHA’s success at organism classification. With the funding provided by the LDRD reserve, we were able to meet our stated milestones. A draft manuscript of the application (intended for submission to Nature Meth- ods) is available upon request, and the manuscript describ- ing the generation algorithm is underway. Impact on National Missions 540

Page 541

Information Science and Technology Exploratory Research Final Report Multi-Phenomenology Explosion Monitoring (MultiPEM) Dale Anderson 20130798ER Abstract Scientific Approach and Accomplishments This research effort developed a general mathematical We have researched source models for the optical framework for two-phenomenology yield estimation energy from either conventional explosions or nuclear for a single-point above-ground explosion, providing a explosions in the form of equations or representative mathematical pattern for multi-phenomenology yield es- data. For optical, we are working with some type of timation and error analysis. The development approach, photo sensor with known sampling rate and detector construction of a statistical likelihood, embeds the efficiency such that we have some initial estimates of mathematics of physical-basis into an error model for these uncertainties from careful calibrations. Whether observed measurements that partitions total error into the sensor is ground based or space based, the opti- bias that can be calibrated, bias that is highly dynamic, cal energy must pass through the atmosphere where, and statistical error. Dynamic bias and statistical error depending on timescales, two effects on the signal apply. are mathematically propagated to provide the error of For very early time features, atmospheric scattering may a yield estimate. The mathematical theory of likelihood impose delays that distort the observed signal. The delay functions is mature. However, research to construct increases for longer paths. This can be modeled for vari- likelihood functions can be transformational, providing a ous atmospheric conditions. The other effect is just that very general research strategy for multi-phenomenology of absorption (attenuation) of the optical energy. This analysis in national security missions. depends on the atmospheric path and can be modeled for different atmospheric conditions such as, look angle, Background and Research Objectives absorption coefficient, aerosol content, cloud extent, To estimate explosive yield, a variety of technologies etc. could be used including, but not limited to, acoustic sig- It is assumed here that each phenomenology has associ- natures, optical signatures, RF signatures, seismic signa- ated models containing descriptions of how an event tures, craterology and various chemical signatures. Each produces an observable signal and how that signal technique has a different accuracy and precision and is propagates from the event location to the receiver loca- subject to various errors that can affect the yield esti- tion. For the acoustic shockwave generated by the blast, mate. The Research challenge is how to mathematically a source model provides an estimate of the overpressure combine these different data types into a single unified from explosive yield, ambient pressure, propagation yield estimate that appropriately accounts for the un- range, and sound speed gradient. The optical signal of certainties in each technology. Focusing on acoustic and interest is the peak luminosity of the fireball that can be optical signatures in this pilot project, our innovation calculated from the explosive yield, conversion effi- is to separate the total error in the statistical likelihood ciency, event duration, and emission region size using an into three components – source, path and sensor error. early time optical transient model. From these models, A multi-phenomenology yield estimate then derived as one can predict the details of a given observation by, the value of yield that maximizes the likelihood. We can derive an expression for the uncertainty in a combined Source Model + Propagation Model = Prediction of maximum likelihood yield estimate in terms of these er- Observation. ror components. In actuality, models are imperfect and observations that provide values for the parameters in a model are imper- 541

Page 542

Figure 1. Variance of two-phenomenology maximum likelihood yield estimate. fect. The result is a prediction that doesn’t match obser- sion monitoring missions, such as the increasing threat vation and a residual defined by the difference. One can of a terrorist event, require a multiple phenomenology separate the contributions to the residual by, approach for all four monitoring functions; source signal detection, source location, identification/discrimination, Source Error + Propagation Error + Receiver/Site Error = and yield estimation. For yield estimation, continuing with Total Error. the single phenomenology approach has the unacceptable potential to give single phenomenology yield estimates In the case that multiple observations of a single type are with incongruous confidence bounds e.g., loosely overlap- made, the different local and propagation errors from each ping bounds. As national missions transform to meet new unique observation provide a means to reduce their influ- requirements, developing capabilities that obviate poten- ence on the estimate of event characteristics. However, tially incongruous technical assessments, and potentially the source errors are common to all observations, and no lead to enhanced results, is essential. number of extra observations from a single phenomenol- ogy will reduce the source errors in the residual. Multiple phenomenologies have the potential to reduce source error. The residuals for each phenomenology can be expressed as a linear combination of random errors produced by the source with variance τ2, propagation effects with variance θ2, and local noise with variance σ2. The residual for a single measurement of a given type is then a multivariate normal distribution with zero mean and covariance matrix that includes tradeoff between source terms of the two phenomenologies, τ12. The likelihood is the mathematical framework for combin- ing observations and uncertainties from multi-phenom- enology measurements, bound by a common source, towards estimating explosive yield and its error. For the two-phenomenology formulation, the variance of the maximum likelihood yield estimate is given in Figure 1. This expression provides the capability to compute a con- fidence bound on a yield estimate that includes statistical and model error. The development properly weights mea- surements from two phenomenologies when estimating yield, and provides a mathematical pattern for multi-phe- nomenology yield estimation and error analysis. The error model formulation can be used to prioritize R&D resources towards the most significant reduction in estimation error (e. g., optimizing sensor network design to reduce local er- ror, or quantifying source or path model improvement). Impact on National Missions Threats in national/homeland security and global explo- 542

Page 543

Information Science and Technology Exploratory Research Final Report Climate Impacts Research through Large Remote-sensing Datasets Timothy M. Kelley 20130800ER Abstract However no globally comprehensive approach and data- Understanding forest mortality—including mortality sets are available now to monitor vegetation dynamics. detection at pan-tropics and global scales—is crucial This lack of data precludes fundamental knowledge gath- for accurate climate modeling. Los Alamos research- ering about where, when, and why vegetation mortality ers recently developed a method, Regional Automated occurs. Disturbance Detection (RADD), for detecting changes Techniques to quantify drought stress and canopy loss in vegetation by using high-resolution remote-sensing from remote-sensing data for 2000-present have been data. The RADD method was implemented in a man- developed and tested at the global scale with Moderate- ner suitable for use on a regional scale; addressing the Resolution Imaging Spectrometer (MODIS); however, pan-tropics problem requires scaling up by six orders of its spatial resolution of 1 km2 is too coarse to capture magnitude. This brief LDRD Reserve project studied the a large fraction of forest mortality. In contrast, Landsat feasibility of scaling up the RADD method, with the ulti- provides 30 m resolution imagery at a similar frequency mate goal of tracking mortality at the pan-tropics scale. as MODIS (> monthly) since 1984. Unfortunately, no We identified two areas that hindered large-scale use: group in the world has yet considered it feasible to ana- the rate at which secondary data products are derived, lyze and interpret global, high frequency Landsat data and the rate at which machine learning classification due to the storage and processing requirements. models are trained. By rethinking and re-implementing these parts of the RADD method, we have demonstrated Creating a global vegetation dynamics monitoring sys- speedups of two orders of magnitude on comparable tem requires answers to three challenges: (1) unsuper- hardware. Our study thus indicates that applying RADD vised vegetation dynamics algorithms in units relevant at the pan-tropics scale is computationally feasible, to the carbon cycle and Earth system models (kg carbon following a straightforward path to scaling up to much km2 yr-1), (2) the ability to attribute causes of changes larger hardware using techniques that are well under- found by (1), and (3) data handling and processing at stood in the high performance computing community. By very large scale. Solutions to (1-2) have been demon- initiating collaboration between ecologists and computer strated for large areas of Southwestern USA. A solution scientists; this project has presented the opportunity to (3) has yet to be attempted. As a first step, we have to consider the architectural requirements needed to studied the computational characteristics of the exist- build a scalable system usable by domain scientists. We ing implementation of the RADD method. We have identified the crucial computational elements for such identified the key roadblocks to greater performance, a system, and illustrated some of these concepts with a and demonstrated methods to work around them. In simple domain specific language and execution environ- particular, we set out to answer three key questions: (1) ment. What rates of processing can we achieve with the RADD method? (2) What scales of storage and I/O bandwidth Background and Research Objectives are required to realize our vision? (3) What architectures Forest mortality is accelerating due to climate warming, will best serve this mission? with massive consequences on economics and climate. Modeling mortality within Earth System Models is criti- Scientific Approach and Accomplishments cal to accurately predict future climate because changes To answer these questions, we began with a careful in the vegetation can modify the global carbon cycle and study of the work of Garrity et al. The RADD method was land surface biogeophysical properties (e.g., albedo). 543

Page 544

developed in MATLAB, a key tool used by domain scientists them in an SQL database, along with metadata on per- for prototyping new methods. It provides ready access to scene basis. While this provided rapid metadata querying, many algorithms, array manipulation, file I/O, and visu- this approach suffered when considering regions smaller alization in an environment that is easily manipulated by than a 1 GB Landsat scene (e.g. subsampling). The third a lone researcher. The performance is not competitive, implementation addressed the granularity mismatch by but the convenience makes up for that in the exploratory storing the data through Pytables, a hierarchical database stages. Better performance is necessary, however, in order for managing large datasets. Data were pre-processed and to work with datasets that are five to six orders of magni- stored in structured arrays, metadata stored on a per-pixel tude larger. Improved performance supports qualitatively basis, and scenes eliminated altogether. The final imple- better results at smaller scales, as well: the improvements mentation stored the data as extendable nodes within an from this study should permit the application of Gaussian HDF5 container. The scenes were mosaicked into contigu- mixture modeling with much greater frequency, enabling ous data structures and were stored based on source, systematic studies for both validation and application of temporal, and spectral groups. the method. From performance profiling, our study identi- The performance of these approaches is compared in fied two aspects of Garrity’s method that are key to per- Figure 1. As expected, the HDF version had the best per- formance: data preparation, and the training of machine formance. By defining the computational workspace as a learning models. node within the HDF5 container, the output of an analysis Data preparation includes the formation of secondary data operation was seamlessly buffered to disk allowing single products, such as the normalized difference vegetation pass analysis at very large scales. The SQL implementa- index (NDVI) and the green red vegetation index (GRVI), tion had the second best performance. Given SQL’s more requiring a balance of file I/O, compute throughput, and powerful metadata querying capabilities, SQL could prove query handling. As with many existing packages, Garrity’s more useful depending on the balance between data MATLAB implementation loaded the multispectral data granularity and query complexity. The use of a structured directly from discrete Landsat scenes—GEOTIFF encoded array with per—pixel attributes in the Pytables approach image files stored on a local file system. Typical Landsat-de- required strided data slicing such that operations on non- rived products are linear combinations of multiple spectral trivial regions were slightly slower than the loading the bands; reading the data dominates the computational cost. entire tile, but could be powerful in the context of sparse These derivations are highly parallelizable, per-pixel opera- metadata driven operations. tions: the time to perform them can be greatly reduced by The HDF5 backend was performance tested on various assigning the work to more processing cores. More com- hardware, including RAIDed mechanical drives, solid- plex operations that require some form of normalization state hard-drives, and a 10 GigE network-attached stor- procedure can still be parallelized across multiple scenes, a age (NAS) system. In order to remove the computational feature not typically exploited by existing methods. Read- bottleneck and to provide a means to scale beyond the ing large datasets from disk thus benefits from careful trial dataset, parallelization via spatial subdivision was handling that balances I/O with processing. implemented: our backend subdivided the domain and ap- To address this balance, we explored ways in which to un- plied multithreaded operations. Despite sublinear scaling derstand and optimize the cost of I/O. In order to provide with thread count (Figure 2), the work quickly became I/O a consistent basis to compare different I/O approaches we bound using either mechanical or solid-state drives. By 48 developed a reasonably complex yet typical processing cores, this implementation had effectively saturated the pipeline. A data view spanning multiple tiles was defined underlying NAS file system; at this point, we were able to and a pipeline consisting of pan—sharpening, NDVI and process scenes two orders of magnitude faster than Gar- GRVI calculation and a Sobel convolution kernel was ap- rity’s implementation. Extrapolating this performance, we plied. believe that it would be possible to perform typical vegeta- tive index preparation for the whole of North America in a We implemented this pipeline with four data storage and matter of minutes. query structures using the Geospatial Data Abstraction Li- brary (GDAL) to convert the Landsat-provided TIF files into The unsupervised classification method developed by numerical arrays. The first approach used flat files, closely Garrity et al. provides a powerful tool for identifying dis- emulating Garrity’s MATLAB implementation with the turbances in vegetation indices. RADD identifies regions notable addition of a lookup table to index and map input of space (pixels, effectively) whose classification changes coordinates to physical files. The second implementation over time and in ways consistent with vegetation mortality. pre-converted Landsat data to binary blobs and stored Classification works by matching each pixel of the satel- 544

Page 545

lite image with a vector (“observation”), finding a model different GMM problems that Garrity et al. had worked that represents that vector data, and then applying that with. The model parameters vary sufficiently that the three model to assign a likelihood of each observation belong- problems, considered together, span the performance ing to each class. In this approach, clusters represent the space. Figure 3 shows the performance of the code on classes in the vector space. The vector spaces are derived these problems compared the original MATLAB: our C++ from the multiple bands of satellite imagery; for example, implementation was up to 60 times faster. Figure 4 shows Garrity used a two—dimensional space spanned by ΔNDVI the absolute performance of the C++ implementation as a and ΔGRVI is used, where each Δ refers to the difference function of thread count. The scaling is reasonable for an between measurements at two points in time of the NDVI OpenMP code. The code has further performance head- and the GRVI indices, respectively. Later, Garrity experi- room via distributed parallelism. As noted, the only com- mented with much higher dimensional spaces—as high as munication costs are reductions during the M-step, and 24 dimensions. scalable methods for performing such reductions are well known. Thus, we expect that application of GMM can scale Garrity’s RADD method used Gaussian mixture models out indefinitely. (GMM) to perform clustering. In GMM, the probability of a vector belonging to a cluster (class) is represented by a Our brief study demonstrated that it is computationally multivariate Gaussian function. Training a model entails feasible to create a system for applying the regional tech- fitting the centroid, weights, and variance matrix of each niques of RADD to data at the pan-tropics scale. However, Gaussian to a subset of the observations. Once trained, almost all of the work to achieve that lies ahead of us. the model is applied to compute the probability of a given As we look ahead, the guiding principle we propose is to vector, and its corresponding pixel, belonging to a cluster. harness the power of high performance computing and Our profiling revealed that almost all of the compute time large scale data handling in a way that remains tractable is spent in the first step, training. For example, in a case in and productive to scientists who are not programming which training the model took 1195 seconds, applying it specialists. This entails creating a system that responds to took 2.8 seconds. the needs of its users while making it possible to ignore as much of the implementation detail as often as pos- GMM’s are trained using the Expectation-Maximization (E- sible. In carrying out this project, we have already identi- M) method. In the expectation step, the conditional prob- fied some of the primary needs. From a computer science ability of observing the present model given each training perspective, these needs include the following: select data observation is computed. In the maximization step, the according to moderately complex user queries at vary- model is evolved in a way that increases the log-likelihood ing granularity; transform data between vector bases; of the problem. Together, this algorithm is guaranteed to form Cartesian products of vector bases; support a small converge (perhaps slowly) to a solution (perhaps not glob- set of vector and matrix operations; perform maps and ally optimal). reductions; iterate all of the above; visualize, monitor, and control these processes; define, partition, and schedule We profiled MATLAB using up to 7 seconds per iteration of work onto available resources. These needs point toward a the E-M algorithm on the hardest problem Garrity at- combination of a compact domain-specific language, and a tempted—a 24 dimensional, 11-cluster model using about runtime environment in which that language can execute. 250,000 training vectors. During profiling, we observed MATLAB using as many as 9 cores: it was already parallel- Toward the end of the project, we developed a simple ized. Our modeling indicated that the work involved about demonstration of these ideas. Exploiting the well-devel- 1010 floating point operations on about 48 MB of data. oped Google Maps API, we created a prototype interface With well-written code, a single core should complete this and processing framework (Figure 5). Using this interface, within a few seconds. Our analysis determined that most the user is able to select a region upon which they wish to of the work in the E-M algorithm could be performed in perform some form of analysis and by clicking VegeDyne, parallel. Little communication would be required—primar- the analysis is executed and the product rendered above ily, a reduction over training observations needs to be the map. (In this case, GRVI was calculated). In order to performed in the M-step. abstract the parallelization a prototype domain specific language was developed such that users could simply We wrote a C++ code that trains and applies GMM’s. The define the arithmetic operation between spectral bands, code was carefully written to permit use of fast Single without specifying any geographical information. In the Instruction Multiple Data features. We enabled multiple example shown in Figure 5, the GRVI operation is defined cores by threading the code with OpenMP. To compare the in our DSL as a JSON page attribute—the procedure is not performance of this code with the original, we used three 545

Page 546

hardcoded into the back end. This is one quick example of how a domain specific language and accompanying runtime can present ecologists access to a computationally powerful environment with a simple interface. Impact on National Missions As discussed above, climate warming is accelerating for- est mortality, with the prospect of participating in large, positive feedbacks. The capability to quantify and monitor vegetation dynamics would fill a crucial gap in the effort to accurately predict future climate. The present work has elucidated the requirements for building a system for the systematic, large—scale study of climate impacts on through remote—sensing data. The avenues explored here for improving data treatment and data analysis, together with our initial explorations of usability and extensibility, form part of a foundation on which such a system can be built. 50 45 40 35 30 25 20 15 10 5 0 Matlab Flat File MySQL PyTables HDF5 546 )ces/BM( htdiwdnaB 1400 1200 1000 800 600 400 200 0 1 2 4 8 16 48 Limit Performance of I/O Methods Method Figure 1. Comparing storage backend implementations with the reference Matlab implementation. As expected, the HDF5 imple- mentation was the fastest. The MySQL implementation handled metadata queries better, but performance was hampered by inflexibility in data granularity. )ces/BM( htdiwdnaB HDF5 Bandwidth Number Cores (Threads) Figure 2. Total bandwidth as a function of the number of threads/cores for the HDF5 implementation. By 48 cores, perfor- mance approaches the hardware limit. 70 60 50 40 30 20 10 0 1 2 3 )1 = baltaM( tuphguorhT GMM fit throughput Matlab C++ Problem Figure 3. Throughput of the C++ Gaussian mixture model train- ing code, normalized to the original MATLAB code, for three representative problems drawn from Garrity et al [10]. For the larger problems 2 and 3, performance is nearly 60 times faster on comparable hardware.

Page 547

1000 100 10 1 0.1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Figure 4. Absolute performance of the GMM training code, expressed as iterations per second. The code was threaded with OpenMP, and run up to 16 cores on a typical HPC cluster node. For the simplest problem, performance saturates at about 8 threads as absolute time approaches 1 millisecond per iteration. For the harder problems 2 and 3, performance increases up to 16 threads. These are reasonable results for OpenMP. Figure 5. Screen capture from notional domain specific language. 547 dnoces rep sno*aretI M-­‐E Throughput on Benchmarks P1 P2 P3 Number of Threads

Page 548

Information Science and Technology Postdoctoral Research and Development Continuing Project Cyber Security of the Smart Grid Russell W. Bent 20120752PRD2 Introduction a deployment of smart sensors and controls to achieve The national interest in a broad class of problems sur- this goal. However, the deployment of these devices has rounding the cyber security of infrastructure systems made the power grid more vulnerable to malicious ac- is rapidly growing. Recent work has shown that it is tors. A successful result on this project will allow for the possible for malicious entities to manipulate the data satisfaction of the Energy Security Mission and ensure provided to control systems and cause operators to that nation’s power systems are safe and reliable as make bad decisions. To date, most research has focused instrumentation of the grid continues. on building better protection and/or detection schemes. This project focuses on the premise that malicious Second, there is strong connection to the information entities can gain access and developing techniques to science mission as this proposal relies heavily on devel- mitigate the consequence of such access. oping new advances in information science. In particu- lar, this proposal relies on producing results in the fields This project focuses on four directions to address these of knowledge extraction, automated discovery, distrib- challenging information science questions: 1) Discovery, uted computing, anomaly detection, and smart sensors, 2) Assessment, 3) Mitigation, and 4) Control. First, we as outlined in the Information Science and Technology will develop efficient algorithms for determining what Grand Challenge. combinations of control system sensors can be compro- mised without being detected by a security system. Sec- Finally, this project is directly related to those federal ond, we will develop metrics for assessing the possible organizations with missions related to cyber security consequences of a successful attack on the combinations including DOE, DHS, DOD, and NIST. discovered in 1). Third, we will develop strategies for combating the attacks, such as redundant sensor place- Progress ment. Finally, we will develop new operations models In FY13, the main task was the development of new ap- that account for security concerns. proaches to mitigate the effect of data integrity attacks on power grids. To achieve this task there were three In order to solve these problems, we rely on interdisci- goals laid out to achieve 1) develop an optimization plinary research at the intersection of power systems, model for siting secure sensors to limit the consequence operations, human factors, probability, networks, of data integrity attacks 2) develop a robust control optimization, and control theory. The approaches will be scheme for responding to data integrity attacks and 3) tested on synthetic, academic benchmarks available in produce a peer reviewed paper on the topic. Goals 1 the open literature. and 3 have already been satisfied. For goal 1) Annarita Giani (the post doc funded by this project) developed an Benefit to National Security Missions algorithm for determining the worst case consequence This project is directly connected to the Energy Security of a data integrity attack. Based on these consequences, mission. Specifically, this project satisfies the mission’s a mixed-integer programming model was developed subtopic “Infrastructure reliability and security.” As the for optimal siting of mitigation devices to counter these demand for energy has grown in recent years, pressure attacks. Two types of siting models were developed. has grown on critical infrastructure, in particular electric The first one determined the least number of mitigation power grids, to operate more efficiently. This has led to devices required to counter all data integrity attacks of 548

Page 549

concern. The second model determined the best locations and K. Poolla. Smart Grid Data Integrity Attacks. 2013. to deploy a fixed set of mitigation devices so as to mini- IEEE Transaction on Smart Grid. 4 (3): 1244. mize the amount of unmitigated consequence. Goal 2 was Giani, A., O. Linda, M. Manic, and M. McQueen. Known satisfied in June 2013 with the submission of the paper Secure Sensor Measurements Concept and Its Ap- entitled “Phasor Measurement Unit Placement for Unob- plication for Critical Infrastructure Systems. 2013. In servable Attack Detection” by Giani, Bent, Pan, and Poolla Systems and Optimization Aspects of Smart Grid Chal- to the journal IEEE Transactions on Smart Grids. lenges. Edited by Marco, S., M. Carvalho, P. Pappu, and S. Vijay. Vol. IV, p. . New York: Springer. This research was also the subject of an earlier workshop paper, “Addressing Smart Grid Cyber Security” by Giani Giani, A., R. Bent, F. Pan, and K. Poolla. Phasor Measure- ment Unit Placement for Unobservable Attack Detec- and Bent. It was published in the proceedings of 8th An- tion. IEEE Transaction of Smart Grid. nual Cyber Secruity and Information Intelligence Research Workshop in Jan. 2013. This research has resulted in Giani, A., R. Bent, M. Hinrichs, M. McQueen, and K. Poolla. requests for invited talks by Annarita at Honeywell and Metrics for Assessment of Smart Grid Data Integrity United Technologies. Annarita was also asked to chair the Attacks. 2012. In 2012 Power And Energy Systems Gen- 1st International Symposium on Resilient Cyber Systems to eral Meeting. (San Diego, CA, July 2012). , p. 1. : IEEE. be held in August 2013 because of her contributions to the field. Research into Goal 2 is on track, and satisfaction of Giani, A., and R. Bent. Challenges in Smart Grid Cyber Se- curity. 2013. In 8th Annual Cyber Security and Informa- that goal is expected at the end of FY13. tion Intelligence Research Workshop. (Oak Rodge, TN, Jan. 2013). , p. 1. Oak Ridge, TM: ACM. Future Work In FY14, the project will focus research efforts on the final Linda, O., A. Giani, M. Manic, and M. McQueen. Multi- task described in the proposal. Specifically, the main task Criteria Based Staging of Optimal PMU Placement Us- ing Fuzzy Weighted Average. 2012. to be submitted to of FY14 is the development of the science to understand 2013 Power And Energy Systems General Meeting. physical consequence and cascading impacts from data integrity attacks. To achieve this task there are three goals Linda, O., A. Giani, M. Manic, and M. McQueen. Multi- we plan to achieve 1) develop a bi-level optimization Criteria Based Staging of Optimal PMU Placement Us- model of physical consequences 2) develop a model of ing Fuzzy Weighted Average. 2013. In 22nd IEEE Inter- cascading impacts and 3) produce a peer reviewed paper national Symposium on Industrial Electronics. (Taipei, on the topic. Taiwan, 28 May - 31 May 2013 ). , p. 1. Taipei, Taiwan: IEEE. Conclusion Wolpert, D., T. Bhattacharya, R. Bent, J. Neil, A. Kent, and This project is driven by emerging technologies in grid A. Giani. Event-driven non-cooperative games. 2013. sensor and control devices and accordingly specifies the Los Alamos National Laboratory, LA-UR-13-22238. technical challenges in securing systems operated by these devices. We focus on a set of problems where we have well-developed Information Science & Technology (IS&T) capabilities, develop an end-to-end security assessment of cyber-physical systems, and thus make a significant impact. The new IS&T and energy security challenges require ad- vances in applied math, algorithms, and system theory. Publications Crespi, V., G. Cybenko, and A. Giani. Attacking and Defend- ing Behavioral Anomaly Detection Methods. 2013. IEEE Journal of Selected Topics in Signal Processing. 7 (1): . Garcia, M., A. Giani, and K. Poolla. Uncertainty Quantifica- tion for Fast State Estimation. To appear in 52nd IEEE Conference on Decision and Control, CDC 2013. (Flor- ence, Italy, December 10-13, 2013). Giani, A., E. Bitar, M. Garcia, M. McQueen, P. Khargonekar, 549

Page 550

Information Science and Technology Postdoctoral Research and Development Continuing Project Edge Traffic Monitoring in Computer Networks Joshua C. Neil 20120767PRD3 Introduction Progress This research will simultaneously enhance two fields Criminal cyber activity is continously increasing. News of interest to the laboratory and the nation as a whole. media contain daily reports of government and corpo- First, the work will result in improved computer network rate networks being compromised and valuable data defense capability. The approach is to model edges being extracted. Detecting these attacks is clearly impor- (communications between pairs of computers in a net- tant. This project concerns anomaly based approaches work) using sophisticated statistical models. This edge to detecting such intrusions. modeling strategy will be combined with the PathScan methodology to serve as the basis for network monitor- Anomaly based detection utilizes baseline models to ing at LANL. PathScan was developed at LANL, as part identify abnormal activity. Motivating our approach is of the Network Scale Analysis and Defense for Cyber that compromised machines will typically share com- Systems LDRD, and has proven effective in detecting the monality in their anomalous activity, and leveraging this lateral movement of hackers within the yellow network. can increase the chance of detecting network intrusions, While the base system is imple mented, the improve- at the same time reducing the false alarms commonly ments to the edge models as part of this project will sig- associated with anomaly based intrusion detection. nificantly advance this important network security tool. So far, this project has considered two types of com- This will benefit both LANL and other government and monality across the compromised machines. The first is industry organizations seeking to better their network commonality in location, resulting in network localized surveillance systems. anomalies. These come about when the compromised machines communicate over the network, such as when Second, the methods being researched have novel an attacker moves between machines in the network, aspects in the field of statis tical modeling. The techni- or when multiple attackers are targeting the same set of cal challenges we face are also common to many other computers. The second type concerns commonality in statistical applications (e.g., social, power grid, sensor features, resulting in what we call host-similarity anoma- networks, financial modeling, etc.) making this work lies. These are motivated by the fact that compromised potentially valuable across many disciplines. machines are likely to share a large number of charac- teristics, such as being recipients of the same phishing Benefit to National Security Missions email, communicating with the same internet server from which the attack is orchestrated, sharing signs of This work is focused on cyber security. As such, it has the same malware being run on the computers, etc. broad relevance to many programs and agencies. The work has already attracted interest at other NNSA labs, The project has developed two approaches to detecting as well as members of DOD and the intelligence com- network local anomalies. In the first, anomaly detection munity. Many of the categories marked A above were is viewed as a statisticial hypothesis testing problem. marked that way since they involve sensitive data which Each computer, and possibly each between computer this project aims to protect through enhanced cyber communication, is associated with a hypothesis to be security. Therefore, this work is relevant to the follow- tested, for example: The computer is acting normally, ing missions: Nuclear Security and National Defense and versus The computer is acting abnormally. When viewed Scientific Discovery and Innovation. in relation to the network traffic, these computer-level hypotheses form a graphically struc- 550

Page 551

tured hypothesis testing problem, where task is to decide activity within computer networks, through the use of between: The network is not under attack; versus: The advanced statistical models. These models will be entirely network is under attack. By formulating models for how novel, and thus will also contribute to the general field of various attacks will perturb the data under attack, the stochastic modeling technology. approach allows for the construction of optimal tests for detecting network intrusions. Novel, and computationally efficient approaches for evaluating this overall hypothesis are developed. In the second approach to detecting network local anoma- lies, a method based on scan-statistics for graphs is de- veloped. In spatial data, a scan statistic is an anomaly score assigned to a local region in space, for instance for detecting local disease outbreaks. However, extending this idea to computer networks requires a notion of distance between computers. While the data can be viewed as producing a graph structure, the classical measure of graph distance does not work when applied to computer net- works. The reason being the presence of highly connected servers, making most computers a most two or three network communications apart. To deal with this issue, we define locality based on proximity measures for graphically structured data. The approach is applied to a real network intrusion at LANL, clearly identifying the attack and uncov- ering interesting attack patterns. Currently, work is in progress on identifying network intrusions via host-similarity anomalies. Much of the data logged concerning computers on a network is categorical, with a large and evolving pool of categories. Examples are visits to new web-pages, or new programs starting up on the machines. Individually, this new activity is typically not indicative of malicious activity, however, compromised machines are likely to have an unusually large number of such features in common. To utilize this we have devel- oped an approach for scoring the commonality of com- puters based on their newly appearing features. These scores are then used to evaluate whether a given cluster of machines have an unexpectedly high degree of similarity, an evaluation which is then combined with computer-level anomaly scores to assess the overall indication of attack. Future Work In the coming year, investigations into multiple, isolated nodes who share common internet behavior will be inves- tigated. Methods for detecting when this behavior indi- cates compromise or pending compromise will be devel- oped, along with models for normal behavior, in order to detect statistically significant deviation. Conclusion Results include highly improved sensitivity to malicious 551

Page 552

Information Science and Technology Postdoctoral Research and Development Continuing Project Mixing and Diffusion in Granular Flows Eli Ben-Naim 20130792PRD2 Introduction flow of granular materials hence has impacts on many This project aims to understand the complexity of force- applied programs. This project aims at the development carrying contacts in a granular medium, which can often of a modeling and simulation capability for describing lead to its destabilization (due to external forcing) along mixing and flow of granular materials. The ability to unexpected fault lines, as commonly observed in geo- better model mixing of granular matter will impact en- physical situations including earthquakes, mudslides, ergy applications such as storage and transportation of and avalanches. The approach will be twofold. First, a materials used in energy applications including coal and model of how granular materials flows will be developed biofuels. Granular materials tend to segregate according for the types of conditions encountered in geophysical to size, a property that can be an obstacle in manufactur- processes. ing and transportation processes. This project addresses precisely mixing characteristics of granular materials and This is a high-potential scientific problem because, in could lead to the design of optimized container design over three decades of research, this most basic question for improved mixing and more efficient transportation. regarding granular materials remains unsettled. Second, The capabilities developed in this project will also be the force-chain network structure of granular materials relevant for other DOE programs and needs including will be obtained from model experiments, being con- waste storage, waste handling and transportation, high ducted at Los Alamos, of a granular medium between explosives, and handling of materials produced in pow- sheared plates responding to the stimulus of their rela- der form. tive motion. The mathematical techniques of network analysis will be used to better understand the dynamic Progress behavior of the medium under such loads. This project is slated to start July 1, 2013. As such, no progress has been made to date. This is a cutting-edge endeavor because few such labora- tory scale experiments and analysis have been carried Future Work out previously. Additionally, new algorithms for simulat- First, we will test the granular flow model of Jop et al. ing universal emergent behaviors (clustering, ordering, (Nature, 2006) against the experiments, being conducted topology) arising from the interplay between discontinu- at Los Alamos, of a sheared granular medium confined ous granular mixing dynamics and microscopic physics between two moving plates. We will determine the will be developed as part of this project. This is a high- extent of applicability of this model, and we will estab- risk undertaking because scaling dynamics of granular lish the corrections that might be needed to capture the materials in geophysical situations to laboratory scale observations. Second, we will analyze the force-chain experiments has yet to be demonstrated. Success in the structure in this granular medium using techniques from project would imply the ability to extend the scaling con- network theory. In particular, we will aim to characterize cepts that have been so successful in fluid dynamics and the network structure and determine basic patterns and aerodynamics to a new area: solid mechanics. laws from the experiments. Our theoretical analysis will be compared with results of experiments on force chains Benefit to National Security Missions in photoelastic granular materials. Granular materials are ubiquitous and are found in a host of applied programs. Understanding mixing and Our theoretical models will be compared to empirical 552

Page 553

laws and relations from geophysics, in order to establish the applicability of the model laboratory experiments to real-world geophysical granular flows. Conclusion The outcome of the first part of this work will be improved prediction of rare and destructive seismic events, which have significant implications for society at large. The outcome of the second part of this work will be improved understanding of how localized processes affect structure formation, which can lead to better prediction of dynamics as diverse as percolation in porous medium and spread of infections on networks. Publications Christov, I., R. Lupetow , J. Ottino , and R. turman . A Study in Three-Dimensional Chaotic Dynamics: Granular Flow and Transport in a Bi-Axial Spherical Tumbler. To ap- pear in SIAM Journal on Applied Dynamical Systems. Christov, I., and H. Stone. Shear dispersion in dense granu- lar flows. Granular Matter. 553

Page 554

Physics

Page 555

Physics Directed Research Continuing Project TeV Jets: Nature’s Particle Accelerators Brenda L. Dingus 20120011DR Introduction Astro2010 Decadal Survey in Astronomy and Astrophys- Nature accelerates particles to energies exceeding 10 ics. This work is closely tied to the High Energy Physics Joules—nearly a billion times higher than terrestrial program of the DOE Office of Science. In addition, the accelerators—yet we do not know the origin of these proposed work builds LANL’s scientific capabilities in particles or the mechanisms that produce them. The computational astrophysics, cosmology and time-domain particles are produced outside our galaxy and they arrive astronomy. This project contributes strongly to NNSA at Earth from random directions due to their electrical national security programs by utilizing the state of the charge and the long distances traveled through magnetic art simulation tools and tests petascale computing codes fields between galaxies and within our galaxy. However, that could have major influences on the future of the these particles produce TeV gamma rays within their ASC program. sources and these gamma rays travel straight to Earth to Progress provide the most direct information about the physical mechanisms in these accelerators. The known extraga- The observational effort of this project has made great lactic TeV gamma-ray sources are produced by jets of progress in this past year. The 300 photomultiplier tubes plasma accelerated to relativistic velocities and many (PMTs) were delivered at the end of last fiscal year. Over exhibit rapid variability. The most advanced models of 200 have been prepared with bases soldered and water these jets suggest that magnetic fields play a critical role proof encapsulation attached. Each of these 200 PMTs in extracting the gravitational energy from the black hole has been tested at multiple high voltages with light and accelerating particles to extreme energies. With this levels ranging from 0.1 to greater than 10,000 photo- project, we will combine and extend LANL’s observation- electrons and an operating voltage selected. These 200 al, theoretical, and computational capabilities to study PMTs have been shipped to Mexico and almost 100 are these extreme sources that test our under-standing of installed in water Cherenkov detectors at the HAWC site. high-energy particle and plasma physics in an environ- The installed PMTs are recording pulses at the expected ment irreproducible on Earth. rates and with the expected time resolution. New tools are required to understand Nature’s highest In addition, we have further developed the data acquisi- energy particle accelerators. We propose to make new tion system to handle these high rates so that all pho- TeV gamma-ray observations to determine the proper- toelectrons are recorded and then a software trigger ties of TeV jets and to make new models of the physical is applied to find timing correlations associated with a mechanisms occurring in TeV jets to compare with the gamma-ray event. When all 300 PMTs are installed, we observations. will need to record 500 MB/sec. We are currently record- ing 151 MB/sec with 4 independent data systems. The Benefit to National Security Missions final system will have 10 independent systems and we Understanding cosmic accelerators is a major challenge have performed tests with simulated data streams that in high-energy particle physics and plasma astrophys- confirm the 500 MB/sec is reasonable. ics, and gamma-ray observations provide some of the The other major part of the experimental effort is the strongest constraints. The acceleration of cosmic par- incorporation of these PMTs into the reconstruction and ticles is listed as one of the 11 key science questions in analysis. This is proceeding well. We have detected the the National Academy of Sciences 2006 report “From shadow of the moon made by cosmic rays and confirmed Quarks to the Cosmos” and is further emphasized in the 555

Page 556

our angular resolution. We are also studying a recent enable new observations of astrophysical particle accelera- gamma-ray flare of the active galactic nuclei Mrk 421. tors. Continuous operations with 1/3 of the detector begin in August 2013 and the full detector will be operational by On the theoretical side of the project, we have performed August 2014. With only one third of the detector, this ob- many large-scale 3 dimensional special relativistic magne- servatory will be more than twice as sensitive as previous tohydrodynamic (MHD) simulations of jet formation and wide field of view, gamma-ray observatories and searches propagation. We have discovered that highly collimated will begin for new sources as well as new flares of known magnetic structures can be formed from magnetic energy sources. We will continue to develop the analysis tools to injection, demonstrating that Poynting dominated regime improve the background rejection and detect extragalactic is essential for jet collimation. Furthermore, we found that sources. We are also creating collaborations to perform the jet is subject to strong non-axisymmetric current-driv- contemporaneous observations at multiple wavelengths en instabilities (CDI) which lead to substantial dissipation that will be triggered by HAWC detections of flares. and jet speed is reduced. However, even with the presence of instabilities, the jet is not disrupted and will continue Next year the theoretical modeling will also advance. We to propagate to large distances. We have analyzed the jet will continue with 3D MHD jet simulations and explore properties and shown that a substantial amount of mag- further the current driven instabilities and their impact on netic energy has been transformed into kinetic energy. energy dissipation. These studies will yield important con- These 3-D MHD simulations have motivated kinetic energy straints on the rate of energy conversation from magnetic dissipation studies that are being carried out in other parts field to particles and the timescale of sudden events, both of this project. of which are being observed by TeV observatories, includ- ing HAWC. In addition, we have started the modeling of Specifically we have performed a series of fully kinetic high energy radiation spectra using our 3-D MHD simula- simulations to study both the formation and dissipation tion results as inputs. This will eventually allow us to make dynamics of thin current sheets. Large-scale 3D simula- comparison with multi-wavelength observations. tions starting with a force-free current layer have revealed a number of new insights into the kinetic-scale dynamics Conclusion of magnetic reconnection. In addition, we have discovered We will build an observatory to study the gamma-ray a new scaling for regimes with extremely strong magnetic emission from transient extragalactic sources, such as fields, which predicts the dissipation rate may be slower black holes of mass exceeding a billion times our Sun’s than previously thought. To study the detailed accelera- that eject plasmas in collimated jets. We also develop new tion mechanisms, we have implemented a particle track- computational tools on petascale computers to simulate ing capability into our simulation code. Finally, we have the variability of the energy flowing in these jets and the started to perform a new type of 3D simulation initialized mechanisms by which this energy can be used to efficiently with a uniform plasma and finite amplitude Alfven waves, accelerate particles. The observations will be compared which interact nonlinearly and generate turbulence. One to the output of the simulations to determine how Nature key feature is the spontaneous formation of kinetic-scale accelerates particles to energies that exceed those of man- current sheets, which may be unstable to reconnection made accelerators by factors of over many millions. and thus give rise to a multitude of acceleration sites em- bedded in the turbulence. These new efforts may provide Publications an important link between the kinetic and large-scale MHD Abdo, A. A., A. U. Abeysekara, B. T. Allen, T. Aune, D. Berley, studies, which also see evidence for these structures. C. Chen, G. E. Christopher, T. DeYoung, B. L. Dingus, R. W. Ellsworth, M. M. Gonzalez, J. A. Goodman, J. Future Work Granot, E. Hays, C. M. Hoffman, P. H. Huntemeyer, B. E. This project has both an observational and a theoretical Kolterman, J. T. Linnemann, J. E. McEnery, A. I. Mincer, component. The observational work requires the con- T. Morgan, P. Nemethy, J. Pretz, E. Ramirez-Ruiz, J. M. Ryan, P. M. S. Parkinson, A. Shoup, G. Sinnis, A. J. struction of new hardware for detecting astrophysical Smith, V. Vasileiou, G. P. Walker, D. A. Williams, and sources of high energy gamma rays, and the theoretical G. B. Yodh. CONSTRAINTS ON THE EMISSION MODEL work involves new computational modeling of the physics OF THE “NAKED-EYE BURST” GRB 080319B. 2012. of particle acceleration and propagation in these gamma- ASTROPHYSICAL JOURNAL LETTERS. 753 (2): -. ray sources. Abdo, A. A., B. T. Allen, R. Atkins, T. Aune, W. Benbow, D. In the next year, the new hardware procured by this proj- Berley, E. Blaufuss, E. Bonamente, J. Bussons, C. Chen, ect will be installed in the High Altitude Water Cherenkov G. E. Christopher, D. G. Coyne, T. DeYoung, B. L. Dingus, (HAWC) observatory to lower its energy threshold and thus D. E. Dorfan, R. W. Ellsworth, A. Falcone, L. Fleysher, 556

Page 557

R. Fleysher, J. Galbraith-Frew, M. M. Gonzalez, J. A. P. A. Toale, K. Tollefson, I. Torres, T. N. Ukwatta, J. Goodman, T. J. Haines, E. Hays, C. M. Hoffman, P. Valdes-Galicia, P. Vanegas, V. Vasileiou, O. Vazquez, H. Huntemeyer, L. A. Kelley, B. E. Kolterman, C. P. X. Vazquez, L. Villasenor, W. Wall, J. S. Walters, D. Lansdell, J. T. Linnemann, J. McCullough, J. E. McEnery, Warner, S. Westerhoff, I. G. Wisher, J. Wood, G. B. T. Morgan, A. I. Mincer, M. F. Morales, P. Nemethy, Yodh, D. Zaborov, and A. Zepeda. On the sensitivity of D. Noyes, J. Pretz, J. M. Ryan, F. W. Samuelson, P. M. the HAWC observatory to gamma-ray bursts. 2012. S. Parkinson, A. Shoup, G. Sinnis, A. J. Smith, G. W. ASTROPARTICLE PHYSICS. 35 (10): 641. Sullivan, V. Vasileiou, G. P. Walker, M. Wascko, D. A. Williams, S. Westerhoff, and G. B. Yodh. OBSERVATION Aluie, H., Shengtai Li, and Hui Li. Conservative Cascade AND SPECTRAL MEASUREMENTS OF THE CRAB of Kinetic Energy in Compressible Turbulence. 2012. NEBULA WITH MILAGRO. 2012. ASTROPHYSICAL Astrophysical Journal, Letters. 751 (2): L29 (6 pp.). JOURNAL. 750 (1): -. Beresnyak, A., H. Xu, H. Li, and R. Schlickeiser. MHD Abdo, A. A., U. Abeysekara, B. T. Allen, T. Aune, D. Berley, E. Turbulence and Cosmic Ray Reacceleration in Galaxy Bonamente, G. E. Christopher, T. DeYoung, B. L. Dingus, Clusters. 2013. Astrophysical Journal. 771: 131. R. W. Ellsworth, J. G. Galbraith-Frew, M. M. Gonzalez, Govoni, F., M. Murgia, H. Xu, H. Li, M. L. Norman, L. Feretti, J. A. Goodman, C. M. Hoffman, P. H. Huntemeyer, G. Giovannini, and V. Vacca. Polarization of cluster C. M. Hui, B. E. Kolterman, J. T. Linnemann, J. E. radio halos with upcoming radio interferometers. McEnery, A. I. Mincer, T. Morgan, P. Nemethy, J. Pretz, 2013. ASTRONOMY & ASTROPHYSICS. 554: -. J. M. Ryan, P. M. S. Parkinson, A. Shoup, G. Sinnis, A. J. Smith, V. Vasileiou, G. P. Walker, D. A. Williams, Guan, X., H. Li, and S. Li. Relativistic MHD Simulations of and B. Yodh. SPECTRUM AND MORPHOLOGY OF THE Poynting Flux-Driven Jets. Astrophysical Journal. TWO BRIGHTEST MILAGRO SOURCES IN THE CYGNUS REGION: MGRO J2019+37 AND MGRO J2031+41. 2012. Liu, Y. H., W. Daughton, H. Li, H. Karamabadi, and V. ASTROPHYSICAL JOURNAL. 753 (2): -. Roytershteyn. Bifurcated Structure of the Electron Diffusion Region in Three-Dimensional Magnetic Abeysekara, A. U., J. A. Aguilar, S. Aguilar, R. Alfaro, Reconnection. 2013. Phys. Rev. Lett.. 110: 265004. E. Almaraz, C. Alvarez, J. D. Alvarez-Romero, M. Alvarez, R. Arceo, J. C. Arteaga-Velazquez, C. Badillo, Liu, Y. H., W. Duaghton, H. Li, H. Karimabadi, and S. P. A. Barber, B. M. Baughman, N. Bautista-Elivar, E. Gary. Do Dispersive Waves Play a Role in Collisionless Belmont, E. Benitez, S. Y. BenZvi, D. Berley, A. Bernal, Magnetic Reconnection?. Phys. Rev. Lett.. E. Bonamente, J. Braun, R. Caballero-Lopez, I. Cabrera, A. Carraminana, L. Carrasco, M. Castillo, L. Chambers, Si, J., S. A. Colgate, H. Li, J. Martinic, and D. Westpfahl. R. Conde, P. Condreay, U. Cotti, J. Cotzomi, J. C. D’Olivo, Data Acquisition in a High-Speed Rotating Frame for E. de la Fuente, C. De Leon, S. Delay, D. Delepine, New Mexico Institute of Mining and Technology Liquid T. DeYoung, L. Diaz, L. Diaz-Cruz, B. L. Dingus, M. A. Sodium αω dynamo experiment. To appear in Review Duvernois, D. Edmunds, R. W. Ellsworth, B. Fick, D. W. of Scientific Instruments. Florino, A. Flandes, N. I. Fraija, A. Galindo, J. L. Garcia- Luna, G. Garcia-Torales, F. Garfias, L. X. Gonzalez, M. Skillman, S., H. Xu, E. Hallman, B. O’Shea, J. Burns, H. M. Gonzalez, J. A. Goodman, V. Grabski, M. Gussert, Li, D. Collins, and M. Norman. Cosmological MHD C. Guzman-Ceron, Z. Hampel-Arias, T. Harris, E. Hays, Simulations of Galaxy Cluster Radio Relics: Insights L. Hernandez-Cervantes, P. H. Huntemeyer, A. Imran, and Warnings for Observations. 2013. Astrophysical A. Iriarte, J. J. Jimenez, P. Karn, N. Kelley-Hoskins, D. Journal. 765: 765. Kieda, R. Langarica, A. Lara, R. Lauer, W. H. Lee, E. C. Linares, J. T. Linnemann, M. Longo, R. Luna-Garcia, Xu, H., F. Govoni, M. Murgia, H. Li, D. Collins, M. H. Martinez, J. Martinez, L. A. Martinez, O. Martinez, Norman, and R. Cen. Evolution and Comparisons of J. Martinez-Castro, M. Martos, J. Matthews, J. E. Cosmological MHD Galaxy Cluster Simulations to Radio McEnery, G. Medina-Tanco, J. E. Mendoza-Torres, Observations. 2012. Astrophysical Journal. 759: 40. P. A. Miranda-Romagnoli, T. Montaruli, E. Moreno, M. Mostafa, M. Napsuciale, J. Nava, L. Nellen, M. Newbold, R. Noriega-Papaqui, T. Oceguera-Becerra, A. O. Tapia, V. Orozco, V. Perez, E. G. Perez-Perez, J. S. Perkins, J. Pretz, C. Ramirez, I. Ramirez, D. Rebello, A. Renteria, J. Reyes, D. Rosa-Gonzalez, A. Rosado, J. M. Ryan, J. R. Sacahui, H. Salazar, F. Salesa, A. Sandoval, E. Santos, M. Schneider, A. Shoup, S. Silich, G. Sinnis, A. J. Smith, K. Sparks, W. Springer, F. Suarez, N. Suarez, I. Taboada, A. F. Tellez, G. Tenorio-Tagle, A. Tepe, 557

Page 558

Physics Directed Research Continuing Project IMPACT: Integrated Modeling of Perturbations in Atmospheres for Conjunction Tracking - A New Orbital Dynamics and Drag Model to Avoid Collisions in Space Alexei V. Klimenko 20120036DR Introduction us to provide accurate collision probabilities for avoid- The goal of this project is to develop a new orbital ance maneuvers and the ability to track space objects dynamics and atmospheric drag model with accurate during geomagnetic storms as well as better satellite quantification of uncertainty. Objects in LEO (Low Earth lifetime and reentry estimates. Nuclear non-proliferation Orbit) experience drag due to atmospheric friction, depends heavily on satellite observations. In addition, which makes it hard to predict their future paths. Cur- we will build underlying capability for understanding rent atmospheric density models used for orbital dynam- weapons effects on the upper atmosphere, with rel- ics are all finely tuned to long-term averages. They show evance to DP programs. good statistical agreement with densities derived from Progress actual spacecraft data during quiet times, but when the upper atmosphere starts changing, all bets are off. The Density modeling and data assimilation with GITM models do not help use understand the physics going on We have completed the data assimilation integration in the upper atmosphere. Furthermore, orbital drag is with the GITM model and have tested the sensitivity of averaged over a whole orbital period not accounting for model output to the model drivers. Currently the model local density variations, say from one side of the Earth is still using the solar radio flux parameter F10.7 as input to another. Current density model errors can lead to but will upgrade the parameter input to EUV flux and prediction errors of kilometers within a very short time. assess the coupling function of EUV and heating/loss Our goal is to develop a new physics-based atmospheric processes in the upper neutral atmosphere. density model by focusing on accurately describing the In parallel to the physics-based atmospheric density energy deposition and cooling during space weather modeling, we are also exploring a non-linear regression events and updating this model to better account for ob- type approach. We have tested a support-vector ma- served changes in satellite tracks. This approach will en- chine algorithm and found promising first results. We able us to answer important physics questions regarding used solar wind and geomagnetic indices as input and the coupling between space weather, upper atmospheric are working on developing a predictive tool and compar- density, and orbital drag. ing the results to the measured atmospheric densities from CHAMP and GRACE satellites. Benefit to National Security Missions Space Situational Awareness (SSA) is one of the core mis- Drag coefficient modeling sions of Los Alamos as outlined in the Strategic Invest- The use of fixed drag coefficients to infer atmospheric ment Plan. SSA is essential to ensuring stability in space mass densities from the orbital decay of satellites has and sustainability of our space activities and we will di- resulted in extensive biases in empirical atmospheric rectly address this particular need with rigorous science models. Physical drag coefficients can help reduce these and innovative technology. Our results will let SSA evolve biases by accurately modeling the interaction of gas par- from today’s forensic operational mode to a predictive ticles with a satellite surface. We are using the NASA pro- mode and transform it so that we will receive advanced vided DAC code and our own test particle Monte Carlo warnings of possible collisions instead of being forced code that are both validated with closed-form solutions to investigate the cause in the aftermath (see Iridium- for the drag coefficient in free molecular flow for simple Cosmos collision). Specifically, our project is striving to geometries. Both the test particle Monte Carlo (TPMC) advance the ability to monitor and track space objects by and the Direct Simulation Monte Carlo (DSMC) methods creating a new atmospheric drag model. This will allow 558

Page 559

accurately account for the changes in the incident veloc- tion, drag coefficient, etc. It employs a light-weight service ity distribution function due to multiple reflections with oriented architecture which lets the user enter and select the satellite surface. We have studied the sensitivity of parameters such time window, density model selections, the drag coefficient and orbit propagation due to changes subset of the satellite catalog. Data is then being processed in the atmospheric density, chemical composition, and on a high-performance computer and results are sent back temperature. to the browser. Observations Future Work A problem at tube manufacturer for the camera has Observations caused a significant delay in the procurement process. The Before the end of this FY, we will have a fully automated tube has finally arrived (after a 1 year delay) and the cam- data pipeline in place. More of our data will be delivered to era team is currently testing the device and finishing the our AFRL Kirtland collaborators for evaluation. In the com- assembly of the camera. In the meantime, we have used a ing FY, we will complete the dedicated IMPACT camera and camera on loan to us and performed several observational continue the series of observational campaigns. campaigns. The data processing pipeline is now complete including detailed documentation. The camera and pro- Density modeling and data assimilation with GITM cessing pipeline have been tested and validated against We will establish fine-tuning of solar/solar wind drivers of know objects with precision ephemeris including WAAS the GITM model to get a better forecast of neutral densi- satellites. The accuracy of the camera has been quantified ties at high altitudes. We will include more appropriate and data has been delivered to our collaborators at AFRL EUV forecasting model replacing the F10.7 radio flux. for testing and validation. Collisional Probability and UQ Collisions Statistics and UQ We will complete the uncertainty quantification of all We have completed the development and testing of full modules in the integrated code. We will submit a paper on Monte Carlo modeling and importance sampling for col- the collisional probabilities based on atmospheric densities lisional probabilities. The UQ framework improves on with the GITM model. previous methods that focus on instantaneous collision probabilities based on Gaussian state assumptions. The Integration new framework considers the distribution of the entire We will complete the integration of our work into one sin- tracks for any pair of objects. The distribution for the rela- gle end-to-end model. This model will allow us to calculate tive distance between two objects over a fixed period of collision probabilities using our physics-based atmospheric time is approximated with a distribution for functions. This density model. approximation accounts for non-Gaussian state estimates Technology Demonstrations and predictions from an orbital propagator. We have We will present our end-to-end model to potential stake- tested and validated the importance sampling method and holders to validate mission relevance. studied the effect of atmospheric density model assump- tion on the collision probability. Conclusion Atmospheric Density Reconstruction using Satellite Orbit The results from this project will revolutionize the physi- Tomography cal understanding of the upper atmosphere. We will build We have developed a new method of estimating density a new physics-based atmospheric density model that is using a tomography-based approach, inspired by X-ray tightly coupled to changing solar flux and space weather computed tomography from the medical imaging field. The conditions. Our new physics-based atmospheric density change in specific mechanical energy of the orbit is used as model will describe the thermospheric response to energy the measurement, which is related to the integrated den- input, recirculation, and cooling. By feeding back informa- sity over the orbit. Using several such measurements from tion from satellite tracking, it will provide both atmospher- a number of satellites, one can estimate the density scale ic density and relative velocity for orbital drag calculations. factor (i.e. a correction to an assumed density model). Results from our project will have a clear technical impact on understanding the upper atmosphere and transform Integration to End-to-End Model (E2E) the techniques of calculating drag and collision probabili- We have been working towards a first end-to-end model ties. that incorporates all aspects of the IMPACT project. The integrated model will let us assess the uncertainty con- Publications tribution of observations, density models, orbit propaga- Elvidge, S., H. Godinez, M. Angling, and J. Koller. Improved 559

Page 560

Mmodelling of Upper Atmospheric Densities Using Model Developed using Direct Simulation Monte Multi-model Ensembles. To appear in 3rd IMA Carlo (DSMC) Method. 2013. AIAA/AAS Space Flight Conference on Mathematics in Defence. (Malvern, UK, Mechanics Conference. Oct 2013). Mehta, P. M., and C. A. Mclaughlin. Drag Coefficient Godinez, H. C., B. Nadiga, A. Ridley, J. Koller, E. Lawrence, Modeling for GRACE using Direct Simulation Monte and D. Higdon. Atmospheric Density Specification with Carlo Drag Coefficient. 2013. AIAA/AAS Space Flight the Global Ionosphere-Thermosphere Model (GITM) Mechanics Conference. using the Ensemble Kalman Filter. 2012. : 2012. Mehta, P. M., and M. Shoemaker. Satellite Drag Coefficient Godinez, H. C., J. Koller, and A. J. Ridley. Ensemble Data Modeling. 2012. SWX Reports. Assimilation for the Global Ionosphere-Thermosphe Model (GITM). 2013. J. Geophys. Res.. : in preparation. Mehta, P., and C. McLaughlin. Drag Coefficient Modeling for GRACE using Direct Simulation Monte Carlo. To Kelecy, T., M. Shoemaker, and M. Jah. Application of the appear in Advances in Space Research. Constrained Admissible Region Multiple Hypothesis Filter (CAR-MHF) to Initial Orbit Determination (IOD) of Shoemaker, M. A., B. Wohlberg, and J. Koller. Atmospheric a Break-up. To appear in 6th European Conference on Density Reconstruction using Satellite Orbit Space Debris. (Darmstadt, Germany, April 2013). Tomography. To appear in Journal of Spacecraft and Rockets. : submitted. Klimenko, A., D. Palmer, D. Thompson, R. Linares, and E. Rabin. SSA Data Analysis Framework 1 Introduction 2 Shoemaker, M. A., B. Wohlberg, and J. Koller. Atmospheric Fenton Hill Observatory Setup Overview 3 Input Data Density Reconstruction Using Tomography of Satellite 4 Telescope Pointing. 2013. LANL Reports LACP-13- Orbits. 2013. AAS/AIAA Space Flight Mechanics 00008. Meeting. Lawrence, E., and D. Higdon. An Importance Sampling Walker, A.. Drag coefficients computed by the Direct Method for Computing Collision Probabilities. 2013. Simulation Monte Carlo method. 2013. LANL Reports. Journal of Spacecraft and Rockets. : in preparation. Walker, A., P. Mehta, and J. Koller. A Quasi-specular Drag Linares, R.. Angles-Only Orbit Determination For Electro- Coefficient Model using the Cercignani-Lampis-Lord Optical Sensors. 2012. SWX Reports. Gas-Surface Interaction Model. To appear in Journal of Spacecraft and Rockets. Linares, R., J. L. Crassidis, C. Wetterer, K. A. Hill, and M. K. Jah. Astrometric and Photometric Data Fusion for Walker, A., P. Mehta, and J. Koller. Different Mass and Surface Material Estimation using Refined Implementations of Diffuse Reflection with Incomplete Bidirectional Reflectance Distribution Functions-Solar Accommodation for Drag Coefficient Modeling. Journal Radiation Pressure Model. To appear in Air Force Maui of Spacecraft and Rockets. Optical and Supercomputing Site Conference. (Maui, Wetterer, C. J., R. Linares, J. Crassidis, T. Kelecy, M. Ziebart, HI, September 2013). P. Cefola, and M. Jah. Distribution Functions. 2013. Linares, R., M. Shoemaker, A. Walker, P. Mehta, D. Palmer, AIAA/AAS Space Flight Mechanics Conference. D. Thompson, J. Koller, and J. L. Crassidis. Photometric Data from Non-Resolved Objects for Space Object Characterization and Improved Atmospheric Modeling. To appear in Air Force Maui Optical and Supercomputing Site . (Maui, HI, Sept 2013). Linares, R., and J. L. Crassidis. Sigma Point Transformation for Gaussian Mixture Distributions Applied to Attitude Estimation. 2013. AIAA/AAS Space Flight Mechanics Conference. Mehta, P. M., A. Walker, C. A. Mclaughlin, and J. Koller. Comparing Physical Drag Coefficients Computed with Direct Simulation Monte Carlo Using Different Gas- Surface Interaction Models. 2013. Journal of Spacecraft and Rockets. : submitted. Mehta, P. M., and C. A. Mclaughlin. GRACE Drag Coefficient 560

Page 561

Physics Directed Research Continuing Project Hydrodynamical Mix Studies at the National Ignition Facility (U) Michael J. Steinkamp 20120072DR Introduction Benefit to National Security Missions The overarching goal of this project is the development The products of this research would advance the state- and validation of computational sub-grid models, that of-the-art Inertial Confinemet Fusion (ICF) modeling capture the coupled physics of turbulent mix, Thermo- capabilities in LANL’s Eulerian ASC code suite. Both this nuclear (TN) burn, and Charged Particle Transport (CPT), new computational capability and the new diagnostics and will tightly couple many Inertial Confinement Fusion developed on the National Ignition Facility (NIF) would (ICF)-specific physics packages contained in the Eulerian directly address NNSA’s goal of fostering a scientific Applications Project (EAP) code suite. This goal requires program at NIF that is both of broad scientific interest development of theoretical and experimental tools and of specific utility of the weapons program. Better necessary for validation of our Eulerian computational understanding of turbulence has broad application to sub-grid models. We must accurately capture the physics multiple missions including climate, energy (combusion), and mix morphology associated with TN burn in an ICF forensic particle transport, etc. capsule. The experimental goals include the design of in- novative NIF experiments and their associated diagnos- Progress tics. Foremost among these are efforts to develop and A sub-grid model that uniquely couples the BHR turbu- improve Neutron Imaging (NI) and reaction-in-flight (RIF) lent mix model and the thermonuclear burn package in neutron diagnostics. A more risky part of the experimen- a physical manner have completed and verified. Valida- tal program is the development of prompt radiochemi- tion is currently in progress using a variety of available cal diagnostics for use at NIF. New ideas for background data sets. It has been determined that the most strin- mitigation and debris assay will be required and will gent test of this model is afforded by separated reac- be tested by a series of characterization and develop- tants experiments. In a separated reactant reactants ment experiments using the Omega Laser facility. These experiment a deuterated (CD) capsule shell is utilized include a feasibility study for the tape transport system with a tritium fill. 14 MeV neutrons (D + T -> α + n (14 and gas jet transport system post-shot assay technique. MeV)) will be generated only when the shell material intimately (atomically) mixes with the tritium fill. To this This project represents a completely new approach to end, a collaboration with LLNL has been formed and per- the modeling of the way the physics of turbulent mix, sonnel from this project have been integral participants TN burn and CPT are modeled. This approach has the in 4 NIF shots to date. Additional shots are planned and potential to provide, for the first time, a physically self- will be conducted during FY-13. In addition a dedicated consistent model of TN burn phenomena which has tra- shot on the OMEGA laser has been obtained for late FY- ditionally been treated in an ad-hoc and overly simplistic 13 and the design of an OMEGA size capsule experiment manner. was initiated. Validating the newest ICF relevant pack- ages in the Eulerian code consumed a significant amount This project will utilize both OMEGA and NIF ICF capsule of effort during FY-12. These packages included the 3-T experiments to validate the modeling approach. As such physics, charged particle transport, ion conduction and this project represents a very aggressive comprehensive the PDF thermonuclear burn package. attempt at making ICF relevant to ASC code and model development in a generalized manner. Work that was initiated in FY-12 and continuing in FY-13 is centered on verifying and validating the knock-on physics in the charged particle transport package (neces- 561

Page 562

sary to perform both RIF and CPT radchem calculations) alone. We are currently testing the new model on un- implementing the necessary cross section data bases for stable flows. Successful implementation of this model will CPT radchem, and completing the full implementation of lead to the generation of a physical realization of sub-grid the PDF burn model in the separated reactants portion of structure that is required for fully coupling to the charged the code. Completion of these items is the necessary initial particle transport and for future extensions to other as- step in utilizing the separated reactants data along with pects of sub-grid physics. the CPT radchem and RIF data along with the CPT and RIF In order to attain the high-fidelity experimental data data for validation of the PDF modeling approach. required to validate these computational sub-grid mod- The charged particle radchem development and the devel- els, we are working towards both our original and our opment of diagnostic techniques to measure high energy stretch experimental campaign goals, and they are to: reaction in flight neutrons conducted during the last 12 (1) design and shoot separated reactant experiments at months have proved successful. Both of these techniques Omega, where we have obtained a dedicated shot day for provide independent measures of the amount and state late FY13, as well as separated reactants and Tritium filled (morphology and distribution) of shell mix into thermo- capsules on the National Ignition Facility (NIF) in FY14 (in nuclear fuel. conjunction with the DIME project at LANL); also focusing on how to change the pressure and fill to influence capsule In order to attain the high-fidelity experimental data convergence; (2) develop a prompt radiochemical assay required for TN burn model validation, the original and capability for the NIF; (3) develop concepts in NIF solid stretch experimental campaign goals, that we have accom- debris collection experiments; and (4) develop a reaction plished thus far are: in flight (RIF) diagnostic capability in ignition capsules for better data collection and improved statistics for charged • Executing more mix experiments with a deuterated particle stopping. (CD) capsule shell Conclusion • Simulating CD mix implosion experiments with the EAP Broadly, we identify two main technical products. The first code suite, and the turbulence transport model will be the development of a new diagnostic suite for NIF, • Experimentally observed RIF neutrons at the NIF capable of addressing many issues in high-energy-density physics, but specifically targeting the physics of mixing at • Analyzed prompt assay detector development for material interfaces in ICF capsules. Success of this diag- Omega and are preparing presentations for summer nostic suite would enable the design and fielding of a full conferences experimental mix program at NIF. The second technical product of our proposed research would be an experimen- • Presented prompt radchem concepts to the national tally validated computational subgrid model that couples ICF Diagnostics Working Group the ICF-specific physics packages that model turbulence, mix, TN burn, and CPT, significantly advancing the capabil- • Through RIF neutron measurements, have shown ity of LANL’s Eulerian ASC code. quantum effects are important for charged particle stopping in a degenerate plasma. Publications on this work are in progress. Future Work The sub-grid modeling effort requires more work as we more fully integrate our current sub-grid thermonuclear (TN) burn model into other aspects of the Advanced Simu- lation and Computing (ASC) code’s burn framework. To ex- tend the Probability Distribution Function (PDF) approach, we are developing an approach that allows the fine-scales of turbulent motions to evolve in a natural fashion, rather than being simply slaved to the largest turbulent scales. This extension describes the evolution of the Kolmogorov dissipative scales and requires minimal changes to the code’s mix model and provides spatial or scale information not predicted by either the BHR mix model or by the PDF 562

Page 563

Physics Directed Research Continuing Project Advancing the Fundamental Understanding of Fission (U) Morgan C. White 20120077DR Introduction Benefit to National Security Missions Nuclear fission is the energy source for both nuclear An understanding of the fission process is critical to the weapons and nuclear reactors. 73 years after the dis- Laboratory’s national security missions. Assessment of covery of fission, it remains imperfectly understood. To fission yield from an NTS event or in a post-detonation improved that understanding, we will: situation depends on an accurate assessment of the fission-product yields. The accuracy and utility of simula- • Build a detector system to measure the velocity and tions for device assessment, reactor design and safety, total kinetic energy (TKE) of fission fragments as a and SNM detection depend on the accuracy of the function of incident neutron energy. fission evaluations. An enhanced understanding of the fission process and the correlations in emissions will • Develop fission theory and modeling to enable the greatly enhance our ability to quantify the sources of un- prediction of complete fission-fragment yields as a certainty in our predictions. This project helps keep LANL function of incident energy forming the basis of new and our nation at the forefront of fission science, further evaluation tools for these data. Benchmark these enhancing our reputation in this field, and strengthen- tools based on the new experimental data. ing the science-based deterrence provided by a strong national laboratory system. • Evaluate the Pu239 fission-fragment yields based on the new theory and experimental data. Progress Recent inquiries into discrepancies between nuclear The SPIDER project, LDRD 20120077DR, seeks to ad- weapons yields reported by LANL and LLNL addressed vance the fundamental understanding of fission through fission-product yields. One conclusion was the “con- a better understanding of the moment of scission. We vincing experimental evidence for energy dependence” are seeking a new capability to measure the individual from 0.2 to 1.9 MeV of the Nd147 cumulative yield from fragments as they are emitted at scission and will use Pu239 fission. This single finding brought the two labs these data to validated guide and validate advanced into closer agreement. New experiments are necessary theories describing the same. to confirm the energy dependence. There are no experi- The theory portion of our project is making steady mental data in the energy range from 2 to 14 MeV and progress. We believe one of the strengths of this effort a 25% discrepancy exists between measurements at 14 is its two pronged development. Our phenomenologi- MeV. Confirming the low-energy yield findings, provid- cal model has been completed and used successfully to ing the missing data, and resolving the discrepancy are evaluate fission product yields and total kinetic energy crucial issues within the Nuclear Performance grand as a function of incident neutron energy. This is a sig- challenge. nificant milestone completed within the project. While We will advance the state-of-the-art for fission science. it is not predictive and must be tuned to experimental This impacts defense programs, nuclear energy, and the data, this tool provides a fallback that can be used to science of signatures. We will push deeper into this fron- interpret the experimental data to be taken and provide tier than previous studies, measuring more of the data the evaluated libraries, of the kind that will be used for than ever before, probing with theory why it happens applications and submitted to the US national ENDF/B this way, and gleaning knowledge that will help LANL repository, desired. and the nation stay at the forefront of this critical field. 563

Page 564

The advanced theory efforts have also made significant chamber continues including testing of new silicon nitride progress towards a more predictive evaluation capability. windows to reduce the energy loss of the fragment as it The LANL model describing the potential energy of the enters the detector. Use of accelerated radioactive beams nuclear surface has been used in static analysis to pre- at facilities like NCSL is being studied as a potential way to dict the nuclear masses at the level of uncertainty in the directly measure the pulse height defect identified as the experimental data. Our project seeks to add a capability to largest source of uncertainty in this measurement. track the dynamic evolution of the nuclear potential such that one may follow the nucleus from its original excitation Future Work to the point of scission. During the final year of the SPIDER LDRD project, we plan fulfill, in general, the original goals of our research plan. This year the code for the dynamic evolution has been That is to take data on plutonium-239 fission fragments developed to the point of tracking Monte Carlo samples of using a new 2V,2E detection system and to make evaluated individual evolutions. Unfortunately, the use of the origi- data libraries for the fission fragment yields of the same. nal grid designed for the static interpretation causes the Unfortunately, we will be unable to build a full instrument model evolve into unphysical representations. Significant and these measurements will not have the resolution to effort has been required to refine this grid to provide the measure the distributions over the full range of incident smoothness necessary to allow for the dynamic evolution. neutron energies desired. However, they will demonstrate These efforts are in the final stages and we expect to begin a capability that can be scaled to make these programatti- the last phase of this work by the end of the year. At that cally desired measurements. time, the tool can be used to compare against experimen- tal data and tuning and validation efforts begin. The experimental efforts will focus on completion of a dual-arm instrument that will be fielded on the LANSCE The experimental efforts have been more difficult than an- 1FP12 beam line. These measurements will focus on taking ticipated. Steady progress has been made developing and data on uranium-235 that can be compared to established testing the individual components but putting together standards. Data will also be taken on plutonium-239; these an instrument at the scale originally desired has proved im- measurements will allow us to explore safety and mea- possible. As discussed with our appraisal panel in February, surement issues working with this more difficult isotope. this has led to a reworking of the goals for the project. Our During the year californium-252 measurements will also focus, as recommended by the committee, is now on prov- be made. The U235 and Cf252 measurements will be used ing the capability can be achieved and working to demon- to demonstrate the accuracy of the detector. The Pu239 strate that the final instrument can be built. measurements will prepare the way to make the final ener- gy-dependent measurements of interest. The prototype To this end, we are in progress to build a dual-arm, as chamber will also be used to investigate key challenges in opposed to the desired nine dual-arm pairs, instrument. scaling up to the final nine dual-arm pair instrument. The Fabrication of all of the components is essentially complete LANSCE Program Advisory Committee strongly endorsed and assembly is in progress to be ready to make measure- this experimental plan. ments in this falls LANSCE beam cycle. To reach this point, significant research has been completed on many of the The theory efforts will focus on finalizing the capabil- components that go into the overall system. ity to perform advanced predictions. These can then be compared against the phenomenlogical model results The instrument measures the velocity and energy of both that have already been produced. Eventually these will be fission fragments emitted. The velocity measurement uses tuned and validated with new data but in the meantime a thin foil to produce electrons from the fragment pas- will still be the most accurate estimates of these key fission sage and a fast timing plate to record that signal. Initial quantities and included in future application libraries. studies of carbon foils used for the electron production have shown them to be inadequate and mylar foils with Conclusion sputtered metal coatings have been obtained and shown The fission process plays a fundamental role in nuclear to produce better signals. We are now producing all of the weapons and nuclear energy, and is therefore deeply tied subcomponents for the final dual-arm chamber. to our national security and energy independence mis- The energy of the fragment is measured by an ionization sions. We expect to improve our understanding of nuclear chamber. These chambers are in wide spread use at LANL fission, which will support these important technologies and we have successfully constructed a prototype. Testing and lead to important advances in basic science. The new indicates it is working as expected and optimization of this experimental data and theoretical models will produce 564

Page 565

evaluated nuclear data that feeds into the codes used for Tovesson, F., C. Arnold, A. Laptev, K. Meierbachtol, M. weapon simulations and nuclear reactors, and will thus White, R. Blakeley, A. Hecht, D. Mader, and L. Snyder. improve technical design and performance predictions of SPIDER: a new instrument for fission fragment research at the Los Alamos Neutron Science Center. To appear these systems. in Fifth International Workshop on Nuclear fission and Fission-Product Spectroscopy. (Caen, France, May 28, Publications 2013). Arnold, C. W., F. K. Tovesson, K. C. Meierbachtol, A. B. Laptev, T. A. Bredeweg, M. Jandel, R. O. Nelson, and Tovesson, F., C. Arnold, A. Laptev, K. Meierbachtol, M. M. C. White . SPIDER: A Path to 1 AMU Resolution White, R. Blakeley, A. Hecht, D. Mader, and L. Snyder. of Neutron-Induced Fission Fragments. Presented at SPIDER: A New Instrument for Fission Fragment Yield Division of Nuclear Physics of the American Physical Measurements. To appear in International Conference Society,. (Newport Beach, CA, 24-27 Oct. 2012). on Nuclear Data for Science and Technology. (New York, March 4-8, 2013). Jandel, M.. Neutron-Induced Fission Measurements Using DANCE and LSDS Facilities at LANL. Presented at Dynamical Aspects of Nuclear Fission (DANF2011). (Smolenice, Slovakia, October 17, 2011). Meierbachtol, K. C.. LDRD Review - SPIDER Detector Ionization Chamber Component. 2012. LA- UR-12-25207 Presentation to LDRD Review October 4, 2012. Meierbachtol, K. C., F. K. Tovesson , C. W. Arnold , A. B. Laptev , T. A. Bredeweg , M. Jandel , R. O. Nelson , and M. C. White . A High Resolution Ionization Chamber for the SPIDER Fission Fragment Detector. Presented at 5th International Conference on Fission and Properties of Neutron-Rich Nuclei (ICFN5) [LA-UR-12-23824]. (Sanibel Island, FL, 04-10 Nov. 2012). Meierbachtol, K., C. Arnold , A. Laptev, F. Tovesson, M. White, R. Blakeley , A. Hecht, and D. Mader. Development of an Ionization Chamber for the SPIDER Fission Fragment Detector. To appear in International Conference on Nuclear Data for Science and Technology. (New York, March 4, 2013). Moller, P., J. Randrup, and A. J. Sierk. Calculated Fission Yields of Neutron-Deficient Mercury Isotopes. 2012. Physical Review C (Nuclear Physics). 85 (2): 024306 (6 pp.). Sierk, A. J., J. P. Lestone, and P. Moller. Calculating Fission- Fragment Distributions. Invited presentation at 2012 Tri-Lab Nuclear Data Workshop. (Los Alamos, NM, 05- 07 Mar. 2012). Tovesson, F. K.. LDRD Review - SPIDER Detector Instrument Overview. 2012. LA-UR-12-25411 Presentation to LDRD Review October 4, 2012. Tovesson, F. K., T. A. Bredeweg, A. B. Laptev, K. C. Meierbachtol, C. W. Arnold, M. C. White, A. Hecht, D. Mader, and R. Blakeley. SPIDER: A New Instrument For Fission Yield Measurement. Presented at 5th International Conference on Fission and Properties of Neutron-Rich Nuclei (ICFN5) [LA-UR-12-23883]. (Sanibel Island, FL, 04-10 Nov. 2012). 565

Page 566

Physics Directed Research Continuing Project Physics Beyond the Standard Model with the Long-baseline Neutrino Experiment Christopher M. Mauger 20120101DR Introduction theoretical team at Los Alamos to tackle the large scope The goal of this project is to enhance the physics reach of BSM physics. This team will be unique within the of the Long-Baseline Neutrino Experiment (LBNE) by LBNE collaboration and cement LANL’s leadership role developing an integrated experimental, theoretical, in both the technology and the physics of LBNE. It will modeling, and simulation capability at LANL that makes simultaneously secure LANL’s leadership in large-scale possible science beyond the LBNE baseline. We drawn computing for nuclear and neutrino physics. Detector on our extensive expertise in neutrino oscillations, research at the cutting edge builds capability for national nucleon decay, supernovae explosions, advanced detec- security nuclear sensing. tor development, and rare event searches. LANL also has Progress a strong experimental history in near detector technol- ogy and large Cherenkov detectors, and we will field an On the experimental side, we have done a detailed de- integrated team capable of addressing the full range sign of the liquid argon detector and begun the procure- of physics available to LBNE. In this project, we explore ment process. The two key elements are the cryostat novel methods at the near site to make higher precision and the electronics. The cryostat is the largest expense measurements of the neutrino flux and cross sections and constitutes the box in which all of our experimental than are possible with previous approaches. We will con- activity will have to be carried out. Extensive simula- struct a liquid argon time-projection chamber and carry tions were carried out to assure all scientific goals will be out measurements associated with improving LBNE’s CP met. The electronics are custom devices that required violation sensitivity and supernova neutrino program. extensive evaluation. Most are now in hand with the rest on order. The time-projections chamber is now fully In addition, we carry out a theoretical program to designed. Additionally, significant scientific progress has enhance our understanding of neutrino-nucleus interac- been made on the planning for neutron beam running at tions. We optimize the performance of the far detector LANSCE (Los Alamos Neutron Science Center). We now through a detailed simulation program relating calibra- have two proposals in for engineering runs for the au- tion, light propagation and event signatures of both tumn run period at WNR (Weapons Neutron Research), signal and background processes. Our theoretical studies the high-energy neutron beamlines of LANSCE. Also of proton decay and supernova neutrino physics will also during this year, we have built an extensive and strong enhance the science that can be done with LBNE. collaboration including many top institutions from out- side of LANL. Benefit to National Security Missions This project sits squarely in the LANL BSM Grand Chal- Regarding neutrino flux measurements, we have taken lenge and strongly enhances the growing LANL-Fermilab data via the NA-61 experiment at CERN and begun the strategic partnership. It enhances core capabilities of analysis. Due to a custom magnet configuration, this re- laboratory strength in several groups across two divi- quired thorough magnetic field simulations and study of sions. LANL has project, base and Early Career Award the trigger configuration. Both of these are completed. funding from DOE HEP to develop a conceptual design The analysis will yield proton-carbon cross-sections with for a traditional near detector and project funding to unprecedented sensitivity in one of the many regions lead the calibration effort of the far detector, but we lack of hadron production phase space important for un- a program to develop the physics reach of LBNE. With derstanding neutrino production in the Long-Baseline this proposal we build an integrated experimental and Neutrino Experiment (LBNE). These data will be ana- 566

Page 567

lyzed and planning for future experiments carried out. The Neutrino flux-related measurements at NA-61 experiment current work underway is an extensive calibration of the at CERN will continue. data from several detector systems. Conclusion In this time, extensive theoretical work has been carried We will produce and publish new calculations for neutrino out including advances in the study of supernova neutri- scattering on the deuteron, carbon, and oxygen, col- nos, non-standard interactions (NSI), and neutrino scatter- lect and publish data on novel H2O and D2O targets that ing physics. include cross-section and absolute flux measurements, and produce and optimize the plan for analyzing the fun- During the past year the theory effort has concentrated damental symmetries of nature with LBNE long-baseline on calculations of inclusive electron and neutrino cross neutrino data. Our calculations and simulations of nucleon sections from Carbon-12. A preprint on the 12C electron decay modes and signatures will help improve the far scattering has been prepared, and a calculation of the sum detector for LBNE and squeeze additional scientific results rules relevant to neutrino scattering are underway. We from the currently running experiment, SuperKamiokande. find a significant enhancement of the transverse response, Finally, our supernova neutrino studies will provide the similar to previous calculations and experiments in 4He path to extract fundamental physics from a galactic super- and also experimental results on 12C. Future calculations nova burst. will help determine the impact of these calculations on neutrino scattering. Publications Abazajian, K. N., E. Calabrese, A. Cooray, F. De Bernardis, S. We have also worked on signals of supernovae neutrinos in Dodelson, A. Friedland, G. M. Fuller, S. Hannestad, B. liquid argon detectors. We are studying signals of various G. Keating, E. V. Linder, C. Lunardini, A. Melchiorri, R. types of supernovae and in different epochs, now including Miquel, E. Pierpaoli, J. Pritchard, P. Serra, M. Takada, initial estimates of coherent neutrino-nucleus scattering. and Y. Y. Wong. Cosmological and astrophysical We are examining potential signals in a variety of scenarios neutrino mass measurements. 2011. ASTROPARTICLE with different detector designs, and trying to understand PHYSICS. 35 (4): 177. the impact of eventual siting decisions on the LBNE far detector. Akin, T., and A. Friedland. The 2010 Interim Report of the Long-Baseline Neutrino Experiment Future Work Collaboration Physics Working Groups. . 2011. e-Print: arXiv:1110.6249 . In neutrino-nucleus scattering, we will work with the experimentalists to make the recent article on neutrino Cherry, J., J. Carlson, A. Friedland, G. Fuller, and nucleus scattering as valuable as possible. We will also A. Vlasenko. Neutrino scattering and flavor study continue to study uncertainties. The main work will transformation in supernovae. 2012. Physical Review be extend the present calculations of the deuteron to Letters. 108 (26): 261104. study light nuclei including 4He with the goal of eventually Mauger, C., G. Garvey, E. Guardincerri, D. Lee, Q. Liu, G. producing results on 12C. Mills, J. Mirabal-Martinez, J. Ramsey, K. Rielage, C. Sinnis, C. Taylor, R. Van de Water, and A. Yarritu. The In supernovae neutrinos, we will work to begin to imple- CAPTAIN Detector and Physics Program . 2013. arxiv. ment calculations that include neutrino inelastic scattering org . (scattering from electrons, nuclei, etc.) as well as coherent forward scattering from neutrino-neutrino and neutrino- Shen, G., L. E. Marcucci, J. Carlson, S. Gandolfi, and electron (MSW) effects. During the first year of the project R. Schiavilla. Inclusive neutrino scattering off the we identified this is a potentially very important issue for deuteron from threshold to GeV energies. 2012. supernovae neutrinos. We will continue to work with the Physical Review C (Nuclear Physics). 86 (3): 035503 (18 experimentalists on issues related to the observation of pp.). supernovae neutrinos. The focus of the experimental effort will be the construc- tion and commissioning of a liquid argon time-projection chamber. The detector will have 5 tons of instrumented mass. The instrumentation will be a three-plane wire chamber with cold analog electronics. A YAG laser will be employed for detector calibration studies. 567

Page 568

Physics Directed Research Continuing Project Disruptive Innovation in Numerical Hydrodynamics Jacob I. Waltz 20130005DR Introduction ies; greater detail in discovery-scale simulations; and an This project researches and develops innovative ap- enhanced ability to model realistic 3D features. Our work proaches for the numerical modeling of high-speed will also enable the use of commodity mesh generation material flows that leverage emerging computer archi- software, which could lead to cost savings of several tectures and are applicable to a wide range of scientific million dollars per year, and reduce problem setup time. and national security problems. The technologies under Lastly, our work will allow future compute platforms to consideration are designed to enhance fidelity and be used earlier and more effectively. computational efficiency relative to existing methods, Future Mission especially in the areas of vorticity, time accuracy, and The jump in simulation capability that results from our thermodynamic consistency. The specific focus is high- research will enable the solution of entirely new classes fidelity three-dimensional algorithms that have not been of problems and therefore has the potential to signifi- previously applied to LANL and NNSA problems but have cantly expand the scope of the Laboratory’s simulation significant potential should they be successfully devel- tools. New application areas might include design of oped. blast mitigation structures for urban environments, In addition to the above this project is investigating ap- energetic disablement calculations of Improvised Explo- proaches for emerging computer architectures such as sive Devices, anti-personnel and anti-structural analysis, the Graphics Processing Unit (GPU). Effective use of this and high-resolution studies of mix and ignition in Inertial computer hardware requires a large shift in both the Confinement Fusion targets. design of numerical algorithms and the use of simula- Progress tion tools. The research performed under this project is intended to address both issues and thereby create a In the past year the project has made significant techni- path forward for emerging architectures and Exascale cal progress in the following areas: concepts, for example by emphasizing accuracy over Hydrodynamic algorithms memory. A new class of hydrodynamic algorithm has been devel- The development of transformative numerical algo- oped that significantly improves accuracy and fidelity rithms that are simultaneously optimized for new particularly in the areas of vorticity and shock com- computer architectures carries significant technical risk, pression. This algorithm is in fact the first successful but if successful will enable a dramatic leap forward in algorithm of its type to have ever been developed and predictive simulation capability. therefore represents a significant research advance. Initial results with the algorithm are to be presented at Benefit to National Security Missions a domestic conference in September 2013 (MULTI-MAT We expect our research to lead to orders of magnitude 2013), and a journal article describing the algorithm is in improvements in fidelity and computational efficiency preparation. for work related to NNSA Defense Programs, Non- Advanced architectures: The project has completed ini- proliferation, and Science Campaigns. The impacts of tial testing and implementation on Graphics Processing these improvements will include faster responses to Units (GPUs). The project also has performed research programmatic questions; increased population sizes for on effective approaches for threading. The purpose of Uncertainty Quantification and other sensitivity stud- 568

Page 569

these studies was to inform research and development Canfield, T. R., N. R. Morgan, L. D. Risinger, J. Waltz, and choices for work that will take place in fiscal year 2014. J. G. Wohlbier. Manufactured solutions for the three- Journal articles on these studies are in preparation. dimensional Euler equations with relevance to inertial confinement fusion. Presented at American Society Verification of Mechanical Engineers Verification and Validation The project has completed verification testing of an exist- Symposium. (Las Vegas, NV, 22-24 May 2013). ing hydrodynamic algorithm which was considered a Morgan, N. R., J. Waltz, D. E. Burton, T. R. Canfield, L. risk mitigation for the new algorithm but will serve as a D. Risinger, J. G. Wohlbier, and M. R. J. Charest. A complementary algorithm otherwise. Two journal articles Godunov-like point-centered ALE finite element were prepared. The first documented testing results on hydrodynamic approach. Presented at Conference on existing problems and was published in July (J. Waltz et al, Numerical Methods for Multi-Material Fluid Flow. (San `Verification of a three-dimensional unstructured finite ele- Francisco, CA, 2-6 Sep. 2013). ment method using analytic and manufactured solutions’, Waltz, J.. GPU and multicore acceleration of a 3D Computers & Fluids 81, 57-67 (2013)). The second article unstructured mesh Eulerian-AMR hydrocode. presented a novel set of test problems for the equations of Presented at Nuclear Explosives Code Developers fluid dynamics with specific relevance for Inertial Confine- Conference. (Livermore, CA, 22-26 Oct. 2013). ment Fusion. This article has been submitted to the Journal of Computational Physics and is under review. These test Waltz, J., N. R. Morgan, T. R. Canfield, M. R. J. Charest, L. problems also were presented at a domestic conference D. Risinger, and J. G. Wohlbier. A three-dimensional (ASME V&V symposium) in May 2013. finite element arbitrary Lagrangian-Eulerian method for shock hydrodynamics on unstructured grids. 2013. Future Work Submitted to Computers & Fluids. The major 2014 milestone for this project is to demon- Waltz, J., T. R. Canfield, M. R. J. Charest, N. R. Morgan, L. D. strate multi-material simulations with a new hydrodynamic Risinger, and J. G. Wohlbier. Operator splitting and time algorithm, including convergence analysis, accuracy mea- accuracy in Lagrange plus remap methods. Presented surements, comparison to experiment, and timing studies. at Conference on Numerical Methods for Multi- Material Fluid Flow. (San Francisco, CA, 2-6 Sep. 2013). Additional tasks are to demonstrate and characterize initial performance on Many-Integrated-Core architectures, and Waltz, J., T. R. Canfield, N. R. Morgan, L. D. Risinger, and to demonstrate and characterize initial distributed memory J. G. Wohlbier. Verification of a three-dimensional performance. unstructured finite element method using analytic and manufactured solutions. 2013. COMPUTERS & FLUIDS. Each task will be documented with journal papers, internal 81: 57. reports, and/or presentations at technical conferences. Waltz, J., T. R. Canfield, N. R. Morgan, L. D. Risinger, and J. G. Wohlbier. Manufactured solutions for the three- Conclusion dimensional Euler equations with relevance to inertial The goal of this project is to develop innovative approach- confinement fusion. 2013. Submitted to Journal of es for the numerical modeling of high-speed material flows Computational Physics. that leverage emerging computer architectures and will form the basis of next-generation simulation tools. The Wohlbier, J. G., L. D. Risinger, T. R. Canfield, M. R. J. Charest, impacts of this work will include significant enhancements N. R. Morgan, and J. Waltz. Programming for modern in simulation detail and accuracy; reduced uncertainty in architectures in the CHICOMA hydrocode. Presented at Conference on Numerical Methods for Multi-Material simulation-based responses to programmatic issues; ad- Fluid Flow. (San Francisco, CA, 2-6 Sep. 2013). vances in scientific understanding in the broader numerical modeling community; and increased diversity in the suite of methods that are trusted for LANL and NNSA problems. Publications Canfield, T. R., M. R. J. Charest, N. R. Morgan, L. D. Risinger, J. Waltz, and J. G. Wohlbier. Simulation of multi-material flows using a finite element Riemann solver and adaptive unstructured grids. Presented at Conference on Numerical Methods for Multi-Material Flows. (San Francisco, CA, 2-6 Sep. 2013). 569

Page 570

Physics Directed Research Continuing Project Illuminating the Origin of the Nucleon Spin Ivan M. Vitev 20130019DR Introduction ing upon our existing strategic partnership with Fer- Nucleons (protons and neutrons) are not fundamen- milab (E906, MiniBoone and LBNE), this project will tal constituents of matter, but are instead made up strengthen our fundamental science capabilities, bring of quarks and gluons, the elementary particles of the new high-luminosity polarized target technology to strong interaction. There is compelling experimental LANL and provide a “major physics thrust to follow cur- evidence that the sum of the quark and gluon intrinsic rent commitments to RHIC”. This project will maintain angular momenta only contributes ~1/3 of the total pro- LANL’s leadership position in the field of spin physics and ton spin. Thus, the majority of the proton spin is unac- produce the world’s most accurate polarized Drell-Yan counted for, which has been referred to as the “proton measure-ment in proton-proton reactions. Our project spin crisis”. The missing fraction of the spin is likely car- is a timely and direct response to the DOE Mile-stone ried by the orbital angular momentum of the quarks and (HP13) to “test unique QCD predictions for relations gluons. The long-term goal of this project is to develop between single-spin phenomena in p-p scattering and the experimental capability to measure the spin of the those observed in deep-inelastic scattering”. We antici- proton in terms of contributions from the spins of the pate that this LDRD project will result in DOE Office of quarks, gluons and their orbital angular momentum. Science funding for a LANL-lead spin physics program An equally important goal is to understand the relative at Fermilab. Our integrated experimental and theoreti- significance of these spin contributions in the theory of cal program will allow us to lead in a major advance in strong interactions, Quantum Chromodynamics (QCD), understanding the polarized nucleon structure through and how they manifest them-selves in reactions with po- the only U.S.-based dedicated polarized DY experiment. larized proton beams and/or targets. To this end, we will Providing a polarized target to FNAL will greatly enhance determine the momentum distribution of quarks inside Fermilab’s capabilities and provide a much needed user the proton, transverse to the proton momentum, from facility for spin physics. Our target will also be able to which one can deduce whether quark orbital motion polarize ND3, thus enabling one to extend the spin phys- contributes significantly to the proton spin. ics to polarized neutrons. Furthermore, developing and testing particle detector technology at high luminosity To determine the distribution of quarks and gluons will directly benefit MaRIE, a LANL institutional prior- within a nucleon, we will carry out the world’s first ity. This detector development is also of fundamental measurement of the production of two simultaneous importance to nuclear detection for applied missions like leptons (electrons or muons) from a polarized proton nuclear nonproliferation. Such detectors often operate in target bombarded by a high-energy proton beam. From a high background environment. a detailed analysis of the azimuthal distribution of such di-leptons, one can deduce properties of the polarized Progress nucleon structure. In particular, we will measure both In the first fiscal year of this project, we carried out a the sign and magnitude of the quark Sivers distribution, feasibility study to reduce risk for FY14-16, in which the which is expected to be zero if the quarks have no orbital project will become a full-scale DR. The conclusions of angular momentum. this feasibility study will be validated by an independent peer review scheduled for Aug. 7. Benefit to National Security Missions This work is central to the FY13, FY14 LDRD Strategic Results on the polarized magnet: During the first week Investment Plan of Nuclear and Particle Futures. Build- of January 2013, the LANL polarized target magnet was 570

Page 571

crated and shipped to the University of Virginia for test- transverse momentum distributions (TMDs). A postdoctor- ing. This superconducting magnet has not been used since al search was conducted and an offer made and accepted. 1992, and one element of risk reduction was to demon- Dr. Boram Yoon will join LANL on August 19, 2013. He is a strate that the magnet is still functional. Kun Liu and Andi young expert in lattice QCD. The collaboration has applied Klein flew out to UVa on 1/27 to perform several initial for time at Ntional centers through USQCD and received tests, including bringing the magnet up to full field (5.1 T.) 3.83 million core hours for computing time. We started In the first test, we performed a leak check of the whole tuning and tests of the multigrid method for calculating magnet. Then, the magnet was first cooled with liquid ni- the quark propagator. We now have the manpower and trogen, verifying that the thermal insulation of the magnet computing time in place to evaluate the contribution of was intact. Next, the superconducting coils were success- disconnected diagrams to the TMDs. In perturbative QCD, fully cooled with liquid helium. Finally, the magnet current we have studied the next-to-leading order pQCD cor- was slowly ramped up over 3 hours to achieve full field. rections to the transverse momentum-weighted Sivers We reached the design field of 5.1 T without a quench on asymmetry for semi-inclusive hadron production in lepton- Feb 5th. The magnet was left at full field for 24 hours. After proton deep inelastic scattering (Zhonbo Kang, Ivan Vitev, 24 hours, the magnet was de-energized in order to per- Hongxi Xing). This is an important step toward the accurate form a stringent final check. For this test, we chose a fast extraction of the internal structure of the nucleon. Our current ramp, such that the maximum field was reached paper was published Phys. Rev. D 87, 034024 (2013) and in only 30 minutes. Again, the magnet behaved flawlessly. presented at the QCD Evolution Workshop, 2013. We have The magnet is at Oxford Instruments for superconducting begun the analysis of the quark Sivers distribution (one coil re-orientation. TMD that is central to the goals of this project), using the world’s polarized SIDIS data (Miguel Echevarria, Zhongbo Results on the Letter of Intent: We have updated our phys- Kang, Ivan Vitev). ics performance plots, based on NLO Drell-Yan rates and realistic detector acceptance. The Drell-Yan cross section Future Work was taken from PYTHIA using the CTEQ5M parton distri- Experiment: In the next FY we will design and build a new bution functions and verified against a modern NLO DY Nuclear Magnetic Resonance system for the polarized tar- calculation from Vitev, et.al. Andi Klein, Patrick McGaughey get. Furthermore, we will setup a lab and test area in Area and Xiaodong Jiang worked on the (LOI). We showed that A of LANSCE, where we will setup and eventually test the approximately 110,000 reconstructed Drell-Yan pairs can magnet, when it comes back from Oxford. In collaboration be collected per year, after applying geometry cuts similar with University of Virginia, we will modify the old refrigera- to that of E906. A strong sensitivity to the sign and mag- tor in such a way that it will fit into the new geometry, re- nitude of the Sivers asymmetry was demonstrated for furbish the sensors and He lines,test it at UVa and ship it to non-zero values. The magnitude of the Sivers function can LANL. Once it is at LANL, the refrigerator will be mounted be determined to better than 4%. An international collabo- in the magnet. ration for experiment E1039 with more than 40 members was formed with LANL being the lead and Andi Klein and Theory: In the next FY we will begin the investigation of the Xiaodong Jiang co-spokespersons. Member institutions in- Sivers function evolution to leading logarithmic accuracy. clude ANL, FNAL, U. Illinois, Academia Scinica Taiwan, KEK We will also make progress in understanding the rapidity Japan, Ling-Tung U. Taiwan, LANL, U. Maryland, U. Michi- evolution technique. We will develop a global fitting pack- gan, U. New Hampshire, National Kaohuisung U. Taiwan, age for the Sivers function in semi-inclusive deep inelastic RIKEN Japan, Rutgers U., JLab, Tokyo Institute of Technolo- scattering. We will extract the Sivers function and publish gy Japan, U. Virginia. The Letter of Intent was presented to results. In Soft Collinear Effective Theory we will begin the the Fermilab Physics Advisory Committe on June 5th, 2013. application of the Rapidity Renormalization Group Tech- We were given 30 min for presentation, which we consider nique to the polarized case, Drell-Yan in particular. We will quite encouraging. The PAC enthusiastically embraced our initiate the derivation of the energy evolution equation for project and the Fermilab Director gave our project E1039 comparison to the ones obtained with the pQCD tech- Stage 1 approval. nique. In Lattice QCD we will incorporate a new multi-grid evaluation method in the suite of lattice codes. We will set Results on theory: In lattice QCD, an international col- up the production runs for disconnected diagrams at JLab laboration has been formed between LANL (Rajan Gupta, and begin their evaluation. We will begin work on under- Tanmoy Bhattacharya, Boram Yoon), MIT (John Negele), standing the renormalization of lattice operators. NM State U. (Michael Engelhardt), Jefferson Lab (Alexei Prokudin) and U. Regensburg (Andreas Schafer) to study 571

Page 572

Conclusion Through a synergy between theory and experiment, we will make a fundamental advance in our understanding of the origin of the nucleon spin. In a strategic partnership with Fermilab, we will carry out the world’s most accurate measurement of di-lepton production from a polarized proton target bombarded by a high-energy proton beam. We will develop state-of-the-art theoretical and computa- tional tools necessary to interpret the experimental results that will directly lead to a major breakthrough in our understanding of the structure of matter and the theory of strong interactions. Publications Kang, Z.. Single transverse spin asymmetries in polarized SIDIS and pp scattering. Invited presentation at The 5th Workshop of the APS Topical Group on Hadronic Physics. (Denver, 10-12 April 2013). Kang, Z. B., I. Vitev, and H. X. Xing. Transverse momentum- weighted Sivers asymmetry in semi-inclusive deep inelastic scattering at next-to-leading order. 2013. PHYSICAL REVIEW D. 87 (3): -. Klein, A., P. McGaughey, and I. Vitev. Letter of Intent for a Drell-Yan experiment with a polarized proton target. 2013. Fermilab PAC . Vitev, I.. Transverse momentum-wighted Sivers asymmetry in semi-inclusive deep inelastic scattering at next-to- leading order . To appear in QCD Evolution Workshop 2013. (Newport News, 14-17 May 2013). 572

Page 573

Physics Directed Research Continuing Project Peta-scale Studies of Cosmic Explosions and Supernova Shock Breakout with Palomar Transient Factory Przemyslaw R. Wozniak 20130030DR Introduction Benefit to National Security Missions The next generation Palomar Transient Factory (PTF-II) is This project will build new Astro-informatics competency poised to revolutionize the field of time-domain astro- to address growing national needs in Space Situational physics with great potential for future discovery, espe- Awareness (SSA), develop new capabilities for the Weap- cially in the area of explosive phenomena. This project ons Program, and test them using actual observations. will conduct the first systematic search for supernova We will combine observation, theory, and computation shock breakout events and other explosive transients on to explore the physics of cosmic explosions and con- time-scales of less than a day. In order to succeed, PTF-II tribute to a high-impact time-domain sky survey. This must automatically identify actionable information from project will position us for a future role in the Large Syn- the torrent of imaging data, classify emerging events, optic Survey Telescope (LSST) survey and will open new and optimize the follow-up strategy. To address this chal- opportunities for a major involvement in NASA missions lenge, we will develop an autonomous event broker that and dark energy investigations identified in the strategic integrates cutting edge machine learning algorithms with plan of the DOE Office of Science. We will develop novel high performance computing infrastructure. Our classifi- algorithms for classification and tracking of celestial ob- cation engine will be based on Bayesian belief networks jects relevant to Space Situational Awareness programs and ensembles of decision trees. Follow-up optimization aimed at sponsorship by NASA, Department of Defense, will employ mutual entropy minimization and dynamic and other WFO organizations. Our models of supernova scheduling of available instruments. shock breakout and other explosive transients will test Advanced Simulation and Computing (ASC) capabili- A major innovation of PTF-II is Spectral Energy Distribu- ties in the critically important regime at the transition tion Machine (SEDM), a low-resolution spectrograph between diffusion and free streaming using both flux that will enable rapid spectral classification of all inter- diffusion and implicit Monte Carlo codes. Ultimately, our esting transients detected by the photometric pipeline. nuclear nonproliferation efforts depend on satellite as- Our event broker will identify “young” supernovae and sets that are increasingly vulnerable, driving the need for enable rapid follow-up observations with high-energy Space Situational Awareness. observatories in time to get a glimpse of the shock breakout. Progress The work on transient classification and broker infra- We will also carry out extensive radiation hydrodynamics structure development is progressing according to our calculations of transient emission from hot shocks over a original schedule. For our pixel level classification we wide range of astrophysical scenarios to interpret PTF-II constructed a large training data set with 78000 samples observations, inform transient classifiers, and test the with known REAL/BOGUS labels based on the data from performance of simulation codes in a critically important the PTF survey. Each sample is a vector of 40 features range of physical conditions. Our goals is to build and with additional 3*441 = 1323 features taken directly publish a suite of ~100 explosion models with time-re- from new, reference, and subtracted images delivered by solved spectra and synthetic light curves at ~100 epochs. the image differencing pipeline. We evaluated the utility A detailed comparison of these models with observed of Bayesian networks to classify this data using both light curves will provide valuable constraints on the structure and parameter learning. While the machine structure of the supernova progenitor stars and lead to a learned networks capture essential relationships con- better understanding of stellar deaths. sidered by domain experts, their performance does not 573

Page 574

match that of state of the art algorithms based on decision winds and shell-like outbursts). We launched a web server trees. We obtained a much more reliable pixel level classifi- and a public database supported under the TurboGears cation using random forests, i.e. randomized collections of web framework that will export our models to research- decision trees and voting. We also introduced a novel ap- ers at LANL and the public using a sophisticated graphical proach that draws on recent advances in computer vision. interface. Using sparse representations with learned dictionaries As part of our data intensive computing effort we imple- (over-complete bases) we compute additional image based mented a high-performance database for storing time-re- features. The addition of machine learned features to the solved imaging data based on advanced spherical indexing baseline set of context features improves the figure of (Healpix). The database supports a wide range of spatial merit by a factor of 2x down to about 4.5% missed detec- scales and pixel sizes suitable for a variety of telescopes tion rate at 1% false positive rate. and surveys. The system has been tested using ~1TB of im- We implemented a software package that performs this ages from RAPTOR telescopes. classification in real time using data from an online data- base. The current version of the software (RB3.x) has been Future Work installed at the NERSC super-computing center that hosts An iterative development schedule will be adopted to the iPTF database and is being configured for deployment. synchronize the project with the timetable for the Inter- We launched a successful collaboration with UC Berkeley mediate Palomar Transient Factory (iPTF). Our goal is to and JPL to further develop this system. complete the second iteration of the event broker featur- ing initial versions of algorithms in the summer of 2014 Our spectral classification effort focused on collecting suit- when Spectral Energy Distribution (SED) Machine will be able training data. We completed a preliminary classifica- operating on the Palomar 60-inch telescope. We will fur- tion run using Support Vector Machines (SVM) applied to ther develop the data management infrastructure required several thousand time-resolved medium-resolution spectra to support a variety of automated machine learning clas- of ~450 supernovae from the SNID database. The results sifiers and follow-up optimization algorithms that will be are very encouraging with error rates below ~3% on data gradually integrated with the broker. The work on machine with high signal-to-noise ratio. learning classification will focus on high level classification of astrophysical transients using survey data and any avail- We began the work on event broker infrastructure. The able external context data from cross-correlation analysis. prototype broker is implemented as a set of network An important objective is to implement reliable selection servers communicating over TCP/IP sockets. Each stage of very young supernovae using low-resolution spectra of processing is encapsulated in a job queue that takes its from SED Machine. We will continue to develop automated input from the previous stage and submits jobs to the next follow-up optimization algorithms for the event broker and one. We also implemented a basic event simulator that implement first generation follow-up marshal toolbox to generates a stream of measurements from a simulated support PTF operations. sky survey and submits this data to the broker using XML based VOEvent messages (a Virtual Observatory standard Theory and simulation effort will focus on running more adopted by the community). Currently the broker performs supernova simulations that incorporate new atomic phys- basic spatio-temporal cross-correlation of measurements ics data and modelling capability. We will simulate and with multi-wavelength catalogs and context gathering us- interpret existing PTF observations and generate light ing several external databases, some of which are repli- curves for available explosion models using our SPECTRUM cated locally within the broker. tool chain. We will upload the first batch of supernova models into the online database that we developed and Since February 2013 the iPTF survey is back on sky and dis- extend the web interface to this public archive to begin covered more than 150 supernovae that account for most serving synthetic light curves of explosive transients. These of the world supply. Using RAPTOR telescopes at LANL we light curves will be used to improve supernova selection collected multi-color follow-up observations of a nearby algorithms. type II SN 13aaz discovered by iPTF in M65. Data intensive computing work will proceed in parallel. We The light-curve simulation program is modeling a range of will continue to integrate additional hierarchical partition- transient outbursts ranging from the “kilonova” produced ing and indexing schemes in spherical geometry (Hierarchi- in the merger of two neutron stars in a binary to standard cal Triangular Mesh, pgSphere) with the existing frame- models for type II supernovae. One of our key survey work for parallel hashed tree data structures. We will also suites will focus on the role of circumstellar material (both create a new training data set using iPTF data collected 574

Page 575

since February 2013 to support algorithm development Palaversa, L., Z. Ivezic, L. Eyer, D. Ruzdjak, D. Sudar, M. and testing. Galin, A. Kroflin, M. Mesaric, P. Munk, D. Vrbanec, H. Bozic, S. Loebman, B. Sesar, L. Rimoldini, N. Hunt- Conclusion Walker, J. VanderPlas, D. Westman, S. J. Stuart, A. Becker, G. Srdoc, P. Wozniak, and H. Olusei. Exploring Time-domain astrophysics and the science of cosmic ex- the Variable Sky with LINEAR. III. Classification of plosions are at the forefront of modern astrophysics. The Periodic Light Curves. 2013. The Astronomical Journal. Palomar Transient Factory (PTF) has emerged as the world 146: 101. leader in the exploration of cosmic explosions and the most successful precursor to multi-petabyte sky surveys Sesar, B., Z. Ivezic, S. J. Stuart, D. M. Morgan, A. Becker, of the future such as the Large Synoptic Survey Telescope S. Sharma, L. Palaversa, M. Juric, P. Wozniak, and H. (LSST). The main goal of this project is to develop informa- Olusei. Exploring the Variable Sky with LINEAR. II. Halo tion technology and data intensive computing infrastruc- Structure and Substructure Traced by RR Lyrae Stars to 30 kpc. 2013. The Astronomical Journal. 146: 21. ture for the next generation PTF and open the domain of short time-scales (less than 1 day) for exploration by deep Vreeswijk, P., A. Raassen, A. Smette, A. De Cia, P. Wozniak, wide-area surveys. A. Fox, W. Vestrand, and P. Jakobsson. Time-dependent excitation and ionization modelling of absorption-line Publications variability due to GRB 080310. 2013. Astronomy and Barth, A. J., A. Voevodkin, D. J. Carson, and P. Wozniak. A Astrophysics. 549: A22. Search for Optical Variability of Type 2 Quasars in SDSS Stripe 82. To appear in The Astrophysical Journal. Whalen, D., J. Johnson, J. Smidt, A. Heger, W. Even, and C. Fryer. The Biggest Explosions in the Universe. II. To Frey, L., C. Fryer, and P. Young. Can Stellar Mixing Explain appear in The Astrophysical Journal. the Lack of Type Ib Supernovae in Long-duration Gamma-Ray Bursts?. 2013. The Astrophysical Journal Whalen, D., J. Johnson, J. Smidt, A. Meiksin, A. Heger, W. Letters. 773: L7. Even, and C. Fryer. The Supernova that Destroyed a Protogalaxy: Prompt Chemical Enrichment and Fryer, C.. Compact Object Formation and the Supernova Supermassive Black Hole Growth. 2013. The Explosion Engine. To appear in Classical and Quantum Astrophysical Journal. 774: 64. Gravity. Whalen, D., W. Even, C. Lovekin, C. Fryer, M. Stiavelli, P. Jin, Z. P., S. Covino, M. Della Valle, P. Ferrero, D. Fugazza, D. Roming, J. Cooke, T. Pritchard, D. Holz, and C. Knight. Malesani, A. Melandri, E. Pian, R. Salvaterra, D. Bersier, Illuminating the Primeval Universe with Type IIn S. Campana, Z. Cano, A. J. Castro-Tirado, P. D’Avanzo, Supernovae. 2013. The Astrophysical Journal. 768: 195. J. P. U. Fynbo, A. Gomboc, J. Gorosabel, C. Guidorzi, J. B. Haislip, J. Hjorth, S. Kobayashi, A. P. LaCluyze, G. Wozniak, P. R., D. I. Moody, Z. Ji, S. P. Brumby, H. Brink, Marconi, P. A. Mazzali, C. G. Mundell, S. Piranomonte, J. Richards, and J. S. Bloom. Automated Variability D. E. Reichart, R. Sanchez-Ramirez, R. J. Smith, I. A. Selection in Time-domain Imaging Surveys Using Steele, G. Tagliaferri, N. R. Tanvir, S. Valenti, S. D. Sparse Representations with Learned Dictionaries. Vergani, T. Vestrand, E. S. Walker, and P. Wozniak. GRB 2013. In American Astronomical Society, AAS Meeting 081007 and GRB 090424: The Surrounding Medium, #221. (Long Beach, CA, 6-10 January 2013). , p. Outflows, and Supernovae. 2013. The Astrophysical #431.05. Washington, DC: American Astronomical Journal. 774: 114. Society. Johnson, J., D. Whalen, W. Even, C. Fryer, A. Heger, J. Smidt, and K. J. Chen. The Biggest Explosions in the Universe. 2013. The Astrophysical Journal. 775: 107. Ofek, E., A. Zoglauer, S. Boggs, N. Barriere, S. Reynolds, C. Fryer, F. Harrison, S. Cenko, S. Kulkarni, A. Gal-Yam, I. Arcavi, E. Bellm, J. Bloom, F. Christensen, W. Craig, W. Even, A. Filippenko, B. Grefenstette, C. Hailey, R. Laher, K. Madsen, E. Nakar, P. Nugent, D. Stern, M. Sullivan, J. Surace, and W. Zhang. SN 2010jl: Optical to hard X-ray observations reveal an explosion embedded in a ten solar mass cocoon. To appear in The Astrophysical Journal. 575

Page 576

Physics Directed Research Continuing Project High Performance Atom-Based Sensors for Fields and Rotations Malcolm G. Boshier 20130058DR Introduction which can help meet Intelligence and Defense commu- nity needs for improved sensors. The National Security community needs improved sen- sors for fields and rotations that advance the state of the Benefit to National Security Missions art in performance, size/weight/power requirements, and ruggedness. We will respond to this challenge by Improved rotation sensors, gravity sensors and magne- developing new atom-based sensors for magnetic fields, tometers are of broad interest to many agencies in the for gravity, and for rotations. Intelligence and Defense Communities. Device applica- tions include respectively inertial navigation, detection Our approach harnesses unique LANL atomic physics of underground structures and of oil or mineral deposits, capabilities - the “Painted Potential” for arbitrary ma- and the use of power lines for communication and for nipulation of Bose-Einstein condensate (BEC) matter characterization of activities in inaccessible facilities. waves, and high sensitivity atomic magnetometers - in conjunction with the Laboratory’s world-class quantum The devices based on quantum technologies that are the science expertise and its engineering capabilities. We focus of this project offer potential gains in Size, Weight, will use the Painted Potential system to create a trapped and Power (SWaP) and in sensor performance. Theory BEC atom interferometer to measure gravity and (with predicts that the Atom-SQUID we will build should have different programming and detection) a rotation sensor state of the art rotation sensitivity, in a physics package based on an “Atom-SQUID.” This new device is a matter which is much smaller than the Sagnac atom interferom- wave analog of the well-known superconducting SQUID eters which currently have the best performance. A BEC used for magnetic field sensing. waveguide atom interferometer would measure grav- ity precisely with a compact device that has no moving We will also advance the state of the art in magnetom- parts subject to wear. We will engineer atomic magne- etry by pushing atomic magnetometers to their funda- tometer technology into a portable device, taking atomic mental limits and by combining an atomic magnetom- magnetometry from the laboratory to the field. eter with the Painted Potential to form a high resolution high sensitivity magnetic microscope. We will develop Finally, while our focus in this project is on applications the theoretical description of these devices and also ex- to sensing, the BEC technologies that we are developing plore possibilities for quantum enhanced measurement, are also very relevant to quantum information process- using quantum correlations to obtain measurement ing and hence the Information Science and Technology precision which beats the standard quantum limit (shot Mission. noise). Most of the size/weight/power requirements in Progress these atom-based sensors are associated with optics and electronics, and so we will engineer those components The magnetometer engineering team’s task is to build to minimize their impact on the total sensor package. a version of the laboratory Atomic Magnetometer (AM) We will build a prototype atom magnetometer which lab instrument that can operate out in the field. Tasks can operate outside of the laboratory and beat the per- partially or fully completed to date include the hardware formance of the best commercial magnetometers by at circuit board and software interface to the laser diode least an order of magnitude. controllers, a Helmholtz coil assembly and associated current drivers and low sensitivity magnetic field sensor The outcome of this project will be a new capability all used to cancel the Earth’s magnetic field, the stand- 576

Page 577

alone power supply, the user interface, photodiode pre- these sub-clouds expand into each other in two arms of amplifier and digitizer, and the enclosure. An isolated test the interferometer. The question is whether the interfer- bed has been identified, and initial baseline measurements ence fringes that will presumably form within the two arms have been made to characterize the location with commer- will be correlated, with peaks and troughs of the interfer- cial sensors. Signal sources to be used in the comparisons ence patterns aligned. The team has also investigated have also been identified. the maximum speed permitted by the laws of physics for manipulating matter waves, and how these limits might be The laboratory magnetometer team focused on improving reached in the laboratory. Finally the theory team has also AM performance in both low and high frequency regimes. been examining the static properties of a single impurity A fiber-compatible polarization modulator is being investi- atom immersed in a BEC. We found that the state of the gated as a path to delivering low-frequency performance impurity atom can be extremely sensitive to the interac- closer to fundamental limits. We also analyzed the perfor- tions with the BEC, which are themselves strongly affected mance of typical high frequency magnetometers and found by magnetic fields. We also discovered a deep and un- significant discrepancies between measured performance expected connection to polaron physics: a cold atom BEC and the fundamental limits expected from theory. We are can provide the first host medium to self-localize polarons now working on understanding these discrepancies and both in a bubble state, analogous to electron bubbles in developing strategies to remove them. helium, as well as in a Landau-Pekar state, analogous to the self-localized electron in a dielectric crystal. We found The BEC team, having realized Josephson Junctions for a that the system can be characterized by only two dimen- BEC in a toroidal trap, a so-called “Atom SQUID”, began sionless coupling strengths and we showed that the system experiments to demonstrate rotation sensing with this smoothly crosses over between the two states in the cor- device. There are two possible methods for the sensing responding two-dimensional phase diagram. of rotation by Atom SQUIDs. One is measuring the maxi- mum velocity at which the BEC can flow around the torus Future Work without loss (the critical current) because that turns out The project is developing three types of sensor, all based to be a sinusoidal function of the rotation rate. The other on atomic physics technologies. Two exploit our capability examining which of the allowed quantized rotation states for manipulating Bose-Einstein condensates (BEC), a state are chosen by the BEC when it is created in the torus, of matter formed when atoms in a trapping potential (in because if the experiment is rotating higher rotational our case the Painted Potential) are cooled to temperatures states will be found more often. These two methods have very close to absolute zero. now been tested and proof of principle results obtained for both approaches. At the same time we have been First, we will continue investigating the application of developing a theory of the Atom SQUID based on the well cold atoms to rotation sensing. Here we form a BEC in a established theory of SQUIDs. This preliminary experimen- toroidal trap with two thin potential barriers positioned tal and theoretical work so far supports our proposal that a diametrically opposite each other. Simple theory and our compact highly sensitive rotation sensor may be built with FY13 experiments imply that the maximum matter wave Atom SQUIDs. current able to flow without friction in this “Atom-SQUID” device is a sensitive function of the speed of rotation. In The quantum theory team has been studying the origin FY14 we will perform detailed experiments to investigate of interference in Bose-Einstein condensates. This is an this property of the device. We will also continue devel- important issue for the project because the gravity sensor oping a proper quantum theory for the Atom-SQUID, and to be constructed will be read out using the interference begin to use it to devise optimum measurement strategies of matter waves. The usual experimental setup — with for the rotation measurement to be performed in FY14. two initially separated clouds of BEC — precludes the existence of a preexisting phase as each cloud contains a Second, in FY14 we will start investigating the performance fixed number of atoms (so, by Heisenberg’s indeterminacy of the Painted Potential as an atom interferometer to mea- relations, its phase must be completely undefined). Yet, as sure gravity. A single BEC will be divided into two pieces a pioneering experiment showed, interference fringes do which sit at slightly different heights above the Earth. Re- indeed form when the clouds are allowed to expand into leasing the two pieces of BEC allows them to overlap and each other. The standard explanation of this phenomenon form matter wave interference fringes which can be used attributes appearance of the interference pattern to the to measure gravity. In FY14 we plan to begin construction measurement carried out by photons that scatter from the of a second Painted Potential machine dedicated to these atoms. We are considering a version of the experiment rel- atom interferometry experiments. We will also begin evant to this project, where each cloud is split in two and 577

Page 578

related theoretical investigations. Third, we will continue development of atom magnetom- eters. The engineered atomic magnetometer instrument being designed in FY13 will be completed and tested in FY14. We will study the performance of laboratory atomic magnetometers with the goal of improving their perfor- mance towards the fundamental limits set by shot noise. Conclusion The overall goal of this project is to develop new atom- based sensors for magnetic fields, for gravity, and for rotations that advance the state of the art in performance, size/weight/power requirements, and ruggedness. We will move atomic magnetometers from the laboratory to the field, where they will advance applications ranging from to facility characterization via power line forensics to biophysics. We will start developing cold atom-based rota- tion sensors that would ultimately realize compact high- performance interial navigation systems. We will also build gravity sensors with the potential to out-perform current gravimeters, particularly in applications such as under- ground structure detection. Publications Blinova, A. A., M. G. Boshier, and E. Timmermans. Two polaron flavors of the Bose-Einstein condensate impurity. To appear in Physical Review A. Campo, A. del. Shortcuts to adiabaticity by counter- diabatic driving. 2013. Physical Review Letters. 111: 100502. Campo, A. del, I. L. Egusquiza, M. B. Plenio, and S. F. Huelga. Quantum Speed Limits in Open System Dynamics. 2013. Physical Review Letters. 110: 050403. Ryu, C., P. W. Blackburn, A. A. Blinova, and M. G. Boshier. Experimental realization of Josephson junctions for an atom SQUID. To appear in Physical Review Letters. Torrontegui, E., S. Ibanez, S. Martinez-Garaot, M. Modugno, A. del Campo, D. Guery-Odelin, A. Ruschhaupt, X. Chen, and J. G. Muga. Shortcuts to Adiabaticity. 2013. ADVANCES IN ATOMIC, MOLECULAR, AND OPTICAL PHYSICS, VOL 62. 62: 117. 578

Page 579

Physics Directed Research Continuing Project Quantum Chemistry, Information, Materials and Metrology Robert E. Ecke 20130727DR Introduction Benefit to National Security Missions The goals of this project are to develop collaborative Novel applications of quantum science underpin nu- scientific progress at the interfaces of quantum chemis- merous Laboratory efforts including quantum informa- try and materials, quantum information and communica- tion that can enable secure data communication, new tion, and quantum optics. In particular, we will explore approaches to quantum devices such as a topological the intersection of quantum information, communica- quantum memory, and improved algorithms for simu- tion and computing with new materials functionality lating scientific problems. Secure data communication such as chiral superconductors and topological insula- has implications for classified environments and for the tors and also with novel detection technology such as physical security of infrastructure systems such as the ultra-sensitive scanned probes. We will develop new electrical power grid. Novel concepts in manipulating computational approaches based on algorithms devel- quantum systems will have impact on Laboratory materi- oped in quantum information science to more efficiently als strategy including under-standing the properties of solve numerically quantum condensed matter models. actinides, in particular plutonium. Possible new quan- We will explore the fascinating overlap between cold tum sensors could impact GS programs. We strongly atom physics, with its pure systems and tunable proper- impact DOE Office of Science mission in the fundamental ties, and technologically important materials systems understanding of electronic and magnetic materials with such as superconductors and multiferroic materials. In applications in energy, electronics, and computing. The this latter category, the interplay between electrons and ability to support both a robust fundamental science the underlying field leads to exotic states of matter that capability and nurture new applications will continue to can provide novel properties or functionalities such as drive our work in the quantum arena. magnetic control of transport properties (anomalous Hall effect), strong vortex pinning (high critical currents) Progress in superconductors, and large magneto-electric effects During FY13, we made the following progress: in insulators. Through our theory and simulations we will identify classes of models and materials where this • Demonstrated that magnetic vortex crystals are in- new functionality can be realized. We will use hybrid duced by magnetic field in certain classes of frustrat- quantum chemistry and molecular mechanical methods ed Mott insulators. Such materials, i.e. spontaneous to incorporate solvent and thermal fluctuation prop- crystals of topological spin structures, are attracting erties of real molecular systems and apply this new an enormous interest because they can provide new phenomenology to carbon nanotubes and biological directions for future spin-electronic technology. light harvesting systems. We will investigate fluctuation We also showed how the magnetic excitations, i.e., induced quantum Casimir forces, and corresponding “magnons”, of the fully polarized state a a quantum classical analogs, for equilibrium and nonequilibrium antiferromagnet in a a strong magnetic field form a conditions. Our external collaborations will enable the diverse set of Bose-Einstein condensate states. use of opto-mechanical resonators for quantum commu- • Proposed a scheme using optimal control to achieve, nication, for fundamental studies of quantum friction, in laser cooledatomic gases, temperatures closer to and for atomically-resolved spectroscopic scanned probe absolute zero than those currently possible - Phys. measurements of condensed matter systems such as Rev. A 87, 043607 (2013). quantum phase transitions occurring in novel systems of competing magnetic and superconducting order. 579

Page 580

• Developed a theory for the effects of time-dependent materials; 3) Determine the role of long-lived electronic noise in thermally isolated systems - Phys. Rev. A 87, coherence in photosynthetic proteins that exhibit light- 043607 (2013). harvesting properties. • Predicted novel states of matter characterized by the Quantum Information and Communication coexistence of the integer quantum hall effect and spin We will 1) Develop applications of quantum channel capac- ice and described the presence of novel anyonic excita- ity for quantum communication applications that provide tions in metallic magnets. secure communication; 2) Explore the intersection of quantum information algorithms and condensed matter • Developed an exciton scattering (ES) approach for systems to better compute the properties of magnetic spin efficient calculations of optical spectra in branched systems; 3) Investigate utility of using topological entities conjugated macromolecules, as an extension of refer- for quantum computing applications such as computer ence quantum-chemical (QC) technique. We verified memory. he applicability of the ES approach using the natural atomic orbital (NAO) analysis on ab initio QC computa- AMO and Quantum Optics tions. Excellent agreement for the absorption spectra We will 1) Develop theory and simulations for fluctuation- between the ES approach and the direct quantum induced interactions in nanostructures, including Casimir chemistry was achieved - J. Phys. Chem. Lett. 2012, 3 forces, near-field heat transfer, patch effects, and quantum and other submitted papers. friction; 2) Explore the use of structuring for novel optical and acoustical phenomena in metamaterials, including • Began development of an extremely sensitive quantum sub-wavelength imaging, perfect absorbers; 3) Use cold magnetometer and applied it to the detection of ultra- atom methods as an ultra-sensitive magnetic field sensors. small magnetic fields originating from areas of only few square millimeters. The magnetometer is based Quantum Materials on detecting the dynamics of atomic spins confined in We will 1) Exploit new efficient algorithms for treating a vapor cell kept at close to room temperature. This quantum spins in classical fields to explore novel phases of research lays the foundation for future exploration of fermionic matter that could lead to new functional mate- cold atom magnetometers - Journal of Magnetic Reso- rials; 2) Explore the use of topological entities for imple- nance 231 (2013), p.39-45. menting quantum memory; 3) Investigate, using theory, simulation, and experiment, the properties of strongly cor- • Proposed a novel measurement approach that ex- related electron systems with novel magnetic, multi-ferroic tends the domain of applications of the experimental and superconducting properties technique (first developed at LANL) called spin noise spectroscopy - Appl. Phys. Lett. 102, 202405 (2013). In Conclusion related work, we found a new capability of spin noise Quantum science is an integrating centerpiece of numer- spectroscopy that allows characterization of semicon- ous LANL strategic directions including information science, ductor nanostructures. sensing, and materials. By supporting an interdisciplinary approach that ranges from fundamental science to applica- Future Work tions, our work develops research collaborations among Overall, we will work at the interdisciplinary intersections different quantum science areas, sustains and encourages of the fields of quantum science with an emphasis on broad scientific capability, and assists in identifying novel Quantum Chemistry, Quantum Information and Commu- and transformational science frontiers at the interface of nication, Atomic-Molecular-Optical (AMO) and Quantum national security and discovery science. Optics, and Quantum Materials. We will develop synergies among the topics but within each topic our set of goals is Publications as follows: Barros, K., and Y. Kato. Efficient Langevin Simulation of Coupled Classical Fields and Fermions. 2013. Arxiv. Quantum Chemistry We will 1) improve quantum chemical numerical ap- Behunin, R. O., D. A. Dalvit, R. S. Decca, and C. C. Speake. proaches to incorporate non-adiabatic processes, solvent Limits on the accuracy of isoelectronic gravity and thermal effects, and many atom simulations to better measurements at short separation due to patch potentials. 2013. Arxiv. model complex quantum chemical processes; 2) Model ultrafast relaxation processes in carbon nanotubes which Campo, A. del. Shortcuts to Adiabaticity by Counterdiabatic have important applications in electronics and optical Driving. 2013. Physical Review Letters. 111 (10): 580

Page 581

100502. Lin, S., and D. Roy. Role of kinetic inductance in transport properties of shunted superconducting nanowires. Campo, A. del, I. L. Egusquiza, M. B. Plenio, and S. F. 2013. Journal of Physics: Condensed Matter. 25 (32): Huelga. Quantum Speed Limits in Open System 325701. Dynamics. 2013. Physical Review Letters. 110 (5):
100503. Muniz, R., Y. Kato, and C. Batista. Generalized spin-wave theory: application to the bilinear-biquadratic model. Campo, A. del, J. Goold, and M. Paternostro. More bang for 2013. Arxiv. your buck: Towards super-adiabatic quantum engines.
100504. Arxiv. Pershin, Y., V. Slipko, D. Roy, and N. Sinitsyn. Two-beam spin noise spectroscopy. 2013. Applied Physics Letters. 102 Campo, A. del., T. W. Kibble, and W. H. Zurek. Causality (20): 202405. and non-equilibrium second-order phase transitions in inhomogeneous systems. 2013. Journal of Physics: Pyka, K., J. Keller, H. L. Partner, R. Nigmatullin, T. Condensed Matter. 25 (40): 325701. Burgermeister, D. M. Meier, K. Kuhlmann, A. Retzker, M. B. Plenio, W. H. Zurek, A. del Campo, and T. E. Chern, G., A. Rahmani, I. Martin, and C. Batista. Quantum Mehlstäubler. Topological defect formation and Hall ice. 2012. Arxiv. spontaneous symmetry breaking in ion Coulomb crystals. 2013. Nature Communications. 4: 3291. D’Alessio, L., and A. Rahmani. Thermally isolated Luttinger liquids with noisy Hamiltonians. 2013. Physical Review Rahmani, A., R. Muniz, and I. Martin. Anyons in Integer B. 87 (17): 174301. Quantum Hall Magnets. 2013. Physical Review X. 3 (3):

Impens, F., R. Behunin, C. C. Ttira, and P. Maia. Neto. Non-local double-path Casimir phase in atom Rahmani, A., T. Kitagawa, E. Demler, and C. Chamon. interferometers. 2013. EPL (Europhysics Letters). 101 Cooling through optimal control of quantum evolution. (6): 60006. 2013. Physical Review A. 87 (4): 043607. Intravaia, F., R. O. Behunin, and D. A. Dalvit. Quantum Rahmani, A., and G. Chern. Universal Rényi mutual friction and non-equilibrium fluctuation theorems. information in classical systems: The case of kagome 2013. Arxiv. ice. 2013. Physical Review B. 88 (5): 054426. Kamiya, Y., and C. Batista. Magnetic vortex crystals in Roy, D.. Cascaded two-photon nonlinearity in a one- frustrated Mott insulator. 2013. Arxiv. dimensional waveguide with multiple two-level emitters. 2013. Scientific Reports. 3: 02337. Kato, Y.. Multidiscontinuity algorithm for world-line Monte Carlo simulations. 2013. Physical Review E. 87 (1): Roy, D.. Two-photon scattering of a tightly focused weak 013310. light beam from a small atomic ensemble: An optical probe to detect atomic level structures. 2013. Physical Kim, J. W., Y. Kamiya, E. D. Mun, M. Jaime, N. Harrison, J. Review A. 87 (6): 063819. D. Thompson, H. T. Yi, Y. S. Oh, S. -W. Cheong, C. D. Batista, and V. S. Zapf. Multiferroicity with coexisting Roy, D., C. J. Bolech, and N. Shah. Nature of the zero- isotropic and anisotropic spins in Ca_\\\{3\\\}Co$$\begin{gathered} {}_{b}iasconductancepeakassociatedwithMajorana \\ \{2-x \\ \} \end{gathered}$$Mn_\\\{x\\\}O_\\\{6\\\}. 2013. Arxiv. bound states in topological phases of semiconductor- superconductor hybrid structures. 2013. Arxiv . Koutroulakis, G., T. Zhou, C. D. Batista, Y. Kamiya, J. D. Thompson, S. E. Brown, and H. D. Zhou. Roy, D., N. Bondyopadhaya, and S. Tewari. Topologically Magnetic Phases of the Spin-1/2 Triangular-Lattice trivial zero-bias conductance peak in semiconductor Antiferromagnet Ba3CoSb2O9. 2013. Arxiv . Majorana wires from boundary effects. 2013. Physical Review B. 88 (2): 020502. Li, H., C. Wu, S. V. Malinin, S. Tretiak, and V. Y. Chernyak. Exciton Scattering Approach for Spectroscopic Roy, D., Y. Li, A. Greilich, Y. Pershin, A. Saxena, and N. Calculation of Branched Conjugated Oligomers. Sinitsyn. Spin noise spectroscopy of quantum dot Accounts of Chemical Research. molecules. 2013. Physical Review B. 88 (4): 045320. Li, H., S. Malinin, S. Tretiak, and V. Chernyak. Effective tight- Savukov, I., and T. Karaulanov. Magnetic-resonance imaging binding models for excitons in branched conjugated of the human brain with an atomic magnetometer. molecules. 2013. The Journal of Chemical Physics. 139 2013. Applied Physics Letters. 103 (4): 043703. (6): 064109. Savukov, I., and T. Karaulanov. Anatomical MRI with an 581

Page 582

atomic magnetometer. 2013. Journal of Magnetic Resonance. 231: 39. Shimada, H., J. L. Jacobsen, and Y. Kamiya. Phase diagram and the strong-coupling fixed point in the disordered O(n) loop model. 2013. Arxiv. Torrontegui, E., S. Ibáñez, S. Martínez-Garaot, M. Modugno, A. del Campo, D. Guéry-Odelin, A. Ruschhaupt, X. Chen, and J. G. Muga. Chapter 2 - Shortcuts to Adiabaticity. 2013. In Advances in Atomic, Molecular, and Optical Physics. Edited by Arimondo, E., P. R. Berman, and C. C. Lin. Vol. Volume 62, p. 117. -: Elsevier. Zhang, Z., K. Wierschem, I. Yap, Y. Kato, C. Batista, and P. Sengupta. Phase diagram and magnetic excitations of anisotropic spin-one magnets. 2013. Physical Review B. 87 (17): 174405. 582

Page 583

Physics Directed Research Continuing Project Non-Equilibrium Phenomena in Materials, Fluids, and Climate Robert E. Ecke 20130728DR Introduction Benefit to National Security Missions This project develops collaborative scientific progress at The materials research that we propose is squarely in the interfaces of hard materials, soft condensed matter the mesoscopic regime where the discreteness of, for physics, and fluid dynamics. In particular, we consider example, dislocations at atomistic scales encounters the the structure and properties of disordered materials elastic plastic continuum at macroscopic scales. This with emphasis on the role of topological defects, for area is an important foundation for the MaRIE program example dislocations. We are exploring the connection and is applicable to recent DOE/BES priorities in funda- between the behaviors of materials under shear with mental materials research. Advances in this area would “universal” scenarios developed in the context of glasses lead to better understanding and modeling of materials, and granular materials as well as more recently in the allowing new applications. Improving our understanding context of plastic flow via dislocations and thermally ac- of turbulence is relevant to the weapons program. tivated avalanches. We will measure the pair distribution function of geopolymers using neutron scattering and Progress compare with molecular dynamics simulations. We will During FY13, we have made the following progress: look for similarities between the transition from revers- ibility to irreversibility in materials under shear and soft • Developed a careful mathematical analysis of scale condensed matter systems such as colloidal suspensions. decomposition and nonlinear dynamics in highly Reversible and irreversible behavior and the associated compressible turbulent flows. Our work provides a kinetics are at the heart of the complex hysteresis and rigorous framework upon which earlier published metastability governing the microstructure of materials work rests and shows how a proper density weight- that have been shocked. We are attempting to describe ing allows the unravelling of an inertial scale-range the universal behavior of phase transforming materials over which viscous dynamics and external forcing subjected to extremes by departing from the traditional are negligible. This gives a solid mathematical basis approach of using equilibrium phase diagrams and to attempts at extending Kolmogorov’s potent ideas notions and emphasize the importance of shear. We on a turbulent inertial range cascade to compress- will understand and characterize how granular material ible turbulence. It shows that direct viscous effects properties can affect earthquake dynamics. We will re- on large-scale dynamics do not need to be modeled. late laboratory experiments in rotating and/or stratified It also shows that there is a new cascade mechanism flows to asymptotically exact mathematical predictions that results intrinsically from compressibility effects and to applications in regional climate models such as of density fluctuations and pressure gradients that heat transport in arctic oceans. Our external collabo- are not present in incompressible flows and needs to rations enable the cooperative study of porous media be modeled properly - Physica D: Nonlinear Phe- flows relevant to carbon sequestration, the exploration nomena 247 (1), 54–65 (2013) of turbulent mixing in the oceans, the description and • Showed that a key result by Hannes Alfven (No- characterization of radiation damage in materials, and bel Prize in Physics, 1970) can break down in the the integration of perspectives on elastic-plastic defor- presence of strong nonlinear flow effects such as mation from both hard and soft matter points of view. turbulence. Our work may answer a long-standing These projects support LANL strategic goals in materials, mystery in plasma physics and astrophysics on how in energy and climate, and in materials under extreme magnetic reconnection takes place over fast time- conditions. 583

Page 584

scales forbidden by Alfven’s result yet manifested combination of experimental neutron scattering PDF mea- around us all the time in various forms such as solar surements and molecular dynamics simulations; and 3) flares and coronal mass ejections. The study combined Investigate novel properties and functions of geometrically interdisciplinary expertise in astrophysics, applied frustrated materials. mathematics, fluid mechanics, data management, and Soft Condensed Matter Systems computer science and adopted a fundamentally new We will 1) Develop models of the structural properties of approach to analyzing very large datasets. We conduct- bio-polymers such as DNA; 2) Explore sheared granular ed a state-of-the-art computer simulation, used novel media and its relationship to earthquake dynamics; and 3) ground-breaking database methods combined with Investigate the reversibility transition, discovered in parti- high-performance computing techniques and original cle-fluid systems, for weakly sheared disordered materials. mathematical developments to make our compelling case that nonlinear flow interactions (or turbulence) Fluids and Climate alone can explain the phenomenon of fast magnetic We will 1) Investigate mass transport properties of porous reconnection - Nature 497, 466-469 (2013). media flows, including differences between 2D and 3D ge- ometries and comparisons with numerical simulations; 2) • Developed a novel experimental diagnostic capability Measure heat transport, test asymptotic scaling in rotating to measure simultaneous temperature and velocity in Rayleigh-Bénard convection, and visualize vortical columns rotating convection to characterize the heat transport that transport heat; and 3) Explore how ocean phenomena efficiency of vortical structures. Such structures may at multiple length and time scales accumulate to produce be related to upwelling structures in the ocean that are the large-scale circulation responsible for the heat and important in heat transport efficiency that ultimately tracer transport affecting the global climate. impact climate evolution. We use thermochromic liquid crystals to provide the dynamics of the overall Conclusion temperature field of the vortex, use high precision thermal probes to measure vertical correlations associ- The goals of this project are to develop collaborative ated with thermal vortical structures, and use particle scientific progress at the interfaces of hard materials, soft tracking methods to determine the fluid velocity field. condensed matter physics, and fluid dynamics. We will develop models to connect to fundamental experiments, • Characterized the onset of irreversibility in periodi- build theoretical frameworks, and perform mathematical cally sheared amorphous media. An important aspect analysis and computer simulations. We expect our results of the physics of amorphous solids is the onset of to impact applications such as geophysical fluid dynamics irreversible behavior usually associated with yield. related to climate modeling. Using quasi-static molecular dynamics simulations, we observe a transition from reversible to irreversible de- Publications formation at a critical strain amplitude. For small strain Aluie, H.. Scale decomposition in compressible turbulence. amplitudes the system exhibits a noisy but repetitive 2013. Physica D: Nonlinear Phenomena. 247 (1): 54. limit-cycle. For large strain amplitudes, however, the Ben-Naim, E., and P. L. Krapivsky. Discrete analogue of behavior becomes chaotic (shows sensitivity to initial the Burgers equation. 2012. Journal of Physics A: conditions) and thus irreversible. We suggest that the Mathematical and Theoretical. 45 (45): 455003. chaotic behavior is a result of the shear band insta- bilities that arise for large strains and the convective Chu, H., C. Zhou, J. Wang, and I. J. Beyerlein. An Analytical displacement fields they create. Model for the Critical Shell Thickness in Core/Shell Nanowires Based on Crystallographic Slip. 2013. J. 61 Future Work (11): 2147. The overall structure of our work is in Structural Materials, Ecke, R., and J. Niemela. Heat transport in the geostrophic Soft Condensed Matter Systems, and in Fluids and Climate. regime of rotating Rayleigh-Bernard convection. We seek synergistic efforts to emerge among topics but Physics Review Letters. report our directed tasks in each area: Eyink, G., E. Vishniac, C. Lalescu, H. Aluie, K. Kanov, Structural Materials K. Bürger, R. Burns, C. Meneveau, and A. Szalay. We will 1) Explore complex hysteretic and metastable Flux-freezing breakdown in high-conductivity behavior in driven systems, including the role of topologi- magnetohydrodynamic turbulence. 2013. Nature. 497 cal defects in phase transformations and in elastic-plastic (7450): 466. deformation; 2) Characterize glassy polymers using a 584

Page 585

Gupta, S., and A. Saxena. Negative Gaussian curvature Zhou, C., J. Su, M. Graf, C. Reichhardt, A. Balatsky, and I. distribution in physical and biophysical systems— Beyerlein. Plastic response of dislocation glide in solid Curved nanocarbons and ion-channel membrane helium under dc strain-rate loading. 2013. Physical proteins. 2012. Journal of Applied Physics. 112 (11): Review B. 88 (2): 4513. 114316. Kent, J., J. P. Whitehead, C. Jablonowski, and R. B. Rood. Potential Vorticity: A diagnostic tool for general circulation models. Quarterly Journal of the Royal Meteorological Society. Provis, J., A. Hajimohammadi, C. White, S. Bernal, R. Myers, R. Winarski, V. Rose, T. Proffen, A. Llobet, and J. J. van Deventer. Nanostructural characterization of geopolymers by advanced beamline techniques. 2013. Cement and Concrete Composites. 36: 56. Regev, I., T. Lookman, and C. Reichhardt. Onset of Irreversibility and Chaos in Amorphous Solids Under Periodic Shear. 2013. arXiv.org. Regev, I., and C. Reichhardt. Rheology and shear band suppression in particle and chain mixtures. 2013. Physical Review E. 87 (2): 201. Reichhardt, C., J. Drocco, C. J. Reichhardt, and A. R. Bishop. Statics and Dynamics of Vortex Matter with Competing Repulsive and Attractive Interactions. 2013. Journal of Superconductivity and Novel Magnetism. : 1. Rivera, M., H. Aluie, and R. Ecke. The direct enstrophy cascade of two-dimensional soap film flows. Physics of Fluids. Wang, J., R. Zhang, C. Zhou, I. Beyerlein, and A. Misra. Characterizing interface dislocations by atomically informed Frank-Bilby theory. 2013. Journal of Materials Research. 28 (13): 1646. Wang, X., and J. Whitehead. A bound on the vertical transport of heat in the ‘ultimate’ state of slippery convection at large Prandtl numbers. 2013. Journal of Fluid Mechanics. 729: 103. White, C., G. Kearley, J. Provis, and D. Riley. Structure of kaolinite and influence of stacking faults: Reconciling theory and experiment using inelastic neutron scattering analysis. 2013. The Journal of Chemical Physics. 138 (19): 4501. White, C., J. Provis, B. Bloomer, N. Henson, and K. Page. In situ X-ray pair distribution function analysis of geopolymer gel nanostructure formation kinetics. 2013. Physical Chemistry Chemical Physics. 15 (22): 8573. White, C., K. Page, N. Henson, and J. Provis. In situ synchrotron X-ray pair distribution function analysis of the early stages of gel formation in metakaolin-based geopolymers. 2013. Applied Clay Science. 73: 17. 585

Page 586

Physics Directed Research Final Report CLEAN Detection and Identification of Dark Matter Andrew Hime 20100063DR Abstract Indeed, we face the daunting task of separating a single MiniCLEAN, a first-generation dark matter experiment event in a tonne of target material or more following a with a target (fiducial) mass of 500 kg (150 kg), will be year’s exposure. Noble liquid detectors exploiting liquid the first “single-phase” liquid argon detector including xenon (LXe) or liquid argon (LAr) hold great promise in full 3-D event position reconstruction capability. The realizing this intimidating goal (Figure 1). LANL scientists ability to inject a radioactive spike of 39Ar into the Mini- co-invented and are spearheading a novel approach, CLEAN detector provides a unique opportunity to study dubbed CLEAN for Cryogenic Low-Energy Astrophys- background discrimination at the unprecedented level ics with Noble Liquids, to the direct detection of dark of parts in 1010. The data from MiniCLEAN will inform matter using the unique capabilities of LAr for unprece- the ultimate background rejection capability of next dented discrimination between electromagnetic back- generation experiments and provide the bedrock for grounds (electrons and gamma rays) and nuclear-recoil design of a massive detector capable of extending the events characteristic of the WIMP signature [1]. CLEAN is reach for WIMP dark matter by several orders of magni- unique in its ability to exchange the LAr target with liquid tude. Funding from this LDRD project has culminated in neon (LNe) in the same detector, offering a means to fur- the concept, engineering design and fabrication of the ther distinguish a potential WIMP signal from radioactive MiniCLEAN detector. The detector is presently being as- background and further expanding its scientific portfolio sembled in the Cube-Hall at SNOLAB with collaborators to the detection of low-energy neutrinos. from the United States, the United Kingdom and Canada. Our vision for a single-phase dark matter experiment has from the very beginning been focused on getting Background and Research Objectives to very large-scale detectors. In this approach, a very It is now generally accepted in the scientific community large detector does not just have increased target mass that roughly 85% of the matter in the universe is in a for WIMP detection, it also has fewer sources of back- form that neither emits nor absorbs electromagnetic ground, is easier to model precisely and understand, and radiation. Multiple lines of evidence from cosmic micro- in many ways is easier to build. The cost of increased wave background probes, measurements of galaxy clus- size scales more slowly than the increase in surface area, ter and galaxy rotation curves, strong and weak gravi- hence far more slowly than the increase in target mass. tational lensing and big bang nucleosynthesis all point We therefore believe the single-phase approach achieves toward a cosmological concordance model containing the best WIMP sensitivity for the lowest cost. Figure 2 cold dark matter particles as the best explanation for the shows the basic design of a single-phase LAr detector, universe we see. Alternative theories involving modifica- indicating the central LAr volume and the wavelength- tions to Einstein’s theory of gravity have not been able to shifting layer of tetra phenyl butadiene (TPB) used to explain the observations across all scales. A compelling convert the 128nm UV scintillation photons created in candidate for dark matter is Weakly Interacting Massive the argon to optical light detectable by the 4-π array of Particles (WIMPs) that could be directly detected as they PMTs. scatter from massive, ultra-pure detector targets operat- ing deep beneath the Earth’s surface. Scientific Approach and Accomplishments The challenge to realizing sensitive dark matter detectors As is the case for all of the noble liquids, LAr is naturally is in separating ubiquitous backgrounds from the nucle- transparent to its own scintillation light and can be made ar-recoil events characteristic of the WIMP signature. very pure, leading to long attenuation lengths for the UV 586

Page 587

photons. Most critically, the time profile of the scintillation cal capabilities of the single-phase approach. MiniCLEAN light created by the nuclear recoil signal is dramatically is large enough that it allows us to test our entire suite of different than that for electron-like backgrounds. Ionizing background rejection tools with 3-D event reconstruction, radiation passing through a noble liquid leads to the for- and we will deploy the requisite calibration sources to test mation of dimers in both singlet and triplet states. When our detector model in detail. these states decay, they create ultraviolet scintillation light. Establishing a high statistics measure of the PSD capability LAr is special in that the lifetimes of these states are vastly in LAr that is meaningful for a large dark matter experi- different---6 ns for the singlet and 1.6 μs for the triplet--- ment (~1010 events) is challenging, to say the least. This and the relative amplitudes of these states depend on the will be possible in the MiniCLEAN detector where 1010 type of ionizing radiation. Boulay and Hime recognized decays will be recorded during a dedicated run with a 39Ar in ref.[1] this utility for dark matter detection in LAr. The spike for a period of ~100 days. The 39Ar radioactivity rejection made possible by this timing difference is crucial has been available through LANL’s radioisotope program because beta decays of intrinsic (“internal”) 39Ar are the and will be extracted and purified in Nuclear Chemistry only background that large size cannot mitigate. Division’s “hot-cell”. When injected into MiniCLEAN, the A significant effort has gone into the design, engineering acquired data sample will have utility in informing the ca- and evaluation of the MiniCLEAN detector [2]. The central pabilities and limitations of multi-tonne, second generation detector, shown in Figure 3, is composed of three major detectors such as DEAP-3600 and DarkSide and a potential elements, an Inner Vessel (IV) that contains the liquid third generation detector using a target mass ~100x larger cryogen, an array of 92 Optical Cassettes that are inserted than MiniCLEAN. Ultimately, we want to be able to gener- into the IV and define the inner target region, and an Outer ate an understanding of the PSD capability in LAr as a func- Vessel (OV) that provides secondary containment and the tion of energy threshold and target mass and at what point necessary thermal insulation of the inner cryostat. The it becomes necessary to consider the added expense and detector will be shielded by 150 cm of water with an active complexity of using argon depleted of 39Ar. muon veto and operated 6800 feet underground in the The PSD calibration using the 39Ar spike in MinCLEAN will Cube Hall at SNOLAB. improve our knowledge of the capabilities of LAr by about With a team of engineers, designers and safety experts at two orders of magnitude beyond that demonstrated in our LANL, the collaboration has ensured that all of the Mini- smaller prototype detectors. It is pertinent to emphasize CLEAN detector components can meet the physics goals that MiniCLEAN is the only experiment advertised to per- of the experiment through a safe design. Finite element form dedicated measurements of 39Ar PSD capability with analyses, calculations of loads and pressures, thermal adequate statistics using a radioactive spike. While our aim stresses, and evaluation of fault conditions have been is to inform the design of the much larger, third generation performed for each component. Experts in cryogenic and LAr detector, such measurements will have immediate util- pressure vessels have performed calculations to determine ity for next generation experiments like DEAP-3600, which the necessary relief systems needed on the OV and IV to will utilize a 3600 kg target of LAr and operate adjacent safely contain any system failures that would lead to boil- to MiniCLEAN in the Cube-Hall at SNOLAB. If DEAP-3600 off of the cryogen. In addition, a full hazard analysis plan should see events in the nuclear recoil region-of-interest, has been developed to mitigate all possible hazards: seis- it will not be possible to know, a priori, that it isn’t back- mic, fire, fall, cryogenic and oxygen deficiency, pressure, ground leakage from 39Ar apart from extrapolation from chemical, electrical, radiation, lifting, and drowning. Figure a model for PSD. MiniCLEAN will allow a direct prediction 4 indicates some of the major elements of fabrication and of the background expected in these second generation construction made possible with funding from this LDRD experiments and will provide a critical element in the project. discovery of WIMP dark matter should an experiment like DEAP-3600 claim observation of a signal. The primary risks to the viability of the single-phase ap- proach are associated with the expected light yield (which More broadly, this research will also inform the “dual- affects energy threshold, position reconstruction, and phase” detector program of DarkSide. Dual-phase detec- particle identification), the efficacy of neutron tagging, and tors rely on PSD using the primary scintillation light (S1) the development and measurement of 39Ar background and the ratio of the secondary and primary light (S2/S1). rejection using pulse-shape discrimination (PSD). The The secondary signal is acquired by drifting the electrons upcoming MiniCLEAN run, which is funded in FY14 through produced in the primary ionization component of the the DOE Office of High Energy Physics, provides us with WIMP interaction. An overarching goal of the PI has been an excellent opportunity to demonstrate all salient techni- to develop a model for scintillation and charge yield for LAr 587

Page 588

as a function of drift field. In this way, a complete and gen- and energy threshold. Nonetheless, at the CLEAN scale, eral model of PSD in LAr can be attacked as a function of a LNe target could cover the minimal theoretical phase electric field --- zero field for single-phase versus optimized space of interest (Fig.1). Moreover, with LNe in place of field for dual-phase --- and thus assess in a quantitative LAr, the number of detected WIMP events would drop by manner the most promising route to a large LAr dark mat- a factor of ~8, much like a “beam off” measurement, while ter experiment. Already, preliminary studies indicate some the number of external backgrounds will remain the same, interesting results (Figure 5). apart from predictable differences in detector response between the two targets. It has been assumed, quite generally for dual-phase detec- tors, that the scintillation yield does not depend in any The ability to swap targets in the same detector and to significant way on the applied drift field. The model under test directly a possible signal is a powerful capability of the development by the PI, however, predicts a rather signifi- CLEAN approach. It may be the only way in a massive dark cant dependence in LAr and this dependence has been ob- matter detector to verify a signal, and thus WIMP discov- served recently in LAr (left panel of Figure 5). Furthermore, ery, without simply relying on an independent experiment it is generally claimed that the discrimination power of S1/ for confirmation. While our immediate priority is to carry S2 that is afforded in a dual-phase detector in addition to out the LAr program described above, we will seek funds the S1 signal used alone in a single-phase detector puts the in FY15 in order to carry out a similar demonstration of the dual-phase technique at an advantage. As it concerns the CLEAN approach using a LNe target. overall ability for rejecting electron backgrounds, however, Finally, other fundamental applications of the CLEAN this conclusion is incorrect (right panel of Figure 5). Since approach have come to bare. In particular, it has been the scintillation yield (S1) becomes reduced in an electric recognized that a single-phase LAr detector at the tonne- field and since the PSD capability is such a strong function scale is ideal for detecting, for the first time, the Coherent of this yield, the added power of S1/S2 can never com- Elastic Neutrino-Nucleus Scattering (CENNS) predicted by pensate for that which is lost due to the lower S1 yield. the standard model. This prediction, made over 35 years Indeed, it appears to be the case that the optimum electric ago, has so far eluded detection for lack of a detector field w.r.t. electron background discrimination in LAr is the technology, like CLEAN, capable of achieving the neces- case of zero-field. This is a profound realization and one sary target mass and low-energy threshold. LANL scientists that deserves further study. are now collaborating with FermiLab to propose such an The ability to reject electron backgrounds in LAr has signifi- experiment using the “off-axis” neutrinos produced at the cant impact on the design and scientific reach of next gen- Booster-Neutrino-Beam. eration experiments. On the one hand the results indicate that PSD, and thus the low-energy threshold achievable, Impact on National Missions in LAr is optimized at zero-field (i.e. using the single-phase The direct detection and identification of dark matter is approach). On the other hand, a dual-phase detector one of the most compelling problems facing modern sci- affords superior position reconstruction resolution that ence and constitutes a top priority within the DOE Office of might prove necessary to better reject external back- Science for High Energy Physics through its Cosmic Fron- grounds such as fast neutrons or radon daughter decay tier Program. It bears repeating that the existence of dark from detector surfaces. It therefore begs the question: Is it matter reflects one of few instances where there is direct possible to conceive of a hybrid design wherein the virtues evidence for physics beyond the Standard Model. of both the single-phase and dual-phase capabilities can be realized in the same experiment? It may be possible, for The development of ultra-sensitive, rare event detectors example, to consider a detector that has full, 4-pi photo- such as MiniCLEAN requires the combined efforts and skills multiplier coverage to maximize light yield as in MiniCLEAN of a scientific and engineering team afforded by a National while maintaining the better position reconstruction of a Laboratory. Developing these novel radiation detectors can time-projection-chamber using an optimized and “mini- have direct impact on the non-proliferation program owing mal” applied drift field. to the capability for distinguishing a weak neutron signal in the presence of an intense gamma-ray background. Initial Unique to CLEAN is the possibility to exchange the LAr research into this application has been supported by an target with liquid neon (LNe). LNe has the advantage that NA22 grant to pursue this possibility. it can be made extremely clean and free of internal radio- activity. Owing to its smaller atomic number, however, the predicted WIMP-nucleus interaction rate is about eight times smaller than in LAr for the equivalent target mass 588

Page 589

Figure 3. (Left) Model of the MiniCLEAN central detector with its 4π target viewed by 92 optical cassettes. (Right) The optical cassettes are 30 cm long and consist of a 10 cm thick acrylic plug, the front surface of which is coated with a wavelength-shifting fluor (TPB), and 30 cm light guide leading to the PMT. The inner target defined by the TPB surface contains 500 kg of LAr within a nominal radius of 44 cm. Figure 1. Recent upper limits and projected sensitivities for the spin-independent WIMP-nucleon cross-section as a function of WIMP mass. Allowed regions reported by DAMA (magenta), CoGeNT (light green), and CRESST-II (blue), are under tension from the bounds from CDMS-II and XENON-100. Parameter space corresponding to constrained SUSY models with recent LHC and Higgs constraints is shown as the shaded regions at 68%, 95%, and 99% confidence intervals. A (45-tonne) CLEAN experiment would provide excellent coverage of the most interesting regions of parameter space and would be capable of both LAr and LNe targets. The projected sensitivity for DEAP- 3600, LUX and a generic LXe experiment at the 1-tonne scale are shown for comparison. Figure 4. A collage of MiniCLEAN detector components under assembly at SNOLAB: (top left) the Outer Vessel on its stand and inside the water shield tank in the Cube-Hall; (top right) a view of light-guides as they are pre-fit into the Inner Vessel; (bottom left) a PMT assembled in its holder along with the light-guide and acrylic plug with wavelength shifter; (bottom right) the Inner Figure 2. (left) Schematic version of a single-phase cryogenic Vessel that serves to hold the cryogenic target and 92 optical liquid scintillation detector. The UV scintillation light (128 cassettes. nm) is converted to detectable optical light by a wavelength- shifting layer viewed in 4-pi by a PMT array. (right) Rejection of internal (39Ar beta decay) is via Particle-ID using pulse-shape discrimination of the triplet-to-singlet light ratio. External backgrounds (surface radon progeny and fast neutrons) are rejected by self-shielding and fiducialization. MiniCLEAN will be the first single-phase LAr detector with 3D event reconstruction. Figure 5. The data points are recent measurements by the SCENE collaboration for the relative scintillation yield in LAr as a 589

Page 590

function of applied electric field. The curves are generated from a Feb. 2012). model, under development by the PI. Right: The solid blue curve indicates a projection of electron background leakage from 39Ar Hime, A.. CLEAN Detection of Dark Matter. Presented at CI- beta decay in zero field when the light yield is 10 p.e./keV. The PANP 2012. (St. Petersburg, FL, 29 May-4 July, 2012). corresponding prediction for dual-phase operation using a drift Hime, A.. CLEAN Detection of Dark Matter. Invited presen- field of 1 kV/cm is shown as the solid black curve and is a product tation at Identification od Dark Matter 2012. (Chicago, of discrimination factors from S1 and S1/S2. IL, 23-27 July, 2012). References MacMullin, S., M. Boswell, M. Devlin, S. R. Elliott, N. Fotia-

1. Boulay, M. G., and A. Hime. A Technique for the Direct des, V. E. Guiseppe, R. Henning, T. Kawano, B. H. La- Detection of Weakly Interacting Massive Particles using Roque, R. O. Nelson, and J. M. O’Donnell. Partial gam- the Scintillation Time Dependence in Liquid Argon. ma-ray production cross sections for (n, xn gamma)
2. Astroparticle Physics. 25: 179. reactions in natural argon at 1-30 MeV. 2012. PHYSICAL REVIEW C. 85 (6): -.
3. Hime, A.. The MiniCLEAN Dark Matter Experiment. Invited presentation at proceedings of the DPF-2011 Rielage, K.. Status and Prospects of the MiniCLEAN Dark Matter Experiment. 2012. In 19TH PARTICLES AND Conference. (Providence, RI, 2011-arXiv:1110.1005). NUCLEI INTERNATIONAL CONFERENCE (PANIC11). Vol. 1441, p. 518. Publications Bodmer, M., F. Giuliani, M. Gold, A. Christou, and M. Baty- Seibert, S.. Status of the MiniCLEAN Dark Matter Experi- gov. Design of an active magnetic field compensation ment. To appear in LIDINE 2013. (Fermilab, 29-31 May system for MiniCLEAN. 2013. NUCLEAR INSTRUMENTS 2013). & METHODS IN PHYSICS RESEARCH SECTION A-ACCEL- ERATORS SPECTROMETERS DETECTORS AND ASSOCI- Shoemaker, I. M.. Indirect and direct detection of dark mat- ATED EQUIPMENT. 697: 99. ter. Invited presentation at Aspen Center for Physics. (Aspen, CO, 6-12 Feb. 2011). Caldwell, T.. Characterization of the R5912-02 MOD Photo- multiplier Tube at Cryogenic Temperatures. To appear macmullin, s., M. Boswell, M. Devlin, S. Elliott, N. Fotiades, in LIDINE 2013. (Fermilab, 29-31 May 2013). V. Guiseppe, R. Henning, T. Kawano, B. LaRoque, and R. Nelson. Neutron-induced gamma-ray production cross Gastler, D., E. Kearns, A. Hime, L. C. Stonehill, S. Seibert, J. sections for the first excited-state transitions in Ne-20 Klein, W. H. Lippincott, D. N. McKinsey, and J. A. Nik- and Ne-22 . 2012. Phys. Rev. C. 86: 067601. kel. Measurement of scintillation efficiency for nuclear recoils in liquid argon. 2012. PHYSICAL REVIEW C. 85 (6): -. Gehman, V. M.. Fluorescence efficiency and visible re- emission spectrum of tetraphenyl butadiene films at extreme ultraviolet wavelengths. 2011. Nuclear Instru- ments and Methods. A654: 116. Graesser, M. L., I. M. Shoemaker, and L. Vecchi. Asymmet- ric WIMP dark matter. 2011. Journal of High Energy Physics. 10: 110. Guiseppe, V., K. Rielage, S. Westerdale, S. Elliott, and A. Hime. Radon Progeny Deposition Model. 2011. In 3rd Topical Workshop in Low Radioactivity Techniques. (Sudbury, Canada, 28-29 August 2010). Vol. 1338, p.
4. AIP Conference Proc.: AIP. Hime, A.. The MiniCLEAN Dark Matter Experiment. 2011. In DPF-2011 Conference. (Providence, RI, 8-13, Aug. 2011). , p. 1. DPF Conference Proceedings: arXiv:1110:1005v1. Hime, A.. CLEAN Detection of Dark Matter. Presented at UCLA Dark Matter 2012. (Marina Del Rey, CA, 22-24 590

Page 591

Physics Directed Research Final Report Multi-Messenger Signals from Low-mass Supernovae Christopher L. Fryer 20110032DR Abstract curves. During the course of this project, our supernova This project focused on modeling all the observational light-curve code moved from the first calculations to diagnostics of core-collapse supernovae, from probes of truly a production-level capability, with a fast-growing circumstellar medium (supernova light-curves and spec- user base. This past year alone, we submitted 7 papers tra) to probes of the explosion mechanism (nucleosyn- studying supernova light-curves. Many of these stud- thetic yields and compact remnant masses) to the inner ies focused on understanding better the mass loss from core (neutrinos and gravitational waves). stars and, in doing so, we moved to the latest progenitor models. One of our biggest discoveries with these mod- Background and Research Objectives els is the fact that the stellar structure changes signifi- Supernovae are one of the ideal physics laboratories for cantly with newer algorithms for stellar convection. For matter in extreme conditions produced in nature. This the models produced by the Tycho code with their new project focused on modeling the diagnostics of these ex- mixing algorithm based on multi-dimensional simula- periments, from probes of circumstellar medium (super- tions (e.g. Viallet et al. 2013), we found that above 20 nova light-curves and spectra) to probes of the explosion solar masses, most of the helium is burned into carbon mechanism (nucleosynthetic yields and compact rem- and oxygen (Frey et al. 2013). This expansive burning nant masses) to the inner core (neutrinos and gravita- provides a natural explanation for the high fraction of tional waves). This work led to collaborations between type Ic supernovae as well as the lack of type Ib super- a broad set of scientists both within the laboratory (T, novae associated with gamma-ray bursts. CCS, LANSCE, P, XTD, XCP, ISR) and the international Explosion Mechanism: A main thrust of the project has community (we are now collaborating with over 100 been to better understand the nucleosynthetic yields scientists world-wide). Over the course of the 3 years, of core-collapse supernovae. Nucleosynthetic yields the publication rate of the team skyrocketed and in the have the potential to probe the strength and asymmetry final year over 40 papers were submitted to refereed of the supernova engine. For example, the high 44Ti journals (with a total of 80 papers over the course of the yield in the Casioppeia A remnant. We published over DR). A large fraction of these papers were high impact 30 refereed papers during this project improving our for LANL; 20% of these papers have each garnered over theoretical rates for core-collapse nucleosynthesis. A 15-20 citations per year. During this project, we instigat- further 6 papers were published (most in the last year) ed collaborations with Southwest Research Institute, the using modern rates to calculate nucleosynthetic yields Argentina-led Transient Optical Robotic Observatory of from massive stars. Our project combined development the South (TOROS), the Palomar Transient Factory (PTF) of new experimental tools to calculate nuclear cross-sec- and expanded our collaborations with the Nucleosynthe- tions with new theories to calculate these cross-sections. sis Grid (NuGrid) collaboration and the Joint Institute for LANL’s expertise in fission products is ideally suited to Nuclear Astrophysics (JINA). Below we highlight a few calculating r-process yields and we have focused a lot of the specific results from this project, spanning the 3 of our effort on this important piece of nucleosynthe- different diagnostic categories at LANL. sis. However, the picture of two extreme processes, the r- and s-processes, to explain all the heavy elements Scientific Approach and Accomplishments has been significantly muddied in the past few years. It Circumstellar Medium: Most of our work probing the is now believed that intermediate processes (neutron circumstellar medium focused on supernova light- capture that is not so fast that elements are pushed to 591

Page 592

the drip line, but not so slow that elements are limited ity to bring together a broad set of disciplines to study an to being only one or two neutrons beyond the valley of applied problem (and indication of the breadth is the fact stability) can dominate the nucleosynthetic yields that we that over 40 academy of scientists work in this field and see. The LANL-founded NuGrid collaboration has been scientists from our project personally interacted with over at the forefront of this new paradigm. The first NuGrid 15 of these academy of scientists during the course of the paper summarizing the range of nucleosynthesis processes project). Unfortunately, this project opened up many new was submitted this year (Pignatari et al. 2013, astro- questions for every one we answered, and this field is, if ph/1307.6961). We have integrated LANL rates into the anything, more wide open than when we started. But we NuGrid network framework, tying our experimental and do understand the multi-messenger signals much better theoretical work firmly into the recent work in the astro- now and have established LANL as a leader in this broad, physics community. but integratable, topic. Another probe of the supernova explosion is the mass of Impact on National Missions the compact remnants formed during the stellar collapse. These projects had a lot of synergy with the ASC and Cam- We led a series of papers tying the explosion energy and paign programs within LANL. The multi-physics nature of stellar progenitor to the expected remnant mass. Fryer the supernova problem is akin to the multi-physics prob- et al. (2012) completed a detailed survey of these ties, lems in these programs and most (7 of 10) of the post-docs providing a new set of remnant mass predictions (based in this program were converted into staff in the last 2 years on the explosion mechanism). This high impact paper of this project. Software developed for the astrophysics has, a year and a half after publication, received nearly 40 light-curve program is now used in a wide variety of mis- citations and will in, just a few years, become another high sion science. Finally, the nuclear physics techniques devel- impact (>100 citation) paper for Fryer (according to LANL’s oped in this project facilitate better uncertainty quantifica- experts database, Fryer has more first-author high-impact tion (UQ) and we are actively engaged in this UQ analysis papers than anyone at LANL – including such superstars as in a follow-on project. John Sarrao, Bette Korber, Alan Bishop, …) The high impact again illustrates the broad application and cutting edge References research led in this DR project.

1. Viallet, M., C. Meakin, D. Arnett, and M. Mocak. TURBULENT CONVECTION IN STELLAR INTERIORS. III. Explosion Engine: The final focus of the project was to MEAN-FIELD ANALYSIS AND STRATIFICATION EFFECTS. study the details of the supernova engine itself. Gravita-
2. ASTROPHYSICAL JOURNAL. 769 (1): -. tional waves and neutrinos are the only direct probes of the explosive engine. We carried out a number of projects
3. Frey, L. H., C. L. Fryer, and P. A. Young. CAN STELLAR studying the role of neutrino oscillations (working with ex- MIXING EXPLAIN THE LACK OF TYPE Ib SUPERNOVAE IN perts like George Fuller for which we successfully proposed LONG-DURATION GAMMA-RAY BURSTS?. 2013. ASTRO- a collaborative grant during this project) and neutrino PHYSICAL JOURNAL LETTERS. 773 (1): -. cross-sections. Our neutrino cross-section work ties well with the focus of the Institute for Nuclear Theory in Seattle 3. Fryer, C. L., and K. C. B. New. Gravitational Waves from and we are closely collaborating with this community. Gravitational Collapse. 2011. LIVING REVIEWS IN RELA- Similarly, we revived the LANL effort in gravitational waves, TIVITY. 14: -. organizing a 3 month long workshop on gravitational waves at the Kavli Institute for Theoretical Physics at Santa Bar- Publications bara, taking part (and leading in some cases: Andersson Belczynski, K., G. Wiktorowicz, C. L. Fryer, D. E. Holz, and V. et al. 2013 – CQG, 2013, Fryer & New 201) several reviews Kalogera. MISSING BLACK HOLES UNVEIL THE SUPER- in this field (including invited reviews at the American NOVA EXPLOSION MECHANISM. 2012. ASTROPHYSICAL JOURNAL. 757 (1): -. Physical Society Meeting). Finally, LANL joined the TOROS collaboration to better tie our work to the gravitational Belczynski, K., T. Bulik, C. L. Fryer, A. Ruiter, F. Valsecchi, J. wave community. S. Vink, and J. R. Hurley. On the maximum mass of stel- lar black holes. 2010. Astrophysical Journal. 714 (2): This project made great strides in all of these fields, each

a full discipline in itself. This project places LANL as a firm leader in integrating all these disciplines under a single Bennett, M. E., R. Hirschi, M. Pignatari, S. Diehl, C. Fryer, F. roof and represents a breadth not seen at universities or Herwig, A. Hungerford, G. Magkotsios, G. Rockefeller, national laboratories for that matter. But this is a strength F. Timmes, M. Wiescher, and P. Young. The effect of of the national laboratories and LANL in particular, the abil- ¹²C + ¹²C rate uncertainties 592

Page 593

on s-process yields. 2010. In Nuclear Physics in Astro- cade ; 8-11 March 2010 ; Austin, TX, USA. Vol. 1294, p. physics IV ; 8-12 June 2009 ; Frascati, Italy. Vol. 202, 1 70. Edition, p. 012023 (4 pp.). Fryer, C. L., K. Belczynski, E. Berger, C. Thone, C. Ellinger, Bennett, M. E., R. Hirschi, M. Pignatari, S. Diehl, C. Fryer, F. and T. Bulik. THE POPULATION OF HELIUM-MERGER Herwig, A. Hungerford, G. Magkotsios, G. Rockefeller, PROGENITORS: OBSERVATIONAL PREDICTIONS. 2013. F. Timmes, M. Wiescher, and P. Young. The effect of C- ASTROPHYSICAL JOURNAL. 764 (2): -. 12+C-12 rate uncertainties on s-process yields. 2010. NUCLEAR PHYSICS IN ASTROPHYSICS IV (NPAIV 2009). Fryer, C. L., K. Belczynski, G. Wiktorowicz, M. Dominik, V. 202: -. Kalogera, and D. E. Holz. COMPACT REMNANT MASS FUNCTION: DEPENDENCE ON THE EXPLOSION MECHA- Bennett, M. E., R. Hirschi, M. Pignatari, S. Diehl, C. Fryer, F. NISM AND METALLICITY. 2012. ASTROPHYSICAL JOUR- Herwig, A. Hungerford, K. Nomoto, G. Rockefeller, F. X. NAL. 749 (1): -. Timmes, and M. Wiescher. The effect of C-12+C-12 rate uncertainties on the evolution and nucleosynthesis of Fryer, C. L., and A. Heger. Forming massive black holes massive stars. 2012. MONTHLY NOTICES OF THE ROYAL through stellar collapse: Observational diagnostics. ASTRONOMICAL SOCIETY. 420 (4): 3047. 2011. Astronomische Nachrichten. 332 (4): 408. Dominik, M., K. Belczynski, C. Fryer, D. E. Holz, E. Berti, T. Fryer, C., K. Belczynski, G. Wiktorowicz, M. Dominik, V. Bulik, I. Mandel, and R. O’Shaughnessy. DOUBLE COM- Kalogera, and D. Holz. COMPACT REMNANT MASS PACT OBJECTS. I. THE SIGNIFICANCE OF THE COMMON FUNCTION: DEPENDENCE ON THE EXPLOSION MECHA- ENVELOPE ON MERGER RATES. 2012. ASTROPHYSICAL NISM AND METALLICITY. 2012. ASTROPHYSICAL JOUR- JOURNAL. 759 (1): -. NAL. 749 (1): 91. Dominik, M., K. Belczynski, C. Fryer, D. E. Holz, E. Berti, T. Fryer, C., and K. B. New. Gravitational Waves from Gravita- Bulik, I. Mandel, and R. O’Shaughnessy. DOUBLE COM- tional Collapse. 2011. Living Reviews in Relativity. 14: PACT OBJECTS. I. THE SIGNIFICANCE OF THE COMMON 1. ENVELOPE ON MERGER RATES. 2012. ASTROPHYSICAL Fryer, null.. Mixing in astrophysics. 2011. DOE. JOURNAL. 759 (1): -. Herwig, F., M. Pignatari, P. R. Woodward, D. H. Porter, G. Ellinger, C. I., P. A. Young, C. L. Fryer, and G. Rockefeller. A Rockefeller, C. L. Fryer, M. Bennett, and R. Hirschi. Con- CASE STUDY OF SMALL-SCALE STRUCTURE FORMATION vective-reactive Proton-¹²C combustion IN THREE-DIMENSIONAL SUPERNOVA SIMULATIONS. IN Sakurai’s object (V4334 Sagittarii) and implications 2012. ASTROPHYSICAL JOURNAL. 755 (2): -. for the evolution and yields from the first generations Ellinger, C., P. Young, C. Fryer, and G. Rockefeller. A case of stars. 2011. Astrophysical Journal. 727 (2): 89 (15 study of small-scale structure formation in three-di- pp.). mensional supernova simulations. 2012. Astrophysical Herwig, F., M. Pignatari, P. R. Woodward, D. H. Porter, G. Journal. 755 (2): 160 (34 pp.). Rockefeller, C. L. Fryer, M. Bennett, and R. Hirschi. Con- Eriksen, K. [Los Alamos National Laboratory]., C. [Los vective-reactive Proton-¹²C combustion Alamos National Laboratory]. Fontes, J. [Los Alamos IN Sakurai’s object (V4334 Sagittarii) and implications National Laboratory]. Colgan, null. Zhang, A. [Los Ala- for the evolution and yields from the first generations mos National Laboratory]. Hungerford, C. [Los Alamos of stars. 2011. Astrophysical Journal. 727 (2): 89 (15 National Laboratory]. Fryer, J. [Rutgers University]. pp.). Hughes, R. [Smithsonian Astrophysical Observatory]. Herwig, F., M. Pignatari, P. Woodward, D. Porter, G. Rock- Smith, and null. Badenes. Fe Atomic Data for Non-equi- efeller, C. Fryer, M. Bennett, and R. Hirschi. CON- librium Ionization Plasmas. 2012. DOE/LANL. VECTIVE-REACTIVE PROTON-C-12 COMBUSTION IN Fryer, C. L., A. J. Ruiter, K. Belczynski, P. J. Brown, F. Bufano, SAKURAI‚ÄôS OBJECT (V4334 SAGITTARII) AND IMPLI- S. Diehl, C. J. Fontes, L. H. Frey, S. T. Holland, A. L. Hun- CATIONS FOR THE EVOLUTION AND YIELDS FROM THE gerford, S. Immler, P. Mazzali, C. Meakin, P. A. Milne, FIRST GENERATIONS OF STARS. 2011. ASTROPHYSICAL C. Raskin, and F. X. Timmes. Spectra of type Ia Super- JOURNAL. 727 (2): 89. novae from Double Degenerate Mergers. 2010. Astro- Herwig, F., null. Pignatari, P. [LCSE and Department of physical Journal. 725 (1): 296. Astronomy. Woodward, D. [Minnesota Supercomput- Fryer, C. L., D. J. Whalen, and L. Frey. Modeling Emission ing Institute. Porter, G. Rockefeller, C. [Computational from the First Explosions: Pitfalls and Problems. 2010. Computer Science Division. Fryer, M. Bennett, and In First Stars and Galaxies: Challenges for the next De- R. Hirschi. CONVECTIVE-REACTIVE PROTON-\{sup 12\} 593

Page 594

C COMBUSTION IN SAKURAI’S OBJECT (V4334 SAGIT- Moller, P., A. J. Sierk, R. Bengtsson, H. Sagawa, and T. TARII) AND IMPLICATIONS FOR THE EVOLUTION AND Ichikawa. Nuclear shape isomers. 2012. ATOMIC DATA YIELDS FROM THE FIRST GENERATIONS OF STARS. AND NUCLEAR DATA TABLES. 98 (2): 149. 2011. Journal Name: Astrophysical Journal. 727 (2): Medium: X; Size: 15 pages. Moller, P., W. Myers, H. Sagawa, and S. Yoshida. New Finite- Range Droplet Mass Model and Equation-of-State Holloway, S. T., T. Kawano, and P. Moller. Time Dependent Parameters. 2012. PHYSICAL REVIEW LETTERS. 108 (5): Particle Emission from Fission Products. 2011. Journal 052501. of the Korean Physical Society. 59 (2): 875. Passy, J., O. De Marco, C. Fryer, F. Herwig, S. Diehl, J. Oishi, Ichikawa, T., A. Iwamoto, Mo&uml, and A. Sierk. Contrast- M. Mac Low, G. Bryan, and G. Rockefeller. SIMULAT- ing fission potential-energy structure of actinides and ING THE COMMON ENVELOPE PHASE OF A RED GIANT mercury isotopes. 2012. Physical Review C (Nuclear USING SMOOTHED-PARTICLE HYDRODYNAMICS AND Physics). 86 (2): 024610 (8 pp.). UNIFORM-GRID CODES. 2012. ASTROPHYSICAL JOUR- NAL. 744 (1): 52. Ichikawa, T., A. Iwamoto, P. Moller, and A. Sierk. Contrast- ing fission potential-energy structure of actinides and Pignatari, M., E. Zinner, M. G. Bertolli, R. Trappitsch, P. mercury isotopes. 2012. PHYSICAL REVIEW C. 86 (2): Hoppe, T. Rauscher, C. Fryer, F. Herwig, R. Hirschi, F. 024610. X. Timmes, and F. K. Thielemann. SILICON CARBIDE GRAINS OF TYPE C PROVIDE EVIDENCE FOR THE PRO- Ichikawa, T., Y. Kanada-En’yo, and Mo&uml. Cluster for- DUCTION OF THE UNSTABLE ISOTOPE Si-32 IN SUPER- mations in deformed states for ²⁸Si and NOVAE. 2013. ASTROPHYSICAL JOURNAL LETTERS. 771 ³²S. 2011. Physical Review C (Nuclear (1): -. Physics). 83 (5): 054319 (7 pp.). Pignatari, M., M. Wiescher, F. X. Timmes, R. J. de Boer, F. K. Ichikawa, T., Y. Kanada-En‚Äôyo, and P. Moller. Cluster for- Thielemann, C. Fryer, A. Heger, F. Herwig, and R. Hirs- mations in deformed states for Si-28 and S-32. 2011. chi. PRODUCTION OF CARBON-RICH PRESOLAR GRAINS PHYSICAL REVIEW C. 83 (5): 054319. FROM MASSIVE STARS. 2013. ASTROPHYSICAL JOUR- NAL LETTERS. 767 (2): -. Johnson, J. L., D. J. Whalen, C. L. Fryer, and H. Li. THE GROWTH OF THE STELLAR SEEDS OF SUPERMASSIVE Pignatari, M., R. Hirschi, M. Wiescher, R. Gallino, M. Ben- BLACK HOLES. 2012. ASTROPHYSICAL JOURNAL. 750 nett, M. Beard, C. Fryer, F. Herwig, G. Rockefeller, and (1): -. F. X. Timmes. THE C-12+C-12 REACTION AND THE IM- PACT ON NUCLEOSYNTHESIS IN MASSIVE STARS. 2013. Kawano, T., P. Talou, M. B. Chadwick, S. Holloway, P. Moller, ASTROPHYSICAL JOURNAL. 762 (1): -. and T. Watanabe. Applications of the Hauser-Feshbach Theory to Advanced Nuclear Sciences. 2011. JOURNAL Pignatari, M., R. Hirschi, M. Wiescher, R. Gallino, M. Ben- OF THE KOREAN PHYSICAL SOCIETY. 59 (2, 3, SI): 785. nett, M. Beard, C. Fryer, F. Herwig, G. Rockefeller, and F. X. Timmes. THE C-12+C-12 REACTION AND THE IM- Magkotsios, G., F. X. Timmes, A. L. Hungerford, C. L. Fryer, PACT ON NUCLEOSYNTHESIS IN MASSIVE STARS. 2013. P. A. Young, and M. Wiescher. Trends in <sup>44</ ASTROPHYSICAL JOURNAL. 762 (1): -. sup>Ti and ⁵⁶Ni from core-collapse super- novae. 2010. Astrophysical Journal Supplement Series. Qiang, Yuan, null. Bi Xiaojun [Key Laboratory of Particle As- 191 (1): 66. tro, null. Liu Siming [Department of Physics and Astron, Y. U. Fan Zhonghui [Department of Physics, and C. [Los Mesler, R., D. Whalen, N. Lloyd-Ronning, C. Fryer, and Alamos National Laboratories. Fryer. MODELING THE Pihlstro&uml. Gamma-ray Bursts in Circumstellar MULTI-WAVELENGTH EMISSION OF THE SHELL-TYPE Shells: A Possible Explanation for Flares. 2012. Astro- SUPERNOVA REMNANT RX J1713.7-3946. 2011. Jour- physical Journal. 757 (2): 117 (11 pp.). nal Name: Astrophysical Journal. 735 (2): Medium: X; Size: 9 pages. Mo&uml, , J. &. Randrup, and A. Sierk. Calculated fission yields of neutron-deficient mercury isotopes. 2012. Ramaprabhu, P., G. Dimonte, P. Woodward, C. Fryer, G. Physical Review C (Nuclear Physics). 85 (2): 024306 (6 Rockefeller, K. Muthuraman, P. -H. Lin, and J. Jayaraj. pp.). The late-time dynamics of the single-mode Rayleigh- Taylor instability. 2012. Physics of Fluids. 24 (7): Mo&uml, , W. Myers, H. Sagawa, and S. Yoshida. New Fi- 074107 (21 pp.). nite-Range Droplet Mass Model and Equation-of-State Parameters. 2012. Physical Review Letters. 108 (5): Randrup, J., Mo&uml, and A. J. Sierk. Fission-fragment 052501 (4 pp.). mass distributions from strongly damped shape evolu- 594

Page 595

tion. 2011. Physical Review C (Nuclear Physics). 84 (3): Kuin, S. R. Oates, R. W. Pogge, G. Pojmanski, R. Stoll, 034613 (15 pp.). B. J. Shappee, K. Z. Stanek, and D. M. Szczygiel. The Unusual Temporal and Spectral Evolution of the Type Randrup, J., P. Moller, and A. J. Sierk. Fission-fragment IIn Supernova 2011ht. 2012. Astrophysical Journal. 751 mass distributions from strongly damped shape evolu- (2): 92 (16 pp.). tion. 2011. PHYSICAL REVIEW C. 84 (3): 034613. Ruiter, A. J., K. Belczynski, S. A. Sim, W. Hillebrandt, C. L. Randrup, J., P. Moller, and A. Sierk. Brownian shape mo- Fryer, M. Fink, and M. Kromer. Delay times and rates tion: Fission fragment mass distributions. 2012. In for Type Ia supernovae and thermonuclear explosions CNR*11 - Third International Workshop on Compound from double-detonation sub-Chandrasekhar mass Nuclear Reactions and Related Topics ; 19-23 Sept. ( models. 2011. MONTHLY NOTICES OF THE ROYAL AS- 2011 ; Prague, Czech Republic). Vol. 21, p. 08006 (8 TRONOMICAL SOCIETY. 417 (1): 408. pp.). Ruiter, A. J., K. Belczynski, S. A. Sim, W. Hillebrandt, C. L. Randrup, J., and P. Moller. Brownian Shape Motion on Fryer, M. Fink, and M. Kromer. Delay times and rates Five-Dimensional Potential-Energy Surfaces: Nuclear for Type Ia supernovae and thermonuclear explosions Fission-Fragment Mass Distributions. 2011. PHYSICAL from double-detonation sub-Chandrasekhar mass REVIEW LETTERS. 106 (13): 132503. models. 2011. MONTHLY NOTICES OF THE ROYAL AS- TRONOMICAL SOCIETY. 417 (1): 408. Raskin, C., E. Scannapieco, C. Fryer, G. Rockefeller, and F. X. Timmes. Remnants of Binary White Dwarf Mergers. Scannapieco, E., C. Raskin, M. Della Valle, C. Fryer, J. 2012. Astrophysical Journal. 746 (1): 62 (10 pp.). Rhoads, G. Rockefeller, and F. X. Timmes. Constraining Type Ia Supernova Progenitors. 2013. BINARY PATHS Raskin, C., E. Scannapieco, C. Fryer, G. Rockefeller, and F. X. TO TYPE IA SUPERNOVAE EXPLOSIONS. 281: 275. Timmes. REMNANTS OF BINARY WHITE DWARF MERG- ERS. 2012. ASTROPHYSICAL JOURNAL. 746 (1): 62. Staff, J. E., A. Menon, F. Herwig, W. Even, C. L. Fryer, P. M. Motl, T. Geballe, M. Pignatari, G. C. Clayton, and J. E. Raskin, C., E. Scannapieco, C. Fryer, G. Rockefeller, and F. X. Tohline. DO R CORONAE BOREALIS STARS FORM FROM Timmes. REMNANTS OF BINARY WHITE DWARF MERG- DOUBLE WHITE DWARF MERGERS?. 2012. ASTROPHYS- ERS. 2012. ASTROPHYSICAL JOURNAL. 746 (1): -. ICAL JOURNAL. 757 (1): -. Raskin, C., E. Scannapieco, C. Fryer, G. Rockefeller, and F. X. Staff, J., A. Menon, F. Herwig, W. Even, C. Fryer, P. Motl, T. Timmes. REMNANTS OF BINARY WHITE DWARF MERG- Geballe, M. Pignatari, G. Clayton, and J. Tohline. Do R ERS. 2012. ASTROPHYSICAL JOURNAL. 746 (1): -. Coronae Borealis Stars form from Double White Dwarf Mergers?. 2012. Astrophysical Journal. 757 (1): 76 (17 Raskin, C., E. Scannapieco, G. Rockefeller, C. Fryer, S. Diehl, pp.). and F. X. Timmes. Astrophysical Journal. 2010. Astro- physical Journal. 724 (1): 111. Talou, P., T. Kawano, J. E. Lynn, P. Moller, O. Bouland, and M. B. Chadwick. Recent Advances in Nuclear Fission Roming, P. W. A., T. A. Pritchard, J. L. Prieto, C. S. Kochanek, Theory: Pre- and Post-scission Physics. 2011. JOURNAL C. L. Fryer, K. Davidson, R. M. Humphreys, A. J. Bayless, OF THE KOREAN PHYSICAL SOCIETY. 59 (2, 3, SI): 797. J. F. Beacom, P. J. Brown, S. T. Holland, S. Immler, N. P. M. Kuin, S. R. Oates, R. W. Pogge, G. Pojmanski, R. Talou, P., T. Kawano, J. E. Lynn, P. Moller, O. Bouland, and Stoll, B. J. Shappee, K. Z. Stanek, and D. M. Szczygiel. M. B. Chadwick. Recent Advances in Nuclear Fission THE UNUSUAL TEMPORAL AND SPECTRAL EVOLUTION Theory: Pre- and Post-scission Physics. 2011. JOURNAL OF THE TYPE IIn SUPERNOVA 2011ht (vol 92, pg 751, OF THE KOREAN PHYSICAL SOCIETY. 59 (2): 797. 2012). 2013. ASTROPHYSICAL JOURNAL. 762 (2): -. Tho&uml, , A. de Ugarte Postigo, C. L. Fryer, K. L. Page, J. Roming, P. W. A., T. A. Pritchard, J. L. Prieto, C. S. Kochanek, Gorosabel, M. A. Aloy, D. A. Perley, C. Kouveliotou, H. C. L. Fryer, K. Davidson, R. M. Humphreys, A. J. Bayless, T. Janka, P. Mimica, J. L. Racusin, H. Krimm, J. Cum- J. F. Beacom, P. J. Brown, S. T. Holland, S. Immler, N. P. mings, S. R. Oates, S. T. Holland, M. H. Siegel, M. De M. Kuin, S. R. Oates, R. W. Pogge, G. Pojmanski, R. Stoll, Pasquale, E. Sonbas, M. Im, W. -K. Park, D. A. Kann, S. B. J. Shappee, K. Z. Stanek, and D. M. Szczygiel. THE Guziy, Herna&acute, A. Llorente, K. Bundy, C. Choi, H. UNUSUAL TEMPORAL AND SPECTRAL EVOLUTION OF Jeong, H. Korhonen, Kuba&grave, J. Lim, A. Moskvi- THE TYPE IIn SUPERNOVA 2011ht. 2012. ASTROPHYSI- tin, Mu&ntilde, S. Pak, and I. Parrish. The unusual CAL JOURNAL. 751 (2): -. γ-ray burst GRB 101225A from a helium star/ neutron star merger at redshift 0.33. 2011. Nature. 480 Roming, P. W., T. A. Pritchard, J. L. Prieto, C. S. Kochanek, C. (7375): 72. L. Fryer, K. Davidson, R. M. Humphreys, A. J. Bayless, J. F. Beacom, P. J. Brown, S. T. Holland, S. Immler, N. P. 595

Page 596

Tsang, M. B., J. R. Stone, F. Camera, P. Danielewicz, S. Gan- dolfi, K. Hebeler, C. J. Horowitz, J. Lee, W. G. Lynch, Z. Kohley, R. Lemmon, P. Moller, T. Murakami, S. Riordan, X. Roca-Maza, F. Sammarruca, A. W. Steiner, I. Vidana, and S. J. Yennello. Constraints on the symmetry energy and neutron skins from experiments and theory. 2012. PHYSICAL REVIEW C. 86 (1): -. Veselsky&acute, , A. N. Andreyev, S. Antalic, M. Huyse, Mo&uml, K. Nishio, A. J. Sierk, P. Van Duppen, and M. Venhart. Fission-barrier heights of neutron-deficient mercury nuclei. 2012. Physical Review C (Nuclear Phys- ics). 86 (2): 024308 (8 pp.). Whalen, D. J., C. C. Joggerst, C. L. Fryer, M. Stiavelli, A. He- ger, and D. E. Holz. FINDING THE FIRST COSMIC EXPLO- SIONS. II. CORE-COLLAPSE SUPERNOVAE. 2013. ASTRO- PHYSICAL JOURNAL. 768 (1): -. Whalen, D. J., W. Even, C. C. Lovekin, C. L. Fryer, M. Stiavel- li, P. W. A. Roming, J. Cooke, T. A. Pritchard, D. E. Holz, and C. Knight. ILLUMINATING THE PRIMEVAL UNIVERSE WITH TYPE IIn SUPERNOVAE. 2013. ASTROPHYSICAL JOURNAL. 768 (2): -. Whalen, D. J., and C. Fryer. The Observational Signatures of Primordial Pair-instability Supernovae. 2010. In Deci- phering the Ancient Universe with Gamma-ray Bursts ; 19-23 April 2010 ; Terrsa, Kyoto, Japan. Vol. 1279, p. 116. Whalen, D. J., and C. Fryer. The Observational Signatures of Primordial Pair-Instability Supernovae. 2010. DECI- PHERING THE ANCIENT UNIVERSE WITH GAMMA-RAY BURSTS. 1279: 116. Woodward, null., null. Dimonte, null. Rockefeller, null. Fry- er, null. Dimonte, null. Dai, and R. J. Kares. Simulating Rayleigh-Taylor (RT) instability using PPM hydrodynam- ics @scale on Roadrunner (u). 2011. DOE. Yuan, Q., S. Liu, Z. Fan, X. Bi, and C. Fryer. MODELING THE MULTI-WAVELENGTH EMISSION OF THE SHELL-TYPE SUPERNOVA REMNANT RX J1713.7-3946. 2011. AS- TROPHYSICAL JOURNAL. 735 (2): 120. Yuan, Qiang, Siming Liu, Zhonghui Fan, Xiaojun Bi, and C. Fryer. Modeling the multi-wavelength emission of the shell-type supernova remnant RX J1713.7-3946. 2011. Astrophysical Journal. 735 (2): 120 (9 pp.). Yuan, Qiang, Siming Liu, Zhonghui Fan, Xiaojun Bi, and C. Fryer. Modeling the multi-wavelength emission of the shell-type supernova remnant RX J1713.7-3946. 2011. Astrophysical Journal. 735 (2): 120 (9 pp.). 596

Page 597

Physics Directed Research Final Report Network-centric Quantum Communications Richard J. Hughes 20110046DR Abstract ensuring that control, access and network management Trends in communications networks are presenting new can only be performed by authorized parties. Critical to cyber security concerns that are challenging to meet these functions is the secure distribution of the secret with conventional cryptography. In principle, quantum random number sequences known as cryptographic keys cryptography using single-photon communications, to authorized users: secret keys are used as parameters with its security rooted in inviolable laws of quantum within algorithms to encrypt, authenticate or sign data physics, could meet these challenges, provided it could that must be secure. Protocols to generate a shared evolve from the current point-to-point instantiations to secret key between two parties are therefore of funda- a form compatible with multi-node networks. Previous mental importance to network communications. Many approaches to quantum communications (QC) networks of today’s widely-used protocols base their security based on a mesh of point-to-point links lack scalability, on assumptions about the computational power of an require dedicated optical fiber, are expensive and not adversary, such as the difficulty of factoring large inte- amenable to mass production, only provide one of the gers. Such classical schemes lack “forward security” –an cryptographic functions (key distribution) needed for attacker could record the protocol execution steps, break secure communications, and so have elicited limited the security on faster computers in the future, and com- practical interest. We have invented and experimentally promise past protocol sessions and the communications demonstrated a new approach to information assurance they had secured. Protocols with information-theoretic called network-centric quantum communications (NQC), (forward) security, are therefore highly desirable. This achieving at least a three-year lead over the state-of- is possible with quantum communications (QC) [1, 2], the art in QC research internationally. NQC is scalable, which requires a channel that transmits quantum states resilient, deployable, amenable to mass production, of single-photons (polarization for example), with negli- and ultimately affordable. We have shown that NQC gible loss of quantum coherence. Eavesdropping on QC can solve new network security challenges in the critical can be detected, because of the inevitable disturbance infrastructure control sector, in particular. Long-dura- of quantum states caused by measurements, and attack- tion/high-value network security environments within ers cannot passively monitor the quantum signals owing the NNSA and certain DoD complexes, and hand-held to the indivisibility of elementary particles, or faithfully device security are other NQC application areas that are copy them because of the quantum “no cloning” theo- being explored. We have protected the NQC intellectual rem. property with twenty nine US and foreign patent ap- QC protocols combine single-photon transmissions with plications, and joint development opportunities with US non-secret conventional communications to accomplish industry are under discussion. specific cryptographic tasks. (Figure 1) The state-of-the art in QC is a specific protocol known as quantum key Background and Research Objectives distribution (QKD) [1] for establishing a shared secret Communications networks have revolutionized the way key between two mutually trusting parties. In optical we work, live, operate computer systems and run our fiber QKD transmission ranges of more than 200km have national infrastructure. But hardly a day goes by without been achieved [3]; our team has previously achieved a a report highlighting the pressing need for improved record 144km with the strongest security assurances cyber security technologies to protect our economic and [4]. Small companies in the US and Europe sell point-to- National security. Cryptography provides security for the point QKD products [5, 6] for applications such as secure protection of data in motion and in storage, as well as 597

Page 598

off-site back-up. Research groups in the US [7], Europe QC hardware. Our first-generation, modularly-integrated [8], China [9] and Japan [10] have explored “trusted QKD QKarD is a fiber-coupled device, which incorporates a networks”, formed from a mesh of trusted nodes between distributed feedback laser and modulator, that produces which two-party QKD is performed on dedicated opti- short (< 1ns), single-photon (mean-photon number < 1), cal fibers, distinct from the data network being secured. polarized light pulses at telecom wavelengths (1,550nm), A user can establish a secret key with a distant node by at a 10 MHz repetition rate in our NQC test bed. (Figure transporting it, encrypted using QKD, through intermediate 3) This device is produced by a small-scale manufacturing nodes. Because the key exists “in the clear” within each process and could be incorporated into end devices. We intermediate node, these nodes must be physically secure are designing a next-generation, monolithically-integrated and trusted. This approach is expensive and does not scale QKarD, which will be an order of magnitude smaller in well: adding users requires additional physically-secure each linear dimension, and amenable to much lower cost trusted nodes; and a duplication of network infrastructure. mass production. In our test bed the QC data processing is Furthermore, it is incompatible with existing network ar- handled by PCs at each node, with Trent’s software archi- chitectures. Underlying these shortcomings is the absence tecture capable of accommodating up to 100 client QKarDs of a method for establishing trust between users. Our on that platform, and more than 1,000 with server-class NQC invention [11, 12, 13] overcomes these limitations to hardware [16]. But we have also demonstrated that the bring the unrivalled forward security of QC to networks in processing software in each client can be transitioned to scalable, deployable, robust form and includes advances in a field-programmable gate array (FPGA), providing further key management that bring new capabilities that take QC miniaturization for use cases such as hand-held device beyond key distribution. security. Scientific Approach and Accomplishments Trent, the NQC receiver, incorporates passive polariza- tion analysis with single-photon detectors, operated at a In NQC, quantum communications between each of N cli- typical detection efficiency of ~ 15%. In our NQC test bed, ent nodes and a central server node at the physical layer which we have operated continuously for 3 years, we have support a quantum key management (QKM) layer [14], time-multiplexed Trent with three QKarD transmitters, which in turn enables secure communications functions Alice, Bob and Charlie, over 50 km of single-mode fiber; (confidentiality, authentication and non-repudiation) at the larger numbers of clients could be accommodated with application layer between N2 client pairs [11]. (Figure 2) a combination of temporal and wavelength multiplexing. The NQC “hub-and-spoke” topology is widely encountered Fiber birefringence necessitates polarization tracking and in optical fiber networks, and permits a hierarchical trust compensation [17], which in our test bed is performed in- architecture that allows the server (the “hub”) to act as the band at Trent, temporally multiplexed with the QC signals, trusted authority (TA, “Trent”) in cryptographic protocols resulting in a QC duty cycle of 20%. (When implemented for authenticated key establishment. (This avoids the poor out-of-band we expect our novel tracking scheme [18] will scaling of previous approaches that require a pre-existing compensate even the much higher polarization rates that trust relationship between every pair of nodes.) By making will be encountered in challenging environments such as Trent a single multiplexed QC receiver, and the client nodes optical fibers strung on telephone poles.) (Alice, Bob, Charlie, etc.) QC transmitters, NQC amortizes the cost and complexity of one of the most demanding QC We operate several QC protocols on each client-Trent link components – the single-photon detectors – across mul- in our test bed. We have implemented quantum secret tiple network nodes [12]. In this way the NQC architecture sharing (QSS) [19] over fiber spans as large as 50km is scalable in terms of both quantum-physical resources between a client and a single, logical Trent node. The and trust. vital trusted authority, Trent, is comprised of two physical nodes: an intermediate, integrated-photonics modulation In previous optical fiber QC instantiations information bits node and a QC receiver node [20]. With QSS a client’s key (quantum bits or “qubits”) have been encoded in the opti- is split between the two physical parts of Trent, both of cal phase of single-photon states. This has required inter- which would be required to cooperate to reconstruct the ferometric stability and inevitably necessitates bulky and key. In this way compromise of either one of the two parts expensive hardware. Instead, for NQC we encode informa- of Trent would not compromise the client’s key [14]; an at- tion bits in photon polarization states, allowing the QC tacker would have to successfully compromise both nodes. transmitters – referred to as QKarDs - to be miniaturized (Splitting of the key into additional shares would be pos- and fabricated using integrated photonics methods [15]. sible with more intermediate nodes.)While other research- This opens the door to a manufacturing process with its at- ers have previously demonstrated some elements of a QSS tendant economy of scale, and ultimately much lower-cost 598

Page 599

protocol, our implementation is the first that is complete, shares with Trent through QKD. Trent can calculate V from correct, verifiable and secure. However, for brevity we will S, and provide it to Bob, authenticated with one of the describe a simpler test bed configuration in which a QKD Bob-Trent QKD keys. When Bob receives the signed mes- protocol is performed on the client-Trent links. sage from Alice he can then verify her signature using V. With the inclusion of the digital signature capability, NQC In other advances we have implemented a quantum can provide the full functionality of a public key infrastruc- identification protocol, QID [21], between each client and ture [14]. Trent, providing a mechanism for client device enrollment and revocation [14, 21] on the NQC network. Following Impact on National Missions successful completion by the client of the QID challenge- By transferring key management functions to the quan- response with Trent, a short authentication key is gener- tum-physical from the computation realm, NQC protocols ated enabling client-Trent QKD protocols to be performed. have intrinsically much lower latencies than their conven- These protocols allow users anywhere on the network to tional counterparts. This attribute is of high significance as share keys and perform all of the standard cryptographic a solution to the hard cyber security challenges of critical functions we use on a network such as digitally signing infrastructure protection, which are a top national security documents and authenticating messages or software pack- concern. The NQC project has already led to a follow-on ages with the security guarantees of quantum keys. We project with DOE’s Cyber Security for Energy Delivery have developed and performed analysis of a new type of Systems (CEDS) program. Energy delivery systems such as potential security issue in the implementation of the basic the SmartGrid have requirements for the secure, low-la- QKD protocol, extending the security guarantees in the tency (~ few ms) delivery of data from, and commands to, face of a new, general class of adaptive attacks [22]. Internet-connected control devices. In the electricity trans- mission sector, for example, phasor measurement units Following the generation of shared secret keys between (PMU) provide information 30 times per second to control each client and Trent using QKD (or QSS), NQC protocols centers (phasor data concentrators, PDC) on the phase of at the QKM layer provide keys for client-to-client secure the electricity at locations throughout the grid. There is communications at the application layer [14]. For example, a low-latency data origin authentication requirement for suppose Alice requires a secret key to encrypt a message the multicast of this PMU data to PDCs. These require- to Bob using the AES algorithm [23], and/or authenticate ments are very challenging with present-day cryptography the message using the HMAC algorithm [24]. (Other sym- on standard processors, without compromising either the metric key encryption/authentication algorithms could be security assurances or the quality-of-service: public key sig- used instead.) For definiteness, we assume that this will natures are too slow, for example; while message authen- be a 256-bit key. First, Alice and Trent parse the results of tication codes using a common group key are vulnerable their QKD session into a 256-bit key, KA. Similarly, Bob and to compromise of a single node. (Similar security/latency Trent parse the results of their QKD session into a distinct requirement mismatches exist throughout the electric 256-bit key KB. Trent knows both keys and can provide grid, as well as in Supervisory Control and Data Acquisition non-secret, classical information to Alice, namely a list (SCADA) systems generally.) Newer, lightweight cryptog- of the bit positions in KA where her key bits need to be raphy could meet the requirement, but with conventional flipped to match Bob’s key. This information, which we call approaches the key management would be impractical. a pair key, enables Alice to securely transform her key, KA, However, optical fiber is widely-deployed throughout the into the key KB that she needs for communications with electric grid, and so NQC with its essentially unconstrained Bob. Important security advantages of this QKM protocol ability to supply fresh, forward-secure keys on-demand is are that: it does not require pre-placement of long-term capable of supporting lightweight cryptography for these secret keys shared by a client and Trent; and each new applications. In December 2012 at the US Department of client-to-client key has no algorithmic heritage in any previ- Energy’s Trustworthy Cyber Infrastructure for the Power ous key. Trent could even go offline after publishing a look- Grid (TCIPG) test bed at the University of Illinois Urbana- up table of clients’ pair keys, and the clients would still be Champaign (UIUC) we demonstrated that our NQC tech- able to execute the above key agreement protocol using nology could provide the necessary security, well within previously-generated, and securely-stored quantum keys. the required latency, for PMU communications. We have also invented a quantum digital signature (QDS) NQC can address long-duration/high-value security needs protocol [14] that allows Alice to sign messages to Bob. within many network environments, such as: certain DoD Alice requires a secret signing key, S, to calculate her signa- and DOE/NNSA complexes; between government agencies ture, and Bob requires Alice’s non-secret signature veri- in the Washington, DC area; financial networks; and for fication key, V. In QDS Alice’s S key is one of the keys she 599

Page 600

secure cloud computing. NQC can also provide multi-level security within optical network environments. Other exam- ples include: IAEA treaty monitoring; within a US Embassy compound; or on board a military or commercial aircraft. NQC could as well enable a hand-held QC device to be used for identification, authentication, access control and secure telephone calls. NQC is extensible to other types of networks: an aircraft or satellite acting as a trusted node could establish ad hoc secure networks of ground, sea and air-based users on a continental or even a global scale. Throughout the NQC project we have been mindful that our inventions will only have practical value as cyber security solutions if they can ultimately be transitioned to the private sector. We have been diligent in protecting the intellectual property that we have created, filing twenty nine US and foreign patent applications. NQC technology transfer discussions with several major US corporations are already well-advanced. Figure 2. Network-centric quantum communications (NQC) architecture. (See text for details.) Figure 1. In a QC protocol random bits are encoded at transmit node (“T”) in orthogonal photon states (linear polarization in this case), using either the rectilinear basis (0° or 90°), or the diagonal basis (-45° or +45°). For each photon receive node (“R”) makes a random choice of decoding basis, either rectilinear or diagonal. The detected bits are post-processed with conventional communications. Figure 3. Generation 1 quantum smartcard or QKarD. (See text for details.) 600

Page 601

References 14. Nordholt, J. E., C. G. Peterson, R. T. Newell, and R. J.

1. Bennett, C. H., and G. Brassard. Quantum cryptogra- Hughes. Quantum communications system with inte- phy: public key distribution and coin tossing. 1984. In grated photonics devices. 2013. US patent application Proceedings of the IEEE International Conference on US2013/055430. Computers, Systems and Signal Processing. (Bangalore,
2. Nordholt, J. E., R. J. Hughes, J. M. Riese, C. M. Ahrens, India, 1984). , p. 175. Piscataway, NJ: IEEE. C. G. Peterson, and J. W. Harrington. Scalable software
3. Stucki, D., N. Walenta, F. Vannel, R. T. Thew, N. Gisin, architecture for quantum cryptographic key manage- H. Zbinden, S. Gray, C. R. Towery, and S. Ten. High rate, ment. 2013. US patent application US2013/055356. long-distance quantum key distribution over 250 km of
4. Nordholt, J. E., C. G. Peterson, R. T. Newell, and R. J. ultra low loss fibres. 2009. New Journal of Physics. 11: Hughes. Polarization tracking system for free-space op-

tical communication, including quantum communica- 3. Rosenberg, D., C. G. Peterson, J. W. Harrington, P. R. tion. 2013. US Patent application US20130083925 A1. Rice, N. Dallmann, K. T. Tyagi, K. P. McCabe, S. Nam, B. 17. Nordholt, J. E., C. G. Peterson, R. T. Newell, and R. J. Baek, R. H. Hadfield, R. J. Hughes, and J. E. Nordholt. Hughes. Great circle solution to polarization-based Practical long-distance quantum key distribution sys- quantum communication (QC) in optical fiber. 2013. US tem using decoy levels. 2009. New Journal of Physics. patent application US 20130084079 A1. 11: 045009. 18. Hillery, M., V. Buszek, and A. Berthiaume. Quantum 4. MagiQ Technologies. 2013. MagiQ home page. Secret Sharing. 1999. Physical Review A. 59: 1829. 5. idQuantique, SA. 2013. idQuantique home page. 19. Schmidt, C.. Experimental single qubit quantum secret 6. Elliot, C.. Building the quantum network. 2002. New sharing. 2005. Physical Review Letters. 95: 230505. Journal of Physics. 4: 46. 20. Hughes, R. J., J. E. Nordholt, C. G. Peterson, J. T. Thrash- 7. Peev, M.. The SECOQC quantum key distribution er, R. T. Newell, J. T. Yard, and R. D. Somma. Multi- network in Vienna. 2009. New Journal of Physics. 11: factor authentication using quantum communication. 075001. 2013. US patent application US2013/055410. 8. Wang, S.. Field test of wavelength-saving quantum key 21. Somma, R. D., and R. J. Hughes. Security of decoy-state distribution network. 2010. Optics Letters. 35: 2454. protocols for general photon-number-splitting attacks. 2013. Physical Review A. 87: 062330. 9. Sasaki, M.. Field test of quantum key distribution in the Tokyo QKD Network. 2011. Optics Express. 19: 10387. 22. Announcing the Advanced Encryption Standard (AES). 2001. National Institute of Standards and Technology 10. Hughes, R. J., J. E. Nordholt, and C. G. Peterson. Secure (NIST) FIPS-197. multi-party communication with quantum key distribu- tion managed by trusted authority. 2013. US Patent 23. The keyed-Hash Message Authentication Code 8,483,394 awarded July 9, 2013. (HMAC). 2002. National Institute of Standards and Technology (NIST) FIPS 198. 11. Nordholt, J. E., R. J. Hughes, R. T. Newell, C. G. Peter- son, D. Rosenberg, K. P. McCabe, K. T. Tyagi, and N. Publications Dallmann. Quantum key distribution using card, base Hughes, R. J., J. E. Nordholt, K. P. McCabe, R. T. Newell, C. station and trusted authority. 2013. US Patent Applica- G. Peterson, and R. D. Somma. Network-centric quan- tion US20130101119 A1 published April 25, 2013.. tum communications with application to critical infra- structure protection. 2013. ArXiV. 12. Hughes, R. J., J. E. Nordholt, K. P. McCabe, R. T. New- Hughes, R. J., J. E. Nordholt, and C. G. Peterson. Secure ell, C. G. Peterson, and R. D. Somma. Network-centric multi-party communication with quantum key distribu- quantum communications with application to critical tion managed by trusted authority. 2013. US Patent infrastructure protection. 2013. ArXiV. 8,483,394 awarded July 9, 2013. 13. Hughes, R. J., J. E. Nordholt, and J. T. Thrasher. Quan- Hughes, R. J., J. E. Nordholt, and J. T. Thrasher. Quantum tum key management. 2013. US patent application US key management. 2013. US patent application US 20130083926 A1. 20130083926 A1. 601

Page 602

Hughes, R. J., and J. E. Nordholt. Refining quantum cryp- tography. 2011. Science. 333: 1584. Nordholt, J. E., C. G. Peterson, R. T. Newell, and R. J. Hughes. Great circle solution to polarization-based quantum communication (QC) in optical fiber. 2013. US patent application US 20130084079 A1. Nordholt, J. E., C. G. Peterson, R. T. Newell, and R. J. Hughes. Polarization tracking system for free-space op- tical communication, including quantum communica- tion. 2013. US Patent application US20130083925 A1. Nordholt, J. E., R. J. Hughes, R. T. Newell, C. G. Peterson, D. Rosenberg, K. P. McCabe, K. T. Tyagi, and N. Dallmann. Quantum key distribution using card, base station and trusted authority. 2013. US Patent Application US20130101119 A1 published April 25, 2013.. Smith, G., J. Smolin, and J. Yard. Quantum communica- tion with Gaussian channels of zero quantum capacity. 2011. Nature Photonics. 5: 624. Somma, R. D., and R. J. Hughes. Security of decoy-state protocols for general photon-number-splitting attacks. 2013. Physical Review A. 87: 062330. Ursin, R., and R. J. Hughes. Sharing quantum secrets. 2013. Nature. 501: 37. 602

Page 603

Physics Directed Research Final Report First Direct Measurement of Particulate Evolution in Isochorically Heated Dense Plasma (U) Brian J. Albright 20130082DR Abstract where the plasma has no time to expand and where This one-year project was intended to determine the we know to good accuracy what state the plasma is in. feasibility of a completely new class of high-energy- Such experiments are possible on the LANL Trident laser, density-physics (HEDP) experiments using laser-based, thanks to innovations in recent years (supported by very rapid (i.e., over time scales of trillionths of seconds) LDRD) in laser-generated ion beam technology. The Tri- heating of matter. The primary objective of this work dent short-pulse beam, of order 80 Joules in energy and was to see whether we could study plasma-phase “elec- 0.5 trillionths of a second pulse duration, when focused trostatically enhanced mix” (EEM) at material interfaces, onto a very thin (of order 100 nanometers thickness), a problem of both widespread application and a great planar target, can create an exotic type of plasma within unsolved scientific problem in its own right. The work the target where the electrons are highly relativistic, involved both a theoretical/modeling and experimental with energies of order 100 times their rest mass, and the study. Both were highly successful, enough so that this plasma within the target can enter the “relativistically work motivated an LDRD new start in FY14 to extend our induced transparency” (RIT). RIT enables very efficient pilot experiments and, we hope, lead to the resolution coupling between laser and plasma and enables the of whether and for how long EEM dominates the evolu- formation of large electric fields with amplitudes of up tion of plasma. Proof-of-principle integrated experi- to 100 trillion volts/m. These fields can accelerate ions ments were conducted which involved both creation of and make collimated ion beams of MeV to GeV ions over multi-species plasma and the x-ray imaging of an evolv- time scales shorter than a trillionth of a second. ing plasma interface. Also, radiation hydrodynamics and These ion beams can be used to rapidly heat target plasma kinetic simulations were used to design relevant matter into the plasma state, enabling a brand-new, follow-on experiments that extend this proof of prin- heretofore unexplored class of plasma experiments with ciple. unprecedented control over the initial conditions. The purpose of our project was to assess the feasibility of Background and Research Objectives such an experiment, using the Trident short pulse laser One of the biggest challenges to doing controlled HEDP to create matter and one of the long-pulse beams to experiments on facilities such as the National Ignition make x-rays to be used to image the evolving plasma. Facility at Lawrence Livermore National Laboratory, the A major objective was to test whether the key diagnos- Z Machine at Sandia National Laboratory is that the tic, the x-ray backlighting and gated x-ray imager, was time scale over which plasma can be made (of order capable of operating in the high electromotive pulse ten billionths of a second) is the same time scale over (EMP) environment of a Trident short-pulse experiment which the matter evolves. As a result, what plasma one and if so, whether it could indeed image the evolving makes often has large variations in density and tempera- plasma. Additionally, kinetic simulations using the LANL ture, making it very difficult to design controlled, “unit- VPIC particle-in-cell code and radiation-hydrodynamics physics” experiments that rigorously test one’s physics simulations using the LASNEX code were used to design hypotheses. This is an endemic problem with the use of and refine follow-on experiments to extend this “proof- such facilities. of-principle.” However, if we could somehow create dense plasma on The overarching objective was to demonstrate the real far shorter time scales—trillionths instead of tens of bil- potential to solve one of the great problems in our lionths of seconds—then we could design experiments 603

Page 604

understanding of plasma-phase mix: namely, the role of Our theoretical/modeling accomplishments: large, electrostatic fields at material interfaces in driving We performed a suite of collisional kinetic simulations of the mix process. This is the biggest difference between high-Z/low-Z interface evolution in plasma. These simula- plasma-phase mix and fluid-phase mix, one that has not tions (Figure 2) support the hypothesis that EEM can be a been explored previously. As indicated below, this objec- dominant actor in governing plasma-phase mix at inter- tive was met. faces. Scientific Approach and Accomplishments We have begun a radiation-hydrodynamics design of Our experimental accomplishments (Figure 1 for a sche- experiments to ensure that our follow-on experiments matic of the apparatus): can indeed sample relevant plasma conditions for ad- dressing these problems (Figure 2). Work is underway to We designed and fielded an integrated target holder understand the implications of the use of the novel ion (Figure 1) capable of housing the three key components beams discovered above in place of our carbon and proton of our experiment: the thin foil target that interacts with beams. the incident short pulse beam to create the ion source, the solid-density sample of material to be heated by the Impact on National Missions ion beam, and the x-ray source foil that is irradiated by the This work has three key areas of impact on missions: incident long-pulse beam. Key access ports in this mount were also designed to allow diagnostic lines of sight for Establishing the feasibility of using short-pulse laser-driven the Thompson parabola (to measure ion energy loss in the particle beams for the creation of HEDP plasma in mission- target) and the gated x-ray imaging (GXI) diagnostic. relevant conditions. Every major HEDP facility in the DOE complex has (or has planned) a co-located short-pulse We demonstrated that the GXI diagnostic is capable laser, so this pilot work may open up a brand new set of of operating in the high EMP environment of a Trident experiments and vastly expand the utility of these facilities. short-pulse experiment and that the photon counts in our experiment would be adequate to image the target. Demonstration of “proof of principle” experiments that get This was a nontrivial accomplishment, and in doing this at one of the most important problems in our understand- work we discovered a minor parallax issue that will need ing of mix behavior in plasma media. Indeed, the physics of to be resolved in future experiments of this type. We also plasma-phase mix undergirds much of our uncertainty of identified an x-ray fiducial from the relativistic electron nuclear performance, both in inertial confinement fusion background that we will exploit in future experiments for and other applications, and thus the design of controlled, timing. unit-physics experiments in these areas is vital. We performed initial design of an optical/UV pyrometer If successful, follow-on experiments will be used to extend (for making temperature measurements of our heated the NNSA Advanced Simulation and Computing (ASC) com- plasma) to be used in follow-on experiments. While mon mix framework in the LANL multi-physics codes to preliminary, this allowed us to make a down-selection of incorporate plasma-phase dynamics. different approaches to pyrometry. We fielded an integrated proof-of-principle experiment that demonstrated all of the essential features of a follow- on experiment. This experiment employed the use of a thin tungsten wire surrounded by a plastic foam. Both were irradiated by an laser generated ion beam (carbon and protons) and the evolution of the foam was captured using an x-ray backlighter and the GXI diagnostic. This experiment demonstrated conclusively the feasibility of experiments of this type in resolving the question of the role of EEM in plasma-phase mix (Figure 1). We serendipitously uncovered a new means of improving our ion heater beam and allowing for access of a wider range of plasma temperatures in our sample. This work is being prepared for publication. 604

Page 605

a) d) b) e) c) Figure 1. Experimental apparatus and data. Shown are: a) a schematic of the experiment, including key diagnostics; b) the GXI diagnostic used to measure the expanding plasma; c) a schematic of the target mount; d) multi-component target used in pilot experiments (tungsten wire surrounded by plastic foam); e) x-ray images of the evolving multi-component plasma target, showing position of the plasma interface initially and at 20 ns after heating. Z (cm) Z (cm) 605 )mc( R Z (cm) Z (cm) )mc( R Meadows, C. Wang, T. Ditmire, and J. C. Fernandez. Laser-driven ion acceleration from relativistically trans- parent nanotargets. 2013. NEW JOURNAL OF PHYSICS. 15: -. Jung, D., L. Yin, B. J. Albright, D. C. Gautier, S. Letzring, B. Dromey, M. Yeung, R. Horlein, R. Shah, S. Palaniyap- pan, K. Allinger, J. Schreiber, K. J. Bowers, H. C. Wu, J. C. Fernandez, D. Habs, and B. M. Hegelich. Efficient carbon ion beam generation from laser-driven volume acceleration. 2013. NEW JOURNAL OF PHYSICS. 15: -. Jung, D., L. Yin, D. C. Gautier, S. Letzring, B. Dromey, R. Shah, S. Palaniyappan, T. Shimada, R. P. Johnson, J. Schreiber, D. Habs, J. C. Fernández, B. M. Hegelich, and B. J. Albright. Laser-driven 1 GeV carbon ions from pre- heated diamond targets in the Break-Out Afterburner regime. 2013. Physics of Plasmas. 20: 083103. t ~10 ps t ~ 100 ps density temperature Figure 2. Simulation data. The four left panels are from Lasnex radhydro modeling of a future experiment, indicating the remarkable plasma uniformity made possible by laser-ion- beam-based isochoric heating. The rightmost panel is from a VPIC (collisional) kinetic plasma simulation of the evolution of the “mix layer” near a plasma interface. The presence of large electrostatic fields (i.e., the EEM process) enhances the rate of atomic mixing compared with that of simple diffusive mixing (shown in red). Publications Fernández, J. C., B. J. Albright, F. N. Beg, M. E. Foord, B. M. Hegelich, J. J. Honrubia, M. Roth, R. B. Stephens, and L. Yin. Fast ignition with laser-driven proton and ion beams. To appear in Nuclear Fusion. Hegelich, B. M., I. Pomerantz, L. Yin, H. C. Wu, D. Jung, B. J. Albright, D. C. Gautier, S. Letzring, S. Palaniyappan, R. Shah, K. Allinger, R. Horlein, J. Schreiber, D. Habs, J. Blakeney, G. Dyer, L. Fuller, E. Gaul, E. Mccary, A. R.

Page 606

Physics Early Career Research Continuing Project Understanding Energetic Ion Transport and Loss in Natural and Artificial Radiation Belts at Low Altitudes Misa Cowee 20120715ECR Introduction cause its results will fill a gap in our space weather Energetic particles in natural and artificial radiation belts modeling effort. Energetic protons in the inner radiation can cause damage to the civilian and military assets belt are not currently included in such models, but pose whose orbits intersect those belts. To better predict a significant hazard to spacecraft in orbit. Space weather the threat posed to satellites, we must understand the effects on sensors can compromise our Nuclear Detona- physical processes which govern the formation and loss tion Detection (NDD) remote sensing program, so under- of these radiation belts. Observations indicate that standing the potential hazard from natural or artificial geomagnetic storms can cause energetic (MeV) protons radiation belts from a High Altitude Nuclear Explosion trapped in the natural inner radiation belt to become (HANE) is critical. NOAA provides space weather fore- de-trapped and lost on timescales faster than any known casts, so improved modeling capability in the inner belts physical processes associated with the Earth’s radiation would increase the accuracy of such predictions. NASA’s belts. Recent analytic studies indicate that these pro- goal of advancing space science and our understanding tons may be rapidly de-trapped because their motion is the near-Earth space environment is directly supported modified by sudden changes in the magnetic field during by this research. Predicting natural or man-made space magnetic storms, in a process called magnetic “field line weather hazards are also important for commercial sat- curvature” (FLC) scattering. Our goal is to carry out par- ellites, whose damage could results in severe economic ticle simulation tests for FLC scattering to better under- losses. stand the physics of this loss mechanism and the magne- Progress tospheric conditions which cause it. Additionally, we will use the simulation to investigate the role of FLC scatter- In the last year, we have continued development and ing in the formation of artificial radiation belts produced validation of our particle tracing simulation code. We after a High Altitude Nuclear Explosion (HANE). Obser- have expanded the code to include several magnetic vations of the Starfish nuclear test of 1962 indicated that field models and created new diagnostics. We have car- its artificial radiation belt extended above 10,000 km and ried out an initial study for the de-trapping of inner belt persisted for several years, damaging a large fraction of protons particles due to magnetic “field line curvature” the existing space infrastructure. The extent of this artifi- (FLC) scattering, and compared results to previously pub- cial belt was not predicted at the time and has not been lished observations and analytic models. Our simulation explained since. We will test a new hypothesis that the results have shown that FLC scattering is a viable mecha- ionized fission fragments became de-trapped due to FLC nism to explain the loss of inner belt protons after a scattering in the HANE-perturbed magnetic fields, allow- geomagnetic storm, which had been previously hypoth- ing them to travel to higher altitudes where they sub- esized. The results have helped us to identify possible sequently populated the artificial radiation belt. Thus explanations for the discrepancies between the current the results of this study can be applied towards greater analytic model predictions and the observations of the understanding of both natural and artificial radiation belt inner belt proton structure after geomagnetic storms. dynamics and will also quantify the FLC scattering pro- Our simulation results have shown that the inner belt cess for use in space weather models for the first time. protons which are transported radially in the system due to FLC scattering are not always lost to the atmosphere Benefit to National Security Missions (de-trapped), instead many of them become trapped but This research directly supports the DOE/NNSA/Nuclear at a different radial distance from where they started. Nonproliferation, DOC/NOAA and NASA missions be- 606

Page 607

Such transport effects are not included in current analytic can be included in space weather models. In FY14 we will models, which assume that all such particles are de- also determine if FLC scattering of HANE ionized fission trapped. Compared to the observations, the analytic mod- fragments is a viable mechanism to transport of these els predict a much steeper gradient in the proton fluxes fragments beyond the burst region. We will investigate with radial distance, and they predict that the de-trapping how this process depends on the burst location (altitude, extends to much lower radial distances. Our simulations, latitude, longitude), magnetic cavity configuration, fission therefore, can partially explain these discrepancies, and fragment spatial and velocity distribution, and fission frag- yield results in better agreement with observations than ment mass and charge state. We will compare our simula- the analytic models. tion results to the spacecraft and sounding rocket observa- tions of the artificial radiation belts taken after the HANE Another important result we have found is that the analyt- tests of 1958 to 1962 to quantify this effect so that it can ic models incorrectly assume that there is an easily identi- be included in artificial radiation belt models. fied “trapping limit” which clearly defines which proton populations will be de-trapped by FLC scattering during a Conclusion geomagnetic storm and which will not. Our simulations Our first objective is to understand the loss of energetic show the very concept of a trapping limit may not apply protons in the natural radiation belts due to magnetic for particles with chaotic trajectories, and the values of the “field line curvature” (FLC) scattering and identify what trapping limits given by the analytic models do not even conditions cause it. Understanding this mechanism allows accurately predict the de-trapping behavior in the simula- us to include it in radiation belt models, to better predict tion. the hazard posed to satellites. Our second objective is to understand how FLC scattering of fission fragments pro- Based on our results thus far, we suggest that any predic- duced by a High Altitude Nuclear Explosion (HANE) affects tive model of the de-trapping of the inner proton belt due artificial radiation belt formation. With proliferation of to FLC scattering during geomagnetic storms must account nuclear weapon and missile technologies, it is important for the relocation of trapped protons across the belt in to be able to predict the potential impact on our space addition to their loss. We have presented this research at infrastructure from HANEs. two conferences, the American Geophysical Union meeting in December 2012 and the Geospace Environment Model- Publications ing workshop in June 2013. A manuscript describing the Cowee, M. M., K. Liu, W. Tu, R. H. Friedel, and G. D. Reeves. first results is in preparation and we expect to submit it for “Test Particle Simulations of mu-Scattering in the In- publication by the mid-summer 2013. ner Proton Belt”. Presented at American Geophysical Union Fall Meeting 2012. (San Francisco, CA, Dec 3-7, Future Work 2012). Our goal is to test the hypothesis that magnetic “field line Tu, W., M. M. Cowee, and K. Liu. “Test particle simulations curvature” (FLC) scattering is the mechanism responsible of field-line-curvature scattering in the inner proton for both the recent observations of sudden de-trapping of belt”. Presented at Geospace Environment Modeling inner radiation belt energetic protons during geomagnetic Workshop. (Snowmass, Co, June 16-23, 2013). storms as well as the outward transport of ionized fission fragments after a High Altitude Nuclear Explosion (HANE) which explains the fifty year-old mystery of why the artificial radiation belts produced by the HANE test shots extended to such high altitudes. During FY14, we will use the simulation code we built in FY12 and FY13 to follow the gyromotion of energetic ions and apply it to quiescent and disturbed magnetospheric configurations to under- stand how de-trapping is sensitive to perturbations in the magnetic field. During FY14, we will continue to quantify how FLC scattering and de-trapping of natural inner radia- tion belt proton depends on geomagnetic storm strength and proton energy and pitch angle and compare our simu- lation results with observations from the HEO, SAMPEX and POES spacecraft to quantify the de-trapping conditions and effect on the proton fluxes so that this loss mechanism 607

Page 608

Physics Early Career Research Continuing Project Enabling Temporal and Spatial Shaping of Ultra-intense Laser Pulses for Next-generation, Compact Particle and Radiation Sources Rahul C. Shah 20120721ECR Introduction Progress Intense laser pulses are being studied for application to The most significant outcome of the first year has been next generation compact particle sources. At current, the preliminary work, both calculation and experimen- these pulses have smooth Gaussian envelopes, which in tal, on a multitude of topics allowing us to evaluate and time have duration and rise determined by laser physics, adapt the original ideas described in the proposal. and in space have dimension limited by classical focus- The calculation work began with a post-undergraduate ing optics. In this project we seek to explore possibilities summer student’s efforts to develop a basis for focus- enabled by new means of temporally and spatially con- ing an intense laser pulse to a near diffraction-limited centrating the light and develop the avenues by which to spatial flattop. To avoid non-linear effects we were most achieve them. At the onset, we had hypothesized that interested in an approach to spatial shaping based on plasma shuttering, a highly non-linear process, might be wave-front modification, which previously had been optimized by means of a spatially flat-top focus. While studied only in the geometric limit applicable for much we have preliminarily devised techniques for flattop larger focal dimensions. Because of the intractability of focusing, our particle-in-cell simulation (for Trident type the diffractive case, we turned to a genetic algorithm parameters) shows that the plasma and laser ripple, based approach. The technique proved successful in resulting in laser filamentation. With respect to Trident finding super-Gaussian type foci of the same dimension experiments, we are instead turning to the use of a low as the diffraction limited spot. Furthermore, we found density plasma (realized with nano-pore foams devel- that a single wave-front solution applied for all focusing oped thru this project) to controllably focus and shutter speeds. the light and enhance accelerated particle spectra. Sep- arately, we are also studying a means to manage impact While we seem to have initiated a promising approach of the plasma rippling by use of a multiple layered target to flat-top focusing, our work in simulating the laser- in which one material buffers these effects to the benefit plasma interaction using two-dimensional particle-in-cell of a second layer. shows that, for our conditions, the spatial flat-top does not create the hypothesized planar interaction. Instead, Benefit to National Security Missions the plasma ripples and the laser filaments. Given the A new level of spatial and temporal control over ultra- tendency for the plasma to render a planar interaction intense laser pulses is needed to advance our develop- unstable, we began exploring the use of thin plasma ment of next generation accelerators for laboratory slabs of nearly transparent density so as to optimize applications. The proposed work investigates how sharp the self-focusing and temporal transmission of the light rise and flat top spatial laser pulses can be generated by pulse. As our simulations show that electric fields are a unique combination of relativistic plasma physics and strongly driven on the rear side of the plasma slab, we optical engineering. In doing so, this project opens up extended our design to that of an integrated target, as a rich, first-of-its-kind technological tool which will yield opposed to the concept of a component “plasma optic.” large scientific possibilities for laser-particle accelera- Even with only a small effort towards this optimization tion, laser-x-ray generation, and their applications. We to date, we have designed two targets with ion spectra envision follow-on support from MARIE, Office of Fusion significantly improved (flat and peaked in energy) over Science (OFS), and the Domestic Nuclear Detection Of- typical exponential distributions. Study of the detailed fice (DNDO). underlying mechanisms of the ion spectral improvement 608

Page 609

remains as a task for the coming year. different charge-to-mass ratio will act to buffer rippling Rayleigh-Taylor like instabilities between the laser and To achieve such targets we have engaged LANL’s MST for plasma. In our own simulations we have observed that development of nano-pore foam and aero-gels. While such instabilities rapidly develop with a single species. The similar nano-pore targets have been used elsewhere, to published experiments lacked our ability to directly diag- our knowledge we are the first to have obtained the ~10 nose transparency by means of the reflected light tempo- micron thicknesses necessary for the optimized particle ral phase and envelope. Using our unique diagnostic, we spectra obtained in our simulations. Our up-coming ex- will examine if transparency time changes as we add an periments will diagnose hole-boring velocity, transparency additional 10 nm C layer to Al and Cu targets of 10-100 nm time, transmitted energy, beam foot-print, and ion-spectra. thicknesses. Last year, in lieu of these targets, we explored the use of stacked nm foils in which the first foil shaped the light Following the experiments we will conduct 2D simulations pulse for the second, “primary” target. By means of the to aid in our interpretation of the physics and publication. instantaneous Doppler shift recorded with reflected light If time permits we will revisit the work on spatial flattop field measurements we observed the un-mistakable sig- focusing. nature of a cleaner temporal rise on the second “primary” target foil. (An interferometric technique to measure this Conclusion rise was also developed but was not yet fielded on light We will use the temporal and spatial shaping of Trident transmitted thru an actual target). The reflected light, how- laser pulses in near critical density plasma to optimize ac- ever, also indicated that the intensity focused on the sec- celerated particle spectra. In addition to experimentally ondary target was reduced over the nominal levels. Such a measuring the spectra using state-of-the-art nano-pore combination of nm targets is difficult to optimize since the foam targets, we will use reflected and transmitted light actual experimentally achieved laser rise can create hydro- diagnostics to characterize the dynamics of the interac- dynamic expansion many times greater in scale than the tion for detailed comparison with simulation. We will also thickness of the nm shaping foil. The near-critical density study, in both simulation and experiment, how multiple targets, which our simulations already suggest enable a ion species in the target allow control over transparency richer ability to optimize the ion spectra than the nm foils, time. Outcomes are revised from original (electric field ought also prove far more robust given their much larger measurements in both transverse dimensions and test-bed scale length of microns. flat-top focusing) so as to support the evolved direction. Future Work In the first set of experiments we are examining the con- cept of plasma lensing, or the use of low density plasma slabs to control self-focusing and transmission time for optimization of accelerated ion spectra. These experi- ments are based on our own 2D VPIC simulations indicat- ing plateau spectra (in lieu of typical exponential drop-off) and quasi-mono-energetic spectra could be achieved with ~10 micron thickness and plasma densities about 10-30- fold less than typical solid density films. To implement such targets we have worked extensively with LANL’s MST who have fabricated 7-12 micrometer thick slabs of nano- pore SiO2 aero-gels fabricated at densities of ~100 mg/cc. To our knowledge we are the only researchers in the world with access to such targets. Preliminary results show that the targets can indeed be fielded and are producing flatter ion spectra. Along with ion measurements, we will mea- sure hole-boring velocity via the instantaneous Doppler shift of reflected light and the transmitted laser foot-print as well as temporal field. A second set of experiments has been inspired by re- cent suggestions in the literature that multiple species of 609

Page 610

Physics Early Career Research Continuing Project Investigating the Itinerant to Localized Crossover of the 5f Electrons in Plutonium Alloys and Compounds Paul H. Tobash 20120736ECR Introduction Benefit to National Security Missions Now more than ever, it is apparent that the 94th ele- The information gained from the project directly ad- ment in the periodic table, plutonium, offers challenges dresses the emergent phenomena present in Pu mate- in the areas of metallurgy and physics. Perhaps the big- rials. These are applicable to the goals outlined in the gest question that needs to be answered is how to un- Pu science strategy that are of interest to DOE and the derstand the true nature of the 5f electrons of Pu which Office of Basic Energy Science (BES). These include first have been shown to exhibit both itinerant (delocalized) understanding the electronic structure in Pu materials and localized behavior. The controversy on explaining and secondly, the dynamic behavior of these materials the absence of local magnetism in elemental Pu is the in extreme environments such as temperature, magnetic ultimate issue this work addresses. Electronic structure field, and pressure. With clean electro-refined Pu readily calculations have suggested the ~20% volume change available, this project will allow for a systematic ap- from δ-Pu to α-Pu can only arise from the formation proach to incorporate materials synthesis, analysis, and of magnetic moments (e.g. increase in lattice spacing). characterization to probe the evolution of magnetism The experimental evidence however is not consistent in mission-related alloy materials and ultimately lead to with this picture where no magnetism is detected. Thus, a more fundamental understanding of the 5f electrons we propose to synthesize Pu binary alloys that can be in Pu. This is a novel targeted approach to tracking the directly compared to elemental delta-Pu. Ultimately, we crossover with compositions close to δ-Pu to address the will be able to determine the critical compositions for Pu grand challenge for the dual nature of the 5f elec- a number of LANL mission-relevant binary alloys where trons. magnetism emerges and eventually vanishes, as one moves closer to “delta-Pu”–like compositions. After pre- Progress paring these alloys, we will probe this crossover behav- To date the project is halfway finished and at this point ior by measuring their physical properties, which include at least on track to gain the desired results by April magnetic susceptibility, heat capacity, and electrical 2014. The significant and most noteworthy work will be resistivity. The ability to “freeze out” the complicated completed in the second half of this project where the physics using low temperatures will be necessary in or- Pu-rich materials that were spelled out in the proposal der to study electronic correlations and detect the pres- will be synthesized and fully characterized. The impor- ence of magnetism. Thermal expansion measurements tance of these compounds is recognized from their close will be utilized via a new low temperature dilatometry relation to delta-Pu and other LANL mission related work setup in order to gain insight on volume changes result- in general. A new capability is also on track to come ing from the presence of magnetic moments. This allows online in the next few months and also to be used to for a link into the expected pressure dependence of the measure many of these newly prepared compounds. materials using the Ehrenfest relationship. This would These results will allow us to better probe the interesting be a significant part for the synergy of the experiment/ behavior of the actinide containing materials. theory framework of fundamental Pu physics because the invaluable input parameters for first principle band The current status of our synthetic efforts is halfway structure calculations based on Density Functional through the targeted compositions with all of the Pu- Theory (DFT) methods. lean compounds being made and measured. These compounds include PuGa3, PuSn3, and PuIn3. We have secured additional Pu starting material that will be used 610

Page 611

to prepare the rest of the compositions. We have also to attend 3 upcoming conferences where I will present synthesized a Pu6Fe compound which has a close relation- some of this new work outlined above. Pu Futures 2014, ship to the Pu-rich materials. The final compounds will be MRS2014, and the US/Russia Workshop are all important completed in the newly installed induction furnace at the venues to showcase this new work and possibly establish CMR facility. As a comparison for the delta-like composi- additional external collaborations for LANL. tions we are targeting, we have measured the specific heat and electrical resistivity on a newly cast delta-Pu sample. Future Work This will serve as a baseline for later doping studies in the Many of the goals which were outlined are currently being system as a function of the stabilizing element such as In, worked on. The PI is building and testing the induction fur- Sn, and Ga. nace melter which will be utilized to synthesize the mission relevant alloys that were outlined in the proposal. Once For the new measurement capability, we have finished the furnace is in place, the desired alloys will be prepared assembling the dilatometry probe that will be used to and then their physical properties will be characterized. measure thermal expansion of the alloys and compounds. These measurements will include magnetic susceptibility, The original designer of the cell has been invited to visit heat capacity, and electrical resistivity. Thermal expansion LANL in late August and run calibration tests and also to measurements will also be completed using the low tem- look at a new encapsulation setup for running Pu on the perature capacitance dilatometer which the PI is assem- probe itself. We will be testing the measurement setup in bling and later calibrating. This dilatometer is expected to a cold Quantum Design PPMS and then will be introducing be up and running in the next couple of months if testing is this setup to CMR to ultimately measure the Pu materi- carried out with satisfaction. The PI expects to obtain pre- als. A number of collaborations have also been estab- liminary results by December 2013 with a majority of the lished early on in this project. We have prepared samples alloys being synthesized and characterized. These results of PuSn3 to look at the valence of Pu and the effects of will indicate if the technique proposed is showing a mea- the ligand atoms around Pu. These measurements were sure of success. The alloys which are expected to be made completed through a postdoc working in the chemistry along with the temperature profile to be used (which is division. We also have plans to run some high pressure unique to each one) has been outlined and discussed with work on these compounds in the future that would re- other individuals on the project. The PI also expects to ob- quire the successful funding of a possible ER proposal for tain electronic structure data on many of these alloys from this work. The theorist that is working on the project has the theorist who is staffed on the project. These calcula- completed much of his band structure calculations on the tions will be completed once the samples are synthesized compounds we wanted to look at. We have also performed and the needed properties are measured experimentally. first-principles electronic structure calculations on the family of intermetallic Pu3X compounds (including Pu3Ga, The tasks for next fiscal year include:Synthesize all alloys Pu3In, and Pu3Sn). These compounds are isostructural to and compositions to be studied (Pu-rich line compounds each other, which enables us to understand systematically and delta-like alloys). Characterize these materials using the nature of Pu 5f localization and delocalization. We have powder X-ray diffraction, heat capacity and magnetic sus- established a connection between the lattice parameter ceptibility. Move the tested thermal expansion setup over and the degrees of 5f electron localization and identified to CMR and begin to measure Pu materials once encap- Pu 5f electrons are most localized in Pu3Sn in all three sulation setup has been sufficient to measure radioactive compounds based on the effective band width of Pu 5f materials. Begin to compile preliminary results in order to orbital as estimated from the partial density of states. As a find missing compositions to target magnetism. future plan, we will carefully analyze the relative strength of Pu 5f hybridization with Pu 6d (which is major source for Conclusion delocalization in elemental Pu solids) and further extend Once the targeted compounds are made, the subsequent our theory capability to tackle the most challenge problem measurements to probe the crossover of the 5f electrons of elemental Pu solid doped with Ga, In, Sn impurities for are relatively straightforward and are sufficient enough to the understanding of phase stability. make conclusions for the critical point of where magnetism disappears as we approach δ-Pu. We expect to: A paper on the new ‘115’ compounds and their relation to these compounds is in preparation. This will serve as 1) Find magnetism in the Pu3X compounds based on the a nice comparison to the binary alloys of this project and previous hypothesis for the expansion of the unit cell the ternary intermetallic compounds which were recently volume leading to a stronger coupling of the magnetic mo- synthesized in the last few months at CMR. I am planning ments (more in the localized regime). 611

Page 612

2. Track the emergence of magnetism in the thermal ex- pansion measurements. Publications Bauer, E., P. Tobash, J. Mitchell, and J. Sarrao. Single crystal growth of plutonium compounds from molten metal fluxes. 2012. Philosophical Magazine. 92 (19-21): 2466. 612

Page 613

Physics Early Career Research Continuing Project Deterministically Enhanced Monte Carlo for Radiative Transfer Allan B. Wollaber 20120738ECR Introduction simulations. The challenge goal of this work is to enable Hot problems (supernovae, inertial confinement fusion, first-ever high-fidelity radiation simulations of type 1a oxy-coal combustion) can require an accurate simulation supernovae that are of direct relevance to cosmologists, of radiative heat transfer, but modeling the nonlinear astrophysicists, and work supported by DOE’s Office of exchange of energy between matter and photon radia- Science (both in HEP and ASCR). This work is also ap- tion is a considerable algorithmic and computational plicable to the simulation of inertial confinement fusion challenge. The Implicit Monte Carlo (IMC) method has experiments performed at the National Ignition Facility emerged as the preferred approach for high-fidelity for DOE’s NSSA. Finally, the hybrid deterministic-Monte radiation simulations, but its underlying equations can Carlo approach for radiation transport is relevant to produce nonphysical solutions for large time steps, and other instances of neutral particle transport, such as oc- its stochastic implementation leads to statistical noise, curs in the simulation of nuclear reactors or of oxy-coal particularly in “cold” problem regions. We propose a combustion. novel way to use a low-order linear algebra (determinis- Progress tic) calculation to simultaneously (1) enhance the tem- poral accuracy of the IMC solution, (2) reduce its statisti- We implemented several solver and temperature-cou- cal uncertainty, and (3) accelerate the overall calculation. pling capabilities into Albireo (a standalone executable in Specifically, averaged interaction coefficients are gener- Asterisk). ated during the high order calculation and provided to • Solver options: “only IMC”, “only low order”, or “IMC a dimensionally reduced deterministic solver that then and low order” calculates an estimate of the upcoming radiation and material temperature solutions. The approximate mate- • Temperature coupling options: “just follow IMC” or rial temperature is then exploited to provide time-cen- “use temperature guess” tered (instead of time-explicit) interaction coefficients and a physics-based time-step controller for the upcom- • Provided a fixed point iteration option to nonlinearly ing IMC cycle. The approximate radiation temperature is converge the low order temperature utilized to reduce the statistical uncertainty in the IMC solution, to reduce the time to solution, and to identify • These features have been tested and reported important problem regions to further accelerate the cal- Presented the accuracy-enhancing temperature coupling culation. Thus primed, the IMC calculation is carried out methodology at the biennial American Nuclear Society’s and supersedes the approximate solution. We intend to Mathematics and Computation Division meeting (this is deploy this approach in production software to enable the premier conference for researchers in my field). first-ever, high-fidelity comparisons between telescopic observations of type IA supernovae and simulations of After initially running into a setback regarding the ac- their radiation-hydrodynamical evolution, which should curacy and positivity of the low order, tensor diffusion help solidify our current understanding of the history solution using the standard continuous finite element and structure of the observable universe. methodology (CFEM), we implemented a custom finite volume (FV) spatial discretization. The FV discretization Benefit to National Security Missions produces better results, although this unplanned effort The essential goal of this work is to accelerate and took about a month to implement. enhance the accuracy of high-fidelity radiative transfer 613

Page 614

Implemented an initial variance reduction capability • Alex Long was hired on as a summer student in May. He has already modified Asterisk and Jayenne project software to implement variance reduction (particle splitting and a lower weight window cutoff to control particle weights based on the deterministic, tensor- diffusion solution). Testing is now underway. Future Work During the remaining 10 months of this project (ending April 2014), we intend to: implement a wavefront-detection type algorithm to limit particle splitting: • A naive application of the deterministic solution as the weight-window center can lead to large particle splitting ratios that decreased the overall efficiency in preliminary work for multifrequency problems (see A.B. Wollaber, “Advanced Monte Carlo Methods for Thermal Radiative Transfer” PhD Thesis, The University of Michigan, 2008). Test the methodology on a supernova simulation: • Expose an interface that makes this methodology call- able from a host application • Work with Chris Fryer et. al. to deploy this methodol- ogy on a “real” supernova light curve simulation Conclusion The anticipated result is production software that auto- matically enhances the accuracy and efficiency of high- fidelity radiation-hydrodynamics simulations with respect to current software. Publications Wollaber, A. B., and J. S. Warsa. A New, Coupled Transport- Diffusion Method for Radiative Transfer Calculations. 2013. In International Conference on Mathematics and Computational Methods Applied to Nuclear Sci- ence and Engineering (M&C 2013). (Sun Valley, ID, 5-9 May 2013). , p. 7293. La Grange Park, IL: American Nuclear Society. 614

Page 615

Physics Early Career Research Continuing Project Large-scale Simulations for Nuclei and Strongly-correlated Matter Stefano Gandolfi 20120758ECR Introduction Benefit to National Security Missions The goal of this project is to calculate properties of This project ties to the basic research mission of LANL’s nucleonic matter and neutron stars that are of great rele- nuclear theory program, sponsored by the DOE Office of vance for nuclear astrophysics. The equation of state and Science. Calculations of nuclear properties and those of related properties including the spin- and neutrino- re- nucleonic matter are relevant to DOE facilities including sponse are critical to understand current observations of CARIBU at ANL and the upcoming FRIB facility at Michi- the mass-radius relationship, and the cooling and X-ray gan State University. It is also of direct relevance to DOE emission of neutron stars. The inhomogeneous neutron nuclear physics priorities in nucleonic matter and nuclear matter in the crust can be modeled by neutrons interact- astrophysics. ing via nuclear forces confined in external traps, and the The work is also related to an underlying capability in higher-density homogeneous matter can also be studied nuclear structure and reactions at LANL, e.g. in support through large-scale Quantum Monte Carlo simulations. of the weapons program. It is relevant to theories of A quantitative understanding of homogeneous and nuclear structure and reactions, particularly light-ion inhomogeneous matter from first principles is an essen- reactions. tial first step towards building an energy density func- tional for nuclei, as properly calibrated nuclear energy Progress density functionals must capture the effects of the In the last year we have finished and published several interaction and reproduce experimental results for any results related to this project. nuclear system. The characterization of nuclear energy density functionals is of high impact to study properties We have successfully studied the neutrino- and anti- of neutron-rich nuclei that will be realized in several cur- neutrino-deuteron interaction. We have calculated the rent and upcoming DOE experimental facilities including dynamic response function for neutrino energies up to CARIBU and FRIB. a GeV, and addressed the role of two-body processes in the nuclear currents. We have derived the corresponding We will simultaneously study related strongly-correlated cross-sections. The results have been published in Phys. ultracold Fermi gases. In the dilute regime where the Rev. C. density is low enough, properties of neutron matter are very similar to those of ultra-cold Fermi gases near The energy of the unitary gas in a trap for different num- infinite scattering length. This limit of cold Fermi atoms is ber of atoms has been calculated using Quantum Monte the most strongly-coupled system known, as its scatter- Carlo methods. The results have been compared to the ing cross-section diverges. These systems are realized in case of trapped neutrons, and have been published in experiments through the use of the Feshbach resonanc- Prog. Theor. Exp. Phys. es; their properties are independent of the microscopic In addition, we have calculated the energy-dependence details of the interaction and are thus universal. Because of the unitary Fermi gas to the effective range of the of their universality, ultra-cold Fermi gases are an ideal potential. This calculation is important to understand the system to test microscopic many-body theories and com- difference between low-energy neutron matter and the pare to nuclei and other strongly-correlated systems. unitary Fermi gas. The results have been published in Prog. Theor. Exp. Phys. and in Phys. Rev. A. 615

Page 616

We have calculated the structure factor and the contact Conclusion parameter of the unitary Fermi gas. The accurate predic- Our results will have important impacts in several com- tions obtained from Quantum Monte Carlo simulations munities. The calculation of nucleonic matter will provide have been compared to new experimental measurements a strong support to experimental data to fit energy density of the same quantities. The new theoretical/experimental functionals used to describe stable and unstable neutron- results have been published in Phys. Rev. Lett. rich nuclei, and of the crust of neutron stars. Also, ultra- cold Fermi gas experiments are also beginning to probe in- We have finished and published in Phys. Rev. C the calcula- homogeneous systems in external potentials. We will also tion of the spin response of neutron matter. Using Quan- calculate dynamic properties of these strongly-coupled tum Monte Carlo methods we have calculated selected systems. The calculation of the spin-response allows us to sum-rules that we have used to constrain the spin-re- estimate the neutrinos mean free path in dense neutron sponse function. This is relates to the neutrino interaction star matter, and to study the cooling rate of neutron stars with neutron matter as a function of the neutrino energy. observed by modern X-ray telescopes. The results have been used to calculate the neutrino emis- sivity that is directly related to the cooling of neutron stars. Publications Bogner, S., A. Bulgac, J. A. Carlson, J. Engel, G. Fann, and We have calculated the binding energy of a Lambda parti- others. Comput.Phys.Commun.. 2013. Comput.Phys. cle in selected hyper-nuclei from A=5 to 91. Our calculation Commun.. 184: 2235. showed that the three-body force, i.e. the Lamda-nucleon- nucleon interaction, are dramatically important in these Carlson, J., S. Gandolfi, and A. Gezerlis. Quantum Monte systems. The results have been published in Phys. Rev. C. Carlo approaches to nuclear and atomic physics. 2012. PROGRESS OF THEORETICAL AND EXPERIMENTAL We have calculated the energy of neutrons confined in PHYSICS. (1): -. external traps using different nuclear Hamiltonians. This work is important to constrain nuclear energy density func- Forbes, M. M., S. Gandolfi, and A. Gezerlis. Effective-range dependence of resonantly interacting fermions. 2012. tionals, but also to understand the effect of using different PHYSICAL REVIEW A. 86 (5): -. Nuclear Hamiltonians to study neutron-rich systems. Gandolfi, S.. The equation of state of neutron star matter We have also compared other properties of these systems, and the symmetry energy. 2013. In 11TH INTERNA- i.e. densities, radii, pairing gap and spin-orbit splittings. TIONAL CONFERENCE ON NUCLEUS-NUCLEUS COLLI- The results have been published in Phys. Rev. C. SIONS (NN2012). Vol. 420, p. -. Future Work Gandolfi, S.. Quantum Monte Carlo study of inhomoge- In the next fiscal year we expect to produce new calcula- neous neutron matter. 2012. HITES 2012: HORIZONS OF INNOVATIVE THEORIES, EXPERIMENTS, AND SUPER- tion of trapped Fermions (cold atoms) and neutrons. We COMPUTING IN NUCLEAR PHYSICS. 403: -. will study the role of pairing correlations in these systems that are important to calculate the pairing gap. The neu- Gandolfi, S., J. Carlson, and S. Reddy. Maximum mass and tron pairing in confined geometries will be important to radius of neutron stars, and the nuclear symmetry en- constrain the pairing term in density functional theories for ergy. 2012. PHYSICAL REVIEW C. 85 (3): -. finite nuclear systems. The study of properties of confined unitary Fermi gases is important to fit density function- Gezerlis, A., I. Tews, E. Epelbaum, S. Gandolfi, K. Hebeler, A. Nogga, and A. Schwenk. Quantum Monte Carlo Calcu- als and test their quality to deal with small systems and/ lations with Chiral Effective Field Theory Interactions. or in the presence of external potentials modelling optical 2013. PHYSICAL REVIEW LETTERS. 111 (3): -. lattices. Hoinka, S., M. Lingham, K. Fenech, H. Hu, C. J. Vale, J. E. We will calculate the sum-rules of electromagnetic and Drut, and S. Gandolfi. Precise Determination of the weak interactions in light nuclei. The results will allow us Structure Factor and Contact in a Unitary Fermi Gas. to fit the neutrino-nucleus response for selected nuclei, 2013. PHYSICAL REVIEW LETTERS. 110 (5): -. and to calculate the cross-section. We first will perform the calculation for very light nuclei, like the alpha particles, Illarionov, A. Y., S. Fantoni, F. Pederiva, S. Gandolfi, and K. and then will try to address nuclei up to the carbon. We E. Schmidt. Determination of the finite temperature expect these results to be important in understanding the equation of state of dense matter. 2012. PHYSICS OF ATOMIC NUCLEI. 75 (7): 866. contribution of one and two body currents in the neutrino- nucleus interaction. Lonardoni, D., S. Gandolfi, and F. Pederiva. Effects of the 616

Page 617

two-body and three-body hyperon-nucleon interac- tions in Lambda hypernuclei. 2013. PHYSICAL REVIEW C. 87 (4): -. Lovato, A., S. Gandolfi, R. Butler, J. Carlson, E. Lusk, S. C. Pieper, and R. Schiavilla. Charge Form Factor and Sum Rules of Electromagnetic Response Functions in C-12. 2013. PHYSICAL REVIEW LETTERS. 111 (9): -. Maris, P., J. P. Vary, S. Gandolfi, J. Carlson, and S. C. Pieper. Properties of trapped neutrons interacting with real- istic nuclear Hamiltonians. 2013. PHYSICAL REVIEW C. 87 (5): -. Shen, G., L. E. Marcucci, J. Carlson, S. Gandolfi, and R. Schiavilla. Inclusive neutrino scattering off the deu- teron from threshold to GeV energies. 2012. PHYSICAL REVIEW C. 86 (3): -. Shen, G., S. Gandolfi, S. Reddy, and J. Carlson. Spin re- sponse and neutrino emissivity of dense neutron mat- ter. 2013. PHYSICAL REVIEW C. 87 (2): -. Steiner, A. W., and S. Gandolfi. Connecting Neutron Star Observations to Three-Body Forces in Neutron Matter and to the Nuclear Symmetry Energy. 2012. PHYSICAL REVIEW LETTERS. 108 (8): -. Tsang, M. B., J. R. Stone, F. Camera, P. Danielewicz, S. Gan- dolfi, K. Hebeler, C. J. Horowitz, J. Lee, W. G. Lynch, Z. Kohley, R. Lemmon, P. Moller, T. Murakami, S. Riordan, X. Roca-Maza, F. Sammarruca, A. W. Steiner, I. Vidana, and S. J. Yennello. Constraints on the symmetry energy and neutron skins from experiments and theory. 2012. PHYSICAL REVIEW C. 86 (1): -. 617

Page 618

Physics Early Career Research Continuing Project Precision Predictions for Jet Cross Sections Christopher Lee 20120759ECR Introduction Benefit to National Security Missions The Large Hadron Collider (LHC) is now colliding protons Two key missions of the DOE Office of Science in High- at the highest-ever energies to search for new particles Energy Physics are 1) to improve our understanding of beyond the Standard Model of particle physics. New Quantum Chromodynamics (QCD), the theory of the particles and forces may explain the origin of all matter strong interactions between quarks and gluons, con- and mass in the Universe. Many possible new particles stituents of all ordinary matter; and 2) to search for new are expected to decay quickly to “jets” of hadrons, en- particles and forces beyond the Standard Model that ergetic sprays of bound states of quarks and gluons like will explain the origin of matter, dark matter, and masses those that make up ordinary matter. But purely Standard of elementary particles. In Nuclear Physics, a key DOE Model processes already produce jets ubiquitously. mission is to use heavy ion collisions to understand the Modern strategies to search for new physics have made nature of the quark-gluon plasma and extreme states of it a key priority to identify and precisely predict differ- QCD matter which pervaded the very early Universe. ences in the substructure of jets produced by new or Jets of hadrons produced by energetic quarks and gluons Standard Model particles. are key to achieving these objectives. They reveal the This project will predict Standard Model jet cross sec- behavior of QCD itself, contain evidence of new particles tions (probabilities to produce jets as a function of their that decay to jets, and probe properties of dense QCD energies, shapes, sizes, and substructure) to unprec- matter through which they pass. edented precision in order to discern signatures of new We tackle problems at the forefront of QCD perturba- physics over background in jets. We will first tackle two tion theory. Physically accurate predictions of jet cross unsolved problems in Quantum Chromodynamics (QCD), sections require resumming arbitrarily many soft and the theory of interactions between quarks and gluons. collinear quark and gluon emissions in and from jets. We These problems limit the precision currently attainable will resum correlated emissions of gluons and quarks in predicting jet cross sections. Accurate predictions for across jet boundaries and within very small jet cones, jet cross sections require summing the probabilities of both unsolved problems in QCD. Our new tools will allow emitting arbitrarily many soft (low-energy) or collinear unprecedented precision in predicting jet cross sections (parallel) gluons or quarks in or from jets. Summing only for probing QCD, new physics, and dense matter at the a finite number of emissions yields inaccurate results— Large Hadron Collider (LHC). -”resummation” of infinitely-many emissions is crucial. Techniques for resummation exist for relatively simple Historically, individuals hired for their expertise in this cross sections, but not for the more intricate cross sec- area have gone on to address a range of Laboratory chal- tions used in jet substructure-based strategies to search lenges, including bio and energy security, thanks to their for new physics. We will solve the long-standing prob- broad analytical skills. lems of resumming gluons that split and cross jet bound- aries, and of resumming emissions confined within very Progress small jet cones. We will use our solutions to predict to In the last year, we have explored several candidates for high precision the cross sections used to search for sig- effective field theories to resum non-global logarithms natures of new physics in the detailed structure of jets. (NGLs) and logs of jet radii R. We have evaluated the 618

Page 619

theoretical promise of each and identified two promising sured using jets in DIS. candidate theoretical frameworks we will fully implement in the next year to achieve the logarithmic resummations. Future Work In the next fiscal year we will implement the new effective A key step has been to separate the steps of resumming theory frameworks we have conceived of to resum loga- NGLs and logs of R. Previously we believed they should be rithms of jet radii and of non-global logarithms. We will done together, that is, by first using ordinary soft-collinear calculate predictions to at least next-to-leading-logarithmic effective theory (SCET) to separate hard, hard-collinear, (NLL) accuracy in resummed perturbation theory. We will and soft degrees of freedom from one another to sum logs demonstrate that our resummed predictions match the of R, and then to further factorize the soft sector using a predictions of fixed-order Quantum Chromodynamics second effective theory to resum the NGLs. However, now (QCD) calculations from numerical programs to second- we understand that it is more promising to reverse this order (alpha_s^2) in the strong coupling. order. First, we separate the modes of full QCD (Quantum Chromodynamics) into hard and soft degrees of freedom We will also continue the new direction of research into to sum the NGLs. Only then do we separately factorize high-precision resummed predictions for jet cross sections these hard and soft sectors into hard/hard-collinear and in deep inelastic scattering (DIS) of electrons and pro- soft/soft-collinear modes. The soft/soft-collinear factoriza- tons. We will perform calculations of the fixed-order QCD tion is also a step we have proposed for the first time. We cross sections to next-to-leading order (NLO) and of the are in the midst of completing the theoretical formulation soft function in the soft-collinear effective theory (SCET) of this doubly-factorized effective field theory and imple- factorization theorem to O(alpha_s^2), both of which way menting it to resum logs of R and NGLs. lay necessary groundwork to extend the precision of these predictions to NNLL+NLO accuracy, allowing for higher- In summary, in the past year we have tested and identi- precision extractions of the strong coupling alpha_s from fied the most promising avenues for constructing effective DIS jet data. theories that will achieve resummation of NGLs and logs of R, preparing us well to achieve these goals in concrete Conclusion predictive calculations in the upcoming year. First, we will create new theoretical tools in QCD that make possible the resummation of quark and gluon emis- Also in the past year, we have found a new area of applica- sions from jets in in two unsolved scenarios: correlated tion of the effective field theory methods upon which this emissions splitting across jet boundaries and emissions LDRD is building to resum perturbative expansions of jet confined to very small jet cones. These tools will solve cross sections in deep inelastic scattering (DIS) of electrons long-standing problems at the forefront of QCD perturba- and protons. While the methods of SCET had previously tion theory and make possible precise predictions of many been applied to electron-positron and proton-proton more observables than at present. Second, with these jet cross sections, we realized they had never been fully tools we will achieve the level of precision in predicting jet implemented in electron-proton collisions. We predicted cross sections required to use jets and their substructure jet cross sections in DIS using a variable “1-jettiness” that to distinguish signals of new physics from Standard Model measures how jet-like the final state of a DIS event is, but background. without using measures like jet radii or NGLs. This type of “global” measurement is a precursor to accurately pre- Publications dicting jet cross sections using the more non-global or Kang, D., C. Lee, and I. W. Stewart. Using 1-jettiness to exclusive measures that use jet radii and contain NGLs. We measure 2 jets in DIS 3 ways. 2013. PHYSICAL REVIEW predicted 1-jettiness cross sections in DIS to next-to-next- D. 88 (5): -. to-leading logarithmic (NNLL) accuracy in resummed per- turbation theory for the first time, one order beyond any previous work on DIS jet cross sections. We also discovered the universality of nonperturbative corrections to several different types of 1-jettiness cross sections. This discovery and the NNLL perturbative resummation exhibit the power of the effective theory approach to jet physics. We com- pleted a paper on this work (arXiv:1303.6952) which has been accepted by Phys. Rev. D for publication. This work advances the precision with which the strong coupling alpha_s or the internal structure of the proton can be mea- 619

Page 620

Physics Early Career Research Continuing Project First Principle Study of Relativistic Beam and Plasma Physics Enabled by Enhanced Particle-In-Cell Capability Chengkun Huang 20130744ECR Introduction Benefit to National Security Missions The Particle-In-Cell (PIC) method is a fully kinetic ab- The proposed work will establish a high-fidelity first- initio method widely used in plasma and particle beam principle predictive modeling capability and improve the modeling. Although the standard PIC method with understanding of high current beams interaction with the second order accurate numerical algorithm is well its coherent radiation and the kinetic plasma behavior suited for the study of beam and plasma dynamics at under extreme conditions. We are pushing the enve- low speed, its application to relativistic phenomena is lope of the PIC method which is widely used in the lab severely limited. The major limitations come from the for various DOE/NNSA projects in high energy density dispersive and anisotropic propagation properties of plasma, accelerator beam dynamics, astrophysics and the electromagnetic (EM) wave on the grid, the numeri- space applications, thus yielding many scientific possi- cal instability associated with the unphysical Cherenkov bilities for LANL. The improved modeling capability and radiation, the inaccuracy in the relativistic particle kine- the new insights into the XFEL microbunching instabil- matics and the undesirable noise from under-resolved ity and particle acceleration/deceleration mechanisms dynamics. In this work we will develop, validate and ap- link directly into LANL’s missions of (1) Matter-Radiation ply a new computational framework so as to be able to Interactions in Extremes (MaRIE) Signature Facility, (2) compute and understand cutting-edge problems involv- Nuclear, Particle, Cosmology, and Astrophysics and (3) ing highly relativistic charged particles. Our approach is Nuclear Energy. to address the deficiencies of the standard PIC method and include enhancements with carefully designed high Progress order particle shapes/digital filters, a relativistically ac- The project started in July 2013; therefore, there is no curate particle pusher and most importantly a 4th-order progress to date. accurate and tunable EM field solver developed recently. This enhanced PIC capability will provide, for the first Future Work time, accurate and self-consistent modeling for many The project has a start date of July 1 in FY13. In the rest relativistic beam/plasma problems of importance to of FY13, we expect to characterize the new high order LANL’s mission and to the broader scientific community. field solver and identify the optimal tuning parameters Specifically, we will focus on (1) understanding how the and the strategy of noise control for the micro-bunching microbunching instability can arise from the interactions instability, collisionless shock and fast ignition study. We between a highly relativistic beam and its radiation in will derive the full dispersion relation of the new solver a Free Electron Laser (FEL) bunch compressor and how to guide our tuning process. Then at the end of FY13 or it affects the beam quality and FEL performance; (2) the beginning of FY14, we will begin the study of the investigation of the collisionless particle acceleration micro-bunching instability by first benchmarking with process in astrophysical shock and the deceleration pro- our 2D coherent synchrotron radiation model for a non- cess related to the collective wave-particle interaction evolving beam. For collisionless shock and fast ignition in high-energy-density plasmas. The former application study, we will begin by investigating the wake of a single is directly relevant to the success of the MaRIE X-ray FEL relativistic particle both in free space and in a plasma. signature facility while the latter addresses fundamental After these benchmarking, we will proceed to the study science questions in both astrophysics and a high-risk of the relevant physics in FY14. In particular, by the end yet high-potential fast-ignition technology for inertial of FY14, we expect to conduct detail validated simula- fusion energy. 620

Page 621

tion study of (1) micro-bunching instability relevant to X- ray free electron design design, (2) relativistic collisionless shock for astrophysics and (3) fast ignition concept. Conclusion The technical goal of our work will be a validated new PIC capability that can be effectively applied to relativistic beam and plasma problems with superior accuracy and stability for a wide range of parameters. The scientific goal is a detailed understanding that bridges the gap between present simulations and theories/models for the micro- bunching instability and collisionless particle acceleration/ deceleration processes. Our results will impact the design of the MaRIE X-ray free electron laser, fast ignition inertial fusion experiments and help elucidate the origin of ener- getic particle in the cosmos. Publications Huang, C., T. J. T. Kwan, and B. E. Carlsten. Synchrotron Radiation Near Field In 3D. To appear in 25th Particle Accelerator Conference. (Pasadena, 30 Sept. - 4 Oct. 2013). 621

Page 622

Physics Early Career Research Continuing Project Stochastic Modeling of Phase Transitions in Strongly Interacting Quantum Systems Christopher Ticknor 20130749ECR Introduction There are more simple studies to be done on phase tran- This research explores thermal phase transitions in sitions in this system. So I will further the research on quantum systems. Two sets of codes will be developed, mixtures and produce research articles of high quality. the first is a static, perturbative treatment. This will This work has been extended for the local version to extend a standard treatment for quantum gases at finite the full finite temperature theory. I have added another temperature to study multicomponent, dipolar quantum basis set to better handle the non-local interactions. gases. The second set of codes, is a stochastic method, This will greatly enhance my ability to study the strongly this theory will more fully account for the fluctuating interacting quantum limit. nature of quantum systems. Both codes will be used to study intriguing quantum behavior near thermal phase Future Work transitions. Completely developing the SPGP method In the next year, I will fully develop a perturbative and in 2 years will be a challenge - there are many technical static treatment known as: Hartree-Fock Bogoliubov challenges to be met. These codes will tackle an interest- Popov (HFBP). With this code, I will study the thermal ing problem: thermal phase transitions. Understanding phase transition of a two component immiscible, dipolar such complex systems is a significant challenge, with quantum gas. substantial payoff. Next, I will start the development of a dynamic method Benefit to National Security Missions known as the Projected Gross Pitaevskii (PGP) method. The numerical methods studied here are an important This method will study two thermal phase transitions: step in developing hybrid quantum/classical codes that the miscible/immiscible character of the quantum gas can treat fundamental quantum behavior in high tem- and the phase coherence as temperatures is changed. perature regimes. Solving a full quantum treatment at This method will be able to go the strongly interacting high temperatures is simply not possible, and therefore regime, unlike the first method (HFBP). an alternate method needs to be developed which re- tains aspects of quantum behavior. That is exactly what Conclusion this work does. This work additionally looks at phase The goal of this project is twofold: first, develop two transitions of quantum systems. This will help under- flavors of codes to study finite temperature quantum stand the behavior of finite quantum systems beyond gases. Second, study thermal phase transitions of ideal settings and conditions. quantum gases. It is unknown how temperature alters the miscible/immiscible character of ultracold quantum Progress gases. Additionally, study the thermal phase transitions Since this project started in Jan 2013, I have implement- which dictates the coherence of the quantum gas. This ed the first part of the research. I have produced a mul- work will influence how experiments are run and how ticomponent, linear excitation theory including non-local finite quantum systems are understood. interactions. Furthermore, I have produced a paper with these codes, currently under review of PRA Rapid. Publications Ticknor, C.. Excitations of a trapped two-component This work was not even for the full system (dipolar gas) Bose-Einstein condensate. 2013. PHYSICAL REVIEW because there is so much to understand about mixture A. 88 (1): -. systems. This will lead to a slight diversion in the plan. 622

Page 623

Physics Early Career Research Continuing Project Answer to Heavy Element Production Puzzle by Measuring Neutron-induced Charged Particles at LANSCE Hye Young Lee 20130758ECR Introduction nuclear astrophysics, while developing instrumentation This project is to build a new capability of measuring that can address outstanding problems for multiple neutron-induced charged particle reactions at Los Ala- sponsors. This combination of world-class basic science mos Neutron Science Center (LANSCE). Neutron-induced while simultaneously solving problems for the nation in charged-particle reactions are critical to understand the multiple nuclear science thrusts makes this truly a LANL- currently unsolved heavy-element production puzzle, appropriate endeavor. since these reactions in stars compete with neutron Progress capture reactions, a main process to produce heavy elements. Of importance, the Ne-22(a,n)Mg-25 reaction The conceptual design of the low energy ionization has been studied for decades to provide a key answer to chamber was determined to use a Frisch grid and seg- how the neutron source was generated for successive mented anodes. Multiple design considerations were captures in heavy element production. However, the put into the process of optimizing in low-energy alpha direct measurements contain large uncertainty due to particle detection for nuclear astrophysics interest and the technical difficulties in detecting low cross-section also with the flexibility of converting this ionization reactions and large background. Using this new capabil- volume into one of delta-E detectors for high energy ity, a time-reversed Mg-25(n,a) measurement at LANSCE charged particle detections, which is planned to be com- will shed a new light on the critical region of interest to pleted in the next year. help astrophysics modelers explain how heavy elements The conceptual design of the high energy charged- are produced. particle detection system is to use a stack of silicon strip Developing the efficient detection system, waveform detectors to increase detection solid angle coverage digitizers, and low background environment is the inte- and improve angular resolutions. The choice of most gral part of this project. Along the experimental effort, it advanced silicon strip detectors were under discussion is planned to study reaction processes using the Monte with the company for the feasibility of development and Carlo Hauser-Feshbach method, which was developed a delivery time for next fiscal year. by the T-2 group. We have been collaborating to im- Upon completion of design parameters in the ioniza- prove the theoretical prediction power by comparing the tion chamber, hardware procurements were progressed calculations with the experimental data, and this new in deciding the specifications of electronics. For the capability will certainly enhance the reliability of theo- best results in reducing the background in signals and retical predictions for currently unavailable reactions- improving particle identification capability, I chose two impossible to measure in laboratories due to very short different types of digitizers. VX1720 from CAEN has a lifetime of nuclei. sampling rate of 250 Mega-samples/second with 12-bit digitization resolution, and VX1724 from CAEN has a Benefit to National Security Missions sampling rate of 100 Mega-samples/second with 14-bit Neutron-induced charged particle reactions are impor- digitization resolution. A 9-slot VME crate from Wiener tant to the Weapons Program for the interpretation of and optical link communication are chosen for the appli- radiochemical diagnostics, nuclear forensics, and other cation. All these electronics are expected to be delivered applications. Scientifically, this project is well-matched by the end of July. At the moment, procurement for the to ongoing Los Alamos Neutron Science Center (LAN- gas handling system is underway. SCE) efforts to increase our portfolio and capabilities in 623

Page 624

A Monte Carlo code was developed to simulate the signal can provide additional information for improving resolu- strength in different geometry of a cathode and anodes, tion and performance of detectors. The combination of the detection thresholds, and the detection efficiency. two techniques allows us to improve the signal-to-back- The code written in C++ runs in the platform of the ROOT ground ratio, pushing the current limitation on detecting package. One student from the SARA (Service Academy low energy charged particles to lower energy than the Research Associates) program has worked for 6 weeks to previously achieved. develop this Monte Carlo code. Based on the characterization of the ionization chamber in Monte Carlo simulations, I have discussed with the engineer for the general vacuum chamber design. I have contacted a mechanical engineer to work on machine drawings for manufacturing, which is scheduled for this summer. I am also working on the materials to be used in the ionization chamber and I contacted the providers for the details. All the procurements went reasonably well and I am still working on hardware manufacturing. I have submitted a beam time request proposal to the Weapons Neutron Research (WNR) Program Advisory Committee in April 2013 and a 10-day beam time will be awarded sometime during the upcoming beam cycle (Aug. 13 through Dec. 21, 2013). Future Work For the second fiscal year, the project is planned to com- mission the newly developed low-energy detection system and complete the development of high-energy detection system. First, assemble all the components with proper small element procurements to build an low-energy ionization chamber. Once all the hardware is installed, test signals with a calibration source at the TA53-Building 17, which is our group’s lab space. I also expect to start a paper work to place proper safety concerns like an IWD. With the requested beam time, 10 days will be awarded to test in- beam background and all the functions. Second, for the high-energy detection system, silicon strip detectors with proper preamplifiers will be purchased and additional digitizers. The full system will be commissioned with alpha source. For the in-beam test, another beam time request proposal will be submitted to the WNR Pro- gram Advisory Committee. Conclusion We expect to produce reliable neutron-induced charged- particle reaction data by developing the experimental capability at LANSCE. The new detection system consists of highly segmented silicon detectors for a large detection coverage with good angular information and waveform digitizers with improved timing resolution. Implementing software filters in digitized waveforms, which is traditional, 624

Page 625

Physics Early Career Research Final Report The Dynamic Environment of Iron Core Formation and Collapse Casey A. Meakin 20120530ECR Abstract responsible for producing a shock wave, which eventu- This project has sought to advance our understanding ally tears the entire star apart in a supernova explosion. of how turbulent flow shapes the evolution of stars, Despite decades of intensive theory, however, we still do and in particular those stars thought to end their lives not have a robust or detailed understanding of how this in supernova explosions. Towards this end we have car- iron-core collapse event results in a supernova explo- ried out new multi-dimensional simulations of turbulent sion. While we have identified the source of energy convection and nuclear burning in stellar interiors and needed to power such an explosion (the gravitational have undertaken the most detailed analysis of this type potential energy liberated during the collapse), we do of numerical data yet, with the following main results. not yet understand how this energy is harnessed. Stellar Our analysis has clarified the important role played by evolution, which encompasses the details of iron core several previously ignored physical processes in the formation and collapse, is a strongly turbulent, highly dynamics of turbulent stellar flow including pressure- non-linear phenomenon, which is not directly amenable dilatation, pressure-flux, and kinetic energy dissipation. to paper and pencil theory and therefore depends on This work has also clarified the shortcomings of modern simulation for progress. The first simulations of stellar turbulence models for treating stellar flow; in particular core collapse, produced in the late 1960s, were success- the so-called “down-gradient” and “return-to-isotropy” ful, but were calculated entirely in 1D, from star birth, to closure relationships are shown invalid. Our work iron core formation, to iron core collapse and explosion. provides a framework for assessing future turbulence However, as the input physics used in these simulations models against multi-dimensional simulation data. Fi- improved (e.g., better equation of state, better transport nally, this work has resulted in several follow on projects, physics), 1D models began to fail to explode. Subsequent including both interfacing our results with core-collapse generations of multi-dimensional models of the core col- supernova modeling efforts as well as more detailed lapse event, which continued to use iron cores evolved statistical turbulence analysis work including evolution in 1D as initial conditions, were begun, and despite ini- equations for higher order statistical moments (e.g., for tial successes, these too were eventually found to fail to turbulent fluxes), a scale-by-scale analysis of the simu- produce explosions naturally and robustly. This remains lated flows, and extensions to treat rotating and magne- the current state of the research. tized systems. There are two strongly favored candidate solutions to Background and Research Objectives this problem, either (1) researchers hope that further Massive stars end their lives after they have completed refinement in simulating iron core collapse will lead to successive stages of nuclear burning in their cores, ulti- successful explosions, which would therefore indicate an mately converting their initial hydrogen and helium com- incredible sensitivity to the finest details of the problem, position into a dense central core composed of primarily or (2) finally examining the highly uncertain initial condi- iron-peak elements. Iron peak material reside at the tions, currently based on 1D models of stellar evolution, peak of the nuclear binding energy curve, which means will provide the missing link in our understanding of this that additional energy cannot be liberated through “supernova problem”. continued nuclear burning. This circumstance, together with a gravitational instability, eventually results in the The present work pursues this second avenue, for the collapse of the iron core thus formed. The ensuing col- following reasons. First, if refined simulations of core lapse of the core to nuclear densities is thought to be collapse do result in robust explosions, these results will 625

Page 626

remain contingent upon the uncertain initial conditions of simulations of oxygen shell burning [1,5-7], a phase 1D stellar evolution. Second, we now have very strong evi- of stellar evolution preceding core collapse in massive dence that the 1D initial conditions used for core collapse stars. In particular, we have conducted a resolution modeling are in fact flawed [1,5-7]. study up to 200 million zones for a 25 species reac- tive flow in order to assess the dependence of various Objectives statistical properties on spatial zoning. These new We believe the above background motivates the need for a calculations have also included larger computational better understanding of stellar evolution, which is tanta- domains and longer run times than earlier work [4]. mount to better understanding the details of turbulent These calculations cost ~4.5 Million CPU-hours and convection. Therefore, we pursue this goal in the present were conducted entirely on the NICS Kraken Cray XT5 work through multi-dimensional simulation of stellar inte- at the University of Tennessee. Additionally, each of riors combined with an appropriate analysis framework to these calculations were instrumented with code to interpret the resulting very large data sets. produce running averages needed to best sample the mean-fields described in the next two sections (pub- Scientific Approach and Accomplishments lished in [12]). Our scientific approach has been three pronged, including new large-scale computations, the mathematical devel- 2. We have developed new simulation setups to directly opment of an analysis framework, and data reduction of model the final burning stage prior to core collapse, numerical results within this framework. In the following which is silicon burning. This involved extending our subsections we provide a summary of our approach and nuclear reaction physics and redesigning our compu- accomplishments in these three areas, and then conclude tational domain configuration to include full shell and with a list of the follow-on work that has resulted from this full core calculations [8]. This latter aspect of devel- project. opment involved implementing a “star-in-box” setup, wherein spherical objects are embedded in a Cartesian Computations box of zone with boundary effects ameliorated through On the computational side, we have developed new “sponge-layer” damping [13]. Finally, an important simulation models and have acquired significant compute part of this new set was the calculation of initial condi- resources to carry these out (over 10 million CPU-hours on tions for simulation. These were drawn from a new competitive national facilities, including XSEDE and DOE library of 1D stellar evolution calculations produced platforms). Additionally, our codes have been optimized to with the MESA code [11]. As of the termination of perform efficiently in parallel on over 100 thousand cores this project, we have begun a series of 2D simulations on both Cray and IBM platforms (currently tested up to with 3D simulations imminent. We have ported this ~130k cores with nearly perfect scaling, and limited to this new setup to the NICS Kraken and ANL Vesta/Mira number of cores only due to system availability). IBM BG/Q computing platforms. Our intentions are to complete our 2D survey and perform a capstone 3D All of our simulations have been conducted with the simulation by early Fall 2013 and submit this work for PROMPI code [5-7], an Eulerian, Godunov-based finite- publication at that time. volume hydrodynamics code outfitted with physics ap- propriate for modeling stellar interiors, including a realistic RANS Framework equation of state, multi-species advection, nuclear reac- We have developed the statistical moment equations, of- tion rates and network evolution code, thermal (radiative) ten referred to as Reynolds Averaged Navier-Stokes (RANS) conduction and realistic opacities, and gravity. Our simula- equations [2-3], for the fully compressible, multi-species tions are conducted within the implicit large-eddy simula- equations of hydrodynamics relevant for stellar modeling tion (ILES) framework [4], wherein molecular dissipative in spherical coordinates. The end result is a set of mean- processes and sub-grid scale physics are relegated to the field equations describing the statistical behavior of the inherent dissipation of the hydrodynamics algorithm. An mean flow, thus providing a framework to analyze the assessment of the behavior of this sub-grid physics is part importance of various distinct physical processes. of the subject of our analysis (discussed more in the fol- lowing sections). This framework lends itself to the following: (1) it provides a means to assess the applicability of a broad range of tur- The new simulations undertaken for this project can be bulence closure models in the literature, (2) it allows the categorized as follows: analysis and comparison of heterogeneous sets of turbu- lence data in a unified way, and (3) it provides a means to

1. We have extended and built upon earlier generation asses the behavior of sub-grid scale behavior inherent in 626

Page 627

our numerical simulation scheme, including but not limited concluded the following. First, turbulent dissipation to turbulent dissipation and material mixing, offering fur- is well behaved in our models implying that relegat- ther insight into the ILES approach. ing the action of viscosity to the numerical algorithm (and not treating it explicitly as in a direct numerical Our development of the first order moment equations, simulation (DNS) is justified. In particular, our resolu- and a second order moment equation for turbulent kinetic tion study demonstrates that our results are consistent energy are published in [12]. Additional moment evolu- with the presence of an inertial range and statistical tion equations, including variances for all relevant thermo- behavior that is independent of dissipation-scale (i.e., dynamic quantities and fluxes have also been developed grid-scale) behavior. Second, we have shown that mix- and are in preparation [9]. The latter work also includes ing behavior at convective boundaries (analogous to a (Fourier mode) scale-by-scale decomposition for each the “entrainment” phenomena in geophysics) requires moment, thus providing a means to quantify the contribu- further study and does not show as good conver- tion to each physical process by either large-scale coher- gence as the dissipative phenomena in the bulk of the ent flow components and/or the smaller scale statistically convective layer: some length scale of importance is steady components. not resolved. This motivates a controlled, systematic follow up work, possibly with DNS. Data Analysis Much of the value of a three-dimensional simulation of 3. We have compared the statistical properties of our turbulent flow resides in the ability to perform detailed 3D simulation results with modern closure relation- analysis and calculate statistics from the numerical data. ships, typified by those discussed in Besnard et al. [2] This is in sharp contrast to current experimental work and find poor agreement. The core features of these which is still hampered by generally poor resolution, relations, i.e, “down-gradient” for fluxes and “return- dimensionality (sometimes only 2D slices of 3D fields to-isotropy” for Reynolds-stress, match the data poorly are possible), and limited access to physical fields (often (including sign errors). Preliminary analysis [9] indi- limited to only kinematic quantities, such as the velocity cates this is due to violation of the “scale-separation” field). Numerical simulation provides access to the com- hypothesis, whereby fluxes are governed on scales plete system data for analysis. Towards this end, we have much smaller than the integral scale. We find transport extracted mean fields for all of the terms constituting the dominated by the largest scales, those produced by RANS moment equations discussed above. In order to global scale Rayleigh-Taylor instability [12]. We are achieve the highest fidelity statistics we have instrumented following with a scale-by-scale budget analysis in an our simulation code to calculate running averages which effort to construct hybrid closure models that combine sample each numerical time step thus providing data that large-scale transport with small-scale dissipation. would have been impossible to produce as a post-process due to the data storage requirements. Follow-on Projects Follow-on work that has been initiated by this project Our primary results from this effort include: include:

1. We have identified and characterized the main physical
2. Core collapse simulations of 2D and 3D progenitor processes responsible for the evolution. Of most inter- structures with Fryer (LANL: using the SN-SPH Code), est is our identification of the work done by turbulent Ott (Caltech: using the Zelmani/Einstein Toolkit Code), pressure fluctuations (pressure-dilatation) in a subset and Adam Burrows (Princeton: using the Castro Code): of our models. This process has not been included in We are providing each of these teams the results of any model of stellar convection known to this author, our 2D/3D silicon-burning calculations, to use as initial yet contributes ~40% of the total turbulent kinetic conditions fir core-collapse simulations to assess the energy production in our models. Other noteworthy role of multi-dimensional effects in shaping the super- processes identified in our simulations include pres- nova explosion mechanism (some preliminary work sure-flux and turbulent kinetic energy flux as important published in [10]). process, particularly at convective boundaries where these processes mediate the interaction between the 2. Continued effort on 3D progenitor modeling: we are convection zone and the stably stratified layers that extending simulation and analysis work to look at more confine the turbulent motions. complex phenomena, including magneto-hydrodynam- ics and angular momentum (rotation).
3. Our work has provided insight into the behavior of the simulation method and the ILES paradigm. We have 3. We have developed a strong collaboration between 627

Page 628

Maxime Viallet (Munich), Miroslav Mocak (LANL), and port”, The Astrophysical Journal, in preparation David Arnett (Arizona) on turbulence analysis, which is 10. Ott. C. et al., “General-relativistic Simulations of Three- expected to continue into the upcoming years. dimensional Core-collapse Supernovae”, 2013, The Impact on National Missions Astrophysical Journal, 768, 115 Our work deepens our understanding of the astrophysical 11. Paxton, B., et al., “Modules for Experiments in Stellar sites (massive stars) hosting supernovae and gamma-ray Astrophysics (MESA)”, 2011, The Astrophysical Journal bursts, which have broad implications for NASA funded Supplement, 192, 3 astronomical observatories, including potential impact on mission design. DOE/SC is impacted by our work through 12. Viallet, M., Meakin, C., Arnett, D., & Mocak, M., basic studies of turbulent reactive fluids, and mixing and “Turbulent Convection in Stellar Interiors. III. Moment transport modeling, as well as through the exercise of large Equations and Stratification Effects”, 2013, The Astro- scale computing facilities and the development of new physical Journal, 769, 1 techniques required to execute our proposed simulation work. Our reactive hydrodynamics work builds capability 13. Zingale, M. , et al., “Low Mach Number Modeling of for new approaches to weapons modeling. Type Ia Supernovae. IV. White Dwarf Convection”, 2009, The Astrophysical Journal, 704, 196 References

1. Arnett & Meakin, “Toward Realistic Progenitors of Publications Core-collapse Supernovae”, 2011, The Astrophysical Meakin, C., and D. Arnett. Core-collapse supernova pro- Journal, 733, 78 genitor evolution. I. Silicon-shell burning hydrodynam- ics. 2013. The Astrophysical Journal, in preparation..
2. Besnard, D., Harlow, F., Rauenzahn, R. & Zemach, C., Meakin, C., and D. Arnett. Core-collapse supernova pro- “Turbulence Transport Equations for Variable-Density genitor evolution. II. Iron-core formation hydrodynam- Turbulence and Their Relationship to Two-Field Mod- ics. 2013. The Astrophysical Journal, in preparation.. els”, 1992, LANL LA—12303-MS Mocak, M., C. Meakin, M. Viallet, and D. Arnett. Turbulent
3. Chassaing, P., “Variable Density Fluid Turbulence”, convection in stellar interiors. IV. Scale-by-scale budget 2003, Applied Mechanics Reviews, vol. 56, issue 5, p. equations for turbulent fluctuations and transport. B72 2013. The Astrophysical Journal, in preparation..
4. Grinstein, F.F, Margolin, L.G., Rider, W.J., Editors, Ott, C., E. Abdikamalov, P. Moesta, R. Haas, S. Drasco, E. “Implicit Large Eddy Simulation: Computing Turbulent O’Connor, C. Reisswig, C. Meakin, and E. Schnetter. Fluid Dynamics”, Cambridge UP, 2007, 2nd printing General-relativistic simulations of three-dimensional
5. core-collapse supernovae. 2013. The Astrophysical Journal. 768 (2): 115.
6. Meakin & Arnett, “Active Carbon and Oxygen Shell Viallet, M., C. Meakin, D. Arnett, and M. Mocak. Turbulent Burning”, 2006, The Astrophysical Journal Letters, 637, convection in stellar interiors. III. Mean-field analysis 53 and stratification effects. 2013. The Astrophysical Jour- nal. 769: 1.
7. Meakin & Arnett, “Anelastic and Compressible Simula- tions of Stellar Oxygen Burning”, 2007, The Astrophysi- cal Journal, 665, 690
8. Meakin & Arnett, “Turbulence Convection in Stellar Interiors. I. Hydrodynamic Simulation”, 2007, The As- trophysical Journal, 667, 448
9. Meakin, C. & Arnett, D., “Core-Collapse Progenitor Evolution. I. Silicon-shell Burning Hydrodynamics”, The Astrophysical Journal, in preparation
10. Mocak, M., Meakin, C., Viallet, M., & Arnett, D., “Tur- bulent Convection in Stellar Interiors. IV. Scale-by-Scale Budget Equations for Turbulent Fluctuations and Trans- 628

Page 629

Physics Exploratory Research Continuing Project Foundational Methods and Experiments in Ultracold Molecular Physics Michael D. Di Rosa 20110356ER Introduction Benefit to National Security Missions We will investigate a novel experimental approach to The principles and techniques introduced by this re- cooling and storing stable molecules at ultracold tem- search will have a malleable applicability to fundamen- peratures and high densities. Our method of cooling tal and applied science. In the latter vein, outlets for molecules involves lasers as established by us in 2004, the work are conceivable in areas of national security, adopted recently by a pre-eminent research group for homeland security, remote sensing, secure communica- a prefatory experiment, and featured in the Search & tion, and information analysis. Regarding fundamental Discovery column of the January 2010 issue of Physics science, ultracold molecular matter is rife with research Today. That column highlighted the work of DeMille and opportunities that will help us understand and control colleagues at Yale in deflecting a molecular beam by a the quantum mechanics of reactions and collective transversely directed laser beam—a precursor to laser phenomena, thus providing another window into the de- cooling—by combining the type of diatomic molecule tails of curious material behaviors we wish to exploit or and the laser-induced electronic transitions we had pre- invent. The development of this promising, nascent field scribed. More importantly, the article heralded that laser will be challenging and will take the combined expertise cooling of molecules was perhaps not “impossible” and and resources such as the Laboratory possesses by vir- anticipated far-reaching consequences in quantum and tue of its missions. Its influence on Laboratory programs chemical sciences once molecules were laser- cooled would then come naturally, and most effectively, by our to <1 mK and collected. Molecules have not been laser firsthand knowledge and practice. cooled, however, let alone concentrated in the ultracold, dense ensembles needed for experiments. Our project Progress aims to remedy these fundamental needs of cooling and Our progress to date was summarized by the poster storage so that molecular physics may enter the same “Developing Density of Laser-Cooled Neutral Atoms and ultracold conditions that reinvented atomic physics. Molecules in a Linear Magnetic Trap,” authors J. Velas- quez, III, P. L. Walstrom, and M. D. Di Rosa, presented Our two major experimental goals are the laser cooling by postdoc Joe Velasquez at the 44th Annual Meeting of molecules in their lowest-energy internal state and of the DAMOP Division of the American Physical Society the demonstration of an accumulator that is based on (held 3-7 June, 2013). Joe drew a crowd intrigued by our principles of high-energy physics and suitable for col- inventive research. lecting successive injections of high-density packets of laser-cooled (paramagnetic) molecules or atoms. The The poster reviewed our proposed approach (with two achievements are each noteworthy, laser cooling for supporting numerical studies) for accumulating and its precedence and the accumulator as a new platform storing cold (millikelvin) molecules at high densities, for experiments in ultracold physics and chemistry. In featured our working molecular/atomic beam machine, our research, these two elements will be demonstrated and highlighted some first results we obtained with our separately, with the accumulator proved for the case of “test particle” of lithium. Our concept for an accumula- laser-cooled atoms as proxies for laser-cooled molecules. tor attracted attention as did our variation of a standard Time permitting, we will inject laser-cooled molecules chemical-physics technique for creating monoenergetic into the accumulator to further reinforce this accumula- beams of atoms and molecules. In our beam machine, tor’s capacity to serve future and first experiments in we seed by laser ablation from a solid target the atoms ultracold molecular physics and chemistry. or molecules of interest directly into the dense part of 629

Page 630

a supersonic expansion of a carrier gas. Collisions would We then developed the idea of a linear trap and accumula- ideally force the seeded particles, initially at some tens tor, which we introduced in the poster. So far, our simula- of thousands of kelvin (the sun, for comparison, shines at tions indicate that particles will be absolutely confined in 5,980 K) in the bright plasma plume of laser ablation, to this style of trap, even in its practical construction. It is also adopt the translational characteristics of the carrier gas, far simpler to make than a racetrack. While we thought though we have not found evidence for this in the litera- this to be an original idea for a trap, we found assuring pre- ture. The many chemists using these sources, dubbed cedence of its use (in 2009) in trapping ultracold neutrons. “Smalley” sources after their Nobel-prize-winning origina- The additional use of the trap as an accumulator remains tor, are usually unconcerned with the translational energy. unique, however. Modeling the efficiency of injection into We showed that Li atoms seeded by laser-ablation of a the accumulator is underway. lithium rod into a (heavier) supersonic carrier do indeed equilibrate to the cold, moving-frame translational temper- Future Work ature of the carrier following its expansion to a collisionless Laser cooling of the paramagnetic CaH is a primary goal beam. Interestingly, the Li beam turned out to be colder and will be a highlight of international recognition. Our than expected. Where the argon carrier of the experiment other goal, that of introducing techniques for magnetic- has a temperature of 5 K in the moving frame, lithium had particle accumulation, is however more expediently a measured temperature of 3 K. We expect to see similar proved through use of a surrogate that is more easily laser results when we seed laser-ablated CaH into the expan- cooled. In our case, that surrogate—that test particle—is sion. This result announced that (1) the cryogenic systems lithium, partly because the lasers we use to cool CaH can favored today by the cold-molecule contingent are unnec- be tuned to cool Li. We then make the connection that the essary for delivering cold molecules to experiments and/or accumulator demonstrated for a paramagnetic atom such laser manipulations and (2) seeded supersonic expansions as lithium is applicable to laser-cooled paramagnetic mol- might deliver atoms to any of the favored styles of traps at ecules such as CaH. Though developed with molecules in greater instantaneous and average fluxes than that pos- mind, the accumulator is a general device that could very sible by the conventional choice of an effusive source. well benefit the field of atomic physics. The accumulator can then be used to build sufficient densities of cold mol- A third of the poster reviewed the maturing accumulator ecules or cold atoms or both for experiments that require concept. When the project began, our model accumula- collisions, whether for studies in chemistry or quantum- tor was a “racetrack” with two straightaways connected state scattering, or as a start toward creating a quantum- by 180-degree turns. Injection took place along one of the degenerate molecular gas. straightaways. A simple analysis showed the racetrack was perfect for our purposes, holding the injected particles The following four major goals are set for next fiscal year, and those already stored for an indefinite number of laps. with note that we have built our atomic/molecular beam Our recent simulations, however, showed that a racetrack machine and have demonstrated its performance by send- would not in practice be able to hold our magnetic par- ing bursts of lithium atoms traveling at ~540 m/s along its ticles in stable orbits. The earlier, simple model assumed axis. the magnetic field defined by the racetrack would be a sort of smooth tube within which the particles would be 1. Using standard laser-based techniques, slow and cool guided. Barring no expense, this could be created by using the atomic-beam packets of Li to 30 m/s and 1 mK, superconducting current loops. Practically, however, the conditions that mimic what we expect to obtain when track is more easily and cheaply constructed by piecing cooling packets of laser-cooled CaH. together magnetic segments. Unfortunately, the inevitable 2. Demonstrate (with the slow and cooled Li beam) the gaps between segments leads to ripples along the guid- state-switching concept for injection into an accumula- ing magnetic tube, appearing to the particles like a badly tor. wash-boarded road. The cumulative effect of all the cor- rugations bounced particles out of the simulated racetrack 3. Switch the machine’s source to CaH, optimize that within a lap or two (Walstrom and Di Rosa, IPAC Confer- source, and then demonstrate the laser-slowing and ence Proceedings, pp. 1281-1283, May 2012). We are a cooling of CaH. unique group in atomic/molecular physics group to have at our disposal for the design of traps the sophisticated 4. Install the linear accumulator (having designed and particle-tracking codes developed for high-energy physics; built it during the FY), and demonstrate its operation anything short of this misrepresents reality. with Li atoms. 630

Page 631

Accompanying these goals will be the equally important task of preparing manuscripts for professional publication. Conclusion Materials are extraordinarily different at ultracold tem- peratures. Kamerlingh Onnes, who in 1911 discovered superconductivity, remarked that his finding of a new state of matter through his work in gas liquefaction surpassed his wildest anticipation. A century later, superconductiv- ity underlies multi-billion dollar industries such as medi- cal imaging. We are mindful of applications our research has in new types of sensors. More generally, however, the dense samples of ultracold molecules we seek to generate will give us and others a new view to the inner workings of matter for both discovery and exploitation, of which super- conductivity is but one storied example. Publications Ferguson, S., P. Walstrom, J. Velasquez, III, and M. Di Rosa. Design of a magnetic trap and accumulator for laser- cooled atoms and molecules (LA-UR-13-25214). Pre- sented at LANL Student Symposium. (Los Alamos, NM, 26-27 July 2013). Velasquez, J., P. L. Walstrom, and M. D. Di Rosa. Developing density of laser-cooled neutral atoms and molecules in a linear magnetic trap. Presented at 2013 Meeting of the Division of Atomic, Molecular, and Optical Physics of the American Physical Society. (Quebec City, Que- bec, Canada, 3-7 June 2013). Walstrom, P. L., and M. D. Di Rosa. Beam transport and storage with cold neutral atoms and molecules. 2012. In International Particle Accelerator Conference 2012. (New Orleans, 20-25 May 2012). , p. 1281. New York: IPAC’12 OC / IEEE. 631

Page 632

Physics Exploratory Research Continuing Project Next Generation Earth Models Monica Maceira 20120047ER Introduction some of the receiver functions. While this helps, and The main goal of this project is to develop a fast and allows for inclusion of some additional data con- robust method to derive pictures of the Earth’s structure straints, the bulk of the noisy receiver functions are from a combination of four geophysical data sets. Our not improved through this approach. This suggests multi-parameter inversion will produce more accurate that these data are intrinsically bad and should not Earth models than currently available and will form the be used in the inversion. basis for the next generation of Earth models.The specif- • The thorough implementation and testing of the ic elements of the proposal are to 1) develop an efficient gravity element of our tripartite inversion for crust and robust multi-parameter inversion scheme, 2) de- and upper mantle structure of the western US. I velop the methodology for pre-processing and weighting have implemented two types of filtering for the the data sets to produce the best image, and 3) develop gravity partials, and tested three different velocity/ a detailed, validated model of the Earth’s crust beneath density relationships. So far, shear velocity results the western United States.This project addresses the from these efforts are very similar, indicating that fundamental problem of creating an accurate image of the inversion scheme is robust and not sensitive to the Earth, currently a barrier to deeper understanding of starting models or velocity/density relationships. the Earth’s current state and evolution. • Presentation of our latest results at the American Benefit to National Security Missions Geophysical Union meeting, 12/2012, in San Fran- The research will create unique expertise at the Labo- cisco ratory and will enhance the Laboratory’s capability to address core missions in Ground-based Nuclear Detona- • Preparation of a journal article showing our joint tion Detection within Nuclear Nonproliferation, as well inversion results using two data types: surface waves as missions in geothermal energy, geological carbon and receiver functions. This paper is approximately sequestration, used fuel disposition, and repository sci- half completed as of 6/6/13. Below is the abstract ence under Applied Energy Programs, Fossil Energy, and for our paper, outlining our results. Civilian Nuclear Programs. All of these efforts rely on ac- curate determination and monitoring of Earth structure As datasets for studies of Earth’s crust and mantle be- for success. We also foresee a follow-on LDRD DR aimed come more extensive, we have the unique opportunity at extending this modeling to the global scale using High to test comprehensive analysis methods not previously Performance Computing. The end result would be a possible. The western United States is an ideal test bed world-wide map of strain perturbations for the purpose for such research due to the extensive geophysical data of earthquake prediction. now available. Utilizing these data resources we have implemented a joint, simultaneous inversion of receiver Progress functions and surface wave dispersion to produce an The key accomplishments this year have been: image of the shear wave velocity structure beneath the western U.S. to depths of about 200 km. We use a • Gathering and processing and incorporation of all standard Least Squares inversion method optimized for receiver function data into the inversion. We have sparse large sparse matrices with simple parallelization. also tested the prospect of low pass filtering of the Our results are largely consistent with previous work receiver functions to eliminate excessive noise in for the area, but with increased vertical resolution due 632

Page 633

to methodology. Surface wave dispersion measurements several years and will be used throughout the scientific come from analysis of regional earthquakes using multiple community for understanding the structure of the Earth’s filter analysis, with good geographic coverage from periods crust and mantle. These models will reveal insights into of 8 s to 130 s. Receiver functions are downloaded from how the Earth’s lithosphere is formed and how it evolves the Earthscope Automated Receiver Survey and we com- over time, with impacts on geothermal resources, disposi- pare results using both high and low frequency receiver tion of nuclear waste, and global nuclear explosion moni- functions. Overall, structure beneath the western United toring. States is lower velocity than that for the continental craton to the east of the Rocky Mountains. In particular, low Publications velocities are seen in the mid-crust under Yellowstone, as Modrak, R., M. Maceira, J. van Wijk, and D. Coblentz. Joint well as beneath northern coastal California and coastal inversion of group velocity dispersion estimates and Oregon. In the middle to lower crust we see low velocities gravity data: Implications for the East Africa rift struc- ture. Earth and Planetary Science Letters. beneath the Aspen Anomaly, the San Francisco Peaks, the Cascade Range and the western Basin and Range. At “typi- Steck, L., M. Begnaud, S. Phillips, and R. Stead. Tomogra- cal” Moho depths of 35 km. we see low velocities beneath phy of crustal P and S travel times across the western the Colorado Plateau and the Snake River Plain indicat- United States. 2011. JOURNAL OF GEOPHYSICAL RE- ing thicker crust for those provinces, while low velocities SEARCH-SOLID EARTH. 116: B11304. beneath western Washington likely reflect subduction of ocean crust in that location and resultant thickening of Steck, L., M. Maceira, and C. Ammon. Crust and Upper Mantle Structure of the Western US from Simultane- the crustal stack. To the south of the Colorado Plateau, ous Inversion of Surface Wave Dispersion, Gravity, and beneath the southern Basin and Range, we see higher Receiver Functions. 2012. In Seismological Society of than average velocities. Broad low velocities are observed America Annual Meeting. (San Diego, 17-19 April). Vol. under California and Oregon at depths of 80 km, tapering 83, p. 357. Albany: BSSA. through the Snake River Plain to a terminus at Yellow- stone. At this depth velocities are also lower under the Rio Zhang, H., C. Thurber, M. Maceira, and P. Roux. Joint in- Grande Rift. Broad but less pronounced low velocities are version of body-wave arrival times and surface-wave observed continuing to depths of 200km where they then dispersion for three-dimensional seismic structure increase to more significant levels.” around SAFOD. Pure and Applied Geophysics. Future Work The rest of this fiscal year and the next fiscal year will be spent acquiring and processing the body wave data (the last of the four data types), and performing inversions of all combinations of data types. For the body wave data we will draw on data holdings from our LANL seismic research database. Results will be presented at the annual meeting of the American Geophysical Union and/or the Seismologi- cal Society of America meeting. We anticipate finishing and submitting our first paper on joint inversion of surface waves and receiver functions by the end of this fiscal year. We then will write and submit a manuscript on the joint in- version of gravity, receiver functions and surface wave data for the western United States. Following the incorporation of body wave data into the inversion, we will write and submit a paper on that topic. All papers will be submitted to high profile earth science journals. Conclusion The expected results include a fast multi-parameter inver- sion code for determining high resolution Earth models, a detailed model of the crust and upper mantle of the west- ern US available, and several publications in high impact Earth science journals. The code will be state of the art for 633

Page 634

Physics Exploratory Research Continuing Project Non-Equilibrium Fluctuation-induced Interactions Diego A. Dalvit 20120149ER Introduction with desired nano-electromechanical functionalities (e.g. Fluctuation-induced interactions are ubiquitous inphys- systems with reduced friction), and enhanced near-field ics. Relevant examples are forces between nanostruc- thermophotovoltaic energy conversion efficiencies. tures due to quantum fluctuations of the electromagnet- Benefit to National Security Missions ic field (EM Casimir interactions), andforces oncolloidal particles due to density fluctuations of binary fluid This project ties to LANL missions on quantum science mixtures (critical Casimir interactions). Although much of and nanotechnology. It leverages existing capabilities the equilibrium aspects of these interactions are under- across the laboratory, both in the T and MPA divisions. stood and confirmed by experiment, non-equilibrium We expect to produce new ideas that will help consoli- is still an open territory. The goal of this proposal is to date LANL as a world leader in fluctuation-induced inter- carry out a consorted theoretical and experimental ef- actions. The new capabilities that we will build, including fort to investigate non-equilibrium quantum and thermal analytical tools, numerical algorithms, nanofabricated fluctuation-induced electromagnetic interactions at the surfaces,and scanning thermal microscopy techniques nano-scale, using state-of-the-art analytical, numerical, for engineered heat transfer, will enable LANL to pursue nanofabrication, and measurement tools. future missions in nanotechnology, nanophotonics, and quantum physics. One of the possible external agencies Our approach will enable the understanding of, and that we believe may be potentially interested in our R&D ultimately controlover fluctuation-induced interactions is DARPA, with which we have worked in the past on in dynamical and thermal non-equilibrium systems, such equilibrium Casimir physics. as: (i) non-contact forces and friction between pat- terned surfaces and atoms in relative motion, induced Progress by quantum EM fluctuations; and (ii)nano-scale radiative We have performed theory and numerical studies of heat transfer between complex nanostructures, induced non-equilibrium fluctuation-induced phenomena, both by thermal EM fluctuations. We plan to formulate the in radiative heat transfer and quantum friction. On the first principles, microscopic theory of non-equilibrium first topic, we have computed the radiative heat transfer EM fluctuation interactions using the open quantum between nanostructured gold plates in the framework systems paradigm, and to develop numerical methods of the scattering theory, and predicted an enhancement to compute them exactly for complex nanostructures of the heat transfer as one increases the depth of the out of dynamic and/or thermal equilibrium. We will corrugations while keeping the distance of closest ap- verify the stationary thermal non-equilibrium aspects by proach fixed. This effect was interpreted in terms of the demonstrating, for the first time, tailored nano-scale ra- evolution of plasmonic and guided modes as a function diativeheat transfer between structured surfaces. To this of the grating’s geometry. This work was published in aim we will fabricate dielectric and metallic patterned Physical Review B as a Rapid Communications. In quan- nanostructures,and measure heat transfer between tum friction, we have studied the non-contact drag force them using scanning thermal microscopy. between a moving atom and a surface arising from the exchange of energy and momentum mediated by quan- Our end result will be a suite of theoretical and ex- tum and thermal fluctuating photons. Using fundamen- perimental tools designed to explore non-equilibrium tal theorems of quantum statistical mechanics, such as aspects offluctuation-induced interactions. This novel ca- the fluctuation-dissipation theorem, we have obtained pability will enable the rational design of nanostructures how the quantum frictional force depends on the atom’s 634

Page 635

velocity and the atom-surface separation, and have clari- aration. Even though we were not able to detect variations fied some controversies in the literature about the correct of the heat transfer with sub-micron tip-sample separa- calculation of this effect. We are currently writing up a tion, we could detect lateral variations of thermal conduc- manuscript to submit to Physical Review. In addition, we tivity between tip and sample which would be applicable have developed a quasi-analytical approach to computing to studies of wide range of nanostructured thermoelectric fluctuation-induced interactions (such as Casimir forces materials available at CINT. To improve the resolution and and radiative heat transfer) in nanosctructured surfaces detect near-field effects in heat transport dependence on based on a modal approach. Our method is based on the sub-micron tip-sample separation, we have established col- exact analytical form of the eigenvectors of such structures laboration with PicoCal personnel who will provide novel and the numerical computation of the eigenvalues from a 4-probe RTD tips that will allow at least an order of magni- simple transcendental equation. This was was published in tude enhancement in the microscope sensitivity to ther- Physical Review A. mal conductivity changes. The improved measurements resolution will enable reliable quantification of tip-sample We have designed and fabricated submicron scale gratings separation dependence of the heat transfer between plane for experimental study of radiating heat transfer between and heterostructured metal surfaces. in the plane sample grating-plane and sphere. Our theoretical study shows that geometry. rate of radiative heat transfer increases with the grating’s corrugation depth. A grating design with a periodicity of Future Work 1 µm and 50% filling fraction shows an order of radiation We will study nanoscale radiative heat transfer between enhancement as the grating aspect (depth/width) ratio metallic and dielectric nanostructures, including gratings, approaches to 10. We fabricated a set of gratings using multilayered materials, and metamaterials. Along this standard nano-lithography, etching, and metallization line of research we will design, fabricate, and character- techniques. At this point, theses gratings are fabricated on ize structures in the micron to nano scale size that will be silicon substrates which are subsequently metalized with used in our experiments on radiative heat transfer. We will a film thickness greater than the skin depth. The fabrica- perform the calibration of the experimental set up using tions starts with e-beam lithography on a silicon substrate, the sphere-plane configuration and compare results with next a dry RIE etching is used to obtain desired corruga- theory. tion depth, and finally, the gratings are metalized using a sputtering system. We optimized the recipe for highly Along the other thread of the quantum friction, we will anisotropic silicon etching and also optimized metallization study non-contact forces and friction between patterned parameters for uniform metal thickness around the grating surfaces and atoms in relative motion, induced by quan- structure. After metallization of 60 nm of silver, which is tum EM fluctuations. We will finish our ongoing work on thicker than the skin depth, our gratings have 50% filling quantum friction between atoms and surfaces in relative fraction. We vary the grating depth from 1.5 µm to 4.5 µm motion, and then generalize our results to the case of which gives a range of aspect ratio from 3 to 9. Further quantum friction between two surfaces in relative motion. design modification of gratings will commensurate with measurement of current samples. Conclusion The main expected result of this project is a complete We have tested the performance of our newly developed physical description of non-equilibrium fluctuation-in- variable-temperature UHV scanning thermal microscope duced interactions. Our two main deliverables will be: 1) a by measuring the temperature distribution on gold/silicon predictive theoretical/numerical framework for non-equi- gratings. When such heterogeneous samples are heated librium fluctuation-induced interactions, and 2) the first using electron bombardment in vacuum, inhomogeneity in experimental demonstration of engineered radiative heat material properties lead to nonuniform temperature dis- transfer with fabricated nanostructures. We will perform tribution on the surface. We have shown that our thermal theory, numerics, and experimental validation. microscope is capable of detecting these variations with better then 15 mK resolution. To measure nanoscale heat Publications transfer between the tip and a flat gold surface, we have Behunin, R. O., F. Intravaia, D. A. Dalvit, P. A. Neto, and used the same instrument in thermal conductivity detec- S. Reynaud. Modeling electrostatic patch effects in tion regime. In these studies, commercially available RTD Casimir force measurements. 2012. Physical Review tips are heated using AC currents and heat dissipation due A (Atomic, Molecular, and Optical Physics). 85 (1): to thermal conduction through the vacuum gap leads to tip 012504 (9 pp.). temperature variations as a function of the tip-sample sep- Behunin, R. O., Y. Zeng, D. A. R. Dalvit, and S. Reynaud. 635

Page 636

Electrostatic patch effects in Casimir force experiments performed in the sphere-plane geometry . 2012. Physi- cal Review A. 86: 052509. Behunin, R., D. Dalvit, R. Decca, and C. Speake. Limits on the accuracy of isoelectronic gravity measurements at short separation due to patch effects. Physical Review D. Guerout, R., J. Lussange, F. S. Rosa, J. -. Hugonin, D. A. Dalvit, J. -. Greffet, A. Lambrecht, and S. Reynaud. Enhanced radiative heat transfer between nanostruc- tured gold plates. 2012. PHYSICAL REVIEW B. 85 (18): 180301. Intravaia, F., P. S. Davids, R. S. Decca, V. A. Aksyuk, D. Lopez, and D. A. R. Dalvit. A quasi-analytical modal approach for computing Casimir interactions in periodic nano- structures . 2012. Physical Review A. 86: 042101. Intravaia, F., R. Behunin, and D. Dalvit. Quantum friction and non-equilibrium fluctuation theorems. Physical Review Letters. Intravaia, F., S. Koev, I. Jung, A. Talin, P. Davids, R. Decca, V. Aksyuk, D. Dalvit, and D. Lopez. Strong Casimir force reduction through metallic surface nanostructuring. 2013. Nature Communications. 4: 2515. Rodriguez, A., M. Homer Reid, A. Woolf, D. Dalvit, F. Ca- passo, S. Johnson, and F. Intravaia. Geometry-induced Casimir suspension of oblate bodies in fluids. Physical Review Letters. Rosa, F. S., D. A. Dalvit, and P. W. Milonni. Electromagnetic energy, absorption, and Casimir forces. II. Inhomoge- neous dielectric media. 2011. PHYSICAL REVIEW A. 84 (5): 053813. Sushkov, A. O., W. J. Kim, D. A. Dalvit, and S. K. Lamoreaux. New Experimental Limits on Non-Newtonian Forces in the Micrometer Range. 2011. Physical Review Letters. 107 (17): 171101 (4 pp.). 636

Page 637

Physics Exploratory Research Continuing Project Petascale Kinetic Plasma Simulation of the Interaction Among Laser Speckles in Nonlinear Optical Systems Lin Yin 20120153ER Introduction emergent phenomena may take place. Understanding nonlinear optics of plasmas is a high Benefit to National Security Missions energy density physics grand challenge. Decades of research has shown that this problem plays host to a Our project engages in high-risk/high-payoff scientific wide range of complex physics. This problem is ideal for R&D that has the potential to enable inertial confine- advancing fundamental, discovery-class science in part ment fusion experiments to achieve ignition, justifying because supercomputing power and simulation codes the first element in Agency Relevance. There is also an have co-evolved to the point where ab initio exploration effort in the DOE SC Office of Fusion Energy to consider of the underlying physics is now possible. inertial fusion energy in various IFE approaches, includ- ing conventional ICF and fast ignition ICF; our work also With Petaflop-class supercomputers, the investigators supports that primary mission, so the DOE/SC “A” is have advanced the understanding of the basic physics of appropriate. stimulated Raman scattering in solitary laser speckles, the irreducible “building blocks” of intensity structures Ignition experiments on the NIF have been identified found in laser beams. If laser speckles do not meaning- as vital for SSP and other defense programs missions. fully interact with one another, then a knowledge of sin- Mission relevance “A” should be checked for “Nuclear gle-speckle behavior and the statistics of speckle intensi- Weapons: Safety, Security, and Reliability of the Nuclear ties would be all that’s needed to craft a macroscopic Stockpile” under “Mission Relevance”. model of plasma behavior: a large laser beam with The DOE/SC energy mission driver makes fusion nuclear hundreds of thousands of speckles would act merely as energy an appropriate energy security mission; hence, the simple sum of the many constituent speckles plus a “A” was checked for Nuclear Energy in this category. relatively tame background contribution. HEDP is a priority of the Laboratory that is closely However, since laser speckles interact nonlinearly with aligned with its nuclear security and energy indepen- one another, “meso-scale” modeling becomes essential dence goals; this work underpins the use of leading because small aggregates of laser speckles may exhibit HEDP experimental facilities. A natural customer for this emergent behavior. One way that this could happen is work would be the Science Campaigns, specifically C10 through the exchange of particles and seed wave energy (LANL Program Manager: Steve Batha); this work would among laser speckles. An intense speckle in an environ- also be applicable to ongoing C4 projects. LANL is likely ment with other speckles nearby could “destabilize” to play a leading role in any prospective national inertial those neighbors, resulting in enhanced emission of fusion energy program, a topic presently being assessed particles and transmission of wave energy from those by a National Research Council panel (whose final report speckles back to the original speckle, thus affecting its will be submitted in two years). LPI is a natural problem behavior nonlinearly and nonlocally. area and test bed for improvements to our VPIC code- In our research, we will study LPI in first-principles base, which is being further adapted for GPU multi-core computer simulations of both modest- and large-sized and Exascale supercomputers. ensembles of laser speckles. We will employ the LANL Progress VPIC particlein-cell code to perform simulations in vari- ous plasma conditions in order to identify when and how In laser-driven fusion experiments, stimulated Raman 637

Page 638

scattering (SRS) occurs when electron density fluctuations including the trapped electron effects, we expect to under- are amplified resonantly by the incident laser beams and stand how the intensity thresholds are lowered through scattered light. These laser beams comprise several thou- longitudinal and transverse coupling among speckles via sands of individual laser speckles. In this project, we em- hot electron exchange and scattered light, to predict the ploy the VPIC particle-in-cell (PIC) code to perform simula- threshold for linear to nonlinear transition, and to obtain tions of SRS to study how laser speckles interact through the conditions for emergent phenomena in speckled laser the exchange of particles and seed waves. During the beams. second year of the project, the nonlinear physics governing the kinetic behavior of stimulated Raman scattering (SRS) Publications in large networks of speckles has been identified in the Yin, L., B. J. Albright, H. A. Rose, K. J. Bowers, B. Bergen, trapping regime at kλD~0.3 (here k is the wave number of and R. K. Kirkwood. Selforganized bursts of coherent the electron plasma waves and λD is the Debye length) in stimulated Raman scattering and hot electron trans- port in speckled laser plasma media. 2012. Physical homogeneous and inhomogeneous plasmas. Hot electrons Review Letters. 108: 245004. from intense speckles, both forward and side-loss hot elec- trons produced during SRS daughter electron plasma wave Yin, L., B. J. Albright, H. A. Rose, K. J. Bowers, B. Bergen, R. bowing and filamentation, seed and enhance the growth K. Kirkwood, D. E. Hinkel, A. B. Langdon, P. Michel, D. S. of SRS in neighboring speckles by reducing Landau damp- Montgomery, and J. L. Kline. Trapping induced nonlin- ing. Trapping-enhanced speckle interaction through trans- ear behavior of backward stimulated Raman scattering port of hot electrons, backscatter, and sidescatter SRS light in multispeckled laser beams. 2012. Physics of Plas- waves enable the system of speckles to self-organize and mas. 19: 056304. exhibit coherent, sub-ps SRS bursts with more than 100% Yin, L., B. J. Albright, H. A. Rose, D. S. Montgomery, J. L. instantaneous reflectivity, resulting in an SRS transverse Kline, R. K. Kirkwood, P. Michel, K. J. Bowers, and B. coherence width much larger than a speckle width and a Bergen. Self-organized coherent bursts of stimulated SRS spectrum that peaks outside the incident laser cone. Raman scattering and speckle interaction in multi- speckled laser beams. 2013. PHYSICS OF PLASMAS. 20 Future Work (1): -. In laser-driven fusion experiments, stimulated Raman scat- tering (SRS) occurs when electron density fluctuations are amplified resonantly by the incident laser beams and scat- tered light. These laser beams comprise several thousands of individual laser speckles. In this project, we employ the VPIC particle-in-cell (PIC) code to perform simulations of SRS to study how laser speckles interact through the exchange of particles and seed waves. The primary objec- tive of the work we proposed is to use petascale kinetic plasma simulations to examine the interaction among laser speckles in nonlinear optical systems. With the additional diagnostic, algorithm development, and new boundary conditions for laser injection imple- mented during the first year, we will study SRS in VPIC simulations of large-sized ensembles of laser speckles in the second year of our project. We will perform two- dimensional simulations in various plasma conditions in order to identify when and how speckle interaction may take place in large networks of speckles. These “capability- class” simulations of large networks of laser speckles to will validate our understanding of the basic phenomena discovered so far. Conclusion Through the use of 2D and 3D kinetic simulations at mod- est and large scale, along with theory of reduced model 638

Page 639

Physics Exploratory Research Continuing Project Multiscale Spacecraft Charging Simulations in Support of New Space Missions Gian L. Delzanno 20120190ER Introduction Benefit to National Security Missions This project is aimed to open up a new field of experi- This project represents an opportunity for LANL to lead mental space plasma physics. The space experiment will the NASA community into new exciting science, enabled connect the Earth magnetosphere and ionosphere by by new, pioneering experimental techniques. In addition, emitting an energetic beam of electrons from a magne- this work involves unique computational capabilities tospheric spacecraft and detecting the beam optically in which, together with the significant experimental capa- the ionosphere. Thus, for the first time, dynamic mag- bilities available at LANL, could be united for future mis- netospheric processes will be probed directly by mag- sions. This will allow unprecedented leverage between netic field line mapping over vast regions (thousands of scientific and programmatic pursuits. In this regard, we kilometers) of near Earth space. have already identified program development opportu- nities with the NASA Explorer program or the Innovative The major obstacle to this experimental technique is Advanced Concepts Office. A program development plan how to emit enough electrons from the spacecraft to is ready (the ConnEx proposal to trace magnetic field permit ionospheric imaging without deleteriously charg- lines from the magnetosphere to the ionosphere) and a ing the spacecraft or, in other words, how to control the successful outcome of this LDRD proposal would enable charging of the spacecraft to benign levels which will its pursuit. We also emphasize that the computational allow the electron beam to be emitted. This is the goal tools available, which will be further developed in this of this LDRD project. project, could have an impact in other fields where kinetic plasma simulations are important such as, for The spacecraft charging problems considered here are instance, magnetic fusion energy. extremely difficult to solve computationally, since they involve spatial scales from mm to 10’s of km and time Progress scales from microseconds to seconds. To solve this We have submitted two papers, one to Journal of Com- multiscale problem, we will model the interaction of the putational Physics on numerical methods for calculat- beam-emitting spacecraft with the surrounding magne- ing particle trajectories in a given electromagnetic field tospheric plasma using a state-of-the-art 3d PIC code. in cylindrical geometry and one to IEEE Transaction on The micro and macro scale physics will be simulated for Plasma Science on an overview of the CPIC code with realistic geometries and conditions. application examples on the charging and shielding of an object in a plasma. This work will result in a thorough and hitherto unknown understanding of the magnetospheric conditions in We are working on the parallelization of CPIC, perform- which it is possible to stably emit a beam of electrons. ing domain decomposition of the computational domain This understanding will enable LANL to proceed with and using the MPI library to handle the communication plans to propose a NASA Explorer mission ($200M-amongdifferentprocessors.Thecodehasbeenparal-$500M, named ConnEx) to experimentally study mag- lelized in two dimensions, while the work on the three netic field line connectivity. Ultimately, this LDRD project dimensional version is ongoing. We have successfully may facilitate a paradigm shift in space plasma physics benchmarked the parallel code against the examples from using static magnetic field models to using dynamic that are presented in the paper above. With this up- probing of the magnetic field configuration of the Earth grade CPIC is significantly further ahead of its competi- magnetosphere. tors in terms of flexibility, performance and the ability to 639

Page 640

handle larger and more challenging problems. which contributes to increasing the importance of ConnEx mission goals. The focus group meets twice yearly, with We have continued the work to obtain a static solution for three sessions at the upcoming summer meeting. New the plume near the spacecraft. We have studied the most this year is the utilization of data from the new NASA Van popular plume models (Korsum and Tverdokhlebova, Parks Allen Probes. These data can be used to provide more and Katz, Ashkenazy and Fruchtman) and discovered that realistic parameters for the ConnEx mission. A manuscript they are intrinsically inconsistent and can produce unphysi- is in preparation which tests the mapping during an active cal solutions. Notably the first model is used to obtain geomagnetic storm time and other researchers are actively the plasma contactor plume in studies relevant to the using the new dual probe data for other mapping ques- International Space Station. These models are based on a tions. macroscopic description of the plasma and the choice of a self-similar solution, but do not fully satisfy the original Future Work model equations. We have developed a new self-similar We plan to continue the characterization of the contac- model that mitigates these drawbacks by adding degrees tor plume with PIC simulations in a rescaled parameter of freedom to the solution. In addition, we have developed regime, focusing on (1) plume injection parallel to the a numerical code that solves the full set of model equa- magnetic field, and (2) plume injection perpendicular to tions (without self-similarity) and used it to calculate the the magnetic field. accuracy of the self-similar models. Our new model has the following advantages: does not produce unphysical solu- With the new plume model developed for the injection tions; produces a broader family of static plume models; parallel to the magnetic field, we will continue performing and the accuracy of such model (relative to the full numer- PIC simulations of the spacecraft charging in presence of ical solution) is at least an order of magnitude better and electron beam emission. The goal is to characterize (1) the in some cases the error is just a few percent. A manuscript physics of spacecraft charging in presence of plume and is in preparation on these results. The model is currently electron beam, (2) the scaling laws that govern the amount being extended to include a magnetic field parallel to the of current that can be drawn from the spacecraft, (3) the plume injection. neutralization of the contactor cloud from the surrounding environment. We have performed microscopic (Particle-In-Cell) simula- tions of the plume injection. We have studied the case We plan to perform studies similar to those in (2) but for where the plasma injection is perpendicular to the mag- an injection perpendicular to the magnetic field. These netic field and found that while the plume expands several simulations are significantly more challenging since the instabilities develop. These are (1) the deformation of the development of the plume is affected by plasma instabili- head of the plume with the formation and detachment ties. We will therefore rescale the simulation parameters of large plasma vortices, (2) the development of large to keep the cost of the simulation manageable, and extract amplitude surface waves on the plume flanks and (3) the relevant scaling laws. overall kinking of the plume at later times. These results are important as they indicate that, for plume injection We will continue work with the Magnetospheric Mapping perpendicular to the magnetic field, static models are in- NSF focus group. Activities for the next year will include adequate. In order to interpret the results concerning the facilitating a mapping challenge, likely using new Van Allen flank instability, we have developed a simple linear theory Probes data, and summarizing current mapping techniques model. The theory reproduces qualitatively some aspects and their uncertainties into a review paper. of the simulations, namely the development of two modes We note that the results from these studies will extend with different wavelengths localized at the top and bot- beyond this project, since contactor plasmas are used in tom flanks. It also characterizes the instability as a Kelvin- various space applications and spacecraft charging is a Helmoltz instability, driven by the velocity shear that de- fundamental process that plagues every spacecraft. velops due to charge separation (induced by the magnetic field) in small layers on the plume flanks. A manuscript is in Conclusion preparation on these results. First, this project will provide a thorough understanding of We have continued efforts with the NSF Geospace Envi- the key physical processes that govern the neutralization ronmental Modeling program’s focus group on Scientific of charged bodies (such as a spacecraft) in space. Second, Magnetic Mapping and Techniques. This raises the in- it will quantify the maximum amount of electrons that can ternational magnetospheric community’s awareness of be emitted from a magnetospheric spacecraft. Both goals mapping issues, and understanding of the scientific gaps, will provide the foundation for a new experimental tech- 640

Page 641

nique aimed at dynamic probing of the Earth magneto- vilinear Particle-In-Cell code for plasma-material inter- sphere, which will shed new light on some of the processes action studies. Presented at American Physical Society that govern the Earth magnetosphere-ionosphere. This Division of Plasma Physics meeting. (Providence, RI, 29 Oct -2 Nov 2012). project can also have significant impact on how to protect satellites and spacecrafts from the nearby environment Delzanno, G. L., E. Camporeale, J. D. Moulton, J. E. Boro- and on future space missions. vsky, E. A. MacDonald, and M. F. Thomsen. CPIC: a Curvilinear Particle-In-Cell code for plasma-material Publications interaction studies. To appear in IEEE Transactions on Boella, E., M. Giraudo, and G. L. Delzanno. Numerical in- Plasma Science. vestigation of the oscillations in a virtual cathode. Pre- sented at 40th European Physical Society Conference Delzanno, G. L., E. Camporeale, J. E. Borovsky, E. A. Mac- on Plasma Physics. (Aalto, Finland, 1-5 Jul 2013). Donald, M. F. Thomsen, and E. A. Hogan. Probing the Earth’s magnetosphere with an electron gun. Present- Camporeale, E.. CPIC: a curvilinear Particle-In-Cell code ed at 40th European Physical Society Conference on for plasma-material interaction studies. Presented at Plasma Physics. (Aalto, Finland, 1-5 Jul 2013). American Physical Society April meeting (International Sherwood Fusion theory conference. (Atlanta, GA, 31 Delzanno, G. L., E. Camporeale, J. E. Borovsky, E. A. Mac- Mar - 3 Apr 2012). Donald, M. F. Thomsen, and S. Markidis. Instabilities during the expansion of a plasma in a transverse mag- Camporeale, E., G. L. Delzanno, J. E. Borovsky, E. A. Mac- netic field. Presented at AGU Fall meeting. (San Fran- Donald, and M. F. Thomsen. The expansion of plasma cisco, CA, 9-13 Dec 2013). in a transverse magnetic field. Presented at 2013 Inter- national Sherwood Fusion Theory Conference. (Santa Delzanno, G. L., E. Camporeale, and J. D. Moulton. Adap- Fe, NM, 15-17 April 2013). tive Particle-In-Cell methods. Invited presentation at 40th European Physical Society Conference on Plasma Camporeale, E., G. L. Delzanno, J. E. Borovsky, E. A. Mac- Physics. (Aalto, Finalnd, 1-5 Jul 2013). Donald, and M. F. Thomsen. The expansion of a plas- ma in a transverse magnetic field. Presented at 40th Delzanno, G. L., and E. Camporeale. On particle movers in European Physical Society Conference on Plasma Phys- cylindrical geometry for Particle-In-Cell simulations. ics. (Aalto, Finland, 1-5 Jul 2013). 2013. Journal of Computational Physics. 253: 259. Delzanno, G. L., E. Camporeale, E. Hogan, J. D. Moulton, Giraudo, M., G. L. Delzanno, and E. Camporeale. Theo- J. E. Borovsky, E. A. MacDonald, and M. F. Thomsen. retical and computational studies of the sheath of a Probing the Earth’s magnetosphere with an electron planar, electron emitting wall. Presented at American gun. Presented at 55th Annual Meeting of the APS Di- Physical Society Division of Plasma Physics. (Provi- vision of Plasma Physics . (Denver, CO, 11-15). dence, RI, 29 Oct - 2 Nov 2012). Delzanno, G. L., E. Camporeale, J. D. Moulton, J. E. Bo- Hogan, E. A., G. L. Delzanno, E. Camporeale, J. E. Borovsky, rosvky, E. A. MacDonald, and M. F. Thomsen. CPIC: a E. A. MacDonald, and M. F. Thomsen. Modeling the ex- curvilinear Particle-In-Cell code for plasma-material pansion of a contactor plasma. Presented at American interaction studies. Presented at American Geophysical Geophysical Union fall meeting. (San Francisco, CA, 3-7 Union fall meeting. (San Francisco, CA, 3-7 Dec 2012). Dec 2012). Delzanno, G. L., E. Camporeale, J. D. Moulton, J. E. Boro- Hogan, E. A., G. L. Delzanno, E. Camporeale, J. E. Borovsky, vsky, E. A. MacDonald, and M. F. Thomsen. CPIC: a E. A. Macdonald, and M. F. Thomsen. Modeling the ex- curvilinear Particle-In-Cell code for spacecraft-plasma pansion of a contactor plasma. Presented at American interaction studies. Invited presentation at 12th Space- Physical Society Division of Plasma Physics meeting. craft Charging and Technology conference. (Kitaky- (Providence, RI, 29 Oct- 2 Nov 2012). ushu, Japan, 14-18 May 2012). Zhao, L., and G. L. Delzanno. Charging and shielding of a Delzanno, G. L., E. Camporeale, J. D. Moulton, J. E. Boro- non-spherical object in a plasma. Presented at 55th vsky, E. A. MacDonald, and M. F. Thomsen. CPIC: a cur- Annual Meeting of the APS Division of Plasma Physics . vilinear Particle-In-Cell codes for dusty plasma studies. (Denver, CO, 11-15 Nov 2013). Invited presentation at 13th Workshop on the Physics of Dusty Plasma. (Waco, TX, 20-23 May). Zhao, L., and G. L. Delzanno. Charging and shielding of a non-spherical dust grain in a plasma. Presented at AGU Delzanno, G. L., E. Camporeale, J. D. Moulton, J. E. Boro- Fall meeting. (San Francisco, CA, 9-13 Dec 2013). vsky, E. A. MacDonald, and M. F. Thomsen. CPIC: a cur- 641

Page 642

Physics Exploratory Research Continuing Project Laboratory Study of Cosmically-relevant Collisionless Shocks Scott C. Hsu 20120200ER Introduction providing a recruiting portal for students and postdocs. This project will generate/diagnose collisionless shocks In addition, this project leverages existing DOE/SC and diagnose/characterize shock-particle interactions in investment in the experimental facility, which supports an existing laboratory plasma facility in P-24. Although another NPAC recommendation to promote LANL’s as- collisionless shocks in plasmas were first predicted in trophysics/cosmology portfolio to DOE/SC. Fundamental the 1950’s and discovered in the 1960’s, many research research in plasma turbulence builds capability for some questions relating to the microscopic physics of collision- of the toughest questions in weapons physics. less shock formation, evolution, and shock acceleration Progress of particles to very high energies remain unanswered. The proposed experiments will emphasize the ability to Thus far in FY13, the project has fired more than 1000 (1) to control and scan physics parameters over a range experimental plasma shots of head-on collisions of su- of values and across physics regimes in order to validate personic plasma jets, and we have pushed into collision- physics models, and (2) obtain far more measurements less parameter regimes that are the focus of the project. in both space and time compared to either in situ space We have also completed, and submitted for publication, satellite measurements or astronomical observations. an extensive set of electromagnetic particle-in-cell (PIC) The collisionless shocks will be created via the head-on numerical simulations to guide the experiments, includ- collision of two oppositely directed high velocity plasma ing simulations of jet propagation and both unmag- jets launched by pulsed-power driven plasma rail-guns. netized and magnetized jet merging. Finally, we are A magnetic field will be applied via coils at the jet inter- working on adding additional diagnostic instruments for action region to access magnetized regimes which are characterizing the shock structure and evolution, as well essential for cosmic-relevance. The scale sizes of the jets as a pulsed magnetic coil set for doing magnetized shock and shock thickness will be 30 cm and 1 cm, respectively, experiments in the third and final year of the project. enabling detailed characterization of shock structure and The experiments are showing the formation of a dynam- evolution via relatively simple diagnostics. As described ic interaction region between the two merging jets. We in the proposal, key dimensionless parameters in the are still in the process of obtaining better experimental experiments will satisfy quantitative physics criteria measurements, but we believe that the lower-density for the collisionless shocks to be of cosmic relevance. leading edges of the jets are interpenetrating at early The objectives of this work are to develop insights into times, and then when the higher-density bulks of the collisionless shocks not obtainable via satellite mea- jets meet, a more shock-like interaction initiates and pro- surements or astronomical observations alone, and to ceeds. The scale size of the interaction region is ~10 cm, un-fold the specific physics mechanisms underlying colli- which is smaller than the collision mean-free-path be- sionless shock formation and shock-particle interactions. tween counter-propagating jet ions. Thus, this is already strong evidence of a collisionless interaction (despite the Benefit to National Security Missions lack of such an interaction in the unmagnetized simula- The recent LANL NPAC Strategic Capability Planning tion results). exercise (April, 2010) recognizes that the depth/breadth of theory, experimental, and computational capability We are in the process of writing up the unmagnetized for the study of extreme astrophysical environments is jet-collision results for publication; also, the results will a LANL scientific “crown jewel.” This project will add a be presented by postdoc Auna Moser at the internation- unique laboratory component to that capability while 642

Page 643

al conference Interrelationship between Plasma Experi- Publications ments in the Laboratory and Space (IPELS) in Japan (July, Merritt, E. C., A. L. Moser, S. C. Hsu, J. Loverich, and M. 2013). Gilmore. Experimental Characterization of the Stagna- tion Layer between Two Obliquely Merging Supersonic The simulations have investigated jet propagation over Plasma Jets. 2013. PHYSICAL REVIEW LETTERS. 111 (8): the 1 m distance before they collide, as well as show what -. jet parameters are needed to give rise to the collisionless Thoma, C., D. R. Welch, and S. C. Hsu. Particle-in-cell simu- shock interactions we seek to study experimentally. In lations of collisionless shock formation via head-on highly resolved 1D and 2D simulations, we have shown merging of two laboratory supersonic plasma jets. that collisionless shocks are generated using our measured 2013. PHYSICS OF PLASMAS. 20 (8): -. jet parameters and when immersed in ~1 kG magnetic fields. The magnetic field orientation, as well as jet density gradients, are shown to have a strong effect on the nature of the jet interaction. We have built and installed probe diagnostics to measure local magnetic field, electron temperature/density, and plasma electric potential. We are presently troubleshoot- ing the probes as they are being affected by pulsed power electrical noise pickup. We have also built and operated a schlieren system to provide 2D time-resolved images of shock formation/evolution. We wrote and had accepted an operational IWD for the schlieren system. Right now, the schlieren system optics is being improved to maximize the sensitivity, and thus we have not yet obtained a shock image using the schlieren. We have an independently funded DOE undergraduate student joining us for the sum- mer to build another probe to measure ion particle ener- gies, to diagnose shock-particle interactions in our experi- ment. Finally, we will modify an existing capacitor bank and design/fabricate magnetic coils to add up to a several- kG magnetic field for experiments in the third year. Future Work Tasks for FY14 will focus on executing magnetized shock experiments using newly installed magnetic coils (to be completed in FY13) and based on numerical simulation work completed in FY13. In particular, we will diagnose the interaction region of two hydrogen plasma jets using newly implemented diagnostics (schlieren, interferometry, magnetic and electrostatic probes, and gridded ion energy analyzers) to provided new data on shock structure, evolu- tion, and shock-particle interactions. Conclusion Expected results are fundamental experimental data that will improve our understanding of cosmically-relevant colli- sionless shocks. In particular, the data will (1) demonstrate the formation/existence of collisionless shocks in a labo- ratory experiment, (2) provide characterization of shock structure and evolution as a function of plasma parame- ters, and (3) provide insight into the important mechanism of shock acceleration of particles. 643

Page 644

Physics Exploratory Research Continuing Project Nuclear Quadrupole Resonance: From No Field to Ultra-low Field Michelle A. Espy 20120218ER Introduction Benefit to National Security Missions The ability to non-invasively determine chemical NQR is a powerful method to detect interesting 14N composition is a significant problem for both medi- containing substances (e.g. proteins in the human body, cal, industrial and security screening. For example, the melamine in food, numerous explosives, and illicit ability to ascertain what a material is or whether it is drugs). NQR can provide highly specific chemical infor- counterfeit without opening a package. Methods based mation, is not a trace method, and can be measured in magnetic resonance, nuclear magnetic resonance from a distance (~ 10-20 cm). LANL already possesses (NMR) or nuclear quadrupole resonance (NQR) are very many of the tools required to develop NQR into a power- appealing because they provide a clear path to chemical ful approach. We propose to develop and extend NQR by information that does not require ionizing radiation and use of pre-polarization and combination with ULF NMR can penetrate material. Our goals for this proposal are to readout. We will develop sensitive NQR that can pen- leverage existing LANL expertise in magnetic resonance etrate through metals but does not require uniform or at low magnetic fields (< 100 mT) for NQR-based non- sophisticated magnetic fields (suitable for fielding). We invasive screening of materials. Our focus is to develop will work closely with existing sponsors (NIH, DHS, OBER) methods to enhance NQR sensitivity, and demonstrate to extend ongoing applications from ULF NMR to NQR. applicability to stand-off detection of contraband (explo- But the numerous new applications enabled will provide sives and drugs). significant program development opportunities in global security ranging from stand-off detection of explosives We will demonstrate an NQR capability that can be within IEDs or the human body to transformational combined with our present ultra-low field (ULF) NMR medicine. systems. We will show nearly simultaneous (interleaved) NMR/NQR. Progress We have constructed a fully homemade portable NQR We will enhance the sensitivity of NQR by use of pre-po- spectrometer designed to measure NQR modes from larization, in which spin polarized protons are recruited 400 kHz to 5 MHz in a 100 ml sample volume. The in an external pulsed magnetic field an transfer their spectrometer is composed of a transmit/receive sole- polarization to the 14N nuclei when ramped down past noid coil, variable tuning and matching capacitors, and the NQR frequency. We will then study the use of mag- supporting circuitry. An induced NQR signal is so tiny netic field gradients and/or radiation damping to reduce (nV – µV), thus it is amplified by a gain of ~ 60 dB in last the T1 (spin-lattice relaxation) time of the material which stage. Initially, we tested a sodium nitrite powder sample would speed up the polarization process. because of its relatively high NQR frequency at 4.64 MHz (thus is easier to detect). We have successfully observed Finally, we will investigate the use of cross-relaxation first NQR signal from the sample. 50 data sets were aver- again, this time using the NMR to read out the NQR. This aged to give an SNR of roughly 40. The spectrometer was approach allows for low-frequency readout of the NQR also used to measure relaxation times (T1 and T2) of the signal, enhanced signal strength for low NQR frequency sample with standard pulse sequences. To improve the materials (like TNT) and the combination with T1 disper- SNR, we employed a pre-polarization field applied to the sion contrast from the NMR which could greatly enhance proton subsystem in order to enhance the NQR signal by medical imaging applications. up to one-order of magnitude by using the mechanism of cross-polarization. To demonstrate the polarization 644

Page 645

enhanced NQR, we tested ammonium nitrate because it is NQR and 3) demonstrate NQR in combination with ultra- of significant interest, it has NQR modes at frequencies less low field NMR. than 500 kHz (thus is difficult to detect), and it has both In the past year we were able to achieve Goal #1 showing a 14N and 1H nuclei. Because the Q factor of the spectrom- fieldable system for NQR detection of materials of signifi- eter was measured to be low at such a low NQR frequency, cant interest to potential customers such as ammonium we replaced the T/R solenoid coil with another one with a nitrate and surrogates for contraband. We deliberately higher inductance. The improvement gives higher Q factor chosematerials that were at low NQR frequencies because of ~ 100 (thus the NQR signal is increased). In addition, we they are the most challenging. Our success has demon- improved a pulse by delivering more power to the system strated that we have mastered the technique. In the next at the start of the pulse, so that the pulse rises to a desired year we will continue to refine this technique - and a spe- voltage in a shorter time. That results in making the pulse cific aim will be the ability to accommodate larger sample closer to ideal rectangular shape. Protons in ammonium sizes, and attain single sided detection that is a more prac- nitrate are pre-polarized by an electromagnet placed 1 m tical geometry for medical and screening applications. away from the NQR spectrometer. After demagnetization, the ammonium nitrate sample is quickly moved into the The second goal has been partially achieved by showing solenoid coil in the spectrometer. We have successfully that we can use NMR-like pre-polarization methods to detected a polarization enhanced NQR signal from ammo- cause signal enhancement of the NQR via cross relaxation. nium nitrate with a multipulse sequence. The calculated We have now demonstrated this for ammonium nitrate SNR was roughly 30 at a polarization field of 280 mT and and achieved a > 10X enhancement. We have also dem- a polarization time of 30 s. The enhancement factor was onstrated read-out of the NQR signal through traditional measured to be 13. We also confirmed temperature de- NMR methods via cross-relaxation. A manuscript on these pendence of NQR frequency, polarization time dependence accomplishments is in preparation and the work will be of the NQR signal, and polarization field dependence of the presented as an invited talk at an upcoming conference on NQR signal. In conclusion, we demonstrated the polariza- magnetic resonance based methods for explosives detec- tion enhancement technique in ammonium nitrate. The tion. results above have been presented at the 54th ENC confer- ence in April and the postdoc research day in June this The third goal for the coming year will involve demonstra- year. In the near future, we will publish the results. tion of interleaved NMR and NQR and follows naturally from the second goal. We intend to leverage an existing ul- We were also able to demonstrate acquisition of the NQR tra-low field NMR system based on inductive coil detection signal from ammonium nitrate via proton read-out. We for the NMR portion. We will also investigate methods to saw very interesting behavior in the spectra for T1 vs. shorten the effect of long T1 times on the measurements. magnetic field, as measured by a field-cycling relaxometer. This behavior included the expected “dips” in T1 which Conclusion have been reported in the literature to occur when the We will develop a non invasive technique to characterize fast relaxing 14N interacts with the protons in hydrogen nitrogen content of materials. Nitrogen content in tissue is and shortens T1. However, we also observed peaks in T1 an indicator of the presence of proteins and can be used to vs. magnetic field, indicating regions where the 14N was study diseases in which protein concentrations are affected actually relaxing slower and lengthening the T1 relaxation. such as de mylenation of neuronal tissues associated with To our knowledge, this effect is predicted but has not been multiple sclerosis, ALS, and other neurological conditions. published via proton NMR measurements. We then cross- validated this against compounds where the 14N had been This approach is also useful for the stand off detection of replaced by 15N (no NQR) . energetic materials, contraband, land mines and impro- vised explosive devices. For the remainder of the project we will focus on demon- stration of single-sided detection, demonstration of com- Publications bined NMR and NQR, and demonstration of T1 shortening. Kim, Y., T. Karaulanov, J. Yoder, A. Matlashov, S. Newman, A. Urbaitis, P. Volegov, and M. Espy. Development of Future Work a Pulsed 14N Quadrupole Resonance Technique for Our project goals are to demonstrate 1) a portable NQR Solid Explosives Detection. Presented at Experimental system with sensitivity to small quantities of nitrogen con- Nuclear Physics Conference. (Asilomar, CA, April 14-19, taining materials of interest such as explosives or explosive 2013). surrogate materials; 2) demonstrate polarization enhanced Kim, Y., T. Karaulanov, M. Espy, A. Matlashov, S. Newman, 645

Page 646

A. Urbaitis, P. Volegov, and J. Yoder. Nuclear Quadru- pole Resonance Technique for Solid Explosive Detec- tion. To appear in Journal of Magnetic Resonance. 646

Page 647

Physics Exploratory Research Continuing Project Laser-driven Relativistic Mechanics, Radiation, and Ion Acceleration James A. Cobble 20120259ER Introduction other energetic ions, and 4) what laser to build for Ma- It is well known that relativistic-electron jets stream from RIE for shock physics and materials research utilizing 20 targets illuminated by extremely high-intensity lasers (I – 100-keV x rays. Successful development of high-energy > 10^18 W/cm^2). However, little detailed information backlighters opens doors, for example, the possibility of is available concerning jet properties, and all lasers are phase-contrast imaging with x rays, a possible result of different. The LANL Trident Laser is among the best in this project. the world in beam contrast, energy content, and focus- Progress ability. It is the proper tool to use to fulfill our first goal: to diagnose the relativistic electron jets for various laser In the last 12 months, we validated the ability to attain conditions with a suite of new diagnostics to remove the phase-contrast images of the electron debris ejected present ambiguity related to jet geometry, the current from the rear of targets irradiated by lasers at intensities carried by the jet, and jet stability. Our second goal is exceeding 3x10^20 W/cm^2. We have catalogued time- to model jet behavior through mature particle-in-cell lapse PCI images of the expanding plasma from ~0.5 – 70 numerical simulations that will ultimately be transferable ps after the start of the laser pulse. The imaging probe to an advanced computer platform. Because the unique laser, wavelength 527 nm, has captured wave-front dis- physics environment created by a relativistic plasma jet tortions due to the integrated plasma density, not only produces intense radiation sources – both particles and the relativistic electron jets, sometimes with apparent x rays, one may envision several applications that could filamentation, but also of thermal electrons accompany- benefit human kind. Our third goal is to manipulate ing protons accelerated to tens of MeV by the departure the jet with modifications of the laser drive to establish of the relativistic electrons from the target vicinity. We scaling laws for the applications for laser- accelerated have solved the problem of electron spectroscopy with protons (upwards of 100 MeV) and laser-produced, a magnetic field to reveal the electron energy distribu- multi-tens-of-keV x-ray bursts. This leads to the abil- tion as a function of laser irradiance and the divergence ity to design appropriately-sized larger facilities in the (emittance) of each band of the measured energy spec- future that may implement useful robust sources of both trum. We have made current density measurements, particle and x-ray energy for diverse applications: fusion, which coupled with the divergence measurement, lead cancer treatments, ion acceleration, and flash x-ray to an estimate of integrated electron current. As a radiography. stretch goal for this project, we have fulfilled laboratory safety requirements to deploy a gas-Cherenkov detector Benefit to National Security Missions to monitor gamma-ray production both from prompt la- The inherent relativistic electron jet dominates radia- ser-target interactions and from the impact of accelerat- tion processes for electrons, ions, and x rays in ps-laser ed protons that convert to gammas in a ‘proton catcher’ interactions. Once jet manipulation is understood and downstream one meter. Either of these features could scaling laws are determined, managers have a decision lead to a laser-produced ‘dial-a-gamma-energy’ source point for capability building. We will know: 1) how big for national security programs. The time-resolving GCD and what quality of laser is necessary to pump electron is on loan from the ICF program. Particle-in-cell codes (or ion) fast ignition at the National Ignition Facility, 2) are presently being deployed to model the early-time what is required to accelerate ions to greater than 1-GeV behavior of the relativistic mechanics. As simulations energy for cancer or other uses, whether for medical or become more mature and analysis of the PCI images pro- national security purposes, 3) the feasibility of focusing duces projected electron density maps, we will compare 647

Page 648

code predictions to experimental measurements. So far, jets and the how useful they may be for fusion ignition, for all targets have been synthetic diamond or carbon-based flash radiography of dynamic shock experiments, and for plastics. Finally, we have obtained a set of thin-film metal ion-beam treatment of tumors. targets – 50 – 500-nm thickness – to measure ionic line emission for so-called ‘hollow atoms’ – atoms in the vicin- Publications ity of the electron channel through the target where the Cobble, J. A.. Trident shot 23858 –electron and proton en- spectator atoms have been preferentially stripped of their ergies & Dexter calibration. 2013. P-24:2013-038, LA- K-shell electrons. We plan to monitor spectral content as UR-13-26309. a function of target thickness to observe possible ampli- Cobble, J. A.. Proton converter for GCD on Trident. 2013. fied spontaneous emission – laser-like radiation – from the P-24:2013-022. target plasmas. Cobble, J. A.. SUBJECT: Divergence of E beam in 100-TW Future Work Trident Experiments. 2013. P-24:2013-070, LA_UR- The high-power laser creates a relativistic, multi-MA elec- 13:27148. tron jet through its interaction with a thin-film target. The Cobble, J. A., S. Palaniyappan, D. C. Gautier, Y. H. Kim, D. D. jet will be resolved with temporally gated phase-contrast Clark, R. P. Johnson, T. Shimada, J. C. Fernandez, and imaging using a ps UV laser probe. The jet’s length, diam- H. W. Herrmann. Relativistic transparency experiments eter, duration, and stability will be determined. The beam at the Trident Laser. 2013. In 55th Annual Meeting of current will be measured by charge collection; hence, the the APS Division of Plasma Physics. (Denver, Colorado, current density may be calculated. In like manner, the de- 11-15 November 2013). Vol. 58, No. 16 Edition, p. velopment of the virtual cathode sheath is diagnosed both PO6.00009. Bulletin of the American Physical Society: geometrically and temporally. Proton-beam geometry and American Physical Society. energy and x-ray yield will be measured with radiochromic film and calibrated x-ray detectors respectively. The scaling of radiation with laser focal properties, particularly pulse duration and focal-spot quality, and with target configu- ration will be measured with these detectors. Thus, we determine the laser’s effect on the relativistic jet for pro- cess optimization whether the desired property is electron beam current, ion current or acceleration, or x-ray produc- tion. Matching simulations with real experimental data, we may predict plasma behavior and radiation efficiencies at even higher laser operating conditions. We are presently in the middle of our 4-wk experimental run for FY13, and results are very promising. We have measured current density, electron-beam divergence and energy distribution, and have obtained phase-contrast images of the electron beam as it emerges from the target. Simulation platforms are being developed for use in FY14, during which we shall press forward with investigations of gamma-ray emission and ‘hollow-atom’ spectra — relevant for homeland security, among other things, and novel soft- x-ray laser action. Conclusion Characterization of the laser-driven relativistic electron jets (evidence of their instantaneous current and their geome- try – their divergence and whether or not they filament) is important to the physics of their applications, e.g., electron fast ignition, ion fast ignition, high x-ray flux for radiogra- phy, and general steep-gradient, high-flux ion acceleration. Results will therefore help determine our control of the 648

Page 649

Physics Exploratory Research Continuing Project Monte Carlo for Quantum Transport James E. Gubernatis 20120266ER Introduction focus is the interaction of a few cold atoms with electro- We propose to develop the first significant Monte Carlo static traps and the interaction of an atom with a laser. tool to simulate transport in an extended quantum Accordingly, the new computational tool has multiprobe system. To do so, we need to remove a computational utility. Quantum optics and MaRIE share the same goal bottleneck present in the current approach. We will do of prediction and control. In fact, the integration of this in an insightful fashion and thereby generate a new the proposed research with MaRIE would give MaRIEa computational capability, posed for widespread applica- distinctive look. Additionally, the new computational tion. Simulating non-equilibrium phenomena, such as capability will create new opportunities for high perfor- transport, is difficult whether the system is quantum mance computing. The Monte Carlo method is a natural or classical. For classical systems, Boltzmann equation for exascale computing. and molecular dynamic methods are ubiquitous. These Progress approaches however are not applicable to quantum problems. While a Monte Carlo method, called wave We completed the wavefunction Monte Carlo code function Monte Carlo, works well for quantum systems development started last year, succeeded in understand- with small size, for example, quantum dots and qubits, ing when the steady-state of the simulation should be its computational cost scales exponentially with size. We independent of the initial conditions, and began to argue that this difficultly is not with the core strategy of benchmark the code against published results. We also the method but with the way Monte Carlo is being used resolved the questions we had about how we should in it. We observe the bottleneck is a step analogous to calculate one-dimensional transport. one appearing in other quantum Monte Carlo methods. The code that was under development uses a Monte Thus, by adapting one of these techniques to a new Carlo method instead of a more costly and memory context, we argue that we can eliminate this bottleneck intensive deterministic method to solve the master and do so in a novel manner. For concreteness, we will equation for an open quantum system interacting with target transport in nanoscale materials. Our intent is to its environment. Our ultimate objective is replacing provide theorists with a tool to model these materials current Monte Carlo techniques with ones that execute and to benchmark approximate theories of the behavior even faster. of these models. The new simulation capability will also provide experimentalists with a tool to reduce the trial Of interest is the long-time solution of the master equa- and error in the design of transport devices and to aid in tion, the steady state. In most cases we want the steady the interpretation and design of their experiments. state to be independent of the initial state as being so is more indicative of the general character of the problem. Benefit to National Security Missions For the models we are considering, we established the The objective of this project is the development of a independence of their steady states from the initial con- new computational capability. We emphasize that it has ditions. That they are independent is something simply a broader range of applications than nanotechnology. assumed in the literature. This range umbrellas the same range of activities as the Laboratory’s proposed material flagship, MaRIE, and We want in particular to calculate the heat current will provide useful tools for researchers engaged in this generated in an open quantum system that is in contact project. While the focus of this project is on transport, with heat baths of two different temperatures. Published the seminal ideas came from quantum optics where the 649

Page 650

is a variety of definitions for the current. We decided to We discovered a way to compute the spectral function use one that gave for a one-dimensional current going left of the open system with the wavefunction Monte Carlo to right the sign opposite to the current going right to left. method. The ability to do so enables a direct comparisons This definition fits the physics. Most others did not or else of our simulations with spectroscopic experiments. If we were unclear about what was actually being calculated. have time, we plan to implement this method. In benchmarking the code against published one-dimen- Future Work sional simulations using the wavefunction Monte Carlo We will focus on making the wavefunction Monte Carlo method, we experienced significant disagreements be- method more efficient. Currently, we are using the stan- tween what was published and what we were calculating. dard drift-jump algorithm. A variety of options exists. Be- In particular, predictions of the Monte Carlo code with the cause of the stochastic jumps that occur in the wavefunc- new and the previous definitions of the current did not tion as a function time, we propose to implement a version agree with published results. of what is called the worm algorighm. In it, a random walker (the worm) moves between different space-time To resolve this disagreement, we developed a code to solve points and at each point changes its directions or changes the master equation deterministically. Having this code en- its quantum state. With it we hope to eliminate the need ables us to compare two quite different calculations of the in the present method of compiling statistical information same thing. We focused on comparing the density matrices by multiple restarts of the simulation. With the worm algo- as all measurements derive from it. The deterministic time- rithm, it appears possible to simulate to a fixed time once dependent results for the density matrix should agree with and use the stochastic movement of the worm to sample the Monte Carlo’s ensemble averages as a function of time statistical information from this solution. and the deterministic steady-state results should agree with the long-time average of the Monte Carlo results. With more efficiency, we will then perform an extensive Developing this code took considerable effort because set of one-dimensional transport simulations. By the end we could not use any portion of the wavefunction Monte of the year, we intend to move the simulations and code Carlo code. development to two-dimensions. In higher dimensions, the quantum transport theorems for our master equations We found that only the Monte Carlo ensemble averages suggest establishing independence of the steady state for short times agreed with the deterministic predictions. from the initial state will be easier to achieve. Further, the differences did not, as they should, become smaller as the Monte Carlo ensemble size, the length of Towards the end of the year we will move towards two- integration time, or the accuracy of the differential solvers dimensions in steps, that is, moving from one chain, to were increased. When we turn off the interaction be- two coupled chains, to three coupled-chains, etc., to study tween the heat baths and the open system, which removes the crossover in dimensional behavior. This sequencing the Monte Carlo from the problem, the two approaches will be important because many low-dimensional quan- agree perfectly which implies the problem is not with the tum systems, often called one-dimensional, are actually way we calculate our measurements. If we look at the quasi-one-dimensional, with the length in one dimension currents produced in the open systems by just the conse- greatly exceeding that in the other. We propose to do the quences of the heat bath, which emphasizes the Monte first such study of this physics. More efficient Monte Carlo Carlo portion of the calculation, we see the disagreements. methods are needed to make this type of study possible. The bias, of the order of a few tenths of a percent, appears caused by the off-diagonal elements of the density ma- Conclusion trix becoming small and exhibiting large statistical errors. Without doubt our new Monte Carlo method will be more Common measures of comparing density matrices, such efficient than the present one. The main question is, “By as the trace norm and the entropy, imply a much better how much more?” Unfortunately, the field has few ex- agreement between the two approaches; however, they actly solvable problems that provide benchmarks. It does are insensitive to these matrix elements being small. We though have a few published results for smaller lattices are using a root mean-square measure which is more com- with related models against which we can benchmark the mon to Monte Carlo simulations. We are still working on methods. We will execute these benchmarks and then understanding the problem. It appears that the statistical push the simulation to larger systems, error generated by this particular Monte Carlo method might just be something that is naturally large so reducing Publications it it takes a lot of computer time. Berry, D., R. Cleve, and R. Somma. Exponential improve- ment in precision for Hamiltonain-evolution simula- 650

Page 651

tion. Physical Review Letters. Gubernatis, J. E.. Sign and Phase Problems: Monte Carlo for the unsampleable. Invited presentation at Monte Carlo Methods in the Physical and Biological Sciences. (Providence, RI, 29 Oct - Nov 2, 2012). Gubernatis, J. E.. Wavefunction Mote carlo for transport in open quantum systems. Presented at March Meet- ing 2013, American Physical Society. (Baltimore, 18-22 Mar. 2013). Hu, F. M., T. O. Wehling, J. E. Gubernatis, T. Frauenhein, and R. M. Nieminen. Magnetic impurity affected by spin-orbit cou[ling: behavior near a topologival phase transitions. 2013. Physical Review B. 88: 045106. Ren, J., J. X. Zhu, J. E. Gubernatis, C. Wang, and B. Li. Ther- moelectric transport with electron-phonon coupling and electron-electron interacion in molecular junc- tions. 2012. Physical Review B. 85: 155443. Somma, R. D., D. Nagaj, and M. Kieferova. Quantum Speed- up by Quantum Annealing. 2012. Physical Review Let- ters. 109: 050501. 651

Page 652

Physics Exploratory Research Continuing Project Nonequilibrium Spin Noise in Semiconductors: Physics and Applications Nikolai Sinitsyn 20120340ER Introduction Progress We develop a novel approach for probing material We wrote a manuscript called “Nonequilibrium spin characteristics and new functionalities in commercially noise spectroscopy” (available as LANL internal publica- important semiconductors. The novelty of our approach tion and submitted to Phys.Rev.Lett). It targets the very is in the observation that a recently developed spin basic effects suggested in the early proposal of this LDRD noise spectroscopy technique can probe essentially new ER. In this work we discuss advantages of measuring spin physical phenomena in the non-equilibrium regime. To noise in nonequilibrium regime and estimate several study this regime we will develop the theory that will effects to be experimentally verified. We also discuss interpret our observations of out-of-equilibrium electron applications for precise measurements of spin orbit cou- spin noise in semiconductors. Our preliminary theoreti- pling anisotropy and exploring physics of spin noise. cal studies predict new effects that we will verify. Each We have submitted an extended article with a proposal effect reveals physics that is either difficult or impossible to apply spin noise spectroscopy for characterization of to study with other competing techniques. If demon- nanostructures. We already received positive reviews strated, these effects will open up the path for further of Phys. Rev. B referees. The work shows that spin noise applications, including high precision measurements of spectroscopy method, developed at LANL, has particular a spin-orbit coupling, obtaining properties of an electron advantages for nanostructure characterization. liquid near the metal-insulator transition, characteriz- ing electron transport at extreme conditions of electric We published an article in Applied Physics Letters with breakdown, discovering new types of quantum phase proposal of new measurement strategy, which we call transitions, measuring entanglement entropy and many “Two beam spin noise spectroscopy”. We showed that other effects of importance for the future of several by measuring spin noise by two spatially separated fields. beams it is possible to obtain new information about spin dynamics in semiconducting materials and nanow- Benefit to National Security Missions ires. This project will build Laboratory capabilities in the de- velopment of novel measurement methods that enable On experimental side, first, one article called “Opti- new scientific discovery at LANL. Our project is enabled cal Spectroscopy of Spin Noise” was published in Phys. by recent technological advancements that use the very Rev. Lett. 10, 176601 (2013) by the experimental team large data storage and manipulation capabilities. The in which the fact that measuring beam is selectively project will advance new Laboratory capabilities that coupled to different quantum dots is clearly exposed. underpin a wide variety of Laboratory missions includ- ing Nanotechnology and Quantum Information Science. Several other experiments were performed, which Demonstration of new effects by the nonequilibrium require additional theoretical studies and experiments. spin noise spectroscopy measurements will generate We measured spin noise at high intensities of a mea- worldwide attention and will form a robust platform for surement beam and observed new phenomena, such as a nanoscale spintronics program at LANL. The present anomalous peaks in the noise power, which are currently project will lay the groundwork for future proposals for not explained but the team has developed working external funding to address national problems, includ- hypothesis for their understanding. We also observed ing new functional materials capabilities for MaRIE and features of noise power spectrum that suggest the ap- energy efficient applications (DOE, DARPA). pearance of purely quantum effects such as the quan- 652

Page 653

tum Zeno effect in GaAs. Our current goal is to produce Current Components in Mesoscopic Electric Circuits. better arguments for a decisive conclusions about these 2011. Physical Review B. 84: 245405. observations. Li, Y., N. A. Sinitsyn, D. Smith, D. Reuter, A. Wieck, D. Yakov- lev, M. Bayer, and S. Crooker. Intrinsic spin fluctuations Future Work reveal the dynamical response function of holes cou- We will explore the effects of spin-orbit coupling, and pled to nuclear spin baths in (In,Ga)As quantum dots. interactions on spin correlators in GaAs semiconductors, 2012. Physical Review Letters. 108: 186603 . and obtain specific properties of materials and informa- tion that can be extracted from such measurements. Spin Pershin, Y. V., V. A. Slipko, D. Roy, and N. A. Sinitsyn. Two- noise in electric fields will be studied theoretically by beam spin noise spectroscopy. 2013. Applied Physics Letters . 102: 202405. Green function techniques, which can be considered as a high order Kubo formula and by methods of the quantum Roy, D., Y. Li, A. Greilich, Y. V. Pershin, A. Saxena, and N. A. counting statistics. Some of the previous insight about Sinitsyn. Spin noise spectroscopy of quantum dot mol- nonequilibrum noise in electric cur-rents will be translated ecules. 2013. Physical Review B. 88: 045320. into a theory of spin noise. A complementary theoretical approach, which already allowed us to gain an important Sinitsyn, N. A., Y. Li, S. A. Crooker, A. Saxena, and D. L. insight, is based on the kinetic equation for hydrodynamic Smith. Role of Nuclear Quadrupole Coupling on Deco- herence and Relaxation of Central Spins in Quantum evolution of charge and spin density where interactions Dots. 2012. Physical Review Letters. 109: 166605 . are included in the collision term. In 2014, we will con- centrate on developing theory of high order correlators in Sinitsyn, N. A., Yuriy V. Pershin, Valeriy A. Slipko, and N. spin noise spectroscopy and effects of many-body interac- A. Sinitsyn. Nonequilibrium spin noise spectroscopy. tions. The latter will be based on theory of Luttinger liquids 2013. Physical Review Letters. 111: 067201. supplemented by our methods that we already developed for characterizing spin noise in quasi-1D geometry. Experi- Sykes, A., D. Solenov, and D. Mozyrsky. Bloch-Redfield theory of high-temperature magnetic fluctuations in ments will be focused on verification of our recent predic- interacting spin systems. 2012. Physical Review B. 85: tion of the peak shift effect in electric field and on studies 174419. of spin noise at high intensity, where we observe unexpect- ed phenomena that we are just starting to explore. Zapasskii, V. S., A. Greilich, S. A. Crooker, Y. Li, G. G. Kozlov, D. R. Yakovlev, D. Reuter, A. D. Wieck, and M. Bayer. Conclusion Optical Spectroscopy of Spin Noise. 2013. Physical Re- We will develop the technique that we call the nonequilib- view Letters. 110: 176601. rium noise spectroscopy. It will allow us to look at material properties that simply could not be probed directly by any other approach and it will enable design of new materials for efficient energy transport and electronic devices. Spe- cifically, we will demonstrate high precision measurements of spin orbit anisotropy and the first studies of the metal- insulator transition in GaAs semiconductors. Higher order spin correlations will reveal the fundamental statistical physics of nonequilibrium spin dynamics and will consti- tute a significant a-vance in the field of quantum measure- ment science. Publications Chaudhury, S., J. Cao, and N. A. Sinitsyn. Universality of Poisson Indicator and Fano Factor of Transport Event Statistics in Ion Channels and Enzyme Kinetics. 2013. Journal of Physical Chemistry B . 117: 503. Chernyak, V. Y., J. Klein, and N. A. Sinitsyn. Algebraic topol- ogy and the quantization of fluctuating currents. 2013. Advances in Mathematics. 244: 791. Ganeshan, S., and N. A. Sinitsyn. Fluctuation Relations for 653

Page 654

Physics Exploratory Research Continuing Project Flipping the (Light) Switch on Nano-magnetism: Emergent Photomagnetization in Quantum-confined Semiconductors Scott A. Crooker 20120374ER Introduction Preliminary growth and measurements of these new Manipulation of magnetic metals forms the cornerstone materials (described below) indicate that all aspects of of today’s vast information-storage industry, while in this project are achievable within 3 years. parallel, semiconductors provide the basis for current Benefit to National Security Missions microprocessor technologies. It is therefore not surpris- ing that the marriage of magnetism and semiconductors This project will build Laboratory capabilities in the de- – as typified in today’s magnetic semiconductors — is a velopment of novel materials-synthesis and experimen- thriving research area (briefly, these materials achieve tal techniques that enable new scientific discovery at new functionality by coupling, via spin-exchange inter- LANL. It is based on recent advances in colloidal growth actions, the semiconductor’s electrons and holes with and tunable magnetic doping of nanocrystals that have embedded magnetic dopants). As fundamental research already demonstrated new emergent functionality. Our in both magnetism and semiconductors continues its approach goes well beyond this specific example photo- focus on miniaturization, magnetic semiconductors are induced magnetization in doped nanocrystals: long-term also now pushing the limits of quantum confinement. advantages and applicability of nano-scale heterostruc- Recent work at LANL and elsewhere have demonstrated tures and time-resolved magnetization dynamics are the synthesis of truly “0-dimensional” magnetic semi- very general. conductors in the limit of extreme quantum confine- Progress ment: these new materials are semiconductor nanocrys- tals (of, e.g., ZnSe or CdSe) containing small numbers In the last year we have published a high-profile paper in (<5) of magnetic atoms (Mn, Co, Fe). Remarkably, these Nature Nanotechnology that describes the main find- novel nanoscale materials exhibit unexpected and new ings of the first stage of this ER — namely, establishing emergent properties that simply do not (or cannot) the existence of optically-induced magnetism in these exist in their more conventional 3-D or 2-D forms. New copper-doped ZnSe nanocrystals. properties include the ability to tune magnetic exchange We have also been focusing on new copper indium interactions via quantum confinement and wavefunction sulphide nanocrystals (CIS), a new materials systems that engineering at the nanometer scale, magnetic polaron has been attracting much attention for solar energy light formation at high temperatures, and – especially — a harvesting. Our recent preliminary data suggests that recently-discovered photo-induced magnetism in a new the light emission mechanism in these new CIS materials magnetic semiconductor system based on copper-doped is quite similar to that from copper doped ZnSe (despite nanocrystals. the completely different stoichiometry), and likely has to Our scientific goal is to investigate these new families do with copper defects. We have measured magneto- of nanometer-scale magnetic semiconductors, with an PL as well as magnetic circular dichroism on these CIS emphasis on wavefunction engineering and controlled materials, and they definitely show a slight paramagnetic placement of magnetic dopants within core-shell nano- behavior, indicative of paramagnetic copper atoms. This crystals to achieve new functionality. The work will focus work is currently being written up. on emergent optically-controlled magnetization, tunable Concurrently, our ultrafast OPO laser system is up and spin-exchange interactions, and ultrafast spin dynamics. running; this will very soon be sapplied to copper-doped Colloidal synthesis and time-resolved magneto-optical nanocrystals in order to study their magnetization dynamics. studies form the essential components of this proposal. 654

Page 655

Future Work We will be focusing in the next year on both copper-doped ZnSe nanocrystals as well as new copper indium sulphide nanocrystals (CIS), a new materials systems that has been attracting much attention for solar energy applications. Our recent preliminary data suggests that the light emission mechanism in these new CIS materials is quite similar ot hat from copper doped ZnSe (despite the completely dif- ferent stoichiometry). In the next year we anticipate that time resolved magneto-optical measurements of these CIS materials will also be performed. The next main steps are to i) study the dynamic magnetization of these particles by using ultrafast optical techniques such as time-resolved Faraday rotation, and ii) to measure the optical properties of individual nanoparticles by using a single-quantum-dot microscope. Equipment infrastructure for both of these measurements is already in place. Conclusion Using our spectroscopic methods, we expect to have direct access to the as-yet-unknown dynamics of photo-induced magnetization in engineered nanocrystals. We expect to observe ultrafast spin precession frequencies of electrons and holes (enhanced by sp-d exchange), and picosecond/ nanosecond timescales for collective magnetization align- ment (and much longer timescales for eventual spin-lattice relaxation). Any demonstration of engineered sp-d cou- pling and its associated dynamics will represent totally new data that will be found of significant interest to research- ers working in all areas of nanoscale magnetism, magnetic semiconductors, and spin interactions. Publications Pandey, A., S. Brovelli, R. Viswanatha, L. Li, J. M. Pietryga, V. I. Klimov, and S. A. Crooker. Long-lived photo-induced magnetization in copper-doped ZnSe/CdSe core/shell nanocrystals . 2012. Nature Nanotechnology. 7: 792. 655

Page 656

Physics Exploratory Research Continuing Project A Novel Exploration of Nature’s Largest Explosions W T. Vestrand 20120375ER Introduction Awareness and can address the declared needs of some We propose to deploy an integrated suite of autono- of our programmatic customers. The principal investiga- mous robotic optical instruments, with capability un- tor is a member of the Space Systems Program Office at matched anywhere in the world, to measure the proper- LANL and with be working directly with external cus- ties of prompt optical flashes generated by Gamma-Ray tomers to transition this work to new Work For Others Burst (GRB) explosions. The novelty of our approach (WFO) programs. lies in the use of simultaneous multi-color measure- Progress ments, polarimetry, and high cadence photometry of the prompt optical flash—each for the first time. Used The optical light that is generated simultaneously with with gamma-ray measurements from NASA’s Swift and the x-rays and gamma-rays during a gamma-ray burst Fermi missions, these multi-wavelength observations (GRB) provides important clues about the extreme will probe the physics of gamma-ray bursts to an unprec- nature of the explosions that occur as massive stars col- edented level of detail, both temporal and spectral. To lapse to form black holes. This year we detected a bright interpret these measurements, and explore their impli- optical flash (7th magnitude) and fading afterglow from cations, our observational effort will be complemented a powerful burst, GRB 130427A, that set new records by a theory effort, culminating in a generic numerical for gamma-ray brightness and duration. Bright optical model for the synchrotron and inverse-Compton emis- flashes are rare, and this flash, the second brightest ever sions from relativistic outflows, with the end goal of un- detected, is the first to allow detailed comparison with derstanding the extreme physics present during nature’s the properties of the gamma-ray emission at very high largest explosions. Some of the key science questions energies. We found a striking similarities between the about GRBs that we will answer are: What are the physi- optical light curve and >100 MeV photon flux light curve cal conditions in GRB jets and the radiation mechanisms during the first 7,000 seconds. We attribute this link to that produce the prompt optical and gamma-ray emis- co-generation in an external shock. The simultaneous, sion?; Is the prompt emission highly polarized, as some multi-color, optical observations are best explained at models predict?; What is the precise temporal relation early times by reverse shock emission generated in the between the prompt optical and gamma-ray emission relativistic burst ejecta as it collides with surrounding components? The simultaneous observations made by material and at late times by a forward shock traversing our ground-based optical instruments and NASA satel- the circumburst environment. lites will provide previously impossible measurements of We examined the peaks of 30 optical GRB afterglows and the relativistic jet and central engine. 14 X-ray light curves that display a good anticorrelation of the peak flux with the peak epoch. We investigated Benefit to National Security Missions the ability of two forward-shock models for afterglow The ability we develop to identify, follow-up, and inter- light-curve peaks - an observer location outside the rogate rapid astrophysical explosions will have significant initial jet aperture and the onset of the forward-shock impact for NASA’s goal of Understanding the Origin and deceleration - to account for those peak correlations. Evolution of the Universe and will provide important For both models, the slope of the Fp-tp relation depends context information for NASA’s Swift and Fermi Missions. only on the slope of the afterglow spectrum. We found The persistent monitoring and real-time robotic tech- that only a conical jet seen off-aperture and interacting niques we refine on this exploratory scientific investiga- with a wind-like medium can account both for the X-ray tion will have important applications in Space Situational 656

Page 657

peak relation, given the average X-ray spectral slope βX = the early Universe, when a class of very massive stars col- 1.0, and for the larger slope of the optical peak relation. lapsed to form the first black holes. We will use the spec- However, any conclusion about the origin of the peak flux- tral, temporal, and polarization properties of these black peak epoch correlation is, at best, tentative, because the hole birth announcements to decipher the extreme explo- current sample of X-ray peaks is too small to allow a reli- sion physics and collect new clues about the origin of our able measurement of the Fp-tp relation slope and because Universe. Building this technical capability will also enable more than one mechanism and/or one afterglow param- new approaches to National Security problems related to eter may be driving that correlation. Space Situational Awareness. Future Work Publications Our GRB observational program will continue to focus on Ackermann, M., R. Kiziah, P. Zimmer, J. D. Beason, E. Spillar, the observations that our RAPTOR telescopes our uniquely W. T. Vestrand, and M. Weeks. Alternatives for Ground- capable of providing. The RAPTOR-Z high cadence imager Based, Large-Aperture Optical Space Surveillance Systems. and multi-color follow-up telescopes—RAPTOR-T and To appear in 2013 AMOS Technical Conference. (Maui, RAPTOR-S—will be used to make observations of Swift 10-12 Sept. 2013). localized GRBs. We plan to expand our GRB observational Bode, M., and W. T. Vestrand. Small and robotic telescopes campaign on two fronts: (1) Deploy real-time localization in the era of massive time-domain surveys. 2012. In New software on our hybrid search/follow-up array (RAPTOR- Horizons in Time-Domain Astronomy, International Astro- W) to search for counterparts to Fermi Gamma-Ray Burst nomical Union Symposium 285. (Oxford, England, 19-23 Monitor (GBM) triggers; and (2) search for GRB prompt/ September 2011). Vol. 285, p. 235. Cambridge, England: precursor emission with our full sky persistent monitors in Cambridge University Press. Hawaii and New Mexico. Jin, Z., S. Covino, and M. Della Valle et al. GRB 081007 and In the coming year our theory effort will concentrate on GRB 090424: The surrounding medium, outflows, and su- candidate GRB models and their predicted polarization pernovae. 2013. Astrophysical Journal. 774 (2): 114. signatures. We goal is to conduct a more comprehensive study of the range of possibilities than has been conducted Panaitescu, A., W. T. Vestrand, and P. R. Wozniak. Peaks to date and prepare for interpretation of the GRB polariza- of optical and X-ray afterglow light curves. 2013. Monthly tion we expect to conduct during the second half of our Notices Royal Astronomical Society. 433: 759. program. Panaitescu, A., W. T. Vestrand, and P. Wozniak. The effect New instrument development will be focused on the con- of photon-photon absorption on Fermi-LAT light-curves of struction of the polarimeter. The polarimeter will employ GRB afterglows. Astrophysical Journal. a rapidly spinning Polaroid that is synchronized with a fast frame rate (~9 Hz) electron multiplication CCD camera. Panaitescu, A., W. T. Vestrand, and P. Wozniak. An external- Our schedule aims for final integration of the hardware shock model for GRB afterglow 130427A. Monthly Notices in September 2013 and first light for the polarimeter in Royal Astronomical Society. November 2013. Panaitescu, A., and W. T. Vestrand. The possible ubiquity We will also make routine, densely sampled (twice per of energy injection in gamma-ray burst afterglows. 2012. night), multi-color optical observations of the bright gam- Monthly Notices of the Royal Astronomical Society. 425 ma-ray emitting blazars. And, as the polarimeter is brought (3): 1669. on line, we will begin collecting nightly polarimetry mea- surements of 12 optical bright (mv<15) GeV gamma-ray Vestrand, W. T., J. Wren, A. Panaitescu, P. Wozniak, H. emitting blazars (3C 273, Mrk 421, Mrk 501, CTA 102, etc.). Davis, D. Palmer, G. Vianello, N. Omodei, S. Xiong, M. S. A related goal is the construction of a photometry data- Briggs, M. Elphick, W. Paciesas, and W. Rosing. The bright base that allows access to the large sample of measure- optical flash and afterglow from the gamma-ray burst ments that we are collecting for bright blazars. GRB130427A. To appear in Science. Conclusion Vreeswijk, P., C. Ledoux, A. Raassen, A. Smette, P. Wozniak, W. T. Vestrand, and P. Jakobsson. Time-dependent excita- We will provide a better understanding of the most ex- tion and ionization modelling of absorption-line variability treme explosions known in the Universe. These so-called due to GRB 080310. 2013. Astronomy and Astrophysics. Gamma-Ray Bursts are bright flashes of gamma-ray 549: 22. emission---arriving at Earth today—that were generated in 657

Page 658

Physics Exploratory Research Continuing Project Transport by Thermodynamic Cross-terms in ICF Capsule Plasmas Xianzhu Tang 20120395ER Introduction Benefit to National Security Missions Recent proton radiography measurements reveal that In an Inertial Confinement Fusion capsule under spheri- electric fields of ~10^9V/m are self-generated during cal implosion, the fusion yield is critically dependent on inertial confinement fusion (ICF) implosions. The impact the relative concentration of the D and T fuel ions, which of such enormous fields on the capsule performance have different mass and charge ratio. The projects aims cannot be fully accounted for by radiation-hydrodynamic to develop a fundamental understanding and modeling codes that are mainly used to simulate ICF implosions. In tool for the shock-enhanced relative ion transport. This particular, it has been claimed that the factor of two dis- will have direct impact the NNSA’s ICF program and SC’s crepancy between these codes’ predictions and experi- Inertial Fusion Energy program. The capability developed mentally observed neutron yield is due to the electric will be transferred to both the ICF and IFE program in field enhanced barodiffusion at the front of the rebound- three years. We will work with LANL’s program managers ing shock wave, an effect that cannot be straightfor- in ICF, ASC, and SC/OFES to ensure this transition. wardly captured within the hydrodynamic approach.This observation raises several important research questions. Progress First, the above-mentioned qualitative analysis discusses In a comment to PHYSICAL REVIEW LETTERS Vol- the effect of barodiffusion in the absence of thermodif- ume: 109 Issue: 26 Article Number: 269501 DOI: fusion, as well as viscous effects. The two latter phenom- 10.1103/PhysRevLett.109.269501 published DEC 27 ena may affect the capsule performance as much as the 2012, Kagan and Tang identified a discrepancy between former one and therefore need to be properly evalu- Peter Amendt’s formulation of inter-species ion diffu- ated. Second, baro- and thermo-diffusion are among a sion and ours, which was published in Physics of Plas- subset of plasma transport driven by thermodynamic mas Vol. 19, Issue: 8, Article Number: 082709, DOI: cross-terms. A comprehensive physics understanding 10.1063/1.4745869 in AUG 2012, and resolved it by cor- of all thermodynamic cross-terms in ICF plasmas where rectly applying the thermodynamic approach of Landau strong electric field exists must be developed. Third,a and Lifshitz. In a reply, Amendt et al. consented to our new computational approach, capable of capturing the correction. relevant physics with reasonable computational effort, needs to be developed. To address these issues this In a preprint to be submitted to Physics of Plasmas, we project will develop a consistent analytical theory of ion gave an exposition of how to correctly apply the thermo- diffusion in the front of a shock wave propagating over a dynamic approach of Landau and Lifshitz to the inter- binary mix of plasmas. Insights into the effect of baro- as species ion diffusion problem. well as thermo-diffusion will be obtained. The next-step In a preprint to be submitted to Physical Review Letters, goal is to develop a comprehensive transport theory for Kagan and Tang performed first-principles calculation of ICF shock front plasma taking into account all thermo- thermo-diffusion in a multi-species plasma, and found dynamic cross terms. The final goal of the proposal is to that contrary to the neutral gas case, thermo-diffusion verify analytical theory with two-fluid plasma simula- can become dominant in a plasma due to the long range tions solving the Braginskii equations. In such a code, nature of Coulomb interaction. This contrasts with previ- electron and ion fluids are treated separately, thereby ous findings from LLNL by Amendt and his collaborators allowing evaluating the electric field and the associated that thermo-diffusion can vanish. We also identified the diffusive fluxes. cause of their error. 658

Page 659

In a set of three preprints to be submitted to Physical Kagan, G., and X. Tang. Electro-diffusion in a plasma with Review Letters and Physics of Plasmas, we laid down the two ion species. 2012. Physics of Plasmas. 19 (8): theoretical formulation of reduced Fokker-Planck models 082709. for high energy ion distribution in an ICF hot spot, com- Kagan, G., and X. Z. Tang. Comment on “Plasma Adia- puted the tail ion distribution, and discovered the so-called batic Lapse Rate”. 2012. Physical Review Letters. 109: inverse Knudsen layer effect on fusion reactivity. This work 269501. lead to an innovative experimental design and interpreta- tion framework to separate the effects of ion diffusion and Kagan, G., and X. Z. Tang. Thermo-diffusion in inertially Kundsen-layer reactivity reduction on total fusion yield in confined plasmas. Physical Review Letters. ICF expriments, which provided the scientific foundation McDevitt, C., X. Z. Tang, Z. Guo, and H. L. Berk. A compara- for a DR full proposal on “Plasma Physics Effects on Fusion tive study of tail ion distribution with reduced Fokker- Yield in Inertially confined systems.” Planck models. Physics of Plasmas. A student, Francesco Boffa, has further developed a 2D Srinivasan, B., and X. Z. Tang. The mitigating effect of mag- implosion module within LAPS which uses spectral volume netic fields on Rayleigh-Taylor unstable inertial con- discretization. This sets up the implementation of the ion finement fusion plasmas. 2013. PHYSICS OF PLASMAS. diffusion calculation. 20 (5): -. Future Work Tang, X. Z., C. McDevitt, Z. Guo, and H. L. Berk. Fusion yield recovery by escaping hot-spot fast ions in the neigh- The following tasks/goals are planned for the next fiscal boring fuel layer. Physical Review Letters. year: Tang, X. Z., H. L. Berk, Z. Guo, and H. L. Berk. Reduced Fok- • Complete the theory of inter-species ion diffusion flux ker-Planck models for fast particle distribution across a in terms of all thermodynamic forces transition layer of disparate plasma temperature. Phys- ics of Plasmas. • Carry out simulation and theoretical studies of the weak shock structure Tang, X. Z., Z. Guo, G. Kagan, C. McDevitt, and B. Srinivasan. Plasma effects on thermonuclear yield rate in the pres- • Implementing inter-species ion diffusion flux in a 1D ence of hydrodynamic mix. Journal of Physics: Confer- implosion code ence Series. Conclusion The scientific effort aims to produce three primary out- come. The first is a comprehensive plasma transport theory taking into account all thermodynamic cross terms in spherical implosion with multiple ion species. The sec- ond is a spherically symmetric two-fluid plasma simulation code that captures the multiple ion species dynamics and transport in an imploded target. The third is the applica- tion of simulation code to validate the theoretical model and explore ways to improve the neutral fluid hydrody- namic model for capturing transport physics such as baro- and thermo-diffusion. Publications Hsu, S. C., T. J. Awe, S. Brockington, A. Case, J. T. Cassibry, G. Kagan, S. J. Messer, M. Stanic, X. Tang, D. R. Welch, and F. D. Witherspoon. Spherically imploding plasma liners as a standoff driver for magnetoinertial fusion. 2012. IEEE Transactions on Plasma Science. 40 (5): 1287. Kagan, G., X. Tang, S. Hsu, and T. Awe. Bounce-free spheri- cal hydrodynamic implosion. 2011. Physics of Plasmas. 18 (12): 120702. 659

Page 660

Physics Exploratory Research Continuing Project Co-Evolution of Protoplanets and Transitional Protoplanetary Disks: Pathway to Giant Exoplanet Formation Hui Li 20120399ER Introduction Astronomy and Astrophysics”(“New Worlds” refers spe- The past 15 years have witnessed the astonishing discov- cifically to exoplanet research). The critical physics that eries of 40 exoplanets and they have stimulated an ex- underpins this field include high Reynolds number turbu- plosive growth in studying the basic processes that regu- lence and shear flow instabilities,all of which match with late the planet formation (e.g., Papaloizou & Terquem LANL’s strength, both in theory and large-scale comput- 2006; Lubow & Ida 2010). Giant planets (e.g., Jupiter ing. With the advent of ALMA and EVLA, LANL can make mass) are believed to form farther out in protoplanetary a critical impact now, building on our existing expertise disks where there is enough gas and dust for theirfor- in disk+planet computational studies. mation. An equally dramatic discovery is that some Progress protoplanetary disks show large (tens of astronomical unit) gaps and holes that are believed to be caused by We have developed and implemented a new bi-fluid the forming giant planets (e.g., Williams & Cieza 2011). model for dusty protoplanetary disks for both our 2D The study of the co-evolution of protoplanets and their and 3D codes. The dust particles are treated as a pres- nascent disks is believed to hold the key to understand- sureless fluid and evolved according to the hydrody- ing the planet formation. The primary objectives of this namic equations. The coupling between the gas and dust project are two-fold. First, how do the forming giant in a disk is through the drag force, which causes the dust planets create gaps and holesin transitional protoplan- particles to drift inward in the radial direction. We have etary disks that are consistent with observations? And performed extensive tests of our model and code using second,what are the key physical processes that govern the theoretical results in the literature and our numeri- the gas accretion through the circumplanetary disk cal results are in good agreement with the theoretical around the giant planet that leads to planet formation? prediction. The proposed work is computationally intensive and it In addition, we have implemented dust diffusion due builds upon the unique strength at LANL where we have to the turbulent motion of the gas in a disk. We have developed advanced disk+planet simulation tools. With modified the dynamic equations of the dust so that they the adaptive mesh refinement capability especially, we are consistent with the dust diffusion. This procedure will be able to simultaneously capture the global pro- removes the possible instability caused by the dust diffu- toplanetary disk dynamics together with resolving the sion. We have performed simulations to study the dust accretion through a much smaller circumplanetary disk settling in the vertical direction of 3D disk due to the around the planet. Utilizing LANL’s advanced capabilities dust diffusion and obtained similar results to the theo- in theory and computing, we expect that the proposed retical prediction. research will enable us to make a critical impact now, especially when a new era is anticipated with the advent We have performed many simulations to study the of ALMA and EVLA observations with unprecedented asymmetric dust distribution observed by ALMA and sensitivity and spatial resolution. other telescopes. In particular, we have studied how the banana-shaped dust distribution can be formed via Benefit to National Security Missions the interaction among dust, gas, and the planet. We Understanding the formation of exoplanets is one of the have found the Rossby wave instability (RWI) is a robust grandest challenges for the coming decade in astronomy mechanism to form banana-shaped gas blobs in a disk. and astrophysics, as evidenced by the recent Decadal Furthermore, we have investigated two mechanisms Survey (Astro2010): “New Worlds, New Horizons in 660

Page 661

to generate the RWI. Both of them require relatively low solar system. New observations by the new generation viscosity. One mechanism is using large planets to induce of astronomical observatories will enable us to constrain the RWI around the planet. We use Saturn-, Jupiter-, and current models or lead us to new scenarios of planet-disk two-Jupiter-mass planet in our simulations. In 2D simula- interaction with unprecedented sensitivity. Applying the tions, we clearly see the banana-shaped distribution of state-of-the-art theory and simulation techniques devel- the gas and dust. The banana-shaped density blob can oped at LANL, we will be able to make a timely impact in be maintained for hundreds of orbits. In 3D simulations, this important, thriving field. however, we find that the evolution of RWI and banana- shaped density blobs are short-lived and quickly become Publications ring-shaped distribution. Aluie, H., Shengtai Li, and Hui Li. Conservative Cascade of Kinetic Energy in Compressible Turbulence. 2012. As- We have performed simulations to study the RWI in a disk trophysical Journal, Letters. 751 (2): L29 (6 pp.). without planets. The RWI can be produced with an initial Collins, D., A. Kritsuk, P. Padoan, H. Li, H. Xu, S. Ustyugov, density bump. We study how the vortices are generated and M. Norman. THE TWO STATES OF STAR-FORMING via RWI and how they are merged and combined into one CLOUDS. 2012. ASTROPHYSICAL JOURNAL. 750 (1): 13. big vortex. We also study the RWI dependence on the disk self-gravity. The RWI caused by the initial density bump is Johnson, J. L., and H. Li. Constraints on planet formation short-lived and quickly disappear due to migration or diffu- via gravitational instability across cosmic time. 2013. sion. To have a long-lived RWI vortex, we set up a disk with Monthly Notices of Royal Astronomical Society. 431: non-uniform disk viscosity, where a region of disk has very 972. low viscosity, similar to the dead-zone configuration. Even Johnson, J., and H. Li. THE FIRST PLANETS: THE CRITICAL without initial density bump, the non-uniform viscosity can METALLICITY FOR PLANET FORMATION. 2012. ASTRO- produce a density bump near the low viscosity region in a PHYSICAL JOURNAL. 751 (2): 81. few hundreds of orbits. With small perturbation, the RWI can be produced and vortices can be generated. We can Li, H., and S. Li. 3D Simulations of Type-I Migration in Near- see the banana-shaped density distribution generated and ly Laminar Disks. 2013. European Physical Journal. 46: maintained for a long time. 05003. These studies are helping us to make direct comparisons between our simulations and the new observations. This allows us to constrain the disk properties and provides more detailed understanding on how the planet + disk interactions evolve in a dusty protoplanetary disks. Future Work We will continue the development of our new 3-D hydro- dynamic code, now including the both the gas and dust dynamics. Modeling the dust evolution is a new feature of this code and it will allow us to make more direct compari- sons with observations. We will collaborate with other teams to model the radia- tion intensity and spectra from dusty protoplanetary disks, esp. those from ALMA and other infrared telescopes. We will investigate how the non-axisymmetric features can be produced by various disk properties and/or from disk+planet interactions. Conclusion Understanding the origin of exoplanets is an exciting new frontier in astronomy and astrophysics. It will have pro- found impact on understanding the formation process of 661

Page 662

Physics Exploratory Research Continuing Project A Computationally Efficient Model for Warm Dense Mixtures Didier Saumon 20130244ER Introduction hundred to a thousand times faster, comparably accu- This research will create the theoretical and computa- rate, and tractable for mixtures. tional framework to model mixtures of two (or more) Benefit to National Security Missions materials in the “warm dense matter” regime. This will be a new, state-of-the-art capability to model extreme NASA: The primary applications thrust of our research is states of matter that will support future experiments at astrophysical – with an initial focus on examining hydro- high-energy-density-physics facilities (e.g., the National gen and helium phase behavior in Jupiter and Saturn, Ignition Facility) and numerous fundamental and pro- and understanding heavy-element diffusion in DZ white grammatic applications involving the hydrodynamics of dwarf stars. In a broader sense, this research will create dense plasma mixtures. a robust capability for scientists to develop accurate equations of state and transport coefficients for wide Warm dense matter (WDM) conditions typically occur at varieties of astrophysical phenomena. temperatures of 5000 to 10,000,000 degrees and densi- ties from a tenth to fifty times normal solid densities. Understanding of Materials: Warm dense matter is a WDM is too hot to be treated by the standard methods frontier of materials physics. Because it describes a re- of condensed-matter and liquid-state physics, but too gion of temperature and density space devoid of simpli- cool to be modeled as a weakly coupled plasma; it is fying approximations, all of the physics matters. It is also too dense to be treated as an ionized gas, but not dense a region difficult to probe experimentally. The capability enough to approach simplifying asymptotic limits. It lies to model mixtures in this regime will advance our basic in an intermediate region where all the physics mat- knowledge of materials behavior in extremes. ters and simplifying approximations are elusive. It is a MaRIE: One of the objectives of MaRIE is to probe exper- frontier of high energy density physics of great practical imentally the same regions of temperature and density importance. that we are describing theoretically, so the two programs WDM mixtures lie at the heart of unsolved problems are wonderfully complementary. Our theoretical model in planetary science and stellar astrophysics. Examples could be used to design and guide MaRIE experiments, include the phase behavior of hydrogen and helium in and the experimental results could be used to validate Jupiter and Saturn and gravitational stratification of ele- and improve our model. ments in white dwarf stars. WDM mixtures also occur NNSA/DP & Nuclear Weapons: During a nuclear explo- during the implosion phase of inertial confinement fu- sion, and in inertial confinement fusion, mixtures of sion capsules and in nuclear explosions. warm dense matter can be created. This research po- Our proposed work lies at the forefront of this exciting tentially enhances our capability of modeling materials and rapidly growing field. While large-scale simulations, as they pass though this regime. such as quantum molecular dynamics, are currently the NNSA/Non-proliferation: The ability to make accurate favored approach to study WDM, they are computation- equations of state for a wide variety of warm, dense, ally very expensive – requiring hundreds or thousands mixed materials lends itself to better anticipating and of computing hours for a single result. By contrast, our ameliorating the effects of rogue WMDs. model, based on a modern variant of the “quantum hypernetted chain” theory, holds the promise of being a 662

Page 663

Progress output of results for our first application. In the eight months since the start of this project, we ac- We are on track to achieve our science goals for FY13 and complished the following: have made headway in Task one is complete. We formulated the multiple-com- solving problems that we had originally planned to address ponent mixture theory by expanding that of our single- in Years 2 and 3. component microscopic model. We arrived at a simpler form than we originally anticipated. This set of equations is Future Work not only valid for binary mixtures but is readily extended to Building on our accomplishments during Year 1, we will ad- mixtures of an arbitrary number of (classical) ionic species dress the following tasks in year 2: and (quantum mechanical) electrons.

1. Complete the calculation (started in year 1) of a table This mathematical formalism has been implemented in of diffusion coefficients of relevance to white dwarf a computer code that solves the system of non-linear, stars with helium-rich outer layers with traces of coupled equations for mixtures. Substantial progress has heavier elements (the so-called DZ stars). We will first been achieved as our success in Task #1 led to writing a look at the diffusion of silicon in helium. If such a table computer code for a mixture of any number of elements can be generated in a reasonable amount of time, we (not just two). This code has excellent convergence prop- will also study the diffusion of carbon in helium. The erties and we have demonstrated its ability to solve the diffusion coefficients will be calculated with our new system of equations for ternary mixtures. Thus, we have theory for the calculation of diffusion coefficients, already overcome one of the main technical difficulties in complemented with classical Molecular Dynamics Year 1. Our proposed incremental numerical approach is simulations. no longer necessary as we now have a numerical capability that we did not expect to develop until year 3.
2. Formulate the thermodynamics (pressure, internal energy, entropy, etc) of mixtures within our model to We have initiated the validation of the results for mixtures develop the capability to compute equations of state of carbon and hydrogen (“plastic”) against published quan- of binary mixtures directly. tum molecular dynamics computer simulations based on a more fundamental theory but computationally very costly.
3. Write / test / validate a computer code to calculate the Initial results are encouraging and show good agree- thermodynamics of mixtures based on the results of ment at high densities (>7 g/cm3) and high temperatures task #2. (>80000K).
4. After the completion of task #3, we will turn to applica- Task two is nearly completed. In a very significant way we tions: a) Calculation of an equation of state table for C/ have surpassed our original goal for Year 1. He mixtures in white dwarf stars; b) Calculation of the diffusion coefficients in H/He mixtures in giant planets Furthermore, two new ideas developed under our equa- such as Jupiter and Saturn. This application will push tion of state and plasma physics projects have flowed into our model in a new physical regime where we do not our LDRD work. The microscopic model for the electronic know how well (or how poorly) it will perform. structure of the plasma ions, the electronic screening, and the fluid structure of the plasma is no longer the quantum
5. We anticipate the publication of the work completed hypernetted-chain (QHNC) model (our initial proposed during year 1 in peer-reviewed journals. A first pub- approach), but an original and important variation that has lication will describe our model for mixtures and a proved far superior for modeling warm dense matter of second will present the results of the calculation of one-component systems. This improves the fidelity of the diffusion coefficients in DZ white dwarf stars. microscopic physics of our mixture model. Conclusion We have also developed a new formulation of transport The primary result will be a robust capability to calculate theory valid for weak to moderately strong interacting the equation of state and transport coefficients of warm systems that allows very fast calculation of diffusion coef- dense mixtures using a single, unified approach with no ficients. This will result in a significant economy of com- adjustable parameters. We will develop the theoretical puter time by limiting the need for expensive molecular framework, implement the model in codes, validate the dynamics computer simulations to the cases where the model through comparisons with experimental data and interactions are very strong. We anticipate a much higher quantum molecular dynamics simulations, apply the model 663

Page 664

to study high-impact astrophysical problems related to planetary luminosity and white-dwarf stars, and advance the model to the point where it can be broadly used to generate a variety of equations of state for mixtures relevant to astrophysical, fusion, and national security research. 664

Page 665

Physics Exploratory Research Continuing Project A New Hypothesis to Explain the Variability of the Outer Radiation Belt: Can we Predict Post-storm Fluxes of Energetic Electrons Based only on Pre-storm Fluxes of the Lower-energy Population? Gregory S. Cunningham 20130297ER Introduction distributions during the event. In order to test the The primary goal of the proposed work is to test the hypothesis, we will create a new modeling capability hypothesis that flux levels of mega-electronvolt (MeV) that couples a low-energy code that predicts the source electrons in the outer belt can be predicted on an population to a code that predicts the evolution of the event-by-event basis solely as a consequence of elec- high-energy fluxes. Such a capability, though highly tromagnetic waves accelerating low-energy particles desirable, does not exist in the magnetospheric science that are convected in from the outer magnetosphere. community, making it nationally competitive. The hy- In order to conduct a test of this hypothesis, we need pothesis, if proved true, will enable forecasting of space a category of events for which fluxes of MeV electrons hazards due to energetic electrons in the outer radiation in the outer belt are highly depleted so that we can belt hours to days in advance. Such a capability would predict their buildup over time due to acceleration of a directly impact NASA’s goal to improve our understand- low-energy source population. We have identified High- ing of how the sun impacts the environment in which Speed Streams (HSS) as meeting this need. Assuming satellites operate, e.g. the ‘Living With a Star’ program. we have an initially-depleted state of the MeV electron The DOD, specifically the Air Force Weather Agency, is population in the outer belt, we must be able to predict tasked with alerting owners of national security assets the time-evolving distribution of low-energy electrons in space of impending hazards, and so DOD will directly convected into the outer belt, since the low-energy benefit. NOAA, specifically the Space Weather Predic- electrons are the source that will be accelerated to tion Center, is tasked with alerting owners of commercial higher energies through diffusion. The Ring-Current At- space-based assets of impending hazards in space, and mosphere interactions (RAM) code will be used for this so it will also directly benefit. Finally, the DOE builds purpose, first for HSS events and then for geomagnetic satellite instrumentation that is hosted on satellites storms. It must be demonstrated that RAM is predictive owned and/or operated by the DOD and Intelligence for HSS events, since RAM has been primarily applied Community. The owners/operators of these satellites to storms. Finally, given the time-dependent popula- rely on predictions from AFWA/SWPC, and hence will tion of low-energy electrons, which acts as a boundary also benefit from such a capability . condition for the 3D diffusion code, we must be able to Progress predict event-specific distributions of the amplitude of the electromagnetic waves responsible for diffusing the The 3D diffusion code has been extended to permit the low-energy particles up to higher energy. We will build use of time-varying boundary conditions using in-situ an event-specific empirical wave amplitude database data from the Combined Release and Radiation Effects using data from the soon-to-be-launched Radiation Belt Satellite (CRRES) or an initial condition from the Van Storm Probes (RBSP) mission that will, in its own right, Allen Probes (formerly called the Radiation Belt Storm result in considerable utility to the magnetospheric sci- Probes. or RBSP) mission. The model uses a wave ampli- ence community. tude database conditioned on geomagnetic activity that was constructed prior to the start of the LDRD project. Benefit to National Security Missions The prediction efficiency of the model was quantified This project will test the hypothesis that fluxes of outer during a training interval specified by the Radiation belt energetic electrons after geomagnetic events can Belts and Waves (RBW) focus group of the National be predicted from knowledge of the low-energy source Science Foundation’s Geospace Environment Modeling population prior to the event and electromagnetic wave (NSF-GEM) program. The sensitivity of the prediction 665

Page 666

efficiency to various model parameters, including radial currently doing. Because the NOAA POES satellites offer diffusion coefficient, location of the boundary condition, excellent coverage in MLT and good time resolution, we do and amplitude of the waves inside the plasmasphere, were not have to construct spatiotemporal correlations of the all computed in an effort to tune the model to perform as waves in order to produce global maps of wave intensity well as possible during the challenge interval also specified using point observations from the Van Allen Probes. This by the RBW focus group. The performance of the tuned work is in the process of being written up and will be sub- model was then computed for the challenge interval. The mitted to Geophysical Research Letters. results from this study were submitted to the Journal of Geophysical Research In May 2013, and are under review. Future Work The results will also be presented at the NSF-GEM RBW Quantify spatiotemporal correlations in the waves using focus group meeting in June 2013 and compared to the RBSP data and use them to compute uncertainties on diffu- performance of other models. sion coefficients. The need for an outer boundary condition in the 3D model Run the modified diffusion code for HSS events using RAM was removed by implementing a time-varying magneto- output as boundary condition. pause location that depends on solar wind driving. For particles outside the magnetopause, the lifetime is com- Predict event-specific energization of MeV electrons, and puted as a fraction of their drift period around earth, compute prediction efficiencies by comparing to data. which is energy-dependent (lower energies have a longer Incorporate a substorm injection model to capture the lifetime). Eliminating the boundary condition in the 3D transport of low-energy electrons into the inner magneto- model means that a source for new particles must be sphere during substorms. incorporated into the model. This source will come from a low-energy boundary condition from the Ring-current Conclusion Atmosphere Model (RAM) or from data. We have received We expect to prove, or disprove, the hypothesis that it is an example output of the RAM code from Vania Jordanova the variability of the low-energy electron source popula- and are working to include it as a low-energy boundary tion which, when combined with geomagnetic event-spe- condition in our 3D diffusion code. We will also be work- cific wave distributions, is responsible for the variability of ing this summer to include data from the Van Allen Probes post-event high-energy electron fluxes. If proven true, this and/or the Time History of Events and Macroscale Interac- hypothesis will lay the foundation for predicting post-event tions during Substorms (THEMIS) mission as a low-energy high-energy electron fluxes hours to days in advance, an boundary condition. Our goal is to present this work at the important capability for protecting space-based assets. If American Geophysical Union meeting in December 2013 proven false, intermediate objectives like producing an and prepare articles for inclusion in a special issue of Geo- event-specific empirical wave database and coupling a physical Research Letters on early innovative results from low-energy Ring-Current Atmosphere interactions code the Van Allen Probes. to a high-energy 3D diffusion code will be of value to the Finally, a new approach to constructing event-specific wave scientific community. models for chorus has been investigated. The basic idea Publications behind this approach is to observe the precipitation of Chen, Y., G. Reeves, R. Friedel, and G. S. Cunningham. mid-energy electrons whose anisotropic pitch-angle distri- Remote sensing whistler-mode waves from low-Earth- butions are unstable to the growth of chorus waves. When orbit electron observations: Statistical and case study the wave amplitudes are high, then precipitation fluxes results. To appear in Geophysical Research Letters. will also be high, and so one can infer the wave ampli- tudes from observations of the precipitation fluxes, which Tu, W., G. S. Cunningham, Y. Chen, M. G. Henderson, E. are monitored by a relatively dense network of low-earth Camporeale, and G. D. Reeves. Modeling radiation belt orbiting satellites (the National Oceanic and Atmospheric electron dynamics during GEM challenge intervals with Administration’s Polar-Orbiting Earth Satellite, or NOAA the DREAM3D diffusion model. To appear in Journal of POES). The NOAA POES satellites provide excellent cover- Geophysical Research-Space Physics. age in magnetic local time (MLT) and adequate temporal resolution (due to their 90-minute orbit). The idea is to use POES observations from specific events to estimate the wave amplitudes for those events rather than relying on databases that are binned by magnetic activity as we are 666

Page 667

Physics Exploratory Research Continuing Project A New Approach to Multiscale Plasma Physics Simulations Gian L. Delzanno 20130334ER Introduction physics since (1) the problematic coupling of macro- Plasmas are characterized by a wide range of spatial scopic and microscopic scales needed in many problems and temporal scales, and pose a formidable challenge of interest is naturally handled by the method and (2) it to the numerical modeling and to our ability to perform offers the possibility for much more accurate results at predictive simulations. This project focuses on a new a fraction of the computational cost obtained by con- method that promises to solve some longstanding issues ventional approaches. As such, any application where in computational plasma physics and, in particular, of plasmas are important will benefit from this capability. Particle-In-Cell (PIC) approaches (which are commonly It is immediately relevant to the DOE/NNSA nuclear used to model microscopic plasma phenomena). Ours is weapons mission, but it also includes agencies such as essentially a spectral approach, where the distribution DOE (office of science), NASA (space science) and DOC function in velocity space is not handled by superpar- (NIST and NOAA). It will also impact many areas of LANL ticles as in PIC, but by an expansion in a suitable basis. mission, including nuclear weapons, energy security and The method is then cast as a series of partial differential global security/space situational awareness. equations for the coefficients of the expansion. First, Progress our preliminary results indicate that this approach can achieve machine precision accuracy with a relatively In the first eight months of this project, we have de- low number of basis functions and at a small fraction of veloped a code to study the evolution of a collisionless the computational cost incurred to obtain a (much less plasma in the electrostatic limit, based on an expansion accurate) solution with a PIC code. Second, since the low in Hermite polynomials. order terms of the expansion correspond to the typi- We formulated a second order accurate, semi-implicit cal macroscopic moments (density, mean velocity and time discretization, where some of the operators are energy), the microscopic/macroscopic coupling critical treated with an explicit time discretization and the oth- to multiscale plasma physics is naturally built-in. We ers are treated with an implicit time discretization. The have successfully demonstrated on a test case that we semi-implicit code conserves the total mass and momen- could use very few basis functions in certain regions of tum (but not energy) of the system exactly. The advan- the computational domain (where the plasma behaves tage of a semi-implicit method relative to a conventional as a macroscopic fluid) while retaining many more basis fully explicit method is that it features a much larger functions in other areas (where the microscopic details stability domain at the same computational cost. In the of the plasma are important). Third, the algorithm can examples considered we have proved that the largest make effective use of emerging architectures, such as time step in the explicit method is at least 400 times GPUs, for the optimization of the basis functions. The smaller than the time step that can be used in the semi- goal of this project is to develop the spectral capabil- implicit method. ity just described for general multiscale plasma phys- ics problems. As a case study, we target an application We compared the semi-implicit method against explicit relevant to space weather, an important area for LANL PIC simulations for standard test problems (plasma mission in threat reduction and global security. waves, Landau damping, two-stream instability, ion acoustic wave) and found that the Hermite code is a Benefit to National Security Missions couple of orders of magnitude more accurate and faster. The capability focus of this project could be a game- On a Landau damping study, the Hermite code results changer in the field of computational multiscale plasma 667

Page 668

to be about 180 times more accurate than the PIC code, terms of accuracy and computational efficiency on ex- and 14 times faster. A manuscript is in preparation on the amples that test the disparity of scales typically associated results of 1) and 2). with plasmas, we are focusing on other examples where the plasma behaves macroscopically (requiring a low num- We developed a fully implicit formulation for the Hermite ber of polynoms) in some parts of the domain, whereas it code. By using an implicit second order accurate Crank- requires microscopic physics (high number of polynoms) Nicholson time discretization and a spectral (Fourier) elsewhere. We are exploring techniques to handle the spatial discretization, we have proved that the resulting nu- microscopic/macroscopic coupling by adapting the local merical model conserves the total mass, momentum and number of polynoms automatically (instead of in a pre- energy of the system exactly. This is a major result consid- defined way) to increase the robustness of the code. ering that PIC algorithms that conserve mass, momentum and energy exactly do not exist. Exact conservation prop- Future Work erties, however, are related to the choice of the temporal The proposed work will continue to build on the working and spatial discretization and we are now trying to extend version of the Hermite spectral capability that we devel- it to other popular discretizations schemes like finite differ- oped. ences or h-p finite elements. Another important advantage of the implicit formulation is that the stability domain of • We will continue the extension of the code from one the method further increases relative to the semi-implicit to two dimensions, in order to apply it to more com- case discussed above. plex problems. This is a milestone for the project. We developed an energy-conserving Hermite code based • We will begin the extension to the electromagnetic on the implicit formulation. At each time step the resulting case. nonlinear equations are solved with an unpreconditioned Jacobian Free Newton-Krylov solver. We have successfully • We will explore preconditioning techniques for the applied the implicit Hermite code to the same test cases linear iterations of the Newton-Krylov solver to speed- described in 2). A comparison with a recently developed up the convergence of the method. implicit PIC code [Markidis and Lapenta, Journal of Com- • We will exploit the CPU/GPU infrastructure that we putational Physics 2010] is underway. A manuscript is in have built to study the choice of the optimal base for preparation on the results of 3) and 4). the Hermite expansion. This technique relies on the We developed a C++ project, called Space Weather Simu- assumption that the plasma does not change much lation Framework (SWSF), that has support for launching during few time steps of the simulation so that the and controlling processes across multiple OpenCL (CPUs optimization can be performed asynchronously. There- or GPUs) devices. Additionally, SWSF provides low-level fore we will quantify the level of asynchronism that data structures for representing fields on structured or can be exploited, the code speed-up and the accuracy unstructured meshes using a general mesh specification of the resulting technique. This is another milestone file. We have also developed a C++ design pattern that for the project. allows asynchronous task launch to a user-defined “event” • The code will be verified against standard test cases. handler. This feature is necessary for us to schedule the We will also continue benchmarking against PIC codes update of the basic solver concurrently with the solution of since our metric for success is to prove that our new the optimization parameters. As part of this design pattern, technique can be much more accurate and faster than we have also added an integrated communication manager PIC, which is the community standard. that utilizes MPI to allow the high-level parallelization of various solver events. This model allows serial execution Conclusion for our early development plan, but also supports the full This project will deliver an electromagnetic, parallelized distributed-memory parallelization needed that we aim code that models a collisionless magnetized plasma. The for. Finally, in order to support flexible problem definition, code will be compared against available PIC codes (the SWSF utilizes an XML input format that allows parameter dominant method in the community) and our metric for specification using several basic types. This approach will success will be based on the comparison in terms of ac- allow us to adapt our input deck format to support new curacy, efficiency and code speed-up. We expect that our solver features and characteristics as they become rel- tool will outperform PIC codes, as our preliminary results evant. indicate. If we can indeed confirm the huge savings factors While we have established the strength of the method in of our approach relative to PIC, it will be a breakthrough 668

Page 669

in computational plasma physics that can impact many plasma applications, such as space weather or magnetic fusion energy. Publications Camporeale, E., G. L. Delzanno, B. Bergen, and J. D. Moulton. Fourier-Hermite spectral method for the Vlasov equation. Presented at 55th Annual Meeting of the APS Division of Plasma Physics. (Denver, 11-15 Nov 2013). Camporeale, E., G. L. Delzanno, B. Bergen, and J. D. Moulton. On the velocity space discretization for the Vlasov-Poisson system: comparison between Hermite spectral and Particle-in-Cell methods. Part 1: semi- implicit scheme. 2013. To be submitted to Journal of Computational Physics. Camporeale, E., G. L. Delzanno, B. Bergen, and J. D. Moulton. On the velocity space discretization for the Vlasov-Poisson system: comparison between Hermite spectral and Particle-in-Cell methods. Part 2: fully- implicit scheme. 2013. To be submitted to Journal of Computational Physics. Delzanno, G. L., E. Camporeale, B. Bergen, and J. D. Moulton. Fully-implicit Fourier-Hermite spectral method for the Vlasov equation. Presented at Vlasovia 2013. (Nancy, France, 25-28 Nov 2013). 669

Page 670

Physics Exploratory Research Continuing Project Enhancing Thermoelectric Properties of Topological Insulators through Nanostructuring Nikolai Sinitsyn 20130348ER Introduction perform device fabrication and characterization of Te This project is motivated by the recent discovery of based nanowires under the guidance of Han Htoon and new materials, called Topological Insulators (TI) which Jennifer Hollingsworth. demonstrate a new state of electrons that is topologi- As a complementary approach to the top-down synthe- cally distinct from the conventional band insulators or sis methods provided to the project by our collaborators, metals. One of the manifestations of the topological we are also pursuing novel solution-phase syntheses properties is appearance of conducting electron states, for Bi2Te3 nanoparticles. The colloidal methods afford on the surface of material, which propagate throughout access to ultra-confined sizes beyond the reach of the the sample without elastic scatterings. Our idea is that vapor-phase methods. Thus far, we have successfully nanostructuring of TIs can substantially increase the role replicated a literature approach for synthesizing sub-10 of these states in charge and thermal transport to the nm Bi2Te3 spherical nanocrystals (Scheele et al. Adv. level of obtaining a device structure with record high Funct. Mater. 2009, 19, 3476–3483). This approach thermoelectric characteristics. requires first synthesizing Bi nanoparticles that are then converted to a Bi-Te alloy and, finally, into Bi2Te3 nano- Benefit to National Security Missions crystals. Our goal, however, is to synthesize ultra-small- Major applications of thermoelectrics is to (i) create diameter nanowires using the metal-catalyzed solution- efficient silent and compact solid state refrigerators, liquid-solid (SLS) method. To this end, we are attempting that one day will replace noisy and spacious spinning to modify the literature synthetic method for nanocrys- components of home fridges. (ii) various technologies to tals to be amenable to adaptation to SLS growth. Spe- recover heat energy, e.g. from car engines and micropro- cifically, SLS will require that Bi2Te3 is formed by direct cessors and convert it back to useful electricity. reaction of Bi and Te molecular precursors, rather than (i) and (ii) correspond to DOE goals. Climate impact is reaction through a Bi (or Te) crystalline intermediate. We positive because substances used in modern refrigera- have to date performed numerous reactions between tors are harmful for atmosphere, e.g. they destroy the different Bi and Te molecular precursors in a range of ozone layer. Replacing them by thermoelectric cooling solvents and mediated by a range of different surfac- will solve the problem. tants. These have been characterized by powder x-ray diffraction (XRD) and transmission electron microscopy (iii) NASA application is due to new energy sources for (TEM). While XRD shows that we are creating crystalline satellites that can be built from thermoelectrics. Bi2Te3, the morphology remains non-optimal and/or the nanoparticles are aggregating. We will continue optimi- Progress zation of the procedure and, once obtained, will extend By collaborating with Prof. Shixiong Zhang of Indiana growth to nanowires via SLS, which can afford sub-10 Univerity, we have successfully synthesized nanowires/ nm diameter 1-dimensional structures. nanoribbons of Telluride-based topological insulators On the theoretical side, we developed a completely new such as Bi2Te3 and Ag2Te. We have utilized thermal approach for classification of topological phase transi- vapor transport method to fabricate nanowires and tions, which can be encountered in topological insula- nano ribbons of diameter: 50~300 nm and width 500nm tors. Such a classification is an important question that respectively. A student of Prof. Zhang, Enzhi Xu will visit should help to identify distinct phases of new materials. CINT (both LANL gate way and Sandia Core facility) to 670

Page 671

We noticed that arbitrary dissipation effects make the ef- Publications fective Hamiltonian of a Bloch band system non-Hermitian. Ren, J., and N. A. Sinitsyn. Braid Group and Topological We then showed that the classification of distinct phases Phase Transitions in Nonequilibrium Stochastic Dynam- of a non-Hermitian Hamiltonian follows almost trivially by ics. 2013. Physical Review E. 87: 050101(R) . identifying a pattern of Bloch band structure with a topo- Roslyak, O., and A. Piryatisnki. Sonoluminescence of car- logically equivalent element of the Braid group. bon nanotubes. 2013. Unpublished Document. It has been discussed recently that topological properties Sinitsyn, N. A.. Landau-Zener Transitions in Chains. 2013. of transport can be broken by the inter-band transitions, Physical Review A. 87: 032701. so called Zener tunneling, near the interface of topologi- cal material. Importantly, Zener tunneling in topological Sinitsyn, N. A.. Nonadiabatic transitions in exactly solv- insulators cannot be described by a standard Landau-Zener able quantum mechanical multi-channel model: role formula. This motivated us to explore Landau-Zener transi- of level curvature and counterintuitive behavior. 2013. Physical Review Letters. 110: 150603. tions in more relevant to topological insulators multi- channel scatterings framework. We uncovered an unex- pected phenomenon of exponentially strong suppression of Landau-Zener transition amplitudes even in the limit at which standard Zener tunneling would be strong. Results were published in Phys. Rev. Lett. A side result of this research was an application in a different field – dynamic quantum phase transition, in which we predicted unusual distribution of defects produced during multi-channel Landau-Zener scattering in Bose condensates. Future Work We will develop a theory of thermoelectric effects in thin films and nanowires made of TIs. This goal will require an understanding the role of disorder and quantum confine- ment and electron/phonon interactions on the topologi- cally protected surface states. Our preliminary studies showed complexity of the disorder effects on transport whose understanding promises to reveal unusual proper- ties. Complimentary experimental efforts will be focused on measurements of thermal transport in topological insulator nanowires using state of the art CINT Discovery platform. This will provide a selective probe for conflict- ing processes determining the figure of merit. Theoretical results will be benchmarked against experimental data and used to provide a guideline for subsequent experimental studies. Conclusion We will develop a theory of thermoelectric effects in thin films and nanowires made of TIs. This goal will require an understanding the role of disorder and quantum confine- ment on the topologically protected surface states. Devel- oping a theory of thermoelectric effects and its applica- tion for optimizing the sample parameters to increase the figure of merit is the main theoretical goals of this project. Complimentary experimental efforts will be focused on measurements of thermal transport in topological insula- tor nanowires using state of the art CINT Discovery plat- form. 671

Page 672

Physics Exploratory Research Continuing Project Giving Cold Atoms Weight: creating Heavy Fermions in Optical Lattices Cristian D. Batista 20130385ER Introduction Lattice Hamiltonian. We will explore the phase diagram Lanthanide and actinide based compounds belong to the and dynamics. By guiding cold atom physics towards a family of strongly correlated ma-terials. The coexistence first realization of heavy fermion physics, this project will of localized f-electrons with itinerant s, p or d electrons establish the first one-to-one strongly correlated physics that interact via the so-called Kondo exchange, leads connection for cold atom physics. Moreover, our studies to very unusual phenomena such as exotic magnetic of the dynamics will feed back onto the heavy fermion order-ings, unconventional superconductivity and heavy physics and we will suggest heavy fermion experiments fermion behavior. One of the most intriguing phenom- that can explore the boson-fermion and Kondo physics ena is the non-Fermi liquid (“strange metal”) behavior mapping. This research is very relevant for uncovering that arises near the quantum phase transition or quan- the origin of the complex properties of rare lanthanide tum critical point (QCP) that separates the magnetically and actinide based materials which are relevant for the ordered and the Heavy fermion sates. This “strange nuclear energy security mission. metallic state” originated by the fluctuations of a QCP Progress usually appears in the proximity of unconventional superconducting states. Unfortunately, the experimental In the cold atom toolkit, Feshbach resonances have characterization of this fascinating state of matter is lim- played a unique role in boosting the cold atom capabil- ited by many factors that we discuss below. Consequent- ity of simulating quantum many-body physics: no other ly, there is a need for finding controllable realizations of quantum many-body system has had a knob available this physics that can shed light on the basic properties to tune the particle-particle interactions at will. Re- of the strange metallic state that arises near the QCP. cently, the increased control and accuracy of external We will address this need by modeling a class of atomic magnetic fields has opened up a new class of Feshbach gases known as Bose-Fermi resonant systems in an resonances: narrow Feshbach resonances, width that is optical lattice, and identifying regimes of parameters for typically less than 10 mG (10 milliGauss). While these which the atomic system can be mapped into a lattice resonances are more difficult to control experimentally, of local moments (localized f-electrons) that interact via these resonant interactions also open up new avenues. ex-change with itinerant electrons. Describing and test- The effective range of the narrow resonant binary atom ing this non-trivial mapping requires a description that interactions can be comparable to or exceed the aver- bridges the Angstrom-sized atomic physics scale with the age inter-particle distance, the closed channel amplitude micron-sized optical lattice physics. dominates the coupled channel interaction physics and the energy dependence of the scattering physics can Benefit to National Security Missions select a slice of phase space that is much less than the We argue that the closed channel amplitude physics many-body-relevant phase space volume. These aspects of the Fesh-bach resonance give access to the Kondo directly affect the physics of atoms in an optical lattice physics responsible for heavy fermion behavior in tra- and may allow the simulation of Heavy-fermion like ditional strongly correlated materials. Narrow Fesh- Kondo physics, the topic of this project. bach resonances, which are beginning to be explored On the atomic side of the project, we have accomplished experimentally, give large closed channel amplitudes the following steps in our investigations: (i) We have near resonance. We will describe the narrow resonance derived the energy-dependent binary atom scattering properties and we will explore the validity of the effec- length for narrow resonances and we have shown that tive Hamiltonian that can be mapped onto the Kondo 672

Page 673

its energy scaling depends on only three parameters. We ing of actinide and lanthanide based compounds that are have shown that it is the slowness of the interaction that of strategic relevance to LANL’s mission. characterizes the narrow resonant interactions. (ii) We have solved the problem of two harmonically confined Publications atoms near a narrow Feshbach resonance (providing a Rahmani, A., T. Kitagawa, E. Demler, and C. Chamon. Cool- model of the on-site interactions in an optical lattice). (iii) ing through optimal control of quantum evolution. We have derived the fraction of the collision time that 2013. PHYSICAL REVIEW A. 87 (4): -. the interacting atoms spend in the closed channel state. As this fraction is sensitive to the energy of the collision complex and as the closed channel state is unstable with respect to three-body loss processes, narrow Feshbach resonances can select the energy of atoms that are prefer- entially removed, introducing a new method for magnetic field steered evaporative cooling or heating. (iv) We have tested the exact eigenstates of the harmonically con- fined binary atom system against a two-state model and we found that the two-state model, which is at the heart of the Kondo-Hamiltonian, describes the harmonically confined narrow resonance well near resonance if the frequency of the width is comparable to or less than the trapping frequency. We are in the process of writing up these results. We have also wrote a Density Matrix renormalization group code for simulating the Kondo Lattice model. We are currently analyzing the results for different types of correlation func- tions. Future Work • Develop the Density Matrix Renormalization Group codes for the Kondo Lattice Model described in our project. • Derive the Kondo Lattice model from first principles in the context of atomic gases and express the Ham- iltonian parameters, t, ε, and g as a function of more fundamental parameters of the atomic gas. • Explore closed channel amplitude in narrow Feshbach resonance interactions in scat-tering and in on-site in- teractions. Test validity of effective g/ε model interac- tions in predicting static properties. • Explore tuning options of effective model. Conclusion The successful accomplishment of this project will guide the realization of heavy fermion phases with atomic gases near Feshbach resonances. This extraordinary achievement will open an intriguing new avenue: controlled experimen- tal explorations of the critical regime near the fermionic quantum critical point that separates the magnetically ordered state from the heavy fermion phase. This break- through would have a profound impact on our understand- 673

Page 674

Physics Exploratory Research Continuing Project Topology in Superposition: Quantum Decoherence in Many-body Systems Wojciech H. Zurek 20130409ER Introduction tal elements of computers will reach single atom size this We will investigate decoherence of quantum systems in decade (“Moore’s law”). Understanding what happens realistic models, where the environment causing deco- to superpositions of collective degrees of freedom and herence is a many-body system (i.e., a system composed of relevant sources of decoherence is indispensable of interacting subsystems). In particular, we shall study in these (and similar) applications of interest to LANL. decoherence caused by environments driven through a Moreover, quantum superpositions in cold atom systems critical point of a quantum phase transition. that we plan to study can be used to enhance sensitivity of measuring devices. Capabilities foreseen for MaRIE We will also study superpositions of collective states of seem well suited for condensed matter studies related to many-body systems. Both spin and field theoretic mod- this project (dynamics of phase transitions). els of phase transitions exhibit stable topological defects (monopoles, vortex lines, domain walls, etc.) as well as Progress non-topological but relatively stable configurations such In the past year, we have proceeded towards the goal of as solitons. These objects are the epitome of locality. Yet, understanding decoherence on several fronts: the underlying many-body system is quantum and, thus, abides by the quantum principle of superposition. There- We have studied conditions that allow a quantum state fore, it should be possible to construct superpositions of a macroscopic system to produce many imprints (e.g., where a topological defect is in a non-local “Schrödinger on its environment; this is the main process behind cat” state, with a vortex or soliton simultaneously in two decoherence). The question is technically challenging, places. Such situations are difficult to even imagine or as macroscopic systems are generally in mixed states, represent in terms of effective, mean-field “order param- and interact with other systems. We have established eter” theories. However, there are solvable models that that – even in such circumstances – the condition for the should allow us to study how manifestations of non- states of the system to be reproducible (i.e., to act as vi- locality and other explicitly quantum behaviors (present able “originals” for copies) they must be distinguishable. at the microscopic level) get eliminated from collective (Technical requirement for this is that they must have many-body states on the macroscopic level. support in distinct – orthogonal – subspaces of the Hil- bert space.) A paper that describes this result has been All along we will explore possibilities of experimental just published in Physical Review A. testing of our predictions in cold atom and condensed matter systems with an eye towards applications in ma- We have also studied production of multiple (even if terials science and metrology. imperfect) copies is central to realistic models of deco- herence (that extend the original idea of a monolithic Benefit to National Security Missions environment to a collection of its natural subsystems This research will contribute to condensed matter – e.g., photons in a photon environment). This leads science at Los Alamos, providing critical insights into to the so-called “quantum Darwinism”. With Michael quantum technological applications. These are diverse, Zwolak, Zurek has studied the nature of such an informa- including superconductors, metrology, etc. Computing tion flow between the system and the environment. The may be most significant and urgent. There is the dream conclusion is that in this case the environment acts as of quantum information processing, but even if it turns a communication channel – a quantum communication out to be more distant than some hope, the fundamen- channel that transmits multiple copies of the (effectively 674

Page 675

classical) information about the preferred basis that is Future Work singled out by decoherence. This analogy is reflected in the Our goals for the second year of this project include study appearance of the Holevo quantity (that bounds the capac- of the dynamics of many-body systems suitable for investi- ity of quantum channels to transmit classical information). gation of the dynamics of phase transitions and its rela- Holevo quantity is maximized for the preferred, effectively tion to decoherence. Bose - Hubbard model (which has classical states of the system. Moreover, depending on the a reasonable chance of experimental implementation in choice of the observable of interest, classical information the not too distant future) will be amongst these studied. available from the environment fragments (given by the Study of various models with the relevance for the forth- Holevo quantity) decreases while quantum information coming experiments in Bose-Einstein condensates will be (quantified by the quantum discord) increases. The sum of also undertaken. these two – Holevo quantity plus quantum discord – turns out to be the von Neumann mutual information which is Investigation of information propagation throughout the independent of the choice of the observable. This is a de many-body environments will continue to the second year facto conservation law that expresses complementarity of the project. We have already made significant progress which has profound consequences for the emergence of by establishing complementarity between quantum dis- preferred effectively classical observables, as well as for cord (measure of quantumness of correlations) and Holevo the loss of quantumness. Our paper on this subject has quantity (that characterizes capacity of quantum channels been just published in Scientific Reports. for classical information transfer). Moreover, we have char- acterized constrains imposed by repeatability of measure- With Alexander Streltsov we were able to provide an at- ments on the states of macroscopic quantum systems that tractive view of quantum discord while shedding a new allow them to be read off many times and/or by many light on quantum measurements. The motivation goes observers. This is of importance to quantum Darwinism. back to the forefathers of quantum theory: Niels Bohr proposed that the outcome of the measurement becomes The work on the Chernoff information and its relation to objective and real, and, hence, classical, when its results quantum Darwinism will be pursued. can be communicated by classical means. In this work we Study of the imprinting of “histories” on the many-body revisited Bohr’s postulate using quantum discord — one of environment will be the new focus of our research on the modern tools from the quantum information theory quantum Darwinism. Exploration of the suitability of ob- arsenal. We find a full confirmation of Bohr’s idea: if a jects such as solitons for the the study of superpositions in measurement device is in a nonclassical state, the mea- Bose-Einstein condensates will be also pursued. surement results cannot be communicated perfectly by classical means. In this case some part of information in In the first year we have (with colleagues in Germany) pub- the measurement apparatus is lost in the process of com- lished (Nature Communications, to appear) a joint paper munication: the amount of this lost information turns out on the formation of topological defects in “ion crystals” to be the quantum discord. The information loss occurs — regular formations of ions that are investigated for the even when the apparatus is not entangled with the system purpose of quantum emulation and quantum computing. of interest. The tools presented in this work allow to gener- While the experimental results were obtained in the re- alize Bohr’s postulate: we show that for pure system-appa- gime where the ion chain was effectively classical, we shall ratus states quantum communication does not provide any pursue the research into the quantum regime (with the advantage when measurement results are communicated hope that experiments shall eventually follow). to more than one recipient. We further demonstrate the superiority of quantum communication to two recipients Conclusion on a mixed system-apparatus state and show that this ef- We will characterize decoherence of topological defects fect is fundamentally different from quantum state clon- in different theoretical models describing both magnetic ing. The paper has been just accepted in Physical Review and cold atom systems. In both cases, we will quantify Letters. how a many-body quantum environment interacting with a topological defect leads to the destruction of a coherent Last not least, we have studied appearance of topological quantum superposition state in which a topological defect defects in phase transitions in an ion trap. This work with was initially prepared. This work will significantly advance Adolfo del Campo is (so far) confined to classical regime, fundamental understanding of the role of many-body ef- but ion traps are perhaps the best candidate to study su- fects in the ubiquitous decoherence process responsible perpositions of topological defects that are of interest for for the emergence of the classical “everyday” world out of the proposal. the quantum substrate. 675

Page 676

Publications Pyka, K., J. Keller, H. L. Partner, R. Nigmatullin, T. Burger- meister, D. M. Meier, K. Kuhlmann, A. Retzker, M. B. Plenio, W. H. Zurek, A. del Campo, and T. E. Mehl- staubler. Topological defect formation and spontane- ous symmetry breaking in ion Coulomb crystals. 2013. NATURE COMMUNICATIONS. 4: -. Riedel, C. J., W. H. Zurek, and M. Zwolak. The rise and fall of redundancy in decoherence and quantum Darwin- ism (vol 14, 083010, 2012). 2013. NEW JOURNAL OF PHYSICS. 15: -. Streltsov, A., and W. H. Zurek. Quantum Discord Cannot Be Shared. 2013. PHYSICAL REVIEW LETTERS. 111 (4): -. Tylutki, M., J. Dziarmaga, and W. H. Zurek. Dynamics of the Mott Insulator to Superfluid quantum phase transition in the truncated Wigner approximation. 2013. 21ST IN- TERNATIONAL LASER PHYSICS WORKSHOP. 414: -. Zurek, W. H.. Wave-packet collapse and the core quantum postulates: Discreteness of quantum jumps from uni- tarity, repeatability, and actionable information. 2013. PHYSICAL REVIEW A. 87 (5): -. Zwolak, M., and W. H. Zurek. Complementarity of quantum discord and classically accessible information. 2013. SCIENTIFIC REPORTS. 3: -. 676

Page 677

Physics Exploratory Research Continuing Project Ultra-bright Electron Beam Acceleration in Dielectric Wake Accelerators Evgenya I. Simakov 20130463ER Introduction Considering the long time scale associated with the Ma- The goal of this project is to demonstrate operation of a RIE project, this technology, if demon-strated now, may high-brightness Dielectric Wakefield Accelerator (DWA) provide an approach to boost the energy of the electron with an acceleration gradient above 100 MV/m and less beam feeding the Ma-RIE XFEL from the nominal 12 than 0.1% induced energy spread in the accelerated GeV up to 20.8 GeV with a very low cost upgrade. This beam. We currently project the DWA concept as a per- up-grade would allow a much greater production of 126- formance upgrade for the future LANL signature facility keV photons, now at the third harmonic of 42 keV. With MaRIE. The pre-conceptual design for MaRIE is under- the current 12 GeV MaRIE linac design, generation of way at LANL, with the design of the electron linear ac- the third harmonic photons is suppressed in the wiggler. celerator being one of the main re-search goals. The cost However, photon energy above 120 keV is required for of the linac is significant and the TA-53 space constraints the K-shell ionization of uranium and other actinides, an dictate that the final energy of the electron beam for important MaRIE mission and part of its funding justifi- the X-ray Free-Electron Laser (XFEL) is no higher than 12 cation. An 8.8-GeV DWA afterburner would lead to over GeV. The number and the energy of photons produced an order of magnitude greater production of these high by the XFEL is however strongly de-pendent on the energy photons. electron beam’s energy with the more energetic beam Progress delivering more energetic photons to the user. Although generally the baseline design needs to be conservative There were two the most significant outcomes of the and rely on existing technology, any future upgrade first year of this project. First, our team have completed would immediately call for looking into the advanced the preliminary design of the Emittance Exchanger (EEX) accelerator concepts capable of boosting the electron for shaping the electron beam and started to plan for an beam energy up by a few GeV in a very short distance experiment. Second, we have established all the neces- without degrading the beam’s quality. This work has the sary collaborations to jump start the experiments in potential to advance the DWA technology to a level that Year 2. would make it suitable for MaRIE and also for a number We used the Elegant matrix code developed and dis- of oth-er national security applications. tributed by the Argonne National Laboratory to design the electron beam optics for the EEX which would Benefit to National Security Missions convert the Gaussian electron beam at the input into The proposed work has the potential to advance DWA two bunches at the output: the drive bunch with a technology to a level to make it suitable for a number of double-triangular current distribution and the witness national security applications, including compact ac- bunch with the trapezoidal current distribution. The EEX celerators for warfighter support (e.g. small weaponized optics consists of two dipole doglegs and a tuned pair free-electron lasers) and active interrogation (e.g. small of a deflecting cavity and a fundamental mode cavity and in-expensive electron accelerators as compact front in between. A number of quadruple magnets precedes ends for muon active interrogation sources or alterna- the doglegs, and a number of sextupoles may be used tively to generate bremsstrahlung radiation). Addition- in the system to compensate for non-linear effects. The ally, this technology may be appli-cable for the future role of each element in the beamline is evaluated. The upgrades to increase the energy and luminosity of the non-linear beam shaping mask is designed to produce MaRIE XFEL. the desired current distributions out of a given Gaussian 677

Page 678

bunch. The Gaussian bunch under consideration is the one wakefield accelerators for FELs in December 2013. The that is being produced at ASTA user accelerating facility at main goal of the workshop is to attract the attention of Fermilab. the DWA community towards the specific needs of LANL’s MaRIE FEL. Our T and XCP divisions part of the team successfully con- ducted particle-in-cell (PIC) simulations of propagation of LANL team is planning to present a poster at the Interna- realistic distributions of electrons in a dielectric fibers. PIC tional Conference on Plasma Science in June, 2013 and simulations were found to be in good agreement with the- has submitted 3 abstracts to the US Particle Accelerator ory with respect to frequency, amplitudes and transformer Conference in October 2013. ratio of the excited electomagnetic signal. Resolution of PIC simulations was studied and adequate resolution was Future Work found to ensure accurate results. It was confirmed that the In FY13 we will initiate the project and conduct the nu- wakefield excited by the drive beam is largely insensitive merical and theoretical analysis of the coupled wakefield to the beam’s emittance, radial profile and energy. Simula- interaction (shaping both the drive bunch and the main tions were conducted with the two codes, MERLIN and beam). Preliminary optimized shapes will be determined. LSP-3D, which perfectly agreed with each other. Detailed particle-in-cell (PIC) simulations of the preliminary optimized shapes and geometry will be conducted. Emit- With regards to the planned experiment, we have initiated tance exchanger (EEX) and mask optics will be preliminary collaborations with the Argonne National Laboratory (ANL) designed and verified with PIC simulations. Once the simu- and the Fermilab (FNAL). We have participated in dielec- lations are complete, we will start ordering experimental tric wakefield accelerations experiments, that are being hardware, which includes the dielectric loaded waveguide, conducted by the ANL and Euclid Techlab’s personnel at support hardware and the mask. Brookhaven National Laboratory. The Accelerator Test Fa- cility (ATF) at BNL uses a simple dispersive dogleg to shape Conclusion electron bunches before passing them through dielectric The most important outcome of the proposed research will fibers. We have built several masks and installed them on be the experimental demonstration of a high-brightness the ATF beamline to observe bunch shaping. We have con- dielectric wakefield accelerator with an accelerating gradi- ducted simultaneous modeling of the ATF beamline with ent well above 100 MV/m and less than 0.1% induced Elegant to understand the observed results. Bunch shaping energy spread in the accelerated beam. It is very likely that in a dogleg has been confirmed in an experiment. We also we will be able to demonstrate the following secondary passed a shaped electron bunch through a dielectric fiber outcomes: (1) simultaneous generation of a drive bunch and observed energy modulation in the drive bunch and and main beam for the first time, (2) significant increases energy change in a witness bunch. in a DWA transformer ratio, and (3) significant decreases in the measured energy spread from a main beam acceler- We currently participate in a bi-weekly teleconference with ated through a wakefield process. ANL and FNAL, at which different research group discuss their near-term plans and experiments on dielectric wake- Publications field acceleration. ANL is planning to put and Emittance Huang, C., T. Kwan, D. Shchegolkov, and E. Simakov. Exchanger at their APS facility and conduct some beam Particle-In-Cell Modeling of Dielectric Wakefield Ac- shaping experiments. LANL has provided input for their celerator. To appear in 2013 North American Particle design of the experiment. LANL personnel is planning to Accelerator Conference. (Pasadena, CA, Sept. 29 - Oct. travel to the ANL and participate in the APS experiment. 4th, 2013). LANL has shipped a 1.3 GHz deflecting cavity to FNAL to Shchegolkov, D., E. Simakov, S. Antipov, M. Fedurin, and C. construct the EEX experiment at their ASTA beamline. ASTA Swinson. Beam Pulse Shaping Experiments for Uniform user facility is coming to operation this summer. The PI is High Gradient Dielectric Wakefield Acceleration. To planning to travel to the first ASTA user workshop in July, appear in 2013th North American Particle Accelerator Conference. (Pasadena, CA, Sept.29-Oct.4th, 2013). 2013 to learn more about the timeline of the EEX experi- ment. Right now, all the experimental components for the Shchegolkov, D., and E. Simakov. Design of an emittance ex- EEX experiment are available at ASTA. Once the facility changer for producing arbitrary shapes of the electron itself is operational, the assembly of the EEX beamline is beam current. Physical Review Special Topics - Accel- going to start. erators and Beams. LANL is putting together a mini-workshop on dielectric Simakov, E., C. Huang, T. Kwan, and D. Shchegolkov. Shap- 678

Page 679

ing Electron Bunches for Ultra-bright Electron Beam Acceleration in Dielectric Loaded Waveguides. To ap- pear in 2013 North American Particle Accelerator Con- ference. (Pasadena, CA, Sept. 29 – Oct. 4th, 2013). 679

Page 680

Physics Exploratory Research Continuing Project Wide Field-of-View Plasma Spectrometer Ruth M. Skoug 20130564ER Introduction Benefit to National Security Missions Through this LDRD-ER project, we will develop and The 2PS Spectrometer represents a revolutionary ad- demonstrate a fundamentally new type of space plasma vance for measurement of space plasmas and will realize spectrometer (named the “2PS” Spectrometer for 2π-sr) improved performance with a reduction in required whose field-of-view is nearly 2π-sr using fewer resources size, mass, and power resources. It will also, for the than traditional methods. The enabling component is first time, enable ultra-fast measurements of plasmas to analogous to a pinhole camera having an electrostatic discover and test hypotheses about physical processes energy-angle filter at the image plane. At the end of this that induce dynamic plasma variations on short time project, we will have an optimized design and thorough scales. By allowing investigation of processes at short understanding of the measurement concept as well as time scales (<0.1 s) this work is thus directly relevant to demonstration of the electrostatic energy-angle filter a number of NASA space science goals, including studies and its performance. The 2PS Spectrometer represents of turbulence and reconnection in space plasmas. a revolutionary advance for measurement of space For 50 years, Los Alamos has flown plasma spectrom- plasmas and will realize improved performance with a eters for national security payloads as well as NASA reduction in required size, mass, and power resources. scientific missions to study solar wind and planetary The concept and operation are intrinsically simple, magnetospheres such as Earth and Saturn. Monitoring and enable ultrafast (<0.1 s) measurement of plasma and understanding of the space plasma environment is distribution functions to understand, for the first time, a critical component of the LANL nuclear detonation de- the physical processes that drive the complex plasma tection nonproliferation mission. LANL has expanded its dynamics of the solar wind and the terrestrial magneto- national security mission into space situational aware- sphere. This breakthrough additionally enables robust ness, which entails understanding and predicting the plasma measurements on 3-axis stabilized spacecraft space environment and its impact on the national space and smallsats, which is critical for the Lab’s future infrastructure. We anticipate that the 2PS instrument national security payloads as well as future scientific concept may also form the foundation for future genera- missions. tions of national security payloads. For 50 years, Los Alamos has flown plasma spectrom- The low-resource 2PS spectrometer could also form the eters for national security payloads as well as NASA basis of a solar wind monitoring instrument for space scientific missions to study solar wind and planetary weather forecasting purposes. While space weather magnetospheres such as Earth and Saturn. The Lab monitoring falls under the purview of NOAA, it is of has expanded its national security mission into space interest to many agencies (e.g. DOC, NASA, NSF, DoD, situational awareness, which entails understanding and DOE, etc.) who together have formed the National Space predicting the space environment and its impact on the Weather Program (NSWP) to coordinate efforts in this national space infrastructure. The 2PS Spectrometer will area. enable Los Alamos to continue to lead studies of space plasmas and the space environment, and will position Progress the Lab as an even stronger “go-to” institution for such measurements. We anticipate that this concept may Work on the 2-pi Plasma Spectrometer (2PS) project also be the foundation for future generations of national in FY2013 involved two components: simulation of the security payloads. instrument concept, and development and testing of 680

Page 681

prototype hardware in the laboratory. ate particles from ~1-60 keV, an appropriate range for this instrument. Experimental work began by investigating round holes drilled in a prototype filter plate. The use of round holes Initial test results do indeed validate the instrument rather than stacked plates with cylindrically symmetric slits concept, with particles for a given hole observed at the is significantly easier to fabricate, and thus allows us to expected central energy and incoming angle. However, more quickly validate the instrument concept. We tested there are interesting differences between the prototype both drilled holes and tapped holes, where the tapped instrument and the simulated design, in particular in the holes are intended to prevent particles from specular re- energy resolution of the instrument, where the labora- flection into the detector. Results of these studies using H, tory prototype shows significantly better results than He, and O ions showed that tapping the holes does in fact predicted. With the instrument tuned for detection of reduce scattering by nearly two orders of magnitude. 10 keV ions, we have input ions over a range of 5-60 keV, and find an energy response ∆E/E that is factor of 2 better In a parallel effort, we investigated suitable detectors for than predicted. We are investigating whether this is due to the 2PS spectrometer. For our initial prototype tests, we fringing fields, construction tolerances, limitations of the determined that we could use the existing imaging micro- model, or some other cause. Results of these studies will channel plate (MCP) system in our test laboratory. While inform the second-generation prototype instrument, to be this MCP is too small for a full 2PS instrument, it is suitable constructed during the second year of this project. In par- for the initial prototype testing. Our research this year ticular, we are very interested in determining the reasons shows that 10x10 cm MCPs are available, and that existing why the laboratory prototype has a better energy response electronics could be modified to use these MCPs with a than predicted by the model so that we can exploit these crossed-delay-line (XDL) anode. In addition, collaborator features in future versions of the instrument. Eberhard Moebius has MCPs from a previous project that may be suitable for a next-generation prototype instru- Future Work ment. As the existing XDL sensors exceed 2PS require- Work during the first year has validated the 2-pi Plasma ments, no technology development is required for this Spectrometer (2PS) instrument concept. During the subsystem. We anticipate using this more sophisticated second year of this three-year project, we will continue detectors in future prototype development. optimization of the 2PS spectrometer. We will begin by completing our studies of the initial prototype, and un- We ran extensive simulations using the SIMION software derstanding the differences between this design and the package in preparation for building a prototype detec- simulations. As fringing fields may be (positively) affect- tor. These simulations included modeling of the original ing the prototype response, we will analyze the effects of symmetric instrument concept, as well as simulations fringe fields around the aperture and filter plate channel of a simplified model based on round holes rather than openings. We will perform further simulations to optimize cylindrically symmetric slits. Simulation of the hole-based the instrument performance, including variation of the instrument required a 3D SIMION model, as this instru- filter plate channel length-to-width ratios, ratio of channel ment lacks the cylindrical symmetry of the original. Mod- diameters to the distance from the entrance aperture, and eling showed that the energy and angle response of the entrance aperture diameter. The results of these laborato- hole-based instrument is very similar to the full instrument ry and simulation studies will inform further development concept, and thus that this simplified design is suitable for of the 2PS instrument. We expect to fabricate and test proof-of-concept experiments. Further modeling efforts a second preliminary prototype instrument during year focused on simulating details of the prototype design to two. It is likely that this instrument can also use laboratory allow the instrument to be more easily fabricated. detectors, but we will evaluate the need for XDL hardware at this stage. In addition, we will begin design work for We then constructed and began testing an initial proto- a final prototype instrument. We will ramp up efforts to type instrument based on round holes. This instrument design a stacked-grid filter plate, including simulations and includes 11 holes, for detection of particles at 11 different preliminary studies. This work will be led by collaborator entrance angles. As drilling holes at angles is difficult, the Eberhard Moebius from the University of New Hampshire, design uses appropriately sized counter-sunk holes at the and is expected to involve a student participant. We will entrance and exit of the filter plate, with a larger straight determine detector and electronics needs for this com- hole connecting these. The instrument was fabricated and plete prototype, and then determine available and lead installed in a vacuum chamber for testing with an ion beam times for acquisition of these parts. We expect that by source. As noted above, an existing imaging MCP system is the end of the second year we will be ready to begin final being used as a detector. The ion beam source can gener- 681

Page 682

design fabrication of a full prototype instrument. Conclusion The 2PS plasma spectrometer acquires a complete measurement with a single high voltage sweep, allow- ing measurement of a complete distribution function in less than 0.1 sec. These fast measurements will enable study of ultrafast dynamic processes in the solar wind and magneto-spheric plasma environments. The unique design enables a dramatic reduction of resources (mass, power, size, and cost) compared to existing designs. We expect that the mass and power resources will be less than 1 kg and 2 W for one instrument, suitable for use in either cubesat or traditional satellite applications, as well as on 3-axis stabilized spacecraft. 682

Page 683

Physics Exploratory Research Continuing Project Magnetic Nanomarker Detection and Imaging with SQUIDs Andrei N. Matlashov 20130624ER Introduction one order of magnitude better than any presently avail- We propose a revolutionary new technology and in- able methods suitable for routine and/or repetitive use. struments for ultra-sensitive magnetic detection and This project illustrates LANL’s unique inter-disciplinary imaging of tissue cells with a detection limit of below ability to develop new instrumentation and measure- 10^4. By comparison, conventional x-ray mammography ment technologies. requires over 10^8 cancerous cells for detection. Our Two key components are present at LANL – 1) Ultra-high method will consist of targeting cells using antibody resolution SQUID instrumentation; and 2) Magnetic labeled single-core superparamagnetic nanoparticles, nanoparticle synthesis and purification, surface modifi- followed by detection and imaging of the targeted cation, and specific bioconjugation. area using high-resolution Superconducting Quantum Interference Device (SQUID) gradiometers. Super- The combination with imaging provides a truly unique paramagnetic relaxometry (SPMR) is used for detection approach to address both the presence and location of of targeted cells with very high specificity: Only bonded labels. The long-term relevance to national security sci- (immobilized) nanoparticles will be detected via their ence is both in global health (as an early cancer diagnos- Néel relaxation. The bonding will occur only with can- tic), but also global security (the ability to rapidly detect cer cells because of specific antibodies conjugated to exposure to pathogens for example, or small quantities the nanoparticle surface. Unbound nanoparticles will of molecules in the environment) and in the capabil- not contribute in the SPMR decay signal. By combining ity to provide rare-molecule detection. Success of this SPMR with ultra-low field magnetic resonance imaging project will lead to further development of the method (ULF MRI), using the same instrument, the targeted area using funds from different agencies such as NIH or, for will be imaged to provide anatomical information. The instance, from DoD Congressionally Directed Medi- same magnetic particles will also work as MRI contrast cal Research Programs such as Breast Cancer Research agents, as they can be detected due to their influence Program. on the relaxation rates of 1H nuclear spins, which are an abundant signal source due to the high water content Progress of soft tissue. The combination of ULF MRI and SPMR During the current financial year the project develop- will provide both accurate localization and cell count of ment moved in three parallel directions. First, the exist- the targeted tissue. SQUIDs enable detection via gradi- ing 7-channel SQUID-based system had been modified ometers with unprecedented sensitivity. This approach and upgraded. It allowed use for both ultra-low field will provide a robust diagnostic tool for detection and magnetic resonance imaging (ULF MRI) and super- localization of diseased (e.g., cancerous) tissue targeted paramagnetic relaxation (SPMR) measurements using with magnetic markers at a very early disease stage. This same phantoms. Second, 25-30 nm diameter super-para- technique can be used as an in vivo (inside body) diag- magnetic nano-particles immobilized in agarose have nostic. ULF MRI and SPMR measurements have never been used to make phantoms with different shapes and before been combined in a single device. particles concentration. Third, a new 7-channel system with three additional reference channels was designed Benefit to National Security Missions and currently is about 75% complete. This system will be We propose here a revolutionary advance in the ability used for combined ULF MRI and SPMR measurements to detect and image very small numbers of labeled cells in unshielded environment using phantoms. Potentially or other microscopic objects, with the resolution at least 683

Page 684

this system can be also used with small animal models or We demonstrated the first time ever the possibility of com- human tissue samples in medical research facilities. bining magnetic relaxometry and ULF MRI in a single in- strument. An image showing the influence of the nanopar- The existing 7-channel system consists of axial second- ticles as a contrast agent was obtained, and a plausible order gradiometers 37 mm diameter and 60 mm baseline. fit for the location and strength of a magnetic dipole was They are positioned in parallel one in the middle and six obtained by magnetic relaxometry. Future work will focus others surrounding it in a hexagonal pattern with 45 mm on using the MRI to constrain multiple dipole fitting. The separation between the axes. ULF MRI was performed combination of MRI with magnetic relaxometry will clearly using field-cycling and spin-echo protocol. A 2D Fourier improve our ability to accurately estimate and localize imaging protocol was used with a frequency and phase dipoles, with direct impact on the efficacy of the technique encoding gradients. The resulting voxel size was 3×3 sq. as a sensitive cancer diagnostic tool. mm with 160 mm diameter field of view. The 400-turn pre-polarization coil was cooled by liquid nitrogen. It was Future Work placed co-axial with the central gradiometer about 100 During the first year of the project we decided to forgo ULF mm below its bottom pick-up coil. This coil generated a 50 MRI experiments using a single channel system and use a mT pre-polarization field for MRI measurements and 5.5 modified 7-channel system. This approach allowed us to mT magnetizing field for SPMR measurements. demonstrate the first time ever the possibility of combin- ing magnetic relaxometry and ULF MRI in a single instru- Recording of the magnetic field relaxation signal started ment. This system shows about 3 mm MRI spatial resolu- about 12 ms after the magnetizing field was zeroed. The tion with 160 mm field of view. In the next fiscal year we relaxation signal from nanoparticles was masked by large will improve its spatial resolution by a factor of 2 and also transient signal. It was possible to suppress the transient improve its signal-to-noise ratio by upgrading the fields by factor about 5 using a compensation coil. After that a generation coil system. This system will be used inside a baseline was recorded without a phantom and subtracted magnetically shielded room (MSR). from the signal recorded with a phantom in place. This difference reveals the relaxation signal primarily from the We will demonstrate interleaved ULF MRI and SPMR using nanoparticles. Raw relaxation signals were fitted using a a single-dipole phantom that will allow us to compare the logarithmic function in the area of slow signal decay and accuracy of an inverse problem solution with and without a 5th-order polynomial fit of the early relaxation curve for MRI data. We will then do localization procedures using extrapolation to time zero. A dipole approximation was multi-dipole phantoms and distributed source phantoms to used for single vial localization and the magnetic moment solve an inverse problem that would be impossible with- was estimated using an inverse problem solution. out MRI data. We will investigate resolution limits for both SPMR and ULF MRI measurements. The phantoms were prepared using vials (with nano-par- ticles inside) placed in a large dish of water. Agarose with Different type of phantoms will be made. For biomarker uniformly embedded iron oxide nanoparticles at about development we will otimize the multifunctional coating, 7×10E+11 per 1 ml was used to fill the vials. Such amount functionalize the surface with amines or carboxylic acids of nano-particles corresponds to about one million of and attach antibodies to the surfaces using one of sev- tagged cells or 1 mm size tumor. We assumed that agarose eral robust and well-known techniques, such as EDC/NHS keeps some fraction of nanoparticles immobilized allowing coupling. only Néel relaxation in the case of magnetic relaxometry. In the case of ULF MRI, the same nanoparticles work as a Unshielded measurements: a new multi-channel system contrast agent. will be built and tested for experiments in unshielded environment. We performed the first ever combined ULF MRI and SPMR measurements. Nanoparticles efficacy as a contrast agent Conclusion was clearly demonstrated. Small vial with 1 ml nanoparti- We will build instruments and demonstrate a proof of cles was used for SPMR measurements. It was successfully principle that magnetic nanomarker detection and imag- localized using an inverse problem solution and a single ing with SQUID-based gradiometers can lead to a creation dipole approximation. These preliminary results have been of revolutionary new cancer diagnostic methods at a very submitted to International Superconducting Electronics early disease stage. We will 1) demonstrate interleaved Conference (ISEC 2013) and chosen as an invited oral pre- ULF MRI and SPMR measurements using a single device; sentation. It was also invited as a peer-reviewed publica- 2) reach 2×2×4 mm3 ULF MRI system spatial resolution; tion in IEEE Xplore. 3) reach SPMR detection limit of below 10,000 cells; 4) 684

Page 685

design and build a multichannel system capable of doing measurements in unshielded environment; 5) develop de- sired optimized superparamagnetic nano-markers; and 6) experimentally demonstrate performance of the method using phantoms and mice. Publications Magnelind, P., Y. J. Kim, A. Matlashov, S. Newman, P. Vo- legov, and M. Espy. Superparamagnetic relaxometry for early cancer detection using a low-Tc SQUID-gradiom- eter array. To appear in Eleventh European Conference on Applied Superconductivity. (Genova, Italy, 15 - 19 Sep. 2013). Matlashov, A., P. Magnelind , H. Sandin, M. Espy , A. An- derson , and H. Mukundan. SQUID instrumentation for early cancer diagnostics: combining SQUID-based ultra-low field MRI and superparamagnetic relaxation . 2013. In Fourteenth International Superconducting Electronics Conference . (Cambridge, 7 - 11 July 2013). , p. 231 . Cambridge: IEEE Xplore. Matlashov, A., P. Magnelind, Y. J. Kim, H. Sandin, M. Espy, A. Anderson , and H. Mukundan. SQUID-based ULF MRI and superparamagnetic relaxometry for early can- cer diagnostics. 2013. Superconductivity News Forum (Global Edition). (25 ): ST 342. 685

Page 686

Physics Exploratory Research Continuing Project Beyond the Standard Halo Michael S. Warren 20130626ER Introduction ery would have profound consequences. To correctly It is widely anticipated that dark matter will be detected interpret these results, it is crucial that the theoretical in the upcoming generation of solid state and noble liq- uncertainties be well understood. uid direct recoil experiments. To correctly interpret these Modern N-body simulation codes provide the theoretical results — whether positive or negative — and to compare basis for our present understanding of the mass distri- them to other indirect searches (e.g., with cosmic rays bution in the Universe, and are an essential link in the or LHC), it is crucial that the theoretical uncertainties chain which connects particle physics to cosmology. The entering into the predicted scattering rates be reduced aim of our research is to understand in exquisite detail as much as possible. The dominant uncertainty in this the distribution of dark matter within halos, and to con- calculation is the small-scale distribution of dark mat- nect this knowledge with the growing suite of astrophys- ter. Predictions for direct detection experiments are ical and particle physics observations. To achieve this, we often computed with the assumption that our local dark will perform the world’s highest-resolution N-body simu- matter distribution can be treated as a “standard halo,’’ lations with our Hashed Oct-tree (HOT) N-body code and a smooth isothermal sphere. However, observations and our recent 40M CPU-hour per year allocation on the new numerical simulations suggest that this assumption is 38,000 processor LANL Mustang supercomputer. invalid. The software developed to perform the simulations and Although we have compelling measurements of the analyze the data during this project will enhance our large-scale properties of our galaxy’s dark matter halo, capability for modeling and high-performance com- we have no direct information on how this dark matter puting, as well as serve as a proving ground for high- is distributed on small scales. The goal of our research performance computers and data-mining technology. is to understand the small-scale distribution of dark The project will demonstrate advanced computational matter within halos, and apply this knowledge to the techniques which are applicable to a variety of difficult interpretation of experimental and observational results. fundamental problems in hydrodynamics, materials sci- To achieve this, we will perform the world’s highest- ence, biology and chemistry. resolution N-body simulations using our 40M CPU-hour per year allocation on the new LANL Mustang supercom- Progress puter. This combination of observations, theoretical ad- The number of dark matter halos in the Universe of a vances, and numerical simulations provides an exciting given mass is a fundamental statistic called the mass opportunity to identify dark matter, with implications for function. The mass function is sensitive to cosmological some of the most important questions in science, such parameters such as the matter density, the initial power as how galaxies form and what determines the proper- spectrum of density fluctuations, and the dark energy ties of fundamental particles. equation of state. For these reasons, the mass function is a major target of current observational programs, and Benefit to National Security Missions provides a basic constraint on the properties of indi- Understanding the nature of the dark matter is among vidual halos. Precisely modeling the mass function is the most important unsolved problems in physics. An an enormous challenge for numerical simulations, since immense amount of theoretical, observational, and both statistical and systematic errors conspire to prevent experimental activity is underway in anticipation of the emergence of an accurate theoretical model. The the dark matter particle being unveiled. Such a discov- 686

Page 687

dynamic range in mass and convergence tests necessary to compensates for the lower forces in the interior and serves model systematic errors require multiple simulations at dif- to reduce the bias in the force calculation. Our tests con- ferent resolutions, since even a very large simulation does firm these conclusions, and we use Dehnen’s K1 compen- not have sufficient statistical power by itself. sating kernel for our simulations. We provide the first mass function calculated from a simu- We published a paper in the Proceedings of Supercom- lation using the new standard Planck 2013 cosmology, with puting ‘13 where we report on improvements made over a 4096^3 particle simulation and six 2048^3 simulations the past two decades to our adaptive treecode N-body completed and shared with our collaborators within 30 method (HOT). A mathematical and computational ap- days of the publication of the Planck 2013 results. Changes proach to the cosmological N-body problem is described, in the parameters from the previous WMAP7 model are with performance and scalability measured up to 256k large enough that extrapolations from the other cosmolo- (2^18) processors. We present error analysis and scientific gies are likely subject to systematic errors which are large application results from a series of more than ten 69 billion compared to the statistical precision of our results. (4096^3) particle cosmological simulations, accounting for 4 x 10^20 floating point operations. These results include We find that Tinker (2008) underestimates the mass func- the first simulations using the new constraints on the stan- tion at scales of 10^15 solar masses by about 5% when dard model of cosmology from the Planck satellite. Our compared with the older cosmological model it was cali- simulations set a new standard for accuracy and scientific brated against. For the Planck 2013 cosmology, the Tin- throughput, while meeting or exceeding the computational ker08 mass function is 10% low at large scales, due to the efficiency of the latest generation of hybrid TreePM N-body added systematic effect of non-universality in the underly- methods. ing theoretical model. Future Work We identify a previously unidentified systematic er- Our simulation campaign is enabled by our large alloca- ror stemming from the improper growth of modes near tion of Institutional Computing time, and the extraordinary the Nyquist frequency, due to the discrete representa- performance of our N-body code. In the past, numerical tion of the continuous Fourier modes in the ideal input cosmology groups have performed a single world-class power spectrum with a fixed number of particles. This simulation (in the range of 1-5M CPU hours) every few is a resolution-dependent effect which is most appar- years. With our present resources, we can perform a ent when using particle masses larger than 10^11 solar world-class simulation every month. Multiple simulations masses(corresponding to using less than 1 particle per provide an invaluable check on the statistical and conver- cubic Mpc/h). Uncertainty in the appropriate correction gence properties of our simulations. and consequences of this effect appear to be the dominant source of systematic error in our results. If uncontrolled, We have previously explored cosmologically relevant this discretization error confounds convergence tests volumes (300 Mpc or more with particle masses greater which attempt to isolate the effects of the starting redshift than 10^8 solar masses. In FY2014 we will progress to of the simulation, since the error becomes larger at higher higher resolution using a combination of re-simulations at starting redshifts. higher resolution of interesting objects we have identified in current simulations, as well as performing new simula- We are in direct conflict with recent results (Watson, 2012) tions with hierarchical initial conditions (where bound- which find the spherical overdensity mass function to be ary conditions are provided by more massive particles, lower than the the Tinker08 result at high masses. Poten- concentrating the computational effort on high-resolution tial explanations would be insufficient force accuracy of sub-regions). Our work in this area will be further enabled the CUBEP3M code, with a secondary contribution from through the use of PANPHASIA initial conditions based on initial conditions that did not use 2LPT corrections. the work of Jenkins (2013). We expect to perform 6-10 simulations with ~ 100 billion particles and mass resolu- The standard practice in cosmological N-body simula- tion between 10^4 and 10^8 solar masses during FY2014. tions is to smooth the forces at small scales, usually with We submitted an INCITE computing proposal during the a Plummer or spline kernel. We have implemented these summer, which if successful will allow us to capstone our smoothing kernels in our code as well as the additional simulations with a trillion particle run using the Titan su- kernels described by Dehnen (2001). Dehnen concludes percomputer at ORNL during calendar year 2014. that the optimal softening method uses a compensating kernel, with forces that are higher than the Newtonian We will address what our simulation results imply for the force at the outer edge of the smoothing kernel, which sensitivity of the experiments. The vast majority of sensi- 687

Page 688

tivity projections are currently performed with extremely restrictive simplifying assumptions. We will avoid these assumptions, and translate the results of current experi- ments into specific predictions. Conclusion We intend to simulate the evolution of matter and to resolve dark matter structures over a spatial range of a million. We will pursue both the dissipationless aspects of structure formation (i.e. dark matter only’) and consider baryons — the gas which can shock and dissipate energy, form stars, and generally modify the dynamical evolution of the dark matter on small scales — with improvements to the smoothed particle hydrodynamic (SPH) component of our code. Connecting fundamental cosmological theories with observational dark matter data can only be accom- plished with numerical experiments such as those we propose here. Publications Hamaus, N., B. D. Wandelt, P. Sutter, G. Lavaux, and M. S. Warren. Cosmology with Void-Galaxy Correlations. 2013. arXiv preprint arXiv:1307.2571. Song, J., J. J. Mohr, W. A. Barkhouse, M. S. Warren, K. Dolag, and C. Rude. A Parameterized Galaxy Catalog Simulator for Testing Cluster Finding, Mass Estimation, and Photometric Redshift Estimation in Optical and Near-infrared Surveys. 2012. The Astrophysical Journal. 747 (1): 58. Sutter, P., G. Lavaux, B. D. Wandelt, N. Hamaus, D. H. Wein- berg, and M. S. Warren. Sparse sampling, galaxy bias, and voids. 2013. arXiv preprint arXiv:1309.5087. Warren, M. S.. 2HOT: An Improved Parallel Hashed Oct- Tree N-Body Algorithm for Cosmological Simulation. 2013. arXiv preprint arXiv:1310.4502. Warren, M. S., and B. Bergen. Poster: The Hashed Oct-Tree N-Body Algorithm at a Petaflop. 2012. In High Perfor- mance Computing, Networking, Storage and Analysis (SCC), 2012 SC Companion:. , p. 1442. 688

Page 689

Physics Exploratory Research Continuing Project Coherent Diffractive Imaging of Ultrafast Ejecta Processes Cynthia A. Bolme 20130632ER Introduction Progress This project will use an ultrafast laser to create a shock Efforts in FY13 have included experimental design, lab wave at a metal sample surface. The metal sample will setup, equipment procurement, and code development. have a periodically roughened surface, and the interac- We reiterate here that only a few similar systems exist in tion of the shock wave with the structured surface will the world (none to our knowledge in the western hemi- create instabilities in the material that will evolve in sphere) that can produce enough coherent, soft x-ray small jets of ejecta particles. This process will be imaged flux for single shot coherent diffractive imaging (CDI). At using ultrafast soft x-rays that are created by the same the time of writing this proposal, about half way through laser system that drives the shock. Using coherent soft the fiscal year, we feel we are on track to make our mile- x-rays, we will provide resolution of the ejecta formation stone for this FY of having the soft x-ray source up and process that is 20 times better than has been previously running and initial attempts at static single shot imaging. performed. This FY, we have focused our efforts on setting up the We will also perform molecular dynamic simulations of tabletop, single shot high harmonic generation system. this ejecta forming process that are perfectly matched in The laboratory that housed the laser system proposed length and time scales to the experiments. By compar- for this project had not been used in many years, so ing the simulation with the experiments, we will gain much of our effort in the first months has been to get new insight into instability formation and evolution. By the lab ready, update the laser interlocks, and upgrade comparison of these results with previous results of the laser system. LUMOS team postdocs and students larger scale ejecta formation, we will learn about the im- from WX-9 and LUMOS also played valuable contribu- portance of time and length scales on instability growth. tions in these efforts. This effort also included a full upgrade of the first stage laser amplifier. Benefit to National Security Missions Concurrently with these efforts, we have performed a The development of this tabletop source will immedi- thorough design of the optical relay system, soft x-ray ately address core mission problems and provide impor- generation, and experimental geometry. These efforts tant validation of physics-based material models as the have included conducting literature searches for similar spatial and temporal resolutions achievable with this im- laser development and soft x-ray generation efforts. Ad- aging diagnostic will be orders of magnitude better than ditionally, numerical calculations and optical modeling current dynamic diagnostic alternatives. This diagnostic were performed to try to reduce risk and ensure a suc- ability was specific called out in the report from the Re- cessful experimental design. Two specific design chal- search Needs for Material Mixing at Extremes workshop lenges that were overcome had to do with the optical stating, “developing diagnostics for HED flows that will design. First, there was a challenge in properly designing allow us to penetrate the mixing regions with greater the up to 7 meter long focus of the intense (up to 100 resolution than ever before achieved.” This project will mJ per pulse, 45 fs) near-infrafred (IR) laser pulses into also greatly advance the state-of-the-art in dynamic ma- a long, low pressure gas cell for optimal phase matching terials imaging using coherent x-ray diffractive imaging, a conditions. This design included folding the beam back key diagnostic proposed in the Matter-Radiation Interac- on itself inside an evacuated chamber to avoid nonlin- tions in Extremes (MaRIE) signature facility. ear effects in air such as self focusing. Achieving this phase matching condition is critical in order to ensure 689

Page 690

the brightest and most coherent beam for CDI. A second the shock properties. These efforts may include perform- design aspect that was challenging given somewhat limited ing diagnostics studies on the shock drive beam and on space in the laboratory was in separating the collinear soft shock breakout on the metal surface. The modeling effort x-ray and mid-IR light. Our design incorporates a beam will attempt to secure computational resources to begin splitter that will allow us to remove most of the unwanted the simulation of 3-dimensional surfaces, i.e. sinusoidal IR light that needed to be placed sufficiently downstream pits or wells. These structures will attempt to replicate that the IR pump light would not damage it. the surface conditions present on the experimental set-up that we propose. Additional analysis of the simulations In addition to experimental design and cleaning, we also performed in FY13 will be done to correlate with the exist- performed an IWM safety review and wrote and reviewed ing solid and liquid break-up models. The experimental the IWD documentation including laser safety. We are cur- dynamic images will allow us a direct comparison with the rently beginning to set up the coherent soft x-ray genera- 3D MD models and will allow us to refine them. tion system and vacuum chamber and are planning the soft x-ray optical we will perform static imaging experiments Conclusion of surfaces in vacuum. Another large effort in these first This project’s goals are to image ultrafast laser shock months has been the identification and purchasing of nec- driven ejecta particles and to compare the imaging data essary equipment including vacuum components, optics, with molecular dynamics simulations. This project will soft x-ray detectors, and in vacuum motion control. We demonstrate the first single shot coherent x-ray diffraction have currently received several vacuum components and imaging of a dynamic process and will validate molecular the soft x-ray detector and have purchases in the works for dynamics simulations. Additionally, we will provide an in vacuum motion controllers and optics. important new experimental capability, x-ray imaging of dynamic events, and will create data on instability forma- This FY, we have also been further developing and refin- tion with unprecedented time and length scales. ing our 2-dimensional molecular dynamics simulations of ejecta in Al, with a specific focus on the process of evolu- tion of the instability and break up of the material into par- ticulates. MD simulations of ejecta formation have been performed on solid state copper looking at the effect of surface morphology in a quasi-2D sense. Additional work has been done in the selection of an Al potential. We have observed that the surface morphology plays a significant role in the resulting structures formed. As an example a pure sinusoidal wave with a kh0=1.0, we observe in copper a spike height of 57.3 nm whereas other surfaces with the same kh0 range from 151 nm to 15 nm. Analysis is under- way to determine how well equation (3) of Ref [Dimonte, et. al. PRL, 107, 264502 (2011)] holds or if it needs to be modified to capture the surface morphology effect. Future Work After successful setup of the tabletop soft x-ray system and initially performing single shot coherent diffractive imaging (CDI) of static objects if FY13, efforts in FY14 will focus on moving towards actually performing the pump-probe dy- namical imaging of the shock breakout on the roughened metal surface. We will optimize static imaging with the co- herent soft x-ray source for brightness, stability, and image resolution, and we will incorporate the single shot imaging with ultrafast laser shocked samples. Samples for these studies will be lithographically patterned samples for reso- lution determination and in order to determine the beam characteristics on a shot to shot basis. Further efforts will include optimizing the laser shock drive and characterizing 690

Page 691

Physics Exploratory Research Continuing Project In Search of Light WIMPs Alexander Friedland 20130637ER Introduction ties of the DOE Office of Science. The aim of this project Astrophysical and cosmological observations over the is to enhance the capabilities of the Majorana detector last two decades have proven that ordinary matter con- to make it into the world’s best probe of dark matter stitutes only a few percent of the energy density of our in the light mass window. Additional strategic impact Universe. Most of the energy density is contained in dark derives form the capability of the Majorana detector matter and dark energy. While the dark matter is know in low-background counting. Improving this capability to hold the galaxies together, its physical nature remains by going to lower energies will enable to us to study a a mystery. Laboratory experiments aiming to detect the wider range of samples. We presently count samples dark matter particle form one of the cornerstones of the for programs in HEP and NP that require fabrication of worldwide campaign to discover new physics beyond the experimental apparatus from radio-pure parts. We have Standard Model. The search for dark matter is one of the also counted a sample for NIF-related efforts. stated priorities of the DOE Office of Science. Progress Traditionally, direct detection experiments have been We have carried out our first study of the capabilities of guided by the so-called WIMP (Weakly Interacting Mas- the Majorana Demonstrator as a light WIMP detector. sive Particle) paradigm. In this paradigm, one expects We have shown that, for a cross section near the current the dark matter particle to be in the mass range of experimental bound, the Majorana Demonstrator should 100-1000 times the mass of the proton. Yet, in the past collect hundreds or even thousands of recoil events. This few years, a contradictory picture has emerged that ap- opens up the possibility of simultaneously determining pears incompatible with the minimal WIMP framework. the physical properties of the dark matter and its local Several experiments, specifically CoGeNT, CRESST-II, and velocity distribution, directly from the data. We have DAMA, have claimed evidence for dark matter, whereas further investigated this possibility and found that allow- CDMS, XENON10, and XENON100 have reported null ing the dark matter velocity distribution to float consid- results. To untangle this paradox urgently requires simul- erably worsens the WIMP mass determination. We have taneous advances in both theory and experiment. traced this result to a previously unexplored degeneracy between the WIMP mass and the velocity dispersion. We will attack the physics of dark matter on two fronts: This establishes that knowledge of the dark matter by scrutinizing theoretical models that go beyond the velocity distribution through numerical simulations is minimal WIMP framework and by performing a low- essential for the mass and cross section determination threshold search for dark matter using the Majorana of a light WIMP. Demonstrator experiment. The experiment will be car- ried out in the timeframe of this proposal and will result The paper with these results has just been accepted to in the world’s best sensitivity in the light dark matter Physics Letters B (vol 724, pp 183-191, 2013). window. The theory effort will analyze the experimental results in the broad context of existing searches and will Future Work translate them into the physical properties of dark mat- The prototype module for the MAJORANA DEMON- ter in various scenarios of new physics. STRATOR is presently (July 2013) being populated with detectors in our underground laboratory at Sanford Un- Benefit to National Security Missions derground Research Facility (SURF). This module should The search for dark matter is one of the leading priori- be commissioned within a couple months and will then 691

Page 692

operated for science runs. Data from the runs of that mod- ule will form the basis for determining the sensitivity of the DEMONSTRATOR to dark matter signals. We will be analyz- ing data with that question in mind. The LANL team will also be looking at data from a test bed in our laboratory at WIPP to develop analysis tools for studying the low energy spectrum. The detectors have been running there for many months and have provided that opportunity. On the theoretical side, we will build on our initial study (carried out and published in FY13) of the dependence of the Majorana sensitivity on the dark matter velocity distribution. We will incorporate the results of the numeri- cal N-body simulations to obtain the physical uncertainty on the mass and cross section determination. We will also explore models in which the mediator particle generating the WIMP-quark interactions is heavier than 1 GeV, but leads to momentum and velocity-dependent WIMP-quark interactions. Separately, we will study dark matter a mass in the ~10 MeV to 100 MeV range. We will also analyze the expected data from the running direct and indirect detection experiments and from the Large Hadron Collider. Lastly, we will study the reach of the IceCUBE experiment, with special attention paid to the oft-neglected prompt two-neutrino dark matter annihilation mode. Conclusion The project aims to carry out the world’s most sensi- tive search for dark matter in the mass window of 5-100 proton masses, using the Majorana detector. It also plans to develop a theoretical framework for understanding the particle physics nature of dark matter, given the latest data from other direct detection experiments, LHC, IceCUBE, Fermi, and other relevant searches. The experimental and theoretical parts will be tightly integrated and the analysis of data will be done jointly. Publications Friedland, A., and M. Shoemaker. Integrating In Dark Mat- ter Astrophysics at Direct Detection Experiments. 2013. Physics Letters B. 724: 183. Giovanetti, G. K.. A Dark Matter Search with The MAJORA- NA Low-Background Broad Energy Germanium Detec- tor. Invited presentation at TAUP 2013. (Asilomar, CA, USA, 9 Sep. 2013). 692

Page 693

Physics Exploratory Research Continuing Project Electron Capture Spectroscopy for Neutrino Mass: Isotopes, Science, and Technology Development Michael W. Rabin 20130679ER Introduction capture spectroscopy technique as a route to determin- Measuring neutrino mass is an extremely challenging ing neutrino mass. As the techniques and capabilities major science goal with intense current interest. Labo- improve, we anticipate this effort will be highly competi- ratory measurements of the electron energy spectrum tive for a major initiative for a large scale experiment for from tritium beta decay set an upper bound for the neutrino mass measurement. For nuclear forensics, we electron antineutrino mass of ~2.2 eV (95% CL), which is are already fully engaged with US government and inter- only about four parts-per-million of the electron mass. national community. The encapsulation of radionuclides This conventional measurement method will soon reach directly into our sensors is a potential new technique for its technological zenith in the enormous KATRIN experi- these nuclear materials analysis with special advantages. ment with a mass sensitivity of ~0.2 eV (by about 2019), Progress but with no hope of significant improvement. Alterna- tive methods are therefore necessary to confirm, refute, We have made progress in three specific areas: isotope refine, or surpass the expected results of this ultimate production, chemical purification, and measurement Goliath of electromagnetic spectrometers. But no known with microcalorimeter sensors. Isotope production for method can do the job today; there are no other suf- this project is done through nuclear transmutation. We ficiently sensitive, technologically mature methods for have investigated the production based on multiple laboratory-based, model-independent neutrino mass methods of nuclear transmutation, principally neutron measurement. Our goal is to develop specific, key sci- and proton irradiation. Preliminary modeling and calcu- ence capabilities central and necessary for one of the lation in this area has been completed. Analysis suggests proposed methods: electron capture spectroscopy that neutron irradiation of enriched erbium has a higher (ECS) of the unusual, rare and low decay-energy isotope production rate, though at cost in isotopic purity of the 163Ho. To determine the neutrino mass, the ECS method intermediate product. A manuscript covering this area requires rare isotope production, radiochemical purifica- has been prepared and recently accepted for publication tion, encapsulation of the Ho inside ultra high resolution in Nuclear Instruments and Methods. Related material cryogenic microcalorimeters (with Delta(E)1-2 eV at was presented at the NuMass workshop where key rep- E3 keV), and spectral analysis near the reaction end- resentatives of each of the significant research programs point. To achieve ~0.1 eV mass sensitivity, a large-scale, working on Ho-163 isotope production participated. This highly demanding ECS experiment will need significant work has also addressed the purity requirements relative method development, sensor arrays with ~10,000 pix- to both short-term method and measurement technique els, array-compatible readout, and years of operation. development and long-term neutrino mass measure- Though ~0.1 eV mass sensitivity is our long-term ambi- ment. Chemical purification based on high-pressure (or tion, it is well beyond the scope of the present project, high-performance) liquid chromatography has also made and we will limit ourselves to developing key methods progress. The development of HPLC has focused on the necessary for ECS success. chemical separation of target holmium from proton- irradiated dysprosium parent material. Small scale HPLC Benefit to National Security Missions using milligram quantities of material has shown the first The development of these techniques is directly ap- successful results with nonradioactive materials. The key plicable to basic nuclear physics and nuclear forensics. result is temporal separation between holmium and dys- For basic nuclear physics, the key goal will be to build prosium in HPLC chromatograms, an essential first step. consensus within DOE OOS for this work electron Subsequent work is expected to demonstrate scaling up 693

Page 694

of this process and assessment of feasibility within a fume MENTS & METHODS IN PHYSICS RESEARCH SECTION hood. Related work on chemical separation has considered B-BEAM INTERACTIONS WITH MATERIALS AND ATOMS. potential for HPLC separation from neutron-irradiated 311: 131. erbium, but this path is not currently emphasized in the laboratory due to time order of element elution. The third area of progress is in measurement with surrogate isotope Fe-55. We have successfully gone from having zero sensors in this category to a series of progressively more successful experiments. The key objective is to demonstrate success- ful electron capture spectroscopy with an easily available material during the period when we have no purified Ho-163. We have conducted a new sensor microfabrication and introduced Fe-55 into this new sensor type. Reso- lution-limiting factors are being assessed now. Multiple methods for the incorporation of the electron-capture- decaying species have been investigated. Future Work We plan to produce high specific activity 163Ho utilizing the IPF at LANSCE. The Ho will be chemically recovered from the target material matrix, leveraging the extensive team expertise at the TA-48 hot cell facility. Purified Ho product will subsequently be converted into aqueous samples suitable for incorporation into microcalorimeters. Detector design, fabrication and testing will begin with our successful TESs, but with intent to optimize for the encapsulation of radioactive materials into very low heat capacity structures. For detector development in year 1, objectives will be demonstrate that we can achieve high resolution spectra and develop the chemical purification process under cold (nonradioactive) conditions. This will allow us to confidently execute a single proton irradiation and a single round of hot cell chemistry in year two. Conclusion We have three specific project goals. Make 163Ho at high rate with high isotopic and chemical purity. Encapsulate radionuclides, measure ECS with elta(E)~1-2 eV, at target rate 1000 cps. Use high resolution ECS to measure 163Ho decay energy (Q). An important feature of each of these goals is method scalability, which is what makes sub-eV mass sensitivity possible. A definitive measurement of Q directly addresses ultimate scale required. Combining these results creates an integrated 163Ho ECS capability directly useable with any thermal-type cryogenic detector, independent of thermometer choice, readout method, or multiplexing scheme. Publications Engle, J. W., E. R. Birnbaum, H. R. Trellue, K. D. John, M. W. Rabin, and F. M. Nortier. Evaluation of Ho-163 produc- tion options for neutrino mass measurements with microcalorimeter detectors. 2013. NUCLEAR INSTRU- 694

Page 695

Physics Exploratory Research Final Report Jet Probes of New Physics at RHIC and at the LHC Ivan M. Vitev 20110098ER Abstract Collider (RHIC) and at the Large Hadron Collider (LHC). High-energy colliders today test not only theories of To this end, we proposed to advance the theory of jets - particle and nuclear physics, but indeed, of the very collimated showers of energetic subatomic particles that first moments of the evolution of our Universe itself. are abundantly produced at these facilities when particle Experiments at the Large Hadron Collider (LHC) advance beams collide. RHIC and LHC explore a broad spectrum both of these frontiers by colliding protons and nuclei at of exciting novel phenomena that range from the forma- unprecedentedly high energies and thereby explore the tion of the quark-gluon plasma to the creation of new origin of mass, possibly the nature of dark matter, and forms of matter that are believed to make up the fabric the unexpected properties of new phases of matter, such of our Universe. Understanding the Quantum Chromo- as the strongly interacting quark-gluon plasma (QGP). dynamics (QCD) theory of jet production and propaga- The Relativistic Heavy Ion Collider (RHIC) continues tion in the vacuum and in matter having extremely high to probe the properties of the QGP at lower energies, energy density is essential to elucidating both the unex- complementing the LHC heavy-ion effort. This project pected properties of the QGP and the TeV-scale physics developed novel quantitative theoretical tools that that is Beyond the Standard Model (BSM). We proposed bridge the gap between particle and nuclear science and to identify experimental jet observables of new physics help decipher the flood of experimental data for signa- in proton-proton (p+p) and nucleus-nucleus (A+A) reac- tures of new physics. It addressed the topmost priority tions and evaluate them using modern techniques, such for both the particle and nuclear physics communities by as effective field theory. Our goal was to incorporate developing and advancing the theory of jets (collimated the theoretical advances in jet simulation packages and showers of energetic subatomic particles) in reactions employ the resulting tools to interpret the anticipated with protons and nuclei to a new level of precision and copious data from RHIC and the LHC. We expected that used them as probes of new physics at RHIC and at the our project would establish the correct physics mecha- LHC. The team published 6 refereed journal articles (in- nisms of jet-plasma interactions and help eliminate the cluding 4 Physical Review Letters and Physics Letters B) vast model dependence that currently hinders the pre- and two more are in preparation. Results from this work cise determination of the plasma properties. We would was presented in 9 invited conference and workshop identify search strategies for and/or place constraints on talks and summarized in 3 refereed conference pro- the Standard Model Higgs boson, as well as on models ceedings. The start-up of the LHC, the rapidly growing of supersymmetry, which predict massive new particles interest in jet physics, and the expertise of the research that decay into jets. The start-up of the LHC and the rap- team has ensured a significant scientific impact of this idly growing interest in jet physics would make it timely project. This work supported the Laboratory Mission to for this project to have a significant impact. enhance the nation’s scientific capabilities, as defined by Scientific Approach and Accomplishments the DOE Office of Science. Within LANL, our project was well-aligned with the Beyond the Standard Model (BSM) One of the main tasks of the LHC is to discover the Higgs Grand Challenge. boson and complete the Standard Model of particle physics. In 2012 the ATLAS and CMS experimental col- Background and Research Objectives laborations announced the discovery of a Higgs-like par- The overarching goal of this project was to develop mod- ticle with a mass around 125 GeV. The evidence for this ern theoretical and computational tools to identify and particle was found in two final states, namely di-photons characterize new physics at the Relativistic Heavy Ion and di-bosons. Since the time of the discovery these 695

Page 696

experiments have gathered more data and confirmed that can easily be contaminated by pile-up - debris produced the couplings of this particle to photons and di-bosons are from other proton-proton collisions (i.e., not from the one in good agreement with the predictions of the Standard producing the Higgs boson). While our analyses currently Model. But in the Standard Model the Higgs boson is also include methods to filter or prune away the particles pro- predicted to decay into pairs of fermions, such as a tau duced from pile-up, we still need to simulate the pile-up lepton pair or a pair of b-quarks. Here the evidence for in order to establish that our methods are robust against the Higgs boson decaying into these final states is weak, such effects. Pile-up can also affect other important ob- though still not in disagreement with the Standard Model servables, such as the missing energy (MET) of the event predictions. The reason for the disparity is that the discov- and whether a lepton is isolated. A detector simulation is ery channels for the Higgs boson – di-photons and di- primarily needed to model the effects of energy mismea- bosons decaying to leptons – are experimentally extremely surement on the jet mass, jet pT and MET of the event. clean. In contrast, the b-quark or tau-lepton final states are Another major component of the LHC program is to carry messy due to the hadronic nature of their decays and large out ultra-high energy heavy ion reactions. Better under- Standard Model backgrounds. In fact, it turns out that if standing the physics of jets in these reactions requires the the mass of the Higgs boson is below O(135 GeV), then construction of a modern effective theory for jet propa- large statistics are required to discover the Higgs particle gation in matter. As a part of this project, I. Vitev and G. in the b-quark or tau-lepton final states when conventional Ovanesyan constructed the Lagrangian for such theory, analyses are applied. which we called Soft Collinear Effective Theory with Glau- New jet physics techniques to discover the Higgs boson ber gluons (SCETG). We derived the Feynman rules for this decaying to fermion pairs, using less statistics and having Lagrangian for different types of scattering centers in the different systematics, are therefore highly desired. Cur- medium, with which an energetic jet may interact, and for rently M. Graesser, along with LANL post-doc G. Ovanesyan different choices for the degrees of freedom of the gluons and external collaborators J. Thaler and T. Chan (MIT) are (the mediators of the strong force) that enter our calcula- finishing a paper that proposes a new method of analyzing tions. We presented the first proof that the jet broaden- the Higgs boson data to do just that [1]. Specifically, we ing and energy loss results do not depend on this artificial are investigating the so-called “tauN subjettiness”, which choice of “gauge”. This fact, known as “gauge invariance”, was originally introduced to study boosted top quarks or allows us to put calculations of jets in heavy ion reaction boosted W bosons. The idea is to look for specific number on firm theoretical footing. By virtue of SCETG, we extend- of subjets in a big protojet and track back this number to ed the evaluation of medium-induced radiation beyond a characteristic Higgs boson decay channel. In our forth- the soft gluon approximation. We also showed how the coming paper we are investigating applications of tauN for process-dependent medium-induced radiative corrections discovering a boosted SM Higgs boson in its decay mode factorize from the jet production cross section. Our paper to b-quark jets. The main new result of our work is that “An effective theory for jet propagation in dense QCD mat- the tauN method and previous analyses have comparable ter: jet broadening and medium-induced bremsstrahlung” efficiencies for retaining the Higgs mass peak. We find is published in the Journal of High Energy Physics [2]. We however that the best discriminating observable is the have given several invited talks on the subject at major ratio tau2/tau1, which provides higher efficiency – of order conferences. In our second publication we applied this new 30% - than the currently best method, called the BDRS al- effective theory SCETG to calculate all the remaining quark gorithm. This is illustrated in Figure 1, where the Higgs and and gluon splitting probabilities in the medium [3], for the vector boson signals stand out above the QCD background. first time without assumption that either of the split par- tons is soft (published in Physics Letters B). The remaining outstanding work to be done before pub- lishing our results is the following. A larger sample of In a third publication [4] I. Vitev and external collaborators Standard Model backgrounds must be simulated to ensure B.W. Zhang and Y. He presented first results of high accu- our results are not biased by the statistical fluctuations of racy (known as next-to-leading order) for jet production our Monte Carlo samples. We also have to include pile-up in the recent LHC lead-lead run. Specifically, we focused in our event generation (pile-up refers to multiple proton- on the suppression for the single and double inclusive jet proton collisions within a single bunch crossing) and pass cross sections in heavy ion collisions. Our analysis in- all events through a detector simulation. Simulating pile-up cluded not only final-state inelastic jet interactions in the and detector effects are particularly important in order quark-gluon plasma created in such collisions, but also to further establish confidence in our findings. The big initial-state cold nuclear matter effects. We demonstrated protojet used in our analysis is pretty fat, and therefore how an enhanced di-jet asymmetry in central Pb+Pb reac- 696

Page 697

tions at the LHC, recently measured by the ATLAS and this case is whether these final-state particles follow a par- CMS experiments, can be derived from these results. We ticular angular ordering pattern. For example, if the angle showed quantitatively that a significant fraction of this of each subsequent branching is smaller than the angle of enhancement may be related to the collisional interactions the previous one, the phenomenon is called angular order- between the quarks and gluons in the jet and the strongly- ing. If the angle of each subsequent branching is larger interacting medium and pointed to a suite of measure- than the angle of the previous one, the phenomenon is ments that can help build a consistent picture of parton called angular anti-ordering. We found that in all cases shower modification in heavy ion collisions at the LHC. the amount of radiation that leaks out of angular ordered (Such stopping power calculations are also of interest to cones corresponds to no angular ordering. In fact, standard traditional plasma physics and Warm Dense Matter). Figure parton cascade based on simple superposition of binary 2 shows the comparison of the di-jet imbalance measured branchings reproduces very well the distribution of the by ATLAS and CMS to our calculations. Our paper “Next- final-state partons. This work is published in the Journal of to-leading order analysis of inclusive jet and di-jet produc- High Energy Physics [7]. Or results have been presented in tion in heavy ion reactions at the Large Hadron Collider” is two invited talks. published in Physics Letters B. We have given one invited talk on the subject and published refereed conference Impact on National Missions proceedings. This project supports the Laboratory Mission to enhance the nation’s scientific capabilities, as defined by the De- To gain further insight into the mechanisms that control partment of Energy (DOE) Office of Science. Within LANL, the modification of jets in strongly-coupled plasmas, Vitev this project is well-aligned with the Beyond the Standard and collaborators (LANL post-doc and Director’s Fellow B. Model (BSM) Grand Challenge and responds directly to Neufeld and external collaborators B.W. Zhang and W. Dai) the LDRD Program Office call to “develop new theory to made predictions for the momentum imbalance distribu- search for BSM physics” and to “further our understanding tions of photon-tagged and Z0-tagged jets at the LHC. The of quantum chromodynamics” at the Large Hadron Col- advantage of these channels is that the photon and the lider and the Relativistic Heavy Ion Collider. Research along Z0 boson do not interact strongly with the quark-gluon these lines is now part of the Nuclear & Particle Futures plasma. Consequently, they escape the region of hot dense pillar. Within the scientific community at large, this project matter unscathed and their experimentally measured addresses top priorities established by the DOE and the momentum can serve as a benchmark with respect to National Science Foundation Offices of High Energy Phys- which the momentum degradation of the away-side jet can ics and Nuclear Physics in the 2008 High Energy Physics be evaluated. The net observable effect is a modification Advisory Panel P5 and the 2007 Nuclear Science Advisory of the away-side jet distribution, as illustrated in Figure 3, Committee reports, respectively. There is significant inter- where theoretical simulations are compared to the CMS est in the stopping power of strongly-coupled plasmas experimental results. The exact shape of the medium- for charged particles in the DOE Office of Fusion Energy modified momentum imbalance distribution depends on Sciences, especially in the area of Warm Dense Matter. the details of the experimental measurements (known as Physics at the nuclear scale and below is an underlying experimental cuts). Theoretical simulations were solicited capability for the weapons program. by the ATLAS collaboration and an excellent agreement between our theory and the LHC results was shown at the Quark Matter 2012 conference in Washington DC. Our theoretical predictions have been published in two high- impact Physical Review Letters [5,6]. We have delivered 3 invited talks on our results at major international confer- ences in the field. Our work was selected as a jet physics highlight of Quark Matter 2012, the most recent biggest conference in the field of heavy ion physics. Last but not least, we (I. Vitev, LANL post-doc G. Ovanesyan and external collaborator M. Fickinger) worked toward achieving an even higher accuracy in the calculations of jet production in heavy ion reactions. For this purpose, one needs to evaluate the processes where a quark or a gluon evolves into a 3-parton final state. The central question in 697

Page 698

tau 2 (cid:31) (cid:30) 0.3 and B tagging 1 Analysis a tau 1 40 (cid:31) (cid:30) ttbar (cid:31) (cid:30) 30 V(cid:29)jets VV V(cid:29)Higgs 20 10 Figure 1. Example of how N-subjettiness techniques help isolate the Higgs and vector boson signals at the LHC above the QCD 0 background (jets). 0 50 100 150 200 4 3 2 1 0 698 Ad/σd )σ/1( J ATLAS central Pb+Pb data NLO pQCD theory, p+p NLO pQCD, A+A, R = 0.2 NLO pQCD, A+A, R = 0.4 NLO pQCD, A+A, R = 0.6 3 2 1 0 0 0.2 0.4 0.6 0.8 1 A J Ad/σd )σ/1( J γ g = 2 med 25 GeV < E < E T2 T1 100 GeV < E T1 Central Pb+Pb CMS central Pb+Pb data NLO pQCD theory, p+p NLO pQCD, A+A, pmin = 0 GeV T NLO pQCD, A+A, pmin = 20 GeV T 50 GeV < E < E T2 T1 1/2 LHC s = 2.76 TeV 120 GeV < E T1 Figure 2. Comparison of theoretical simulations with collisional and radiative energy losses to the ATLAS and CMS dijet asymme- try results in lead-lead collisions at the LHC. σ σ ψ π γ Figure 3. Comparison of the theoretically predicted photon- tagged jet momentum imbalance distribution to CMS data in lead-lead collisions at the LHC. References

1. Ovanesyan, G., M. Graesser, T. Chan, and J. Thaler. Tau_N subjettiness methods for Higgs discovery. 2013. Manuscript in Preparation.
2. Ovanesyan, G., and I. Vitev. An effective theory for jet propagation in dense QCD matter: jet broadening and medium-induced bremsstrahlung. . 2011. Journal of high energy physics. 2011 (6): 80.
3. Ovanesyan, G., and I. Vitev. Medium-induced parton splitting functions from Soft Collinear Effective Theory with Glauber gluons. 2011. Physics Letters B. 706: 371.
4. He, Y., I. Vitev, and B. Zhang. Next-to-leading order analysis of inclusive jet and di-jet production in heavy ion reactions at the Large Hadron Collider. 2012. Phys- ics letters B. 713: 224.
5. Vitev, I., and R. B. Neufeld. The Z0-tagged jet event asymmetry in heavy-ion collisions at the CERN Large Hadron Collider. . 2012. Physical Review Letters. 108:
7. Dai, W., I. Vitev, and B. W. Zhang. Momentum imbal- ance of isolated photon-tagged jet production at RHIC and LHC. 2013. Physical Review Letters. 110: 142001.
8. Fickinger, M., G. Ovanesyan, and I. Vitev. Angular distributions of higher order splitting functions in the vacuum and in dense QCD matter. 2013. Journal of High Energy Physics. 1307: 039.

Page 699

Publications Vitev, I.. NLO analysis of inclusive jet, tagged jet and di-jet Dai, W., I. Vitev, and B. W. Zhang. Momentum imbalance production in Pb+Pb collisions at the LHC. 2011. Jour- of isolated photon-tagged jet production at RHIC and nal of Physics G. 38: 124087. LHC. 2013. Physical Review Letters. 110: 142001. Vitev, I.. Electroweak boson-tagged jet event asymmetries Fickinger, M., G. Ovanesyan, and I. Vitev. Angular distribu- at the LHC. 2013. In Quark Matter 2012 . (Washington, tions of higher order splitting functions in the vacuum August 2012). , p. 701c. Ridge, NY: Nuclear Physics A. and in dense QCD matter. 2013. Journal of High Energy Vitev, I.. Jet quenching . Invited presentation at QCD Struc- Physics. 1307: 039. ture I. (Wuhan, China, October 2012). He, Y., I. Vitev, and B. Zhang. Next-to-leading order analysis Vitev, I.. Hadron, photon, and jet production at the LHC. of inclusive jet and di-jet production in heavy ion reac- Invited presentation at High pT at he LHC . (Wuhan, tions at the Large Hadron Collider. 2012. Physics letters China, October 2012). B. 713: 224. Vitev, I., and R. B. Neufeld. The Z0-tagged jet event Huang, J., Z. Kang, and I. Vitev. Inclusive b-jet production asymmetry in heavy-ion collisions at the CERN Large in heavy ion collisions at the LHC. To appear in Physics Hadron Collider. . 2012. Physical Review Letters. 108: Letters B. 242001. Ovanesyan, G.. An effective theory for jet propagation in dense QCD matter: jet broadening, radiative energy loss and LHC phenomenology. 2011. Poster for Quark Matter 2011, Annecy, France. Ovanesyan, G.. Effective theory approach to jet propaga- tion in dense QCD matter. Invited presentation at Boston jet physics workshop, Harvard University, Cambridge, January 2011. (Cambridge, January 11-14. 2011). Ovanesyan, G.. Effective theory description of jets in dense QCD matter. Invited presentation at Winter Workshop on Nuclear Dynamics 2011. (Winter Park, February 6-13). Ovanesyan, G.. Effective theory approach to jet propaga- tion in dense QCD matter. Invited presentation at SCET workshop 2011. (Pittsburg, 6-8 March, 2011). Ovanesyan, G.. Effective theory for jets in medium. Invited presentation at Department of particles and fields meeting. (Providence, RI, 9-13 August, 2011). Ovanesyan, G.. Medium-induced Splitting kernels from SCETG . 2013. In Quark Matter 2012. (Washington, Au- gust 2012). , p. 981c. Ridge, NY: Nuclear Physics A. Ovanesyan, G., M. Graesser, T. Chan, and J. Thaler. Tau_N subjettiness methods for Higgs discovery. 2013. Manu- script in Preparation. Ovanesyan, G., and I. Vitev. An effective theory for jet propagation in dense QCD matter: jet broadening and medium-induced bremsstrahlung. . 2011. Journal of high energy physics. 2011 (6): 80. Ovanesyan, G., and I. Vitev. Medium-induced parton split- ting functions from Soft Collinear Effective Theory with Glauber gluons. 2011. Physics Letters B. 706: 371. 699

Page 700

Physics Exploratory Research Final Report Daylight Imaging with Seismic Noise Xiaoning Yang 20110108ER Abstract noise cross-correlation (NCC) is related to what we call In this project, we developed a new methodology to the Green’s function. It is the seismic signal generated by extract the Earth’s attenuation property from seismic an impulsive source at one location inside the Earth, and noise. In recent years, seismic background noise has recorded by a seismic sensor at another location. The been increasingly exploited to glean useful information Green’s function between the two locations contains in- about the Earth. Most of the research, however, is di- formation about the strength of the source and proper- rected at getting the seismic velocity property, i.e., how ties of the Earth medium in which the signal propagates. fast a seismic wave travels, of the Earth. Our research Specifically, it contains, in its travel time between the represents a breakthrough in that the new method al- two locations, information about the seismic velocity of lows us to obtain the seismic attenuation property of the the Earth, and in its amplitude, information about the Earth, i.e., how the Earth absorbs seismic-wave energy, Earth’s attenuation. It can then be used to probe these from seismic background noise. This is a far more dif- properties and develop velocity or attenuation images of ficult problem. the Earth through tomography. The technique to image the Earth using noise is known as “daylight imaging”, in In addition to developing the new method, we con- analogy to the illumination of shaded areas by scattered structed a tomographic attenuation map of the western daylight. U.S. using the method. The map correlates well with the geology and with other published results. We also Traditionally, we obtain the Green’s function by record- developed a new theoretical formula linking the noise ing the signal from an energetic seismic source such as measurement with properties of the noise field and an earthquake. Using NCC, we cross-correlate noise re- those of the Earth. This permits a straightforward map- corded by two seismic sensors at two different locations ping between observations and their underlying causes. to obtain an approximation of the Green’s function be- The new formula has been tested with numerical simula- tween the two locations without a source. The approxi- tions. mation is called the empirical Green’s function (EGF). This gives us a new tool to investigate the Earth since we The new methodology that we developed has direct do not rely on when and where earthquakes occur and applications to problems in global security and energy we can design our sensor deployment to suite our spe- security missions of the Laboratory. The new capabilities cific needs. It is particularly important for regions where acquired during this project advanced the standing of earthquakes are scarce. In addition, by investigating how the Laboratory in this field and have attracted new fund- and why we can extract the Green’s function from NCC, ing to the Laboratory. we have a better understanding of seismic noise genera- tion and propagation within the Earth, which is a basic Background and Research Objectives science question. Seismic background noise is usually treated as a nui- sance that is either ignored or suppressed. In recent In the last decade, the science to derive the velocity years, however, studies have revealed that there is a property of the Earth from the travel time of EGF has wealth of information contained in seismic noise that been relatively well developed theoretically [1-3] as well we can extract. By processing the noise using the cross- as in laboratory experiments [4, 5] and in field applica- correlation method, we can bring out useful signals that tions [6-8]. To obtain the attenuation property from EGF are buried in the noise. The signal that emerges from the amplitude, which dictates how seismic energy decays 700

Page 701

with distance and in different directions, however, re- This, however, is not the case for attenuation measure- mained a challenge. This is because the amplitude of the ments from NCC. Because the noise-field intensity mainly EGF is affected not only by the Earth medium, but also by affects the amplitude of EGF, attenuation measurements the locations and strengths of noise sources, which are not from these amplitudes are biased from the true values. very well known. To date, most of the studies using seismic This poses a major obstacle to reliably measuring attenua- noise to extract EGF deal with velocity measurements. Very tion from EGF. few studies tried to extract attenuation from noise. Our In this research, we attack the problem from two direc- research is the first attempt to make accurate attenuation tions. We investigate novel NCC calculation methods that measurements from NCC and to develop a tomographic minimize the variation of noise-field intensity. We also attenuation map from the measurements. conduct numerical and theoretical analysis to characterize In this research, our objectives are to 1) develop new how the noise field behaves. theory and methodology to accurately measure Earth’s Since the theory requires a perfect diffuse noise field for attenuation from EGF and 2) construct a two-dimensional NCC to work, any attempt to make the seismic noise field (2D) attenuation map of the western U.S. using noise data more diffuse should help with reducing the bias caused recorded by the USArray, which is a dense and regularly by the noise-field variation. To accomplish this, we turn spaced seismic sensor array deployed throughout the con- our attention to the coda wave of the EGF. In general, the tinental U.S. including Alaska. Through these objectives, Green’s function in the real Earth is not just a single pulse we advance the Earth science by achieving better under- representing the signal traveling from A to B. It is usually standing and utilization of seismic noise. composed of a long wave train, whose amplitude tapers We have accomplished both of these objectives during the off with increasing time. The later part of this wave train is course of this project. We have developed a novel method- called coda. It is made up of waves that are reflected off of ology to accurately measure Earth attenuation from noise heterogeneities – small-scale material elastic and anelas- cross-correlation. We have also made theoretical develop- tic property variations that exist throughout the Earth. ments to characterize the behavior of seismic noise field. Because these heterogeneities as sources of the coda wave Using the new method we developed, we constructed a are more diffused spatially than the source for the direct tomographic attenuation map for the western U.S. In addi- signal of EGF, we should be able to take advantage of this tion, we also conducted numerical simulations to confirm and create a smoother effective noise-field intensity. To do our theory. this, we use the coda wave of EGF in a second iteration of NCC calculation. We call the new technique the correlation Scientific Approach and Accomplishments of coda of correlation, or C-cubed (C3). Lobkis and Weaver [1] first put forward a theory in 2001 Figure 1 is an illustration showing how C3 works. To cal- stating that the signal from cross-correlating noise gener- culate C3 between a pair of seismic sensors R1 and R2, ated by a diffuse noise field, meaning that noise sources we select another group of sensors surrounding R1 and are uniformly distributed in space, and recorded by two R2. We call these sensors coda sensors. We first calculate sensors A and B at different locations is equivalent to the NCC between each of these coda sensors S and R1, and Green’s function between the two locations. The onset between S and R2. We then calculate C3 between R1 and time of the Green’s function is the signal travel-time from R2 by cross-correlating the coda waves of NCC previously sensor A to sensor B, which can be used to measure the calculated between S and R1, and S and R2. C3 from all wave velocity of the medium between A and B. The ampli- coda sensors are then added together to further improve tude of the Green’s function is a product of the noise-field the noise-field diffusivity and to improve the signal-to- intensity and the medium-attenuation effect between A noise ratio. and B. If the noise field is fully diffused, noise-field in- tensity is a constant and can easily be removed from the Figure 2 gives an example comparing EGF amplitude amplitude. The corrected amplitude can then be used to decay as a function of inter-station distance with that of measure the medium attenuation. signals from an earthquake. The map on the left shows the locations of seismic sensors (circles) used to calculate When seismologists applied this theory to seismic noise, the cross-correlation and an earthquake (red star). NCC they discovered that the seismic noise field is not fully is calculated between the first sensor (black circle) and diffused. Instead, the seismic noise-field intensity changes other sensors (yellow circles) in a linear array to obtain EGF with location, direction and time. Nevertheless, scientists between the sensors. EGF amplitudes are then compared obtain reliable velocity measurements from NCC that are with signal amplitudes from the earthquake in the plot as accurate as those measured from earthquake signals. 701

Page 702

on the right. Green-circle sensors are coda sensors used inversion. Our numerical simulations confirm the feasibility to calculate C3. The amplitude decay of the earthquake of this approach. signal is due to the Earth attenuation and amplitudes of So far, we have published four peer-reviewed papers EGF should follow the same decay rate if there is no bias. documenting our accomplishments. The first three papers, Apparently amplitudes from simple NCC calculations decay published in the Comptes Rendus Geoscience, in the Jour- slower. Amplitudes obtained using the C3 technique, on nal of Acoustical Society of America and in the Geophysical the other hand, have a decay rate similar to that of the Journal International respectively, describe our theoretical earthquake signal. development and our numerical simulation results. We To establish the statistical significance of our comparison, report the development of the C3 method in a paper pub- we select a suite of linear sensor arrays at different loca- lished in the Journal of Geophysical Research: Solid Earth. tions and in different directions from among USArray sen- Our tomography results will be published in another paper sors for further testing. Based on the slope of the straight in the near future. lines shown in Figure 2, we estimate a parameter called attenuation coefficient that reflects how strongly the Earth Impact on National Missions attenuates the energy of waves propagating through it. Our research has important implications to multiple na- The larger the value is, the more strongly the Earth attenu- tional missions of the Laboratory. One of those missions ates the wave energy. Figure 3 compares the attenuation is global nuclear security. As a critical component of the coefficients estimated from earthquake signals with those mission, we monitor the world for (potentially clandestine) estimated from noise, using the suite of sensor arrays and nuclear-explosion tests using seismology. To do a good job selected earthquakes. It is apparent that using C3, we sig- using the seismic method, we need to know the detailed nificantly reduce the bias (average difference) of traditional property of the Earth in which seismic waves propagate. NCC estimates (from -5.5×10-4 km-1 to -0.8×10-4 km-1). Traditionally, we obtain this information using seismic The variation of the difference (spread) is also reduced waves generated by earthquakes or man-made sources (from 3.2×10-4 km-1 to 1.2×10-4 km-1). such as underground explosions. The technique of using seismic noise to image the Earth allows us to create better After developing the C3 method, we use the method to images of the Earth, which in turn improves our capability construct an attenuation map of the western U.S. as the to monitor the globe for potential nuclear explosions. This second objective of our project. We select crisscrossing is particularly important for regions where we do not have arrays of seismic sensors from among the USArray sen- many earthquakes, and for cases in which we need specific sors to calculate C3 EGF. Amplitudes measured from these sensor configurations to obtain good images of the Earth. EGF are then used in a tomographic inversion procedure to produce a 2D map of attenuation variations. Figure 4 Another important application of daylight imaging is in shows the map from the inversion. Attenuation patterns natural-resources exploration, which is part of the energy correlate well with known geological features, both large security mission. As the technology advances, dense, and small. Examples include strong attenuation in Colum- large-scale, 2D seismic sensor arrays are deployed in many bia Plateau, along the Sierra Nevada mountain range and exploration campaigns. Seismic noise recorded by these in Yellowstone area, and weak attenuation in Colorado sensors can be used to construct detailed images of the Plateau. The whole Basin and Range province has generally Earth on reservoir scales. It eliminates the need to use ac- strong attenuation except for some small areas such as its tive explosion sources, which is extremely expensive. Since northern end. The map also compares favorably with other the sensor arrays can be put in place for a long period of published attenuation information for the western U.S. [9]. time, noise signals from these sensors can also be used to monitor the changes of Earth medium properties over In addition to developing a new NCC processing method time. This is important for fluid-migration monitoring as and constructing an attenuation map for the western U.S., well as carbon-sequestration monitoring. we also conduct theoretical and numerical investigations to constrain the behavior of seismic noise field. Based on Our achievement in this area puts the Laboratory at the the theory that elastic-wave fields, including seismic noise forefront of the field. Our work has attracted the atten- field, should obey the so-called radiative transfer equa- tion of Department of Energy/National Nuclear Security tion, we develop a new relationship that links the EGF Agency funding officials. They have awarded a Broad Area amplitude with the Earth’s attenuation and a specific form Announcement contract to us in collaboration with Uni- of noise-field representation. With this new formula, both versity of Illinois at Urbana Champaign to continue the the Earth’s attenuation and the noise-field intensity can development of the methodology and to apply the method be teased out of EGF amplitudes through a tomographic to other regions of interest. The new capability that we de- 702

Page 703

veloped in this project will allow us to open new research 106 areas in the Nuclear Detonation Detection program at the Laboratory. We are also actively pursuing the opportunity to apply the new methodology to energy security prob- 105 lems such as reservoir and carbon-sequestration monitor- ing and geothermal exploration. 104 a) 103 100 200 300 400 500 600 700 800 900 earthquake−to−sensor or inter−sensor distance (kilometer) R1 S R2 Figure 2. A comparison between earthquake amplitude decay, amplitude decay from tradi- tional NCC, and C3 amplitude decay. The map on the left shows the locations of seismic sensors (circles) used to record the earthquake (red star) and to calculate the noise cross- correlation (NCC and C3). NCC and C3 are calculated between the black sensor and yellow sensors. Green circles are coda sensors. We use sensors along the coast as coda sensors because they generate C3 with better signal-to-noise ratios. The plot on the right shows the amplitude-decay comparison. Straight lines are best linear fits to the data. b) NCC of S and R1 coda Figure 1. An illustration showing the correlation of coda of correlation (C3) calculation. a) Schematic locations of sensor NCC of S and R2 pair R1 and R2, and coda sensors S (yellow triangles). b) Noise cross-correlation (NCC) signal (empirical Green’s function, or EGF) coda between S and R1, and between S and R2. Coda waves, which are highlighted in red, are used to do another correlation calculation time (s) to obtain C3 EGF shown in c). EGF signal c) shows up in both positive and negative times C3 due to noises that propagate in opposite directions. The more symmetric appearance of C3 EGF indicates that the noise field is made more diffuse through the C3 procedure. time (s) Figure 1. An illustration showing the correlation of coda of correlation (C3) calculation. a) Schematic locations of sensor pair R1 and R2, and coda sensors S (yellow triangles). b) Noise cross-correlation (NCC) signal (empirical Green’s function, or EGF) between S and R1, and between S and R2. Coda waves, which are highlighted in red, are used to do another correlation calculation to obtain C3 EGF shown in c). EGF signal shows up in both positive and negative times due to noises that propagate in opposite directions. The more symmetric appearance of C3 EGF indicates that the noise field is made more diffuse through the C3 procedure. 703 edutilpma 50° N 45° N 40° N 35° N earthquake amplitude NCC amplitude C3 amplitude 30 1 ° 2N5 ° W 120° W 115° W 110° W 105° W Figure 2. A comparison between earthquake amplitude decay, amplitude decay from traditional NCC, and C3 amplitude decay. The map on the left shows the locations of seismic sensors (circles) used to record the earthquake (red star) and to calculate the noise cross-correlation (NCC and C3). NCC and C3 are calcu- lated between the black sensor and yellow sensors. Green circles are coda sensors. We use sensors along the coast as coda sensors because they generate C3 with better signal-to-noise ratios. The plot on the right shows the amplitude-decay comparison. Straight lines are best linear fits to the data. 10 5 0 −5 −5 0 5 10 attenuation coefficient×10−4 (kilometer−1) from earthquakes Figure 3. Attenuation coefficients estimated from noise versus attenuation coefficient estimated from earthquake signals. The average difference between C3 estimates and earthquake estimates is -0.8×10-4 km-1 with a spread of 1.2×10-4 km-1. The average difference between traditional NCC estimates and earthquake estimates is -5.5×10-4 km-1 with a spread of 3.2×10-4 km-1. esion morf )1−retemolik( 4−01×tneiciffeoc noitaunetta attenuation coefficient from NCC attenuation coefficient from C3 Figure 3. Attenuation coefficients estimated from noise versus attenuation coefficient estimated from earthquake signals. The average difference between C3 estimates and earthquake estimates is -0.8×10-4 km-1 with a spread of 1.2×10-4 km-1. The average difference between traditional NCC estimates and earthquake estimates is -5.5×10-4 km-1 with a spread of 3.2×10- 4 km-1.

Page 704

8. Prieto, G. A., and G. C. Beroza. Earthquake ground mo- tion prediction using the ambient seismic field. 2008. GEOPHYSICAL RESEARCH LETTERS. 35 (14): -. Columbia 9. Phillips, W. S., and R. J. Stead. Attenuation of Lg in the Yellowstone Plateau western US using the USArray. 2008. GEOPHYSICAL RESEARCH LETTERS. 35 (7): -. B a sin a Sierra N evad n d R a n g e P W u ea b i v n l e i fe c r, r a R e ti n . L c o . e . n O o s f n a t tt h e e n a u m ati p o li n tu s d , e sp s e o c f i fi co c r i r n e t l e a n ti s o iti n e s s a a n n d d t h si e te a Colorado factors from ambient noise. 2011. Comptes Rendus Plateau Geoscience. 343: 615. Weaver, R. L.. On the retrieval of attenuation and site amplifications from ambient noise on linear arrays: further numerical simulations. 2013. GEOPHYSICAL JOURNAL INTERNATIONAL. 193 (3): 1644. Figure 4. Attenuation-coefficient map of the western U.S. from the tomographic inversion of EGF amplitudes measured using the Weaver, R. L.. Retrieval of Green’s function in the radiative C3 method. Warm color indicates stonger attenuation compared transfer regime. 2013. JOURNAL OF THE ACOUSTICAL with cool color. SOCIETY OF AMERICA. 133 (2): 792. Figure 4. Attenuation-coefficient map of the western U.S. from the tomographic inversion of EGF amplitudes measured using the C3 method. Warm color indicates Ref st e on r ge e r a n tte c nu e ati s on compared with cool color. Zhang, J., and X. Yang. Extracting surface wave attenuation from seismic noise using correlation of the coda of cor-
9. Lobkis, O. I., and R. L. Weaver. On the emergence of relation. 2013. Journal of Geophysical Research: Solid the Green’s function in the correlations of a diffuse Earth. 118: 2191. field. 2001. JOURNAL OF THE ACOUSTICAL SOCIETY OF AMERICA. 110 (6): 3011.
10. Snieder, R.. Extracting the Green’s function from the correlation of coda waves: A derivation based on sta- tionary phase. 2004. PHYSICAL REVIEW E. 69 (4): -.
11. Roux, P., K. G. Sabra, W. A. Kuperman, and A. Roux. Ambient noise cross correlation in free space: Theo- retical approach. 2005. JOURNAL OF THE ACOUSTICAL SOCIETY OF AMERICA. 117 (1): 79.
12. Weaver, R. L., and O. I. Lobkis. Ultrasonics without a source: Thermal fluctuation correlations at MHz fre- quencies. 2001. PHYSICAL REVIEW LETTERS. 87 (13): -.
13. Larose, E., P. Roux, and M. Campillo. Reconstruction of Rayleigh-Lamb dispersion spectrum based on noise ob- tained from an air-jet forcing. 2007. JOURNAL OF THE ACOUSTICAL SOCIETY OF AMERICA. 122 (6): 3437.
14. Campillo, M., and A. Paul. Long-range correlations in the diffuse seismic coda. 2003. SCIENCE. 299 (5606):
16. Bensen, G. D., M. H. Ritzwoller, and N. M. Shapiro. Broadband ambient noise surface wave tomography across the United States. 2008. JOURNAL OF GEOPHYS- ICAL RESEARCH-SOLID EARTH. 113 (B5): -. 704

Page 705

Physics Exploratory Research Final Report Balanced and Unbalanced Turbulent Cascades Balasubramanya T. Nadiga 20110150ER Abstract a side effect of which is to render small scale dissipation Despite its importance in the understanding and model- ineffective in directly dissipating balanced mesoscale and ing of climate variability, the energetics of ocean circula- large-scale energy. Thus a fundamental question regard- tion is still not well understood. After identifying gaps in ing the energetics of ocean circulation is as to how the current understanding in the field, we focus our research ocean equilibrates in the presence of continuous large- efforts on the movement of energy across scales in two scale forcing and an inverse cascade back to large scales fundamental and prototypical, but distinct oceanograph- at intermediate scales. An analysis of the problem points ic contexts: a) Effects of basin geometry on geostrophic to the need for better understanding how energy moves turbulence and b) The interaction of balanced and across scales and reservoirs in the multiscale system that unbalanced motions in the interior ocean. Our findings constitutes ocean circulation. lead us to propose fundamental modifications to the Scientific Approach and Accomplishments state-of-the-art understanding of how energy moves across scales in ocean circulation in a series of articles Our research focuses on the movement of energy across in leading journals in the field. In the context of the scales in two prototypical, but distinct oceanographic DOE and LANL mission to model terrestrial climate, the contexts: a) Effects of basin geometry on geostrophic improved understanding of energy pathways in ocean turbulence and b) The interaction of balanced and un- circulation achieved under this LDRD project is expected balanced motions in the interior ocean. to lead to improvements in predictive modeling of cli- Effects of basin geometry on geostrophic turbulence mate variability. We first point out that even in the classical regime of geostrophic turbulence, the question of energy transfer Background and Research Objectives across scales is not well understood in oceanographically Differential solar heating of the atmosphere-ocean relevant settings. This is because much of the theory of system between the tropical and polar latitudes is the geostrophic turbulence has been worked out assuming ultimate driver of climate on earth. In particular differen- zonally or doubly periodic geometries that are relevant tial heating of the atmosphere sets up large scale wind in the atmospheric context. patterns and latitudinal temperature gradients that then drive the general circulation of the world oceans (Figure Accomplishments: In geometries of relevance to the at- 1). The latter in turn crucially controls the variability of mosphere, energy transfers between scales occur solely climate on the interannual and longer timescales. in association with nonlinearity. In a basin setting that is the primary geometry of relevance to the oceans, this is The large-scale ocean circulation that results from the not the case; meridional boundaries allow for the linear large-scale atmospheric forcing itself develops instabili- beta-term to also transfer energy across scales (see ties and gives rise to intermediate-scale eddies. The [4,5]). (Note that meridional boundaries play an impor- large-scale circulation and the resultant eddies are both, tant role in the dynamics in each of the oceans including however, in approximate geostrophic balance---a bal- the Antarctic.) ance between pressure gradients and rotational effects. An important aspect of turbulence in the context of such We find the classic picture of energy cascades in geo- balanced dynamics (called geostrophic turbulence) is an strophic turbulence to be an oversimplification in the inverse cascade of energy. Such an inverse cascade of basin setting ([6]): Unlike in the classic picture of inverse energy leads to a confinement of energy to large scales, 705

Page 706

cascade of energy---a crucial aspect of geostrophic tur- non-hydrostatic Boussinesq equations and companion bulence, the scale at which the vertically-uniform mode balanced integrations of the quasi-geostrophic equations, receives energy (from vertically-varying modes) is not we find evidence for a local route to dissipation of bal- fixed at the deformation scale. Instead it is variable and is anced energy directly through interior turbulent cascades related to the scale at which bottom drag removes energy as opposed to boundary and surface related processes. In from the vertically-uniform mode (Figure 2). particular, we identify two phases of imbalance. Energetic consequences of the initial phase are minimal and charac- Unlike in the classic picture of geostrophic turbulence, we teristics of this phase of imbalance are similar to those of find that the nonlinear inverse energy cascade does not the kind of imbalance considered in previous studies. The extend between a deformation scale source and a larger second phase of imbalance has larger energetic conse- scale sink (see [6]). Instead, it is part of a robust double quences and evidence, including the robust formation cascade of energy. This newly identified mechanism results of cyclonic spiral structures (Figure 4, left panel), suggests from a self-organization of the flow and leads to a local that ageostrophic baroclinic instabilities play an important recycling of energy in scale space. In physical space, this role in setting up this interior dissipation route. Finally, we mechanism is related to multiple, alternatingzonal-jet demonstrate, for the first time, an exponential scaling of structures can develop in the ocean (in addition to the interior dissipation with Rossby number (Figure 4, right usual gyre circulation and mesoscale eddies) in responseto panel). purely large-scale and steady-wind forcing (Figure 3). Related work on parameterizations and statistical mechani- Balance-imbalance interactions in the interior ocean cal approaches resulted in two other publications [9, 10]. The inverse energy cascade of large-scale, quasi-two dimensional (anisotropic) rotating, stratified turbulence is Impact on National Missions in contrast to the forward energy cascade of small-scale, This research directly supports one of the aims of the Bio- three dimensional, isotropic turbulence. The scales and logical and Environmental Research arm of the DOE Office phenomena that span these two asymptotic regimes of of Science as relates to understanding of complex environ- turbulence and the interactions of the flow with boundar- mental systems across many spatial and temporal scales ies are then expected to hold, in part, the answer to the by coupling theory, model, and simulations. In particular, ocean-equilibration question above. The range of such DOE seeks the ability to model climate variability towards processes include submesoscale and inertia-gravity wave providing environmental sustainability and stewardship. processes and they are further broadly categorized into Proper representation of energy budget in the ocean boundary and surface related processes, inertia-gravity model component will lead to a more realistic representa- wave processes, and other instability processes (e.g., see tion of ocean circulation and thus enhance its predictive [2] and references therein). A great amount of research ef- capabilities. fort in the recent past has focused on surface and bottom boundary related processes (e.g., see [3] and references therein). Since atmospheric forcing of ocean circulation is mediated by the top surface, the importance of under- standing the dynamics of this region cannot be over-em- phasized. Nevertheless, a fundamental question remains as to whether surface-related processes are necessary to the activate pathways that lead to a forward cascade of balanced energy. Indeed, the bulk of the balanced energy resides in the vertical interior and there is the alternative Figure 1. Global ocean circulation is forced at the large scales (spatially) more local route to dissipation directly through and controls climate variability. It is a combination of a) surface- intensified wind-driven circulation, as exemplified by the Gulf the interior turbulent cascade. It is research into this latter Stream and the subtropical gyre of the North Atlantic (shown on alternative route to dissipation that led us to significant the left), and b) large-scale overturning circulation that is buoy- breakthroughs. ancy- driven and popularly called the “conveyor-belt’’ circulation (schematic on right from Wood Hole Oceanographic Institution). Accomplishments: We consider the nonlinear evolution of an interior baroclinic front in a regime of rotating stratified flow that is of direct relevance to circulation in the oceans- —a regime characterized by small large-scale Rossby and Froude numbers and a small vertical to horizontal aspect ratio (see [7, 8]). Using high-resolution simulations of the 706

Page 707

However, since they cannot be captured by state-of-the-art ocean models, our results suggest a parameterization of the effects of these instabilities. Such a parameterization will lead to improve- ments in predictive modeling of earth’s climate. References

1. Richards, K. J., N. A. Maximenko, F. O. Bryan, and H. Sasaki. Zonal jets in the Pacific Ocean. 2006. GEOPHYS- Figure 2. Schematic on left shows currently accepted phenom- ICAL RESEARCH LETTERS. 33 (3): -. enology of geostrophic turbulence. In particular injection to vertically-uniform (barotropic) mode is at a fixed scale. Figure
2. Molemaker, M. J., J. C. McWilliams, and I. Yavneh. on the right shows our finding that this injection scale varies Baroclinic instability and loss of balance. 2005. JOUR- with the bottom drag scale showing that energy is injected into NAL OF PHYSICAL OCEANOGRAPHY. 35 (9): 1505. the vertically-uniform mode at the horizontal scale that is most energetic.
3. Thomas, L. N., A. Tandon, and A. Mahadevan. Subme- soscale processes and dynamics. 2008. In Ocean mod- eling in an eddying regime . , p. 17. Online: Wiley.
4. Nadiga, B. T.. On zonal jets in oceans. 2006. GEOPHYSI- CAL RESEARCH LETTERS. 33 (10): -.
5. Nadiga, B. T., and D. N. Straub. Alternating zonal jets and energy fluxes in barotropic wind-driven gyres. Figure 3. In an idealized setting of the wind-driven circulation,
6. OCEAN MODELLING. 33 (3-4): 257. broad, sweeping flow in the interiors accompanied by swift return western boundary currents are expected (as on the left).
7. Straub, D. N., and B. T. Nadiga. Energy fluxes in the Instability of such a circulation gives rise to mesoscale eddies (as quasigeostrophic double gyre problem . Journal of seen in the western central region in the picture on the right). But Physical Oceanography. we find for the first time that even with steady and large- scale winds alone, distinct multiple zonal jets—jets that are narrow in
8. Nadiga, B. T.. Imbalance on a baroclinically unstable the latitudinal direction and elongated in the zonal direction— interior front . Journal of Fluid Mechanics. can appear (center). These jets are an explanation of recently observed alternating jets in the world oceans. For example, the
9. Simon, G., and B. T. Nadiga. Instability of a periodic right panel (from [1]), shows zonal velocity at 400 m. This study flow in geostrophic and hydrostatic balance . Physics of used POP, developed at LANL (http://climate.lanl.gov), and is Fluids. forced by unsteady winds and buoyancy.
10. Nadiga, B., and F. Bouchet. The equivalence of the Lagrangian-averaged Navier-Stokes-α model and the rational large eddy simulation model in two dimen- sions. 2011. Physics of Fluids. 23 (9): 095105.
11. Loxley, P., and B. Nadiga. Bistability and hysteresis of maximum-entropy states in decaying two-dimensional turbulence. 2013. Physics of Fluids. 25: 015113. Publications Figure 4. Snapshot of potential vorticity on a horizontal surface (left) from one of the high resolution runs that uses the most Connaughton, C., B. Nadiga, S. Nazarenko, and B. Quinn. accurate description of dynamics. State-of-the-art ocean model- Modulational instability of Rossby and drift waves and ling studies are conducted at far coarser resolutions and using generation of zonal jets. 2010. JOURNAL OF FLUID ME- various other approximations. Imbalanced instabilities in this CHANICS. 654: 207. highly accurate simulation give rise to a spiral eddy. These eddies Godinez, H., and B. T. Nadiga. Effects of coupling strength are ubiquitous in space imagery of the world oceans, but are not and stochastic parameterizations on data assimilation. captured in state-of-the-art ocean models because of computa-
12. In EGU General Assembly Conference Abstracts. tional limitations. Right: Energetic consequences of the imbal- (Vienna, 22-27 Apr. 2012). Vol. 14, p. 3413. Munich: anced instabilities as a function of Rossby number. As Rossby EGU. number increases, these instabilities become very important. 707

Page 708

Loxley, P., and B. Nadiga. Bistability and hysteresis of maximum-entropy states in decaying two-dimensional turbulence. 2013. Physics of Fluids. 25: 015113. Nadiga, B. T.. Stochastic approaches for multiscale model- ling in geosciences. 2011. In European Geophysical Union, General Assembly 2011. (Vienna, Austria, 3-8 Apr. 2011). Vol. 13, p. 13064. Munich: European Geo- physical Union. Nadiga, B. T.. Flux of energy across scales in two oceano- graphic flows. 2011. In Balance, Boundaries, and Mix- ing in the Climate Problem. (Montreal, 28-30 Sep. 2011). , p. —. Montreal: Centre de Recherches Math- ematiques. Nadiga, B. T.. Imbalance on a baroclinically unstable inte- rior front . Journal of Fluid Mechanics. Nadiga, B. T., and D. N. Straub. Alternating zonal jets and energy fluxes in barotropic wind-driven gyres. 2010. OCEAN MODELLING. 33 (3-4): 257. Nadiga, B. T., and D. N. Straub. Zonal jets and energy cas- cades in the baroclinic ocean double gyre problem. 2011. In American Geophysical Union Fall Meeting. (San Francisco, CA, 5-9 Dec. 2011). , p. OS41A. Wash- ington D.C: AGU. Nadiga, B. T., and F. Bouchet. When can we expect statisti- cal mechanics to help predict large scale atmospheric and oceanic features?. 2010. In American Geophysical Union, Fall Meeting 2010. (San Fransico, 13-17 Dec. 2010). , p. B1209+. Washington D.C.: American Geo- physical Union. Nadiga, B. T., and T. J. O’Kane. Predictability of Low-Fre- quency Regime Transitions in the Barotropic Vorticity Equation Driven by Weak Stochastic Forcing. 2011. AGU Fall Meeting Abstracts. : B1487. Nadiga, B. T., and W. R. Casper. Energy cascade in the re- gime of realistic ocean circulation. 2011. In Annual Meeting of the American Physical Society Division of Fluid Dynamics . (Baltimore, MD, 20-22 Nov. 2011). Vol. Bulletin of the APS, 56, p. L1.008. New York: American Physical Society. Nadiga, B., and F. Bouchet. The equivalence of the Lagrang- ian-averaged Navier-Stokes-Œ± model and the rational large eddy simulation model in two dimensions. 2011. Physics of Fluids. 23 (9): 095105. Simon, G., and B. T. Nadiga. Instability of a periodic flow in geostrophic and hydrostatic balance . Physics of Fluids. Straub, D. N., and B. T. Nadiga. Energy fluxes in the quasi- geostrophic double gyre problem . Journal of Physical Oceanography. 708

Page 709

Physics Exploratory Research Final Report Time-dependent Quantum Molecular Dynamic Simulations of Dense Plasmas Supporting Thomson X-ray Scattering Experiments Jerome O. Daligault 20110228ER Abstract capability that solves the real- time dynamics of strongly We have developed a computational capability that interacting quantum electrons and classical ions in warm solves the real-time dynamics of strongly interacting dense matter and dense plasmas. The capability enables quantum electrons and classical ions in dense plasmas. the calculation of truly dynamical and non-equilibrium This capability allows the calculations of truly dynamical properties that are so far inaccessible with state-of-the- and non-equilibrium properties that are inaccessible to art quantum molecular dynamics (QMD) approaches. state-of-the-art quantum molecular dynamics approach- Scientific Approach and Accomplishments es. We have begun to use this unique tool to develop a fundamental physics understanding of correlation and Simulation capability quantum effects on the properties of dense plasmas, to As described in the proposal, we briefly recall that the foster and validate better models of dynamical processes new simulation tool solves a mixed quantum-classical in dense plasmas, and to assist and motivate experi- dynamics that involves an approximation of the time- ments. dependent DFT (TDDFT) equations for the electrons and the classical equations of motion for the ions; TDDFT Background and Research Objectives extends the basic ideas of equilibrium DFT implemented The understanding of strongly coupled quantum plas- in the state-of-the-art simulation codes to the treatment mas is an important challenge in modern physics. The of time-dependent phenomena. It allows for a truly new generation of X-ray light sources, such as the Linac dynamical treatment of the electrons, with no reliance Coherent Light Source at Stanford, holds the promise on the Born-Oppenheimer approximation. While the to profoundly advance our fundamental understand- high computational cost of TDDFT has so far limited its ing of dense plasmas. X-ray scattering experiments application to small atomic systems (e.g. molecules in will be performed with far greater spatial and energy intense laser fields), we believe that the next generation resolution, and will provide access to measurements computational resources will bring its application to bulk that have so far been impossible. As suggested by early systems within reach. experimental efforts, the ability to model dynamical and We refer to our approach as the Time-Dependent non-equilibrium systems must be developed if we are Thomas-Fermi approximation (TDTF) since it can in- to maximize the scientific impact of the wealth of data deed be regarded as the extension to time dependent from those experiments. Indeed, in addition to the in- phenomena of the popular Thomas-Fermi model used situ measurements, the coming experiments will make to model the equilibrium properties (e.g., equations of it possible to measure genuine dynamical properties of state) of dense plasmas. The electrons are described as dense plasmas. Moreover, by their nature dense plasma a (quantum) distribution function f(r,p,t) in phase-space, experiments generally produce highly time-dependent which contains not only information on the electron par- (i.e. non- equilibrium) situations. Finally, time-dependent ticle density but also on its dynamics, e.g. the electronic simulations are required to support application-driven current. The distribution function evolves according to experiments, e.g. to assess the effectiveness of energy a Vlasov-like kinetic equation that is coupled to the clas- deposition by charged projectiles in fast ignition physics. sical ion dynamics RJ(t). This traditional structure of the The main goal of this project is to enable first-principle kinetic equation is very advantageous since it allows us- simulations of dense plasmas both in and out of equilib- ing the numerical methods of traditional plasma physics. rium. To this end, we have developed a new modeling Accordingly, the numerical methods that we have been 709

Page 710

implementing make use of ideas of the particle-in-cell (PIC) where f0 is known (either analytically or numerically) and and molecular dynamics (MD) methods, combining par- “delta f” represents the part of the distribution that is be- ticle-particle and particle-mesh calculations. To solve the ing calculated with test particles. This greatly reduces the Vlasov-like equation, we thus decided to use a so-called noise inherent to P3M simulations. We have developed a “test particles method”, where the distribution function version of the delta-f method that is suitable to our prob- f (r, p, t) is sampled with a swarm of Ntp test particles lem and have successfully implemented it. The number centered at rk(t) and momentum pk(t) By direct substitu- of numerical particles needed is reduced by at least one tion, the solution of (1) amounts to the solution of coupled order of magnitude from that needed by the brute-force Newton’s equations, which lies in the realm of classical implementation that we had originally envisioned. molecular dynamics simulations. Because it involves the We have also developed a separate code to construct the dynamics of long-range interacting particles, the problem initial conditions of the main capability. Given an ionic con- is particularly adapted to the molecular dynamics capabil- figuration at initial time t0, one needs to know the initial ity previously developed by the PI Daligault especially for electron distribution function f(r,p,t0), which is strongly this type of interactions. The latter uses the state-of- the- spatially inhomogeneous. While testing the main code with art Particle-Particle-Particle-Mesh (P3M) algorithm, which fixed ionic impurities as originally planned in the proposal, combines high-resolution of close encounters (PP) and we realized that our original scheme for the initial condi- mesh-based long-range force calculation (PM), and the tions was not accurate enough, and therefore we decided particle dynamics are resolved using the Verlet algorithm to implement an alternative, but much more complicated in periodic boundary conditions. In its current version, method. The development of this auxiliary code called for the code allows the simulation of large, complex purely significantly more time and efforts that we had anticipated. classical Coulomb systems, including mixtures, Coulomb It uses an approach recently presented in Phys. Rev. E 73, and screened Coulomb systems containing up to several 016403 (2006) and solves the equilibrium Thomas-Fermi millions of particles, and over long time scales; in the past, equations using techniques developed in density func- the code was successfully used to address investigate key tional theory (including Car-Parrinello, Fourier transform microphysical properties of dense plasmas of importance based field solver…) We have successfully implemented to laser-produced high-energy-density plasmas and astro- this approach and validated it by comparison with results physical plasmas. in the literature. As a by-product, this code gives us access A flowchart of the algorithm is shown in Figure 1. The algo- to equilibrium properties of dense plasmas. We have made rithm combines both molecular dynamics techniques (as contact with equation-of-state experts at Los Alamos (D. used in material sciences) and particle-in-cell (PIC) simula- Saumon, XCP-5), who have shown great interest in our new tion techniques (as used in traditional plasma physics). We numerical capability. Together with the main code, both note the interesting fact that the test-particle method is a equilibrium and nonequilibrium properties will be acces- particle-mesh approach similar to the standard PIC, except sible within the same simulation capability (e.g. calculate that in traditional applications of PIC each test particle self-consistently the equation of state, electron conductiv- represents many real particles whereas here each test par- ity and ion viscosity of a dense plasma). ticle represents a fraction of a physical electron. The code Finally, we have developed our own average-atom code in is written in Fortran 95 and implemented in parallel using order to build various electron-ion interaction potentials Message Passing Interface (MPI)). We are currently test- needed in the main code. In the average atom model, the ing the performances of the algorithm as a function of the plasma surrounding an atom is smeared out and replaced various purely numerical parameters (> 10 parameters), by a boundary condition; the model amounts to consid- including the number of numerical particles, the mesh size ering a neutral atom of finite size a, where a is the inter- and various parameters involved in the P3M method. We ionic distance. There are two main kinds of potentials have determined the domain of stability of the code as a depending on whether one considers all the electrons in function of the numerical parameters, where stability is the plasma on the same footing or whether one discards measured by the time over which a simulation conserves the tightly bound electrons. In all electron calculations, energy, entropy and fermionic character. which are required when dealing with fully ionized plas- A salient development in this year (that we had not initially mas, the electron-ion potential should in principle be the envisioned) has been the implementation of the delta-f bare Coulomb potential shown in blue. However, it is well method, a state-of-the-art method developed in plasma known that the TF density is ill behaved and diverges at physics to accurately solve the gyro-kinetic equations. In the ionic position. This is illustrated in Figure 2 by the blue the delta-f method, the distribution function f=f0+delta f, line that shows the TF electron density as a function of the 710

Page 711

distance from a nucleus. To deal with this we regularize rates and the wave amplitude remains finite (neglecting the electronic density below some cutoff radius rc, and use collisions). The question arises as to how these classical this density to get a regularized potential by inverting the wave-particle couplings are modified in quantum plasmas, TF equation, which is illustrated here by the red line. In the as occurs when the electron Fermi energy EF(n and m are second case, a nucleus and its cloud of tightly bound elec- the electron density and mass) exceeds the thermal energy trons are described by an effective, pseudopotential. kBT. Here, the Fermi-Dirac (FD) statistics should substitute the Maxwell-Boltzmann (MB) statistics and one should Simulation results ascribe wave-like attributes to the electrons. Understand-

1. Conservation of fermionic character: a very strong ing the properties of quantum plasmas (e.g., warm dense constraint on the numerical scheme is that it must con- matter) is a frontier of high-energy density physics with serve the quantum fermionic character of the distribution relevance to laboratory experiments and astrophysics, function f used to describe the electrons. The evolution which has recently been emboldened by new experimental equation that is being solved is a Liouville equation: it facilities and high-performance computing. preserves the phase-space volume and therefore pre- serves the number of electrons per unit of phase-space Until now, most theoretical and numerical studies have volumes imposed by the Pauli principle. Another way of relied on linear response theory, whereby the fluctuation- saying the same thing is to say that the evolution equation dissipation theorem connects equilibrium properties to solved by the new code does not create entropy. However, transport properties such as electrical conductivities or numerically, due to the discretization of the equations, one Thomson scattering cross sections. Nonlinear effects such actually solves the latter approximately only. One solves an as trapping have been ignored, although laboratory experi- equation that contains a spurious collision term due to all ments often create non-equilibrium conditions. With our the glitches caused by the numerical approximations. This new tool, we have investigated the onset on electron trap- collision term irreversibly lead to the ”Boltzmannization” of ping in quantum plasmas across the degeneracy regimes, the distribution function. It is important that this Boltzm- from fully-degenerate, zero temperature to classical, annization occur as late as possible. We have tested the high-temperature plasmas. We assessed for the first time numerical scheme on the jellium model, where a uniform the onset of particle trapping in electron plasma waves in background that neutralizes the electron system replaces quantum plasmas. In addition to the characteristic classical ions. Figure 3 shows the occupation number, i.e. the distri- time scales, namely the plasmon frequency, linear Landau bution of energy states, for a partially degenerate electron damping rate and bounce frequency in the electrostatic gas at three different times in units of inverse plasma wave trough, quantum trapping depends on an additional frequency. Over this time scale, the quantum statistics is time scale related to the recoil of electrons after absorp- well conserved and spurious collisions have not ruined the tion/emission of plasmons. dynamics yet. Figure 4 shows the results of simulations for the time evo-
2. Landau Damping and the Onset of Particle Trapping lution of the Fourier component of the electric field E(k,t) in Quantum Plasmas. The notion of wave-particle inter- as obtained with an initial standing wave of (dimension- actions, the coupling between collective and individual less) amplitude epsilon=0.005 for a number of initial wave- particle behaviors, plays a fundamental role in modern vectors k around the trapping threshold value k* discussed plasma physics. Landau damping, the collisionless damping before for reduced densities rs=1 and degeneracy param- of an electron plasma (Langmuir) wave, is a paradigmatic eters EF/kBT = 0.1 (degenerate) and 7.5 (classical regime); illustration of the effectiveness of wave-particle couplings Here rs=n-1/3/aB is the reduced density, EF is the Fermi in plasmas. Resonant electrons with velocities sufficiently energy, T the temperature . In both cases, for wavevectors close to the wave phase velocity experience a nearly con- k>k*, we observe that, after a fast transient time due to stant force and so can efficiently exchange energy with the phase mixing, the electric field oscillates and its amplitude wave. For Langmuir waves of vanishing amplitude about a envelope decays exponentially in time with frequency and uniform equilibrium, the energy exchanges overall result damping time in excellent agreement with linear theory in the exponential damping of the wave amplitude in time (see blue crosses in the figure). For long-wavelengths k< at a rate g(k) (k is the wave number). In general, this linear k*, while the wave amplitude display a preliminary ex- theory prediction breaks down after a time O(1/\w(k)) ponential Landau decay (again in remarkable accordance beyond which particles can get trapped and oscillate at a with linear theory as shown in Figure 4) then the latter bounce frequency wB(k) in the potential troughs of the saturates: the amplitude starts oscillating on a small fre- electrostatic wave. Landau damping is effective provided quency time scale, the amplitude of which increases with g(k)<< w_B$, while when g(k)<< wB(k), the damping satu- decreasing k. For k<< k*, in the time asymptotic limit, the 711

Page 712

trapping oscillation in the electric field envelope disappear 6,4+()47(+,."" and the wave goes on propagating at constant amplitude. !+6!!-0'47#%-#)'.+,!v ps These are signatures of trapping: nonlinear effects stop the !86!”#$0!!RI,PIdampingatathresholdinagreementwiththetheory,even!36!"$”.+,“9’!7”0)“84.#$!:4$3.#$!forsmallamplitudewaves,chancethattrappedparticle!76!"$”.+,“9’!-0’47#%-+.3,‘!r!i ,p i dynamics play an important role in the evolution of the !/”#,-.” system long before the wave has had a chance to damp by !”#$!$!&!!v(lr) a substantial amount according to linear theory. ii !”#!%!'%!!!!!&!!v(lr) ps !'%!%!'%!!!!!!&! vH+vxc Impact on National Missions A new and unique simulation capability in modeling ex- !!"#%&’”&()&)(+,-." treme states of matter at LANL has been developed. The !"#!%!”#!&!!v(sr) t t+dt ii capability will allow significant scope for exploration of ← !"#!%!’%!!!!!&!!v(sr) ps non-equilibrium conditions in dense plasmas and materials !’()’+,!-#)'.+,! under extreme conditions (e,g, radiation damage, ion slow- ing-down, non-equilibrium phase transitions). This work 89*(+$,-"$#“0$+$,.” enhanced our basic science capabilities, enabling LANL !!R I (t),P I (t),r i (t),p i (t) to better meet its mission challenges, including the sci- !+—,2!;<=! ence campaigns, Advanced Scientific Computing (ASC) and national High Energy Density Laboratory Plasma (HEDLP) /’(-%‘0’,1-”2”34(5,-+&-." !'’/“‘01!‘)*#-21!34**')05! programs. This exploratory research has strong ties to two LANL grand challenges, namely “Materials: Discovery Sci- ence to Strategic Applications” and “Nuclear Performance”, and to the areas that will be explored in the future MaRie Figure 1. Flowchart of the algorithm implemented. Here are the facility at LANL. We are exploring potential funding oppor- salient characteristics of the numerical scheme. We consider periodic boundary conditions so that the main simulation cell tunities with the interested program managers. represents a small but representative chunck of a plasma. First a few words about the ions. They interact through the bare, Overall, this project advances methods and predictive long- range Coulomb interaction and require an Ewald summa- power for the burgeoning area of warm dense matter, with tion to account for the interaction with periodic replicas. A very potentially significant impact on high-energy density phys- good algorithm to deal with this is the so-called particle-particle ics modeling, including inertial-confinement fusion physics. particle-mesh or P3M algorithm, which is a sophisticated way of Those applications aside, the theoretical understanding of dealing with Ewald summation. It combines high-resolution of and tools for simulation of dense plasma physics are highly individual encounters and rapid, mesh-based, long-range force cross-disciplinary as dense plasmas are exactly where the calculation. Briefly, the Coulomb potential is split into two part, methods of condensed matter physics and plasma physics for instance with the help of an error function: a short-range part each lose traction. that contains the singularity at the origin, and the remaining part that is long-range and is smooth enough so that its Fourier transform vanishes very rapidly with increasing wavevector. The resulting short-range forces are treated using traditional mo- lecular dynamics techniques such as nearest-neighbor lists and constitute the PP part of the P3M algorithm. As for the long- range forces, they are calculated on a mesh in a way that is not unlike what is done in particle-in-cell code. First the particles are assigned on the mesh, and the force field is calculated on a mesh by solving a Poisson equation using fast-fourier transforms. The force field is interpolated back onto the particles. 712

Page 713

Orbital-free approaches require specific effective, electron-ion potential !”#/07#01$23,+Z¯/.\ast 04)2$#/)IN-72#SK)) v ps (r)= − r θ(r − r c ) !"#%&‘6),++L/+/20#%)M’07),)($)\ast +,(-.$/%#01$23,+$”’%),%&)>:)?/#,7) )))))))))))))))))))))))0$),B$’&)&/%-‘04)&‘B/#(/%2/),0)%“2+/’)${-}^{\prime }5$%-)) @74-:)A/B:)C)DE1)FGEHFD)IJFFEK) NB/#,(/),0$.).$&/+) TF− 1 )))O%/“0#,+),0$.)$8)P%‘0/)-‘Q/), ’ R)) $7-$23,+% $+$81(02%(,/-,+%/26-1"% 0.006 r 0.003 , ') r c c 0 -0.003 -0.006 -0.009 !"##”%&’%()*+,-.,)-./,#/&)"01) 2 # 7 /* ,# +, ( 2 / / )% & / )3 " 4 0 ) # , , ) + 3 '0 4 ” ) / % #/ 6) ) ) ) 0 100 200 300 400 t 5 ω 00 600 700 800 900 1000 ai=(3/4πni)1/3 TF Al , 5 eV , 2.7 g/cm3 p Figure 2. Electron-ion pseudo-potential calculation using the Thomas-Fermi average atom model. Figure 3. Distribution of energy states, for a partially degener- ate quantum electron plasma at three different times in units of inverse plasma frequency (conditions: rs=a/aB=0.1, degeneracy EF/kBT =0.1, where a is the inter-particle distance, aB the Bohr radius, kBT the temperature, and EF the Fermi energy.) The blue line is the exact Fermi-Dirac distribution function for these condi- tions, and the red dots show the occupation number extracted from the code at different times. 713 )t( E k k=0.96 k 4 s k=0.75 k 3 s k=0.69 k 2 s k=0.61 k 1 s Figure 4. Damping of electrostatic plasma wave in a quantum plasma. References

1. Daligault, J.. Nonlocal orbital-free noninteracting free- energy functional for warm-dense matter. To appear in Physical Review B.
2. Daligault, J.. Time-Dependent Thomas-Fermi Simula- tions of Dense Plasmas. Invited presentation at in- ternational meeting “Orbital-Free Density Functional Methods. (Paris, Sept. 2012).
3. Daligault, J.. Non-Adiabatic Electron-Ion Coupling in Quantum Molecular Dynamic Simulations. Invited presentation at Workshop on “Nonequilibrium Phe- nomena, Nonadiabatic Dynamics and Spectroscopy”. (Tellurides, CO, July, 2013). Publications Daligault, J.. Time-Dependent Thomas-Fermi Molecular Dynamics Simulations of Dense Plasmas. Invited pre- sentation at Computational Challenges in Warm Dense Matter, long program Computational Methods in High Energy Density Plasmas held at the Institute for Pure and Applied Mathematics (IPAM). (Los Angeles, 21-25, 2012). Daligault, J.. Orbital-Free Molecular Dynamics Simulations of Dense Plasmas. Invited presentation at CECAM workshop on “Orbital-free approach for high energy density physics”. (Paris, France, 5-7 Sept. 2012). Mithen, J. P., G. Gregori, and J. Daligault. Comparative mer- its of the memory function and dynamic local field cor- rection of the classical one-component plasma. 2011. Physical Review E. 85 (5): 056407. Mithen, J. P., J. Daligault, B. J. B. Crowley , and G. Gregori. Density fluctuations in the Yukawa one-component

Page 714

plasma: an accurate model for the dynamical structure factor. 2011. Physical Review E. 84: 046401. Mithen, J. P., J. Daligault, and G. Gregori. Applicability of the hydrodynamic description of classical fluids. 2011. Physical Review E. 83: 015401(R). Mithen, J. P., J. Daligault, and G. Gregori. Molecular Dy- namics Simulations for the Shear Viscosity of the One- Component Plasma. 2012. Contributions to Plasma Physics. 52 (1): 58. Mithen, J. P., J. Daligault, and G. Gregori. Onset of nega- tive dispersion in the one-component plasma. 2012. Proceedings of the Strongly Coupled Ultra Cold and Quantum Plasmas conference in Lisbon, held in Lisbon in September 2011.. 2012 (1421): 68. 714

Page 715

Physics Exploratory Research Final Report Precision Measurement of Atomic Parity Violation in Trapped Yb+ Martin M. Schauer 20110258ER Abstract tivity due technical issues revolving around laser devel- Atomic parity non-conservation (PNC) experiments have opment and certain systematic effects. Later work by provided constraints on physics beyond the standard Schacht showed how to address the most troublesome model that rival those of accelerator-based experiments systematic effects [5]. This development along with the such as the LHC but in a very different energy regime. more easily accessible laser wavelengths for Ytterbium PNC measurements in trapped ions have the potential to ions led to the proposed work discussed here. The goal improve the sensitivity of such experiments by an order of this research was to demonstrate the viability of this of magnitude or more. We have developed an experi- approach in Yb+ ions. ment based on single, trapped, Ytterbium ions with Scientific Approach and Accomplishments demonstrated sensitivity sufficient to begin to search for PNC effects. The largest effect of the weak interaction in atomic sys- tems is the mixing of certain atomic states. In the case Background and Research Objectives of Yb+ the effected states are the 2S1/2 and 2P1/2. The A single trapped ion provides a superior system for result of the weak interaction is that the ground state studying all aspects of atomic structure and numer- no longer has completely well-defined parity. Hence ous fundamental questions about quantum mechanics, the transition amplitude for the 2S1/2 to 2D3/2 state but it has yet to be fully exploited for studying nuclear now has a small dipole component in addition to elec- structure, the Weak interaction, or extensions to the tromagnetically-allowed quadrupole component. These Standard Model. Parity non-conservation is uniquely components will have different spin dependences, and sensitive to the relatively small signals due to the weak this will result in different AC Stark shifts, or light shifts, interaction, since all other known interactions conserve for transitions between different mj magnetic sublevels parity. A measurement of atomic PNC to 0.1% would be of the atomic states. We seek to measure the size of this sensitive to new heavy gauge bosons with masses up to spin-dependent shift, since it can provide constraints on about 2.5 TeV [1][2]. physics beyond the standard model of particle physics. PNC has been measured in atomic system with preci- Broadly speaking, there are two key areas which will sions of about 1% [3], but these measurements are determine our ability to make this measurement: the ac- currently limited by sensitivity, systematic effects, and curacy with which we can measure the frequency of the atomic structure unknowns. A trapped ion provides a hyperfine transitions and the size of the laser-induced system for making the same measurement with the AC Stark shift. We are now able to measure the 12.6 possibility of more than an order of magnitude improve- GHz center frequency of the magnetic-field-insensitive ment in sensitivity and more easily understood system- F=0, mf=0 to F=1, mf=0 transition with a precision of atic uncertainties. In addition, using an ion trap allows 0.05 Hz/hr1/2 and the F=0, mf=0 to F=1, mf=±1 transi- identical measurements of a number of nuclear isotopes tions with a precision of better than 10 Hz/hr1/2. of the same atom which would dramatically improve We have developed a laser source to drive the 436 nm interpretation of the results by factoring out difficult-to- S-D transition, thereby providing the necessary light calculate aspects of atomic structure. shift, which regularly operates at 70 mW of laser power. Measurements on Ba+ ions based on a method pro- A 100 mW laser beam focused to a 10 μm spot will pro- posed by Fortson [4] did not reach the necessary sensi- duce a 1 GHz spin-independent quadrupole shift and a 1 715

Page 716

Hz spin-dependent, PNC-induced dipole shift when on res- experiment are directly applicable to other LANL efforts, onance. Given this laser power, the 10 Hz/hr1/2 sensitivity most notably the recently approved work on Thorium we have achieved is therefore sufficient to distinguish the metastable nuclear states. This work, if successful, could PNC-induced shift from the quadrupole shift within about lead to clocks and sensors of unprecedented sensitivity. 100 hours of observation time, well within the capabilities Development of the high power 436 nm laser involved of our apparatus. doubling of an 871 nm Ti::Sapphire laser through the use We have measured light shifts of several Hz with the 436 of nonlinear crystals in a high finesse resonant cavity. The nm laser detuned from resonance. With these relatively thorium work requires the generation of vacuum ultra- small shift,s we are only sensitive to the quadrupole- violet (VUV) light (<200nm) to drive atomic transitions, as induced component of the light shift. In order to increase well as the nuclear transition, through the same sort of the magnitude of the light shift, the laser must be brought techniques albeit in a much more challenging regime. The closer to resonance with the S-D transition frequency. This experience gained on the 436nm laser will be directly ap- is in itself not difficult, but must be done in a deliberate plicable to the VUV system. way to understand systematic effects such as laser polar- Identification of the Thorium nuclear state will be done ization and frequency stability. It is possible that additional through pump/probe experiments, the same technique frequency stabilization of the laser will be required, and used to measure the transition frequencies of the atomic although this is not trivial, such stabilization is a well estab- levels in Ytterbium mentioned above. Hence, the optimiza- lished technique. tion techniques developed here will be indispensable to The above-mentioned precision was achieved with no extracting the small signals one can expect to encounter magnetic shielding. This is quite remarkable, since the in the Thorium work. More generally, interpreting those ΔmF=±1 transitions suffer from linear Zeeman shifts. This signals will require a thorough understanding of the atomic precision was achieved through careful attention to the physics at play, just as in the Ytterbium experiments. trap environment, but more importantly, by rigorous sta- At a wider level, the techniques developed to deal with tistical analysis of the pump-probe process that lies at the systematic effects and the surprisingly high sensitivity heart of these measurements. We are not aware of any should reinvigorate the PNC experiment community. It is other ion trap experiments that have incorporated such interesting to note that there are a large number of experi- an analysis, and this may prove to be a valuable technique ments that fall into the pump/probe class and that the for numerous other experiments. Several manuscripts statistical techniques developed here may be of interest to describing this work are under preparation. Furthermore, this broader national community. a measurement of the lifetime of the D-state, which is important for the eventual interpretation of any PNC mea- Finally, our techniques are potentially applicable to the surements, was made using these methods, and a manu- analysis of picoscale samples for forensics and non-prolif- script describing this work is under review. eration missions. Some increased precision can be achieved by shielding References the trapped ion from ambient magnetic fields. We have constructed and started to test a new trap that has mag- 1. Rosner, J. L.. Atomic p[arity violation and precision netic shielding incorporated into the vacuum system. This electroweak physics. 1999. Physical review D. 61 (1): should provide magnetic field suppression at the loca- 016006. tion of the trapped ion by a factor of at least 1000 over 2. Rosner, J. L.. Atomic parity violation as a probe of new the unshielded trap. However, this would only result in physics. 1990. Physical Review Letters. 65 (24): 10. about a factor of two improvement in the linewidth, since fluctuations in the intensity of the 436 nm laser presently 3. Wood, C. S., S. C. Bennet, D. Cho, B. P. Masterson, J. L. contribute a roughly equal amount to the linewidth as Roberts, C. E. Tanner, and C. E. Weiman. Measurement magnetic foiled noise does in the unshielded trap. Further of parity nonconservation and an anapole moment in improvement in the sensitivity would require intensity sta- cesium. 1997. Science. 275 (5307): 1759. bilization of the laser or the use of a standing wave at the ion location. Full implementation of the new trap has not 4. Fortson, N.. Possibility of measurinr parity nonconser- been achieved as of this writing. vation with a single trapped atomic ion. 1993. Physical Review Letters. 70 (16): 2383. Impact on National Missions The tools and infrastructure developed in the course of this 5. Schacht, M.. Precision measurement of p-odd neutral 716

Page 717

currents in trapped ions. 2011. LDRD ER proposal. Publications Schacht, M., J. R. Danielson, S. Rahaman, J. R. Torgerson, J. Zhang, and M. M. Schauer. 171Yb+ 5D3/2 hyperfine state detection and F=2 lifetime. Physical Review A. Schacht, M., and M. M. Schauer. Shelving and probe effi- ciency in trapped ion experiments. Physical Review A. 717

Page 718

Physics Exploratory Research Final Report The Largest Cosmic Implosions: Formation of Supermassive Black Holes Hui Li 20110300ER Abstract accretion, though its strong radiation will tend to regu- The past decade has witnessed the astonishing discov- late how much material can accrete and how fast mate- ery that supermassive black holes (SMBHs) reside in rial can accrete. The more massive the seed is, the more the centers of nearly all galaxies exist at redshift z ~6.5, likely it can reach supermassive black hole early in the less than one billion years after the Big Bang. To under- universe, as indicated by the observations. stand the formation of supermassive black holes in the During the past several years, we have made important early universe, we have performed extensive numerical progress in addressing the black hole formation prob- simulations to understand the formation of massive stars lem. Some critical questions we investigated include: in the early universe and their impact on their surround-

1. what is the likely mass of the “seeds” of SMBHs? 2) ings. We showed that, at a high accretion rate, mas- Will the formation of massive “seeds” eventually over- sive stars could be turned into massive seeds for SMBH come the radiation feedback to form SMBHs? 3) What formation. Various SMBH growth history and models are regulates the angular momentum transport during the calculated and compared with available observational collapse leading to compact object formation? LANL has constraints. In addition, we have investigated the pro- the unique advantage of being able to model the col- cesses of angular momentum transport in the formation lapse of cosmological halos together with the central processes. These studies have provided critical informa- object’s implosion and explosion. The results obtained tion for future strategies of major observatories such as thus far have placed LANL at a strong position for the WFIRST. Capabilities built through this project (including up-coming revolution in astronomy to be initiated by the postdoc training) have contributed well to programmatic James Webb Space Telescope and WFIRST in the coming research and recruiting. decade. Background and Research Objectives Scientific Approach and Accomplishments To form SMBHs so early in the Universe, the center of a We have completed one major study on understanding galaxy must accumulate about 1065 baryons in a region the formation and growth of massive “seeds”. We inves- the size of our solar system on a short timescale. The tigated the collapse of baryons into extremely massive physics of the SMBH formation has become an impor- stars (with masses higher than 10,000 solar masses) in tant mystery, given the important roles of SMBHs in a small fraction of proto-galaxies. We have determined cosmology, in cosmic re-ionization, and in shaping the the maximum masses such stars can attain by accreting subsequent cosmic structure formation. Understanding primordial gas, which depends on how fast material can such systems thus poses one of the most exciting chal- accrete. We find that, at relatively low accretion rates, lenges in modern astrophysics where gravity, magneto- the strong ionizing radiation of these stars will limit the hydrodynamics and radiation-hydrodynamics of baryons growth of these seeds. On the other hand, our extensive must work collectively to accommodate such an extreme numerical simulations of protogalactic collapse have outcome. shown that the accretion rates unto these primordial The primary scenario we studied in this project is that stars can be quite high. In fact, under these high rates, the gravity of matter in a high density region of the the massive star mass can reach even about 1,000,000 primordial background plasma will eventually lead to the solar masses, almost 1% of a SMBHs. This is because formation of a massive “seed” at the center of baryon the radiation produced by the forming massive star is and dark matter halo. This seed then grows its mass by confined deep within the protogalaxy so its feedback is 718

Page 719

limited and de-coupled from the rest of host galaxy. In ad- which has appeared in the June 12, 2013 issue of New dition, our detailed analyses led us to conclude that initial Scientist (The Supernova that Blew up a Galaxy). Our BH seeds may have been as massive as 100,000 solar mass- numerical models have culminated in a number of Journal es. This conclusion is strongly in favor of the direct collapse publications. scenario of SMBH seed formation, in which a supermas- sive primordial star forms in a region of the universe with Impact on National Missions a high molecule-dissociating background radiation field, The work has addressed one of the most exciting and fun- and collapses directly into a 10,000 – 1,000,000 solar mass damental questions in modern astrophysics and cosmol- seed BH. These results corroborate recent cosmological ogy, namely how the supermassive black holes (SMBHs) simulations and observational campaigns that suggest that form at the centers of galaxies in the early Universe. The these massive BHs were the seeds of a large fraction of the project has also advanced LANL’s capabilities in: a) large- SMBHs residing in the centers of galaxies today. scale computational astrophysics by studying multi-dimen- sional gas collapse physics, b) cosmology by simulating the Next, we have completed a large number of simulations formation of SMBH from galaxy scale to black-hole scale, to understand the impact of the birth and death of mas- nuclear and particle physics by investigating the neutrino sive stars that provide insights into the early stages of the cooling signatures during the SMBH formation. Through SMBH growth. The numerical studies were carried on sev- making important contributions in this forefront scientific eral fronts: First, we simulated the effects of the explosion challenge, we have enhanced LANL’s scientific reputation of one such massive star, to understand whether it helps and attracted top scientists to the Laboratory. One post- fuel the rapid growth of a massive black hole by changing doc has been converted to be a staff member, working in the cooling processes in the surrounding gas. Second, we a programmatic area. The other postdoc has also gotten investigated the blast wave from these explosions in terms involved in programmatic research as well, supporting the of how it can deposit dense metal-enriched material into nuclear security mission. the surroundings. We have modeled the light curves and spectra of the first supernovae in the universe with the Publications radiation hydrodynamics code RAGE with various types Fryer, C. L., and A. Heger. Forming massive black holes of explosions: core-collapse, pair-instability, Type IIn and through stellar collapse: Observational diagnostics. supermassive thermonuclear events. 2011. Astronomische Nachrichten. 332: 408. Overall, our extensive theoretical and numerical stud- Guo, F., S. Li, H. Li, J. Giacalone, J. R. Jokipii, and D. Li. ON ies have not only enabled a deeper understanding of the THE AMPLIFICATION OF MAGNETIC FIELD BY A SUPER- formation processes of SMBHs, but also produced several NOVA BLAST SHOCK WAVE IN A TURBULENT MEDIUM. 2012. ASTROPHYSICAL JOURNAL. 747 (2): 98. testable predictions. In fact, because the radiations from the growing star are trapped, they are likely to be repro- Johnson, J. L., H. Li, D. J. Whalen, and D. E. Holz. Supermas- cessed into strong Balmer series emissions, which may sive Seeds for Supermassive Black Holes. 2013. Astro- be observable by the James Webb Space Telescope. This, physical Journal. 771: 116. strong He II with wavelength at 1640 λm, and continuum emission are likely to be the key observational signatures Johnson, J., D. J. Whalen, W. Evan, C. L. Fryer, J. Smidt, A. of the progenitors of SMBHs at high redshift. Furthermore, Heger, and K. Chen. The Biggest Explosions in the Uni- verse. To appear in Astrophysical Journal. we have discovered that these ancient explosions will be detectable by the James Webb Space Telescope (JWST) and Johnson, J., D. J. Whalen, W. Even, C. L. Fryer, A. Heger, J. the Wide-Field Infrared Survey Telescope (WFIRST) at the Smidt, and K. Chen. The Biggest Explosions in the Uni- earliest epochs in the universe. These simulations are also verse. To appear in Astrophysical Journal. having a real science impact on the astronomy community. Lead scientists who are leading the development of NASA’s Johnson, J., D. Whalen, C. Fryer, and H. Li. THE GROWTH OF next major observatory WFIRST have consulted us on filter THE STELLAR SEEDS OF SUPERMASSIVE BLACK HOLES. 2012. ASTROPHYSICAL JOURNAL. 750 (1): 66. designs for WFIRST to detect these ancient events in the near infrared, and have solicited contributions from LANL Li, S.. High-order time integration with local Lax-Wendroff on the science case for WFIRST to be submitted to NASA. procedure for central scheme on overlapping grid. We have also been contacted by several independent 2012. Journal of Computational and Applied Math. groups of astronomers for assistance in devising searches 236: 4756. for the earliest supernovae in the universe, Population III SNe. Reporters have also contacted us about our work, Li, S.. Code comparison on MHD wave propagation prob- lems. 2012. In Numerical Modeling of Space Plasma 719

Page 720

Flows: ASTRONUM2011. (Valencia, Spain, 13-17 June, 2011). Vol. 459, p. 353. Valencia, Spain: ASP Confer- ence Series. Si, J., S. Colgate, H. Li, J. Martinic, D. Westpfahl, J. Slutz, B. Klein, P. Schendel, C. Scharle, T. McKinney, R. Ginanni, I. Bentley, T. Mickey, and J. Lacy. Data Acquisition in a High-Speed Rotating Frame for New Mexico Institute of Mining and Technology Liquid Sodium αω dynamo experiment. To appear in Review of Scientific Instru- ment. Whalen, D., A. Heger, K. Chen, W. Even, J. Smidt, C. Fryer, M. Stiavelli, H. Xu, and C. C. Joggerst. Supermassive Population III Supernovae and the Birth of the First Quasars. To appear in Astrophysical Journal. Whalen, D., J. Johnson, J. Smidt, A. Heger, W. Evan, and C. Fryer. The Supernova that Destroyed a Protogalaxy: Prompt Chemical Enrichment and Supermassive Black Hole Growth. 2013. Astrophysical Journal. 774: 64. Whalen, D., L. Frey, W. Even , J. Smidt, C. Fryer, M. Stia- velli, D. Holz, A. Heger, S. E. Woosley, C. Lovekin, and A. Hungerfort. Finding the First Cosmic Explosions I: Pair-Instability Supernovae. To appear in Astrophysical Journal. 720

Page 721

Physics Exploratory Research Final Report “Listening” to the Noise of a Single Electron Spin: Methods for Fast, Ultra- sensitive, Non-perturbative Noise Spectroscopy Scott A. Crooker 20110306ER Abstract precession frequencies, decoherence times, and inter- Measurement of electron and nuclear spins forms the particle couplings), without ever having to excite, pump, basis of modern magnetic resonance technologies such or drive the spin ensemble away from thermal equilib- as NMR, MRI, and electron spin resonance. Driven by rium. Such a noise-based approach to spin resonance is the desire and need for ever-increasing sensitivity and in marked contrast to conventional nuclear- or electron- resolution (e.g., for MRI), the number of detected spins spin resonance measurements, which necessarily per- necessarily decreases. The goal of this project is to reach turb the spins. Nevertheless, this approach is guaranteed the limit of single quantum-mechanical spins – that is, by the Fluctuation-Dissipation Theorem. Importantly, a single electron — and to measure the intrinsic fluc- noise-based measurements become viable alternatives tuations of that single spin. At present, however, the to conventional (dissipative) methods as the few-particle direct magnetization measurement of a single spin is a limit is approached (N→1), since noise signals fall only as tremendously difficult proposition, due to the extreme sqrt(N) rather than as N. sensitivity required (conventional NMR or MRI lacks the The goal of this project is therefore to measure the ran- sensitivity by about a factor of a trillion!). Nonetheless, dom spin fluctuations of electron and hole spins that are from viewpoints of both fundamental measurement trapped in semiconductor quantum dots, with an aim to and basic science, single-spin detection represents a eventually reach the truly quantum-mechanical limit of truly landmark achievement in measurement science. a single electron (or hole) spin. Measuring spin dynamics Exploiting recent advances in optical detection and ef- via intrinsic noise signatures (rather than via perturba- ficient digital signal processing, coupled with alternative tive pump-probe techniques) provides a completely measurement approaches based on passive spin noise alternative and arguably more meaningful approach to detection (rather than conventional perturbative stud- ascertain intrinsic spin dynamics. We therefore develop ies), we have demonstrated a viable route for measure- an ultra-sensitive magnetometer capable of “listening” ments of a single quantum-mechanical entity – a single to the tiny spin fluctuations of truly quantum-mechanical electron spin. objects: single electron spins trapped in semiconductor quantum dots. New instrumentation for efficient digital Background and Research Objectives signal processing based on configurable hardware (FP- Not all noise in experimental measurements is unwel- GAs) formed an essential component of this work. come. Certain fundamental noise sources contain ex- tremely valuable information about the system itself – a Scientific Approach and Accomplishments classic example being the inherent voltage fluctuations We demonstrated that our scheme for using lasers to that exist across any resistor (Johnson noise), from which sensitively measure spin fluctuations works — and works temperature can be determined. In magnetic systems, very well — for electrons trapped in semiconductor quan- fundamental noise exists in the form of small, random tum dots. We completed our first set of measurements, spin fluctuations. wherein we showed that not only are spin fluctuation For example, statistical fluctuations of N paramagnetic signals measurable using the optical techniques that we spins should generate “spin noise” of order sqrt(N), even proposed, but that the data are of sufficient quality to in zero magnetic field. Crucially, the frequency spectra achieve a detailed understanding of spin relaxation and of these tiny fluctuations - if measurable - reveal the coherence for the case of electron spins in semicon- important dynamic properties of the spins (such as spin ductor quantum dots. This initial work was written up 721

Page 722

and published in the high-profile journal Physical Review for nanoscale spintronics at LANL. Letters, and the paper was especially highlighted by the journal Editors. Publications Cooper, F., B. Mihaila, J. F. Dawson, C. Chien, and E. Tim- Based on this work, we received numerous invitations mermans. Auxiliary-field approach to dilute Bose to present invited talks at international conferences. For gases with tunable interactions. 2011. Physical Review example, we presented an invited talk at the SPIE meeting A (Atomic, Molecular, and Optical Physics). 83 (5): in August 2012 (San Diego), the 4th Nano-MRI conference 053622 (15 pp.). (Switzerland, July 2012), the 2012 Quantum Dot confer- Cooper, F., Chih-Chun Chien, B. Mihaila, J. Dawson, and E. ence (May 2012). We have also been invited to give semi- Timmermans. Composite-field Goldstone states and nars/colloquia at Rice University, Texas A&M, Ohio State, U. Higgs mechanism in dilute Bose gases. 2012. Physical Oregon, and Argonne Nat’l Lab. Also, our postdoc Yan Li Review A (Atomic, Molecular, and Optical Physics). 85 has landed a prestigious faculty job as an assistant profes- (2): 023631 (5 pp.). sor at the University of Utah (Dept. of Physics), which she started in July 2013. Dawson, J., B. Mihaila, and F. Cooper. Josephson relation for the superfluid density and the connection to the As a result of our data, our theoretical component has Goldstone theorem in dilute Bose atomic gases. 2012. made significant advances in our understanding of how PHYSICAL REVIEW A. 86 (1): 013603. single electron spins couple to a bath of nuclear spins Li, Y., N. Sinitsyn, D. L. Smith, D. Reuter, A. D. Wieck, D. R. (directly addressing the “central spin problem” — a topic of Yakovlev, M. Bayer, and S. A. Crooker. Intrinsic Spin hot theoretical interest). This work was also published in Fluctuations Reveal the Dynamical Response Function Physical Review Letters. of Holes Coupled to Nuclear Spin Baths in (In,Ga)As Quantum Dots. 2012. PHYSICAL REVIEW LETTERS. 108 Again, as a result of our data, our close collaboration with (18): 186603. the Dortmund and Bochum University groups of Profes- sor Manfred Bayer and Dirk Reuter continue. We have just Mihaila, B., F. Cooper, J. Dawson, C. Chien, and E. Timmer- published a joint theoretical/experimental paper on the mans. Analytical limits for cold-atom Bose gases with use of these new experimental techniques as a high-reso- tunable interactions. 2011. Physical Review A - Atomic, lution spectroscopic tool. Again, this paper was accepted Molecular, and Optical Physics. 84 (2): 023603. at the prestigious journal Physical Review Letters. Mihaila, B., J. Dawson, F. Cooper, C. Chien, and E. Timmer- mans. Auxiliary field formalism for dilute fermionic In the last year we finished construction of the new ultra- atom gases with tunable interactions. 2011. PHYSICAL stable single-quantum dot microscope, which works very REVIEW A. 83 (5): 053637. well. Yan has shown that we can measure the absorption from a single quantum dot, which is a remarkable achieve- Roy, D., Y. Li, A. Greilich, Y. Pershin, A. Saxena, and N. Sin- ment. We have measured noise signals from this quantum itsyn. Spin noise spectroscopy of quantum dot mol- dot – thereby achieving our goals. We are currently writ- ecules. 2013. Physical Review B. 88: 045320. ing up these results for publication. Sinitsyn, N., Y. Li, S. A. Crooker, A. Saxena, and D. L. Smith. Role of Nuclear Quadrupole Coupling on Decoherence Impact on National Missions and Relaxation of Central Spins in Quantum Dots . This project has built Laboratory capabilities in the devel- 2012. Physical Review Letters. 109: 166605. opment of novel measurement methods that enable new scientific discovery at Los Alamos National Laboratory. It is Zapasskii, V. S., A. Greilich, S. A. Crooker, Y. Li, G. G. Kozlov, based on recent technological advancements that enable D. R. Yakovlev, D. Reuter, A. D. Wieck, and M. Bayer. the use of FPGAs and on the very large data storage and Optical Spectroscopy of Spin Noise. 2013. Physical Re- manipulation capabilities that have now become avail- view Letters. 110 (17): 176601. able. Our basic approach to noise spectroscopy goes well beyond the specific example of electron spin noise that will be investigated here and the advantages of noise spectros- copy as an experimental probe are very general. The proj- ect advances new Laboratory capabilities that underpin a wide variety of Laboratory missions. Demonstration of single spin noise spectroscopy measurements has generat- ed worldwide attention and has formed a robust platform 722

Page 723

Physics Exploratory Research Final Report The Terahertz Quantum Hall Effect Rohit P. Prasankumar 20110311ER Abstract and apply it to probe QHE physics in novel regimes. Two-dimensional electron gases (2DEGs) in semicon- Through comparison to theoretical models, we could ductor heterostructures are a widely used platform for then shed new light on the nature of the quantum phase modern electronics. However, at very low temperatures transitions in the QHE, with the potential to conclusively (T<4 K) and high magnetic fields (B>>1 T), they evolve determine the physics behind this phenomenon. Fur- into a new frontier for quantum processes and a testing thermore, the ability to perform THz spectroscopy in ground for novel theoretical predictions. This is argu- a magnetic field would also impact the study of many ably best exemplified by the quantum Hall effect (QHE), other phenomena, ranging from the QHE in graphene to which has been the subject of intense research world- electromagnons in multiferroics. Here, we will describe wide (and four Nobel Prizes) since its discovery in 1980. our efforts in these directions, and show that this project Despite the many publications devoted to elucidating has resulted in capabilities that will continue to impact different aspects of the static properties of the QHE, LANL research. there remain several unanswered questions regard- Background and Research Objectives ing the underlying physics of this unique phenomenon. For example, the precise nature of the quantum phase The discovery of the integer QHE by von Klitzing, et al1 transitions between conductivity plateaus as a function opened a rich field of research, spurring breakthrough of B field is unclear, particularly in understanding the innovations in many areas of physics. The most remark- importance of electron-electron interactions. This could able characteristic of the integer QHE is a quantized Hall be revealed by measuring the conductivity over a broad conductivity (σxy), which takes exact integer multiples bandwidth (several terahertz (THz)), which should be (ν) of e2/h, leading to its use as a resistance standard.3 possible since researchers have recently predicted that Experimental observations of the QHE employ the same the optical Hall conductivity will have a plateau struc- geometry used to measure the classical Hall effect, in ture that persists even in the AC regime (at terahertz which a magnetic field B is applied perpendicular to (THz) frequencies),1, 2 which we refer to as the “Tera- the direction of current flow in a sample and a voltage hertz Quantum Hall Effect (THz-QHE)”. Experimental is induced orthogonal to both B and the current, from characterization of the THz Hall conductivity could thus which one can then obtain the Hall conductivity. To conclusively resolve a host of anomalies regarding the measure the QHE, a similar experiment is performed on existence of a quantum phase transition in quantum Hall a high mobility 2DEG under much higher magnetic fields systems, while also having wide ranging implications and lower temperatures (T), which cause the density of across many areas of physics. states to clump into a series of discrete Landau levels separated by the cyclotron energy (typically in the THz The goal of this project was to use ultrafast THz spec- range). As the magnetic field is swept, the longitudinal troscopy to gain new insight into the quantum Hall effect conductivity (σxx) oscillates between zero and non-zero by probing regimes inaccessible to conventional electri- values depending on the relative position of the Fermi cal measurement techniques. To accomplish this, we energy EF with respect to Landau level positions for a developed an innovative experimental technique, THz given magnetic field (Fig. 1).4 quantum Hall spectroscopy (TQHS), which would allow us to make non-contact measurements of the complex Although the physical reasons for the quantization of conductivity tensor as a function of magnetic (B) field conductivity are relatively well understood,5 the nature and temperature (T) over a broad frequency (ω) range, of the quantum phase transitions between conductivity 723

Page 724

plateaus as a function of B (RH in Fig. 1) is not yet clear. Scientific Approach and Accomplishments Existing scaling theories of the integer QHE have assumed The scientific directions described above depend on the the interactions between electrons to be irrelevant6; how- design and development of our THz quantum Hall spec- ever, in reality the electrons interact and move in a finite troscopy system, which made this the primary initial focus magnetic field, which can cause deviations from existing of our project. The essence of this technique is its ability theories. Therefore, despite decades of research, the cor- to generate linearly polarized THz radiation and coherently rect theoretical description of quantum phase transitions detect orthogonal components of the radiation transmit- in the QHE is still unresolved. ted through a sample while varying the magnetic field. This facilitates characterization of the THz Hall conductivity in a The most critical piece of data for testing the validity of 2DEG sample by measuring the orthogonal components of different scaling theories is a measurement of the com- the transmitted THz radiation, one polarized in the direc- plex conductivity tensor, σμ,υ(ω,T,B) (where the use of the tion of the incident radiation and the other in the perpen- indices μ and υ includes both the diagonal and the Hall dicular direction. The orthogonal component results from conductivity), over a broad bandwidth of several THz,6, 7 B-field induced THz emission and exhibits the characteris- which is impossible to perform with existing electrical tech- tics of quantum Hall transitions (THz-QHE). These experi- niques. Measurements of the THz Hall conductivity would ments would thus allow us to measure the frequency and therefore offer much insight on QHE physics but have not temperature dependent complex Hall conductivity tensor, yet been performed, impeding progress in this field and σμ,υ(ω,T,B) over a bandwidth of several THz, allowing us to motivating this work. The use of ultrafast THz spectroscopy explore the directions described above. We note that there to study the quantum Hall effect in the AC regime could are fewer than ten systems of this kind in the world, to the conclusively determine the underlying physics behind best of our knowledge. this unique phenomenon, with significant ramifications for applications of the QHE. In addition, non-contact THz In the first year of this project, our efforts thus focused on measurements would provide the ability to both spatially designing and building the novel setup required for mea- resolve the contributions of edge states to the conductivity suring the THz-QHE, as well as on hiring a new postdoc and temporally resolve carrier relaxation back to equilib- to replace the project co-PI, Prashanth Upadhya, who left rium after ultrafast photoexcitation, pushing our under- LANL in 2011 for a professorship at IISER-Kolkata, in India. standing of quantum Hall physics to new frontiers. We spent a significant amount of initial time planning and designing this system, in which the major components Therefore, in this project, our original objectives were to are a femtosecond Ti:sapphire oscillator (already installed employ THz spectroscopy to demonstrate the existence of in our laboratory) and a superconducting magnet. The Hall plateaus in the THz-QHE in 2DEGs embedded within superconducting magnet, which was custom designed to GaAs/AlGaAs heterostructures, and to study: our specifications for these experiments by Oxford Instru- ments, was delivered (after a lead time of ~9 months) in • The validity of different scaling theories of quantum late 2011 and installed in early 2012. Simultaneously, we phase transitions between Landau levels in the QHE, hired a new postdoc, Stephane Boubanga Tombet, to work explored through measurements of frequency-tem- on this project; he experienced some unexpected visa perature conductivity scaling over a broad frequency issues, delaying his arrival at LANL until February 2012. range (0.1–7 THz) and comparisons to theoretical These delays in receiving the superconducting magnet and models hiring a new postdoc to work on the project set us back • Spatially-resolved mapping of the complex conductivity significantly as compared to our original timeline, but once tensor to compare electron properties at the edges of Dr. Boubanga-Tombet arrived we were able to make sub- the system with the properties within the bulk material stantial progress. He spent the next year building the TQHS system, which was completed in the spring of 2013. • The influence of femtosecond optical excitation on the THz response of a 2DEG in the quantum Hall regime Our first data was taken in April of 2013 on a topological insulator sample provided by our collaborator, Professor These experiments would be free from many of the arti- Benjamin Williams at UCLA, who had submitted a CINT facts, such as spatial inhomogeneity and sample geometry, user proposal to perform THz-TDS measurements in a B which complicate electrical measurements of the QHE. field on his samples, since this capability was not readily Perhaps more importantly, THz spectroscopy allows us to available elsewhere. We measured the transmission of the test and extend different theories of the QHE that cannot THz electric field through this sample at a temperature of be studied by any other existing technique. 2.4 K as a function of the magnetic field (Fig. 3). 724

Page 725

These results demonstrated the capability of our TQHS typically in the THz frequency range). Characterization of system to perform THz experiments under an applied B the cyclotron resonance in the time domain has revealed field, which was quite exciting. We then planned to spend useful information on the carrier scattering dynamics and the last part of the project exploring the scientific direc- dephasing mechanisms in the high mobility samples that tions described above, particularly exploring the THz-QHE we tested. This understanding is essential for the ex- in 2DEGs and in graphene, while also working with Prof. perimental realization of the THz QHE. This collaboration Williams and his postdoc Zhijun Liu (who visited CINT) to enabled us to be scientifically productive while waiting for obtain a full set of data on his samples. Unfortunately, we different components of our setup to arrive, and a publica- suffered some additional setbacks that prevented us from tion on our results is in preparation. doing this. Soon after acquiring the data shown in Fig. 3, Furthermore, this collaboration also extended to ultrafast the photoconductive antenna used to generate THz radia- optical and infrared measurements on the canonical metal- tion burned out (a commonly encountered problem). We insulator transition material vanadium dioxide (VO2), in had alternate antennas available, but unanticipated techni- which we were able to study the effects of strain-induced cal issues prevented us from performing B-field-dependent domain formation from growing VO2 films on different experiments (although we were able to acquire data with substrates on carrier dynamics across the metal-insulator no B field). This problem in itself could be overcome by transition. This work has been presented at several confer- ordering new antennas, which was done. However, more ences (APS, Ultrafast Phenomena, and the Gordon Con- importantly, Dr. Boubanga-Tombet accepted an associate ference on Ultrafast Phenomena in Cooperative Systems) professorship at Tohuku University in Japan, leaving LANL and is being prepared for publication. In addition, we also in June of 2013. This limited our progress in the last few examined plasmonic carrier injection in Au nanomesh/ months of this project and prevented us from directly mea- VO2 hybrid samples, which has been submitted to Nano suring the THz-QHE. Letters10. It is important to note, however, that our TQHS system Finally, our theoretical efforts throughout the project is a capability that very few other groups in the world duration focused on expanding upon the novel directions possess, and it will allow LANL to perform a wide variety that we planned to pursue once our experimental setup of experiments in the future. For example, our group at was built. These include the study of percolation clusters LANL has pioneered the study of ultrafast phenomena in in quantum Hall systems through the power law scaling multiferroics8; our TQHS system will enable us to study of the THz conductivity, as well as the use of THz spec- elementary excitations in these materials, known as “elec- troscopy to study edge states of the QHE through a high tromagnons”9, providing more insight on magnetoelectric frequency tail in the THz spectrum. coupling in these systems. Overall, the development of our TQHS system is an important addition to LANL’s suite of Impact on National Missions ultrafast optical and THz capabilities. This project directly addresses several challenges described In addition, it is worth noting that during the first year of in the 2007 BES report “Directing Matter and Energy: Five this project, when waiting for the magnet to arrive, we Challenges for Science and the Imagination,” including initiated a collaboration with David Hilton’s group at the controlling material processes at the level of electrons University of Alabama-Birmingham (UAB) and Steve McGill (which could potentially be done with shaped THz pulses), at the NHMFL in Tallahassee to perform THz cyclotron understanding how the properties of matter emerge from resonance experiments on 2DEGs. Prashanth Upadhya correlations of the electronic constituents (which is the visited the NHMFL and helped construct an ultrafast THz basis of the QHE), and how we characterize and control magnetospectrometer that is based on an Oxford Instru- matter far away from equilibrium (which can be done by ments split coil magnet (similar to the system we built at photoexciting a material with a femtosecond optical pulse the CINT Core Facility). This system allows the measure- and using a THz pulse to observe the associated changes in ment of the conductivity tensor in a quantum confined the time-dependent conductivity as a function of B field). It semiconducting system at THz frequencies as a function of also addresses the new BES initiative in Mesoscale Science, temperature (down to 0.4 K) and magnetic field (0–7 T). particularly in the dependence of the QHE on electron- Experiments were then performed to test and study the electron interactions. In addition, the TQHS system devel- behavior of the 2DEG sample that will be used for our THz oped in this project addresses the Materials Grand Chal- QHE experiments at LANL. Under a perpendicular magnetic lenge at LANL while bringing a new capability to LANL that field, the density of states in 2DEGs clump into discrete has already been utilized in experiments by CINT users and Landau levels separated by the cyclotron energy (which is been leveraged in support of other LDRD projects. 725

Page 726

4. Upadhya, P. C., J. E. Cunninaam, C. K. Tiang, M. Lachab, S. P. Khanna, E. H. Linfield, and A. G. Davies. Cyclotron absorption in two-dimensional electron systems moni- tored by terahertz time-domain spectroscopy. 2007. In 2007 JOINT 32ND INTERNATIONAL CONFERENCE ON INFRARED AND MILLIMETER WAVES AND 15TH INTER- NATIONAL CONFERENCE ON TERAHERTZ ELECTRONICS, VOLS 1 AND 2. , p. 481.
5. Yoshioka, D.. The Quantum Hall Effect. 2002. Figure 1. (from ref. 4). Shubnikov-de Haas oscillations (Rxx) and 6. HUCKESTEIN, B.. SCALING THEORY OF THE INTEGER Hall resistance (RH) of a GaAs/AlGaAs 2DEG heterostructure QUANTUM HALL-EFFECT. 1995. REVIEWS OF MODERN measured electrically on a Hall bar at T=1.5 K. PHYSICS. 67 (2): 357.
6. Damle, K., and S. Sachdev. Nonzero-temperature transport near quantum critical points. 1997. PHYSICAL REVIEW B. 56 (14): 8714.
7. Qi, J., L. Yan, H. D. Zhou, J. X. Zhu, S. A. Trugman, A. J. Taylor, Q. X. Jia, and R. P. Prasankumar. Coexistence of coupled magnetic phases in epitaxial TbMnO3 films re- vealed by ultrafast optical spectroscopy. 2012. APPLIED PHYSICS LETTERS. 101 (12): -. Figure 2. Setup for TQHS measurements.
8. Sheu, Y. M., S. A. Trugman, Y. S. Park, S. Lee, H. T. Yi, S. W. Cheong, Q. X. Jia, A. J. Taylor, and R. P. Prasankumar. Ultrafast carrier dynamics and radiative recombination in multiferroic BiFeO3. 2012. APPLIED PHYSICS LET- TERS. 100 (24): -.
9. Pimenov, A., A. M. Shuvaev, A. A. Mukhin, and A. Loidl. Electromagnons in multiferroic manganites. 2008. JOURNAL OF PHYSICS-CONDENSED MATTER. 20 (43): -. Figure 3. Time-dependent transmission of the THz electric field through a (BiSb)2Te3 topological insulator film as a function
10. Kida, N., Y. Takahashi, J. S. Lee, R. Shimano, Y. Yama- of applied B field. Left panel: THz signal over the full temporal saki, Y. Kaneko, S. Miyahara, N. Furukawa, T. Arima, range. Right panel: B-field-dependent changes at the negative and Y. Tokura. Terahertz time-domain spectroscopy of peak of the THz signal, near a time delay of 5.2 picoseconds (ps). electromagnons in multiferroic perovskite manganites [Invited]. 2009. JOURNAL OF THE OPTICAL SOCIETY OF References AMERICA B-OPTICAL PHYSICS. 26 (9): A35.
11. VONKLITZING, K., G. DORDA, and M. PEPPER. NEW METHOD FOR HIGH-ACCURACY DETERMINATION OF
12. Sushkov, A. B., M. Mostovoy, R. V. Aguilar, S. W. THE FINE-STRUCTURE CONSTANT BASED ON QUAN- Cheong, and H. D. Drew. Electromagnons in multifer- TIZED HALL RESISTANCE. 1980. PHYSICAL REVIEW LET- roic RMn2O5 compounds and their microscopic origin. TERS. 45 (6): 494.
13. JOURNAL OF PHYSICS-CONDENSED MATTER. 20 (43): -.
14. TSUI, D. C., H. L. STORMER, and A. C. GOSSARD. TWO- DIMENSIONAL MAGNETOTRANSPORT IN THE EXTREME Publications QUANTUM LIMIT. 1982. PHYSICAL REVIEW LETTERS. 48 Appavoo, K., N. F. Brady, B. Wang, M. A. Seo, J. Nag, R. P. (22): 1559. Prasankumar, S. T. Pantelides, D. J. Hilton, and R. F. Haglund. Plasmonic electron injection drives ultrafast
15. Jeckelmann, B., and B. Jeanneret. The quantum Hall phase transition by ultrafast phonon collapse I: experi- effect as an electrical resistance standard. 2003. MEA- ment. Presented at March Meeting of the American SUREMENT SCIENCE & TECHNOLOGY. 14 (8): 1229. Physical Society. (Baltimore, MD, 18-22 March, 2013). 726

Page 727

Appavoo, K., N. F. Brady, B. Wang, M. A. Seo, J. Nag, R. P. Prasankumar, S. T. Pantelides, D. J. Hilton, and R. F. Ha- glund. Ultrafast phase transition driven by plasmonic electron injection. Nano Lett.. Appavoo, K., N. F. Brady, M. A. Seo, J. Nag, R. P. Prasanku- mar, D. J. Hilton, and R. F. Haglund, Jr.. Ultrafast metal- insulator transition in VO2 driven by electron injection. Presented at Conference on Lasers and Electro-Optics. (San Jose, CA, 6-11 May, 2012). Appavoo, K., N. F. Brady, M. A. Seo, J. Nag, R. P. Prasanku- mar, D. J. Hilton, and R. F. Haglund, Jr.. Dynamics of ul- trafast interfacial electron injection on phase-transition mechanism. 2013. In 18th International Conference on Ultrafast Phenomena. (Lausanne, Switzerland, 8-13 July, 2012). , p. 03026. London: Oxford University Press. Brady, N. F., K. Appavoo, M. A. Seo, J. Nag, R. P. Prasanku- mar, D. J. Hilton, and R. F. Haglund, Jr.. Dynamics of ul- trafast electron injection in phase-changing vanadium dioxide. Presented at Ultrafast Phenomena in Coopera- tive Systems. (Galveston, TX, 19-24 Feb., 2012). Brady, N. F., K. Appavoo, M. A. Seo, J. Nag, R. P. Prasanku- mar, D. J. Hilton, and R. F. Haglund, Jr.. Ultrafast dy- namics of VO2 grown on different substrates with non- degenerate pump-probe spectroscopy. Presented at Ultrafast Phenomena in Cooperative Systems. (Galves- ton, TX, 19-24 Feb., 2012). Brady, N. F., K. Appavoo, M. A. Seo, J. Nag, R. P. Prasanku- mar, R. F. Haglund, Jr., and D. J. Hilton. Ultrafast dynam- ics of the VO2 insulator-to-metal transition observed by nondegenerate pump-probe spectroscopy. 2013. In 18th International Conference on Ultrafast Phenom- ena. (Lausanne, Switzerland, 8-13 July, 2012). , p. 3006. London: Oxford University Press. Brady, N. F., K. Appavoo, M. A. Seo, J. Nag, R. P. Prasanku- mar, R. F. Haglund, Jr., and D. J. Hilton. Substrate effects on the ultrafast dynamics of the VO2 insulator-to-met- al transition observed by nondegenerate pump-probe spectroscopy. Presented at Frontiers in Optics. (San Jose, CA, 17-18 October, 2011). Brady, N., K. Appavoo, P. Upadhya, M. A. Seo, J. Nag, R. P. Prasankumar, R. F. Haglund, Jr., and D. J. Hilton. Ultra- fast pump probe transmission spectroscopy of VO2. Presented at March Meeting of the American Physical Society. (Boston, MA, 27 Feb.-2 March, 2012). Wang, B., K. Appavoo, N. F. Brady, M. A. Seo, J. Nag, R. P. Prasankumar, D. J. Hilton, R. F. Haglund, and S. T. Pan- telides. Plasmonic electron injection drives ultrafast phase transition by ultrafast phonon collapse II: theory. Presented at March Meeting of the American Physical Society. (Baltimore, MD, 18-22 March 2013). 727

Page 728

Physics Exploratory Research Final Report A Compact, Brilliant, Coherent X-ray Source Based on a Dense Relativistic Electron Mirror Juan C. Fernandez 20110341ER Abstract Under the right conditions, such electrons can preserve World-wide, scientists are searching for the next gen- their ultrathin character (thickness ~ nm) and form a rel- eration of coherent X-ray sources. Free-electron lasers ativistic forward-propagating electron sheet, i.e., a REM, (FELs) supply powerful coherent X-ray beams. However, moving with a relativistic factor γ. When a second, coun- X-ray FELs are extremely expensive and available in only ter-propagating, lower intensity probe laser collides with a few institutions. With comparable power output, the this REM, a coherent and powerful X-ray pulse can be microscale X-ray source based on relativistic electron produced through coherent Thomson scattering (CTS). mirror (REM) may revolutionize X-ray generation by The ratio of X-ray photon energy to laser photon energy providing such capability driven by a relatively com- is called as Doppler shift factor, ideally 4γ2. To date, the pact high-intensity laser. The same is true of a recently technique has been developed extensively in a regime discovered relativistic laser-plasma process known as in which the x-ray contributions from each electron sum Coherent Synchrotron Emission (CSE), which generates linearly, i.e., incoherent scattering. In contrast, coherent coherent laser-frequency harmonics all the way up to scattering, in which the output goes with the square of x-ray energies. In this LDRD project, although the origi- the number of participating electrons, will occur if the nal scope was to examine the former, both mechanisms electrons are bunched with a layer thickness less than have been researched. For the former, the research has that of the output x-ray wavelength. Remarkably, this centered on testing a scheme, developed via simula- bunch layer requirement can be achieved by using a tions, to overcome the known limitations of the REM. fast-rising intense laser pulse to rip an entire relativistic The final experiments on this project on this topic are electron population from an ultra-thin foil target. The proceeding at this moment. Regarding the latter, the laser subsequently traverses plasma and continues to CSE mechanism has been established on experiments propagate forward. There is evidence [2,3] that such thin at the LANL Trident laser facility, explained theoretically, electron sheets are generated in Trident-laser experi- successfully reproduced on simulations, and reported in ments with nanofoil diamond or diamond-like (DLC) tar- a journal publication. This part of the research has been gets. Such experiments have operated in a novel laser- achieved as part of an international collaboration. These plasma regime, the relativistically-induced transparency experiments have demonstrated CSE generation of x- (RIT) regime. In this regime, the laser drives all the target rays extending to the keV range. electrons to relativistic energies. Then, the electron mass in the laboratory frame greatly increases, and therefore Background and Research Objectives the effective plasma frequency decreases, at some point In this project, two mechanisms for laser-driven genera- dipping below the laser frequency. At that point, the tion of coherent x-rays have been studied. The first, the laser drives the electrons volumetrically (not just at the original aim of this project, was light generation based “critical” density surface), and traverses forward beyond on the so-called relativistic electron mirror (REM) [1]. the target. For reasons that are explained in the next The second, pursued subsequently, is based on Coherent Section, the onset of RIT is relatively sudden, occurring Synchrotron Emission (CSE) [2]. in less than the 50 fs time resolution of our diagnostic. That raises the possibility of using RIT as a switch to pro- REM duce a fast rising laser pulse, which is useful for the aims Ultra-intense lasers are now strong enough to push all of this project, as explained below. electrons out of a nanometer foil and accelerate them to relativistic energy during a period of a few-light-cycles. However, as shown in earlier theoretical investigations, 728

Page 729

the laser plasma interaction not only imparts forward mo- of further investigation in this project. The experiments in- mentum to the electrons, but finite transverse momentum stead uncovered a novel mechanism for the generation of as well. As a result, the actual upshift factor is degraded to forward directed harmonics: coherent synchrotron emis- ~ 2γ, and coherence is spoiled, even for an otherwise suf- sion (CSE). The accomplishments of the elucidation of the ficiently thin REM. CSE mechanism and its demonstration of in Trident experi- ments [2] are summarized in the next Section, and may be In this project, we have studied the novel double-foil REM- found in the journal article. CTS scheme [1], illustrated in Figure 1, to overcome these problems with the REM concept. It combines the well Scientific Approach and Accomplishments known approach to up-shift optical frequencies via rela- REM tivistic Doppler shifting from energetic electrons, with a The first accomplishment of this project (goal REM- scheme to recover coherence and the ideal upshift, 4γ2 by 1 above) has been to improve the original theoretical eliminating the transverse electron momentum that would treatment of the double-foil REM-CTS scheme [1] with a otherwise spoil both in a short-pulse laser interacting relativistic treatment extended into the non-linear regime, with a nanofoil target. This is done with a secondary thick i.e., a02 > 1 [4], where a0 is the normalized laser electric foil that reflects back the transmitted laser. That second field of the probe laser to be backscattered and upshifted interaction cannot impart any net directed momentum (not the one that makes the REM). In this treatment, it is to an electron sheet (a well known result that rules out shown that the CTS process saturates when a0 exceeds ~ 1, direct electron acceleration by a laser), but it can cancel as shown in Figure 2(a), by nonlinear effects such as pulse transverse momentum imparted earlier by the same laser contraction and steepening. The second theoretical accom- to those electrons. It is intuitively clear that implementing plishment relates to goal REM-2 above. Although a uniform this scheme would work best with a laser pulse with a rise and stable relativistic electron sheet can be generated by time of a few cycles (~ 1 – 10 fs), rather than the > 500 fs the double-foil target scheme with a linearly polarized Trident pulse. The use of a pre-foil target to steepen the drive laser, the energy and density of the electron layer Trident pulse by exploiting RIT is a possible and attractive are found to be sensitive to the carrier-envelope phases practical alternative to a native ultrafast-rising laser pulse. of few-cycle laser pulses. To circumvent this problem, it has been shown in a separate article [5] that by utilizing The objectives of this project can be summarized as: 1) a circularly polarized laser, the electron layer becomes extending the theory for this improved double-foil REM- completely independent of the phase of the laser, avoiding CTS concept with a more proper nonlinear relativistic the rigorous requirement for phase stabilization in an ultra- treatment, 2) exploring improvements to the original intense few-cycle laser system, making the double-foil scheme, 3) validating experimentally the RIT regime where REM-CTS scheme for coherent x-ray sources more attain- Trident experiments operate, 4) exploring whether an RIT able. This result is shown in Figure 2b. pre-foil target could be used to decrease the rise time of the Trident laser pulse, and 5) testing experimentally the The third goal, REM-3 above, has also been richly fulfilled. ability of the double-foil REM-CTS scheme to reduce (or We have utilized the superior complement of optical diag- eliminate) the transverse momentum from an REM. nostics on Trident, especially our single-shot diagnostics implementing the well-known Frequency Resolved Optical CSE Gating (FROG) scheme, which are capable of measuring This part of the project relates to the fact that, in the pro- fast optical signals with 50 fs resolution in a single high- cess of characterizing and validating the modeling results energy shot, which is unique. These have been applied for the Trident laser plasmas in the RIT regime intended for to measure the incoming laser, the reflected laser light, the REM-CTS experiments, the generation of laser harmon- and the transmitted laser light after the onset of RIT. The ics are expected via the well-known process of relativistic outcome of these measurements include the first demon- oscillating mirror (ROM) prior to the onset of RIT. Basically, stration of relativistic transparency , enabled by ultrahigh the electrons at the critical surface oscillate in response contrast laser pulses interacting with diamond nanofoils, as to the laser field. However, even if the laser field oscillates well as the first demonstration and characterization of the harmonically, relativistic effects make the electron oscilla- RIT regime and unprecedented validation of the simulation tory response anharmonic, generating copious amounts of accuracy [6]. The dynamics are illustrated in Figure 3a. It light at upshifted frequencies. Although these harmonics also shows the rationale for the possibility of exploiting RIT are normally seen only in the backwards direction with as a rectifying light switch. Figure 3b and 3c show repre- thin targets, the questions of whether there is forward sentative traces of the measured reflected and transmit- ROM emission, and whether this source could be another ted laser light. Figure 3b shows the reflected laser pulse viable coherent x-ray source, became convenient subjects 729

Page 730

(3b), which goes down abruptly when RIT is established direction owing to the relativistic ponderomotive force of across the target thickness. Figure 3c shows the onset of a the laser. These nanobunches undergo rapid elliptical tra- transmitted laser pulse (which is normally absent in thick jectories across the plasma-vacuum boundary before being targets), shown to be much shorter than the reference la- emitted through the rear of the foil in the transmitted laser ser pulse. The evaluation of a laser pulse shaped (optically direction. Examining the temporal evolution of radiation rectified) by RIT (goal REM-4) was inconclusive. Although emitted from a given trajectory reveals an intense, ultra- Figure 3c may suggest a slower rise time than expected, fast (<0.1 laser cycle duration) XUV pulse emitted in the the measurement is contaminated by diffracted light from transmitted direction just as the bunch reverses its direc- the whole radial extent of the laser spot. Since the radial tion towards the target. The nanobunches are observed to laser-intensity profile is approximately Gaussian, the RIT have a spatial extent at the resolution limit of the simula- onset varies with radius, which smears out the measured tion of ~10nm. In the experiment, strong suppression of rise time. A later implementation of the diagnostic imple- lower-order harmonics (n<26) is observed with a 200nm- menting image relay has solved that problem, and the true thick diamond target, implying a maximum target density rise time of the transmitted pulse will be determined in the of >252 Nc = 625 Nc at the time of generation, where Nc future. = 1 × 1021 cm-3, the critical density for 1054 nm laser light. The observed absolute energy in a single harmonic The initial plans for utilizing current sheets generated can be estimated as 200 ±50 μJ at 20nm (58th harmonic) with Trident using nanofoils has been ruled out by more for a 200nm DLC foil target and can be considered a lower detailed simulations with the VPIC code. The simulations bound due to the limited detection aperture in our ex- indicate that the Trident pulse with 200 cycles gives not periment. The on-axis cone angle of ~1 keV emission (n ~ one but many mirrors. This is shown in Figure 4a. A first 1,000) is observed to be 10 ± 1 mrad at full-width at half measurement of this has been done utilizing the setup maximum. The significantly slower decay of CSE spectra illustrated in Figure 4b. It is consistent with the predicted (n-4/3) when compared with those produced by the ROM electron-sheet bunching at twice the laser frequency. An mechanism (n-8/3) indicates the potential for substantial integrated double-foil REM-CTS test with such a multiplic- increases in peak intensity, well beyond those predicted ity of REMs would be very difficult to interpret. It shows under idealized conditions for the ROM mechanism. The that a much shorter laser pulse is preferable. ability to generate CSE extending to the X-ray regime in transmission is enabled by the unique electron dynamics To that end, we have established a collaboration with UT driven by the intense laser field at the front surface of the Austin, where much shorter, high-energy laser pulses are relativistic plasma. These dynamics allow for dense nano- available. A subcontract to UT in this 3rd and final year bunches (that is, 10-9 m scale) to be formed and acceler- of the project has been setup to enable the collaboration ated on ultrafast timescales (<< 10-15 s). The observation using these systems for these final experiments. Our initial of harmonic spectra in transmission from ultrathin foils plan was to use a 35fs, 3J laser. Simulations for such a laser with characteristic CSE scaling is clear evidence that the are shown in Figure 5, showing the effect of the reflected unique dynamics provided during relativistic laser plasma laser from the secondary foil in decreasing electron trans- interactions can produce and rapidly accelerate dense verse momentum, as evidence by a change in the electron attosecond electron bunches suitable for the emission of beam divergence. That initial plan was foiled by a breakage bright, coherent XUV/X-ray radiation in a convenient, prac- of key components that disabled the system. The experi- tical geometry. ment has been redeployed in the GHOST laser system, which features similar energy and a 150fs pulse length, still Impact on National Missions much shorter than Trident. Those experiments are ongoing The pursuit of coherent x-ray sources impacts the ability at the time of this writing. to execute the science, defense and global security mis- sions of the DOE and the NNSA. This project supports the CSE LANL Materials Research Pillar, one of the identified broad As part of this project, strongly beamed XUV/X-ray har- range of activities needed to succeed in these missions. A monic emission with scaling consistent with the CSE brilliant X-ray source enables to study the properties and mechanism has been observed in transmission from ultra- structure of materials under extreme conditions of pres- thin solid foils, as seen in experiments at the LANL Trident sure and temperature in dynamic experiments. Investigat- facility with diamond nanofoil targets [2]. The fundamental ing the potentially transformational technologies within mechanism for coherent XUV/X-ray generation in this ge- this project has the potential of fundamentally shaping ometry has been studied using particle-in-cell (PIC) codes. the future signature facility (MaRIE) and improving LANL’s Dense nanobunches of electrons are formed at the front capabilities and scientific outlook for years to come. surface of the target and accelerated into the transmitted 730

Page 731

sity drops abruptly when the plasma becomes relativistically transparent. (c) Incident laser pulse (red solid line) and the laser pulse transmitted through a 10 nm DLC foil (blue solid line) and its temporal phase (blue dotted line). The sharp rise time in the transmitted laser pulse is degraded due to initial self-focusing a nd diffraction. Figure 1. Schematic depiction of laser-based coherent x-ray generation by coherent Thomson backscatter from a relativistic electron mirror. Left: The intense laser incident on an nanofoil Measurement (a) target makes an electron sheet moving forward with a relativistic factor γ, and traverses the target. That sheet would not support coherent scattering due to finite lateral momentum imparted by the laser on first passage. The laser, upon reflection and traversal through the sheet in the opposite direction, nearly cancels the transverse momentum on the sheet, leaving only longitudinal momentum, making it capable of coherent scattering. Right: A relatively weak backward propagating laser is upshifted to a frequency 4γ2, i.e., a coherent x-ray pulse. Figure 2. (a) Saturation of nonlinear scattering. a0 is the normal- ized electric field strength. (b) Relativistic electron mirror density profile snapshots as a function of polarization and carrier-phase relationship for linear polarization (black, linear polarization with phase =0 deg; red, linear polarization with phase of 90 deg.; and blue, circular polarization). There is no dependence of carrier phase for circular polarization. Figure 3. (a) Laser plasma interaction leading to transmitted laser pulse shaping via relativistic transparency. (b) Laser pulse shape (red solid line) and its temporal phase (blue dotted line) reflected from a 100 nm diamond foil. The reflected light inten- 731 (b) e- CTR signal Target Laser CTR foil (c) (d) no relativistic transparency Figure 4. (a) Theoretical prediction with a sub-ps Trident pulse interacting with a nanofoil, showing the generation of multiple current sheets every half cycle. (b) Experimental setup to charac- terize the laser-driven electron sheets with Coherent Transition radiation (CTR). (c) CTR spectrum from a thick target, which does not undergo RIT, showing a broad electron spectrum, as predict- ed by simulations (not shown). (d) CTR spectrum from a nanofoil, which undergoes RIT, consistent with the simulation prediction of m ultiple electron sheets. 120 100 80 60 40 20 0 -90 -60 -30 0 30 60 90 artceps ralugnA Angular spectra of full electron population 3nm+30nm 3nm Angle (degree) Figure 5. Predicted angular electron spectrum resulting from the interaction of a 3 nm DLC foil with a 35fs, 3J laser pulse (black), and from a 3 nm + 30 nm double-foil target. The two simulations show the collimating effect on the electron sheets of the laser- pulse reflected from the 30 nm foil.

Page 732

References Palaniyappan, S., B. M. Hegelich, H. C. Wu, D. Jung, D. C. Gautier, L. Yin, B. J. Albright, R. P. Johnson, T. Shimada,

1. Wu, H. C., J. Meyer-ter-Vehn, J. Fernandez, and B. M. S. Letzring, D. T. Offermann, J. Ren, C. K. Huang, R. Hegelich. Uniform Laser-Driven Relativistic Electron Horlein, B. Dromey, J. C. Fernandez, and R. C. Shah. Layer for Coherent Thomson Scattering. 2010. PHYSI- Dynamics of relativistic transparency and optical shut- CAL REVIEW LETTERS. 104 (23): -. tering in expanding overdense plasmas. 2012. Nature Physics. 8: 763.
2. Dromey, B., S. Rykovanov, M. Yeung, R. Horlein, D. Jung, D. C. Gautier, T. Dzelzainis, D. Kiefer, S. Palaniyp- Palaniyappan, S., R. C. Shah, R. Johnson, T. Shimada, D. C. pan, R. Shah, J. Schreiber, H. Ruhl, J. C. Fernandez, C. L. Gautier, S. Letzring, D. Jung, R. Hoerlein, D. T. Offer- S. Lewis, M. Zepf, and B. M. Hegelich. Coherent syn- mann, J. C. Fernandez, and B. M. Hegelich. Pulse shape chrotron emission from electron nanobunches formed measurements using single shot-frequency resolved in relativistic laser-plasma interactions. 2012. NATURE optical gating for high energy (80 J) short pulse (600 fs) PHYSICS. 8 (11): 804. laser. 2010. Review of Scientific Instruments. 81 (10): 10E103.
3. Kiefer, D., A. Henig, D. Jung, D. C. Gautier, K. A. Flippo, Palaniyappan, S., R. C. Shah, R. Johnson, T. Shimada, D. S. A. Gaillard, S. Letzring, R. P. Johnson, R. C. Shah, T. C. Gautier, S. Letzring, D. Jung, R. HoÃàrlein, D. T. Of- Shimada, J. C. Fernandez, V. K. Liechtenstein, J. Sch- fermann, J. C. FernaÃÅndez, and B. M. Hegelich. Pulse reiber, B. M. Hegelich, and D. Habs. First observation of shape measurements using single shot-frequency re- quasi-monoenergetic electron bunches driven out of solved optical gating for high energy (80 J) short pulse ultra-thin diamond-like carbon (DLC) foils. 2009. EURO- (600 fs) laser. 2010. Review of Scientific Instruments. PEAN PHYSICAL JOURNAL D. 55 (2): 427. 81 (10): 10E103.
4. Wu, H. C., J. Meyer-ter-Vehn, B. M. Hegelich, and J. C. Wu, H. C.. Phase-independent generation of relativistic Fernandez. Nonlinear coherent Thomson scattering electron sheets. 2011. Applied Physics Letters. 99 (2). from relativistic electron sheets as a means to produce Wu, H. C., J. Meyer-ter-Vehn, B. M. Hegelich, and J. C. Fer- isolated ultrabright attosecond x-ray pulses. 2011. nandez. Nonlinear coherent Thomson scattering from PHYSICAL REVIEW SPECIAL TOPICS-ACCELERATORS relativistic electron sheets as a means to produce iso- AND BEAMS. 14 (7): -. lated ultrabright attosecond x-ray pulses. 2011. Physi- cal Review Special Topics-Accelerators and Beams. 14
5. Wu, H. C.. Phase-independent generation of relativistic (7). electron sheets. 2011. APPLIED PHYSICS LETTERS. 99 (2): -.
6. Palaniyappan, S., B. M. Hegelich, H. C. Wu, D. Jung, D. C. Gautier, L. Yin, B. J. Albright, R. P. Johnson, T. Shi- mada, S. Letzring, D. T. Offermann, J. Ren, C. K. Huang, R. Horlein, B. Dromey, J. C. Fernandez, and R. C. Shah. Dynamics of relativistic transparency and optical shut- tering in expanding overdense plasmas. 2012. NATURE PHYSICS. 8 (10): 763. Publications Dromey, B., S. Rykovanov, M. Yeung, R. Horlein, D. Jung, D. C. Gautier, T. Dzelzainis, D. Kiefer, S. Palaniyppan, R. Shah, J. Schreiber, H. Ruhl, J. C. Fernandez, C. L. S. Lew- is, M. Zepf, and B. M. Hegelich. Coherent synchrotron emission from electron nanobunches formed in relativ- istic laser-plasma interactions. 2012. NATURE PHYSICS. 8 (11): 804. Jung, D., R. Hoerline, D. C. Gautier, S. Letzring, D. Kiefer, K. Allinger, B. J. Albright, R. Shah, S. Palaniyappan, L. Yin, J. C. Fern·ndez, D. Habs, and B. M. Hegelich. A novel high resolution ion wide angle spectrometer. 2011. Re- view of Scientific Instruments. 82 (4): 043301. 732

Page 733

Physics Exploratory Research Final Report Majorana Neutrinos Steven R. Elliott 20120293ER Abstract work will ensure that follow-on cryostats containing Ge The Majorana nature of the neutrino (whether or not it detectors fabricated from enriched Ge will be successful. is its own antiparticle) is a key question that impacts the Scientific Approach and Accomplishments fields of particle physics, nuclear physics, cosmology, and astrophysics. In particular this issue is extremely impor- We obtained all of the equipment and supplies to as- tant in understanding the matter/ antimatter asymmetry semble the prototype cryostat. We obtained beneficial of the Universe. The process of neutrinoless double beta occupancy of our underground laboratory and installed decay can only exist if neutrinos are Majorana particles. apparatus, including the shield for the detector. Be- Furthermore, searches for double beta decay are the cause of various timing issues and the availability of the only practical way to seek an answer to this question. underground lab, we decided to assemble the prototype Our project was aimed at demonstrating a prototype there. This permits a more sensitive understanding of module of a new experimental approach to the search the performance of the prototype module. We have as- for double beta decay. sembled the prototype module and performed a num- ber of tests that resulted in changes to the electronics Background and Research Objectives and vacuum design. The Majorana nature of the neutrino (whether or not it The specific scope of this project was to test a string of is its own antiparticle) is a key question that impacts the detectors prior to enriched detectors being deployed in fields of particle physics, nuclear physics, cosmology, and MAJORANA. The string is a group of Ge detectors held astrophysics. In particular this issue is extremely impor- together in a linear arrangement. The string design for tant in understanding the matter/ antimatter asymmetry MAJORANA was tested with this so-called prototype of the Universe. The process of neutrinoless double beta cryostat. Strings will be deployed within the prototype decay can only exist if neutrinos are Majorana particles. module for testing, and the performance of these strings Furthermore, searches for double beta decay are the is critical to the success of detectors made from very ex- only practical way to seek an answer to this question. pensive enriched material. For this string test within the Our plan to address this question is to look for neutrino- prototype module, all parts were bought by LDRD. These less double beta decay in the isotope 76Ge. High purity parts include all the detectors, the data acquisition Ge detectors fabricated with material enriched in 76Ge system, vacuum system and shielding. Software devel- provide an excellent opportunity for these experiments opment was done by a university partner (UNC). Both because the detector material can also play the role of the fabrication and test was supported by LDRD. Figure 1 the source. shows a photograph of the prototype module soon after The MAJORANA Project, will use detectors fabricated assembly and just before insertion into a glove box and from enriched Ge to search for double beta decay. These the installation of a detector string. The work was done detectors are very expensive and it is necessary to test at the Sanford Underground Research Facility in Lead, the planned deployment of those detectors. SD. As a project aimed at a proof-of-principle, we developed The primary goal of this project was to bring a proto- a prototype cryostat filled with non-enriched Ge detec- type cryostat filled with natural Ge detectors to fully tors. This work is a separate activity from MAJORANA operational status. The results from this prototype led to and has its own technical goals. The results from this improvements in the design of cryostats that will house 733

Page 734

Ge detectors fabricated from enriched Ge. This included References aspects of the vacuum and wiring of the string. We identi- 1. Elliott, Steven R.. Recent progress in double beta de- fied and have designed fixes for several internal electronic cay. 2012. Modern Physics Letters. A27: 1230009. and cryogenic issues. Our results allow the MAJORANA project to proceed with construction of the first cryostat 2. Elliott, Steven R.. The Majorana Demonstrator neu- that will house enriched Ge detectors. The analysis of trinoless double-beta decay experiment. To appear in the data from these enriched detectors will lead to much Advances in High Energy Physics. needed improved constraints on neutrino properties, in particular, whether the neutrino is its own antiparticle (a Publications so-called Majorana neutrino). Elliott, S. R.. The MAJORANA DEMONSTRATOR Neutrino- less Double-Beta Decay Experiment. To appear in Ad- Impact on National Missions vances in High Energy Physics. The work has developed science and technology underpin- Elliott, Steven R.. Recent Progress in Double Beta Decay. nings for new program areas. In particular, the path for 2012. Modern Physics Letters A. 27: 1230009. impact on LANL programs are indicated by the Commu- nity support of this science as described by NRC commit- tees and the DOE NP/HEP NuSAG review committee. The 2007 Nuclear Physics Long Range Plan strongly endorses research in this area. The DOE Office of Nuclear Physics and the NSF have been very supportive and therefore the future funding potential is strong. The similarity in require- ments means that the R&D for neutrinoless double beta decay will be applicable to certain environmental monitor- ing applications. The low-background and detector exper- tise capabilities developed by this project have been ap- plied externally-supported projects in homeland security. Figure 1. The prototype module just before a detector string was inserted. The large Cu cylinder to the right is the vacuum cryostat that contains the detectors. 734

Page 735

Physics Exploratory Research Final Report Quantum Hydrodynamics Approach to Non-equilibrium Modeling of Compressed Plasmas Michael S. Murillo 20130242ER Abstract simplifying assumptions due to the disorder caused by Experiments in the high energy-density physics regime higher temperatures. Two theoretical approaches were are notoriously challenging to perform and diagnose; considered and initial coding was completed. One poten- and, such experiments require very large and expen- tial description is through a quantum hydrodynamics ap- sive facilities. For all of these reasons it is prudent that proach, which has computational advantages, but could we rely on simulations to the largest possible degree. be lacking in physics fidelity. We explored how such In many cases, the simulations are quite challenging an approach could be made computationally attractive because the matter has complicated properties, which through a smoothed-particle hydrodynamics approach. makes the subject itself interesting and useful. We have Second, a potentially more refined approach uses quan- proposed a hybrid method of performing microscopic tum kinetic theory, which retains much more informa- simulations of warm dense matter in which very strong tion than does hydrodynamics. We also developed a test interactions between the nuclei and quantum mechani- code for solving such kinetic equations. cal properties of the electrons are both accounted for. Impact on National Missions Our goal is to understand those subtle physics issues as- sociated with inertial confinement fusion experiments. The impact on core nuclear weapons missions and iner- tial confinement fusion campaigns could be large. Such Background and Research Objectives a code capability would allow one to compute various Billions of dollars have been invested in the National physics processes to very high fidelity in the absence of Ignition Facility. The campaign to achieve ignition ended experimental data. about a year ago – without achieving ignition. While the reasons are unknown, and are likely to be complex, one potential error was missing physics in the simula- tions used to design the experiments. Our objective was to build a new simulation capability that could address physics that is known to not be in the simulations used to design last year’s experiments. Our main objective was to combine a molecular dynam- ics approach, which naturally handles strong interactions between nuclei, with a dynamical quantum description of the electrons, including important electric fields that arise form charge separation. The most challenging por- tion is the description of the dynamical quantum elec- trons. Scientific Approach and Accomplishments Our approach was to begin with the electron descrip- tion. In general, computing the time evolution of a quantum many-body system is extremely difficult. How- ever, for high energy-density matter one can make some 735

Page 736

Physics Exploratory Research Final Report Toward Laser Manipulation and Quantum Control of Th-229 Nuclei Embedded in a Solid Xinxin Zhao 20130503ER Abstract The success of this measurement would quickly establish The thorium-229 (t1/2 =7340 yr) nucleus possesses the LANL as a leader in the field. lowest-energy nuclear isomeric state [1-3] in nature. Scientific Approach and Accomplishments This LDRD reserve project aims to preserve the momen- tum of LANL’s recent breakthrough in 229Th isomer The challenges for optical spectroscopy of the isomeric research [4] and work towards the first direct optical transition are very low signal count rate and many spectroscopic measurement of the 229Th isomeric co-existing “dirty effects” that are much larger than nuclear transition using a monochromator. Such an the signal. We have overcome all these difficulties but optical spectroscopy measurement will greatly reduce need to perform the measurements at higher resolution the search range required to directly excite the nuclear over the expected wavelength range that is very time- isomeric transition with a laser. The successes of this consuming because of the low signal count rate and high project would lead to exciting new developments such background from the dirty effect. In the past year, great as a solid-state nuclear clock that will have a transform- progresses have been made in preparation for such a ing impact in quantum optical science, precision mea- measurement. surements, navigation, cosmology, and threat reduction. (1) We have perfected the system and completed the tests of the spectrometer that couples an MgF2 plate to Background and Research Objectives a monochromator for such a measurement. The sys- It has been known for decades that the 229Th nucleus tem collect 229g, 229mTh recoil ions from alpha-decay possesses the lowest-energy nuclear isomeric state of 233U into a MgF2 plate (233U à 229g, 229mTh + α), [1-3] in nature. Two widely accepted indirect measure- and measure the emission spectrum using a vacuum ments of the transition energy [2, 3] place it within reach UV monochromator with an estimated isomer signal to of existing laser capabilities notwithstanding a huge noise ratio of more than 10. Tests show that our system uncertainty in the transition wavelength. There is now has an accuracy of 0.25 nm and sufficient sensitivity to a growing interest in 229Th isomer research because of see the transition, as shown in Figure 1. Determining its potential application in nuclear, atomic, condensed the isomeric emission wavelength to this accuracy would matter and optical physics, quantum information, and be a very important scientific result. metrology, including the development of an optical clock based on this nuclear transition. The laser-accessible (2) Because the MgF2 phosphorescence and scintilla- energy and the long lifetime of the isomeric state results tion backgrounds depends on the temperature, we have in an exploitable Q = f/ δf of up to 1020 for this nuclear stabilized the temperature of the MgF2 plate holder to transition, which combined with the well isolated en- 23.6±0.3 °C year round, despite a very large room tem- vironment of the nucleus provides a nuclear probe of perature fluctuation due to aging facilities (air condition unprecedented precision, and would lead to a nuclear doesn’t work in the summer, not enough heating in the frequency standard with a stability 100 times better than winter). the best atomic clocks. (3) We have obtained 0.95g of pure 233U material from The project aims to eventually pin down the wavelength ORNL, which contains only 3ppb of the 232U contami- of the nuclear transition to a level that is ~2 orders of nant. The 233U will be used for making recoil sources as magnitude better than the current published result, soon as funding is available. The 233U material is more which is a prerequisite for the laser search to succeed. 736

Page 737

than 2600 times purer than what we have been using, and will last for years without the problems caused by the in-growth of 233, 232U decay daughters. This will greatly simplify and speed up the measurement. (4) We have built a larger collaboration across the lab with scientists in T, P and MST divisions for an ambitious ER project. (5) Technical spin-off of the thorium research has already led to an idea approved by LANL for patent application market survey. Impact on National Missions Success of the optical spectroscopy measurement would clear the path for laser excitation of the nuclear isomeric transition that lays the foundation for a nuclear clock. This would result in a wide range of enabling capabilities such as an improved Global Positioning System (GPS), detec- tion of underground structures, or the rise of the sea level through its sensitivity to gravitational red-shifts. LANL is one of only a few institutions that can assemble a tightly- coupled effort combining experimental capabilities in actinides nuclear chemistry, atomic, nuclear and material sciences. Figure 1. Tests of spectrometer system. Plot on the left: Dark counts of our monochromator system as a function of the wave- length. Plot on the right: The 184.95 nm line of a Hg pencil lamp (running at 10 mA) observed from the indirect scattered light through a 20 micron pinhole located one foot away from the monochromator entrance slit. References

1. Kroger, L. A., and C. W. Reich. Features of the low-en- ergy level scheme of 229Th as observed in the gamma- decay of 233U. 1976. Nuclear Physics A. 259: 29.
2. Helmer, R. G., and C. W. Rewich. An excited state of 229Th at 3.5 eV. 1994. Physical Review C. 49: 1845.
3. Beck, B. R.. Energy splitting of the ground-state dou- blet in the nucleus 229Th. 2007. Physical Review Letter. 98: 142501.
4. Zhao, X.. Observation of the deexcitation of 229mTh nuclear isomerPhysica. 2012. Physical Review Letters. 109: 160801. 737

Page 738

Physics Exploratory Research Final Report Materials Study for Small Self-regulating Fission Reactor System for Space Applications Patrick R. McClure 20130767ER Abstract on the overall neutronic response of the reactor. Historically NASA has relied on 238Pu-based power Background and Research Objectives systems for space exploration. Diminishing global supply of 238Pu has reached a critical point and is adversely im- LANL is proposing a reactor for space applications that is pacting future NASA deep space science missions. LANL based on the physics of a fast reactor with heat pipes to is proposing a very simple reactor concept as an alterna- transfer heat from the nuclear fuel to Stirling engines as tive to 238Pu power sources. This simple reactor con- a means of power conversion. cept is a fast spectrum reactor that incorporates a solid Heat pipe/Stirling engine reactors have been proposed uranium core that is highly reflected and has heat pipes because of their inherent simplicity. The simplicity of as a means of heat extraction from the reactor core to these reactor argues for a reliable, safe, easy to operate the power conversion system. Unique features include and easy to manufacture reactor. One of the key fea- low thermal power and self-regulation. tures of these types of reactors is the reactivity feedback This research investigates two issues that require further mechanism that allows for the reactor to be self-con- investigation to advance this reactor concept. The first is trolling. Fast reactors in this size range are controlled by selection of the appropriate materials and the second is thermal expansion and subsequent negative reactivity the physics of self-regulation with emphasis on accident feedback. Thermal feedback lowers reactor power if conditions. less heat is extracted by the power conversion. In self- regulating cores, therefore, a narrow temperature region In the materials portion of the project, we investigated exists over which design power can be delivered to the the alternatives materials configurations for the reactor. power converter; any change in one parameter will nec- This research included selection of the uranium alloy and essarily change the other. Precise understanding of cou- its casting; selection of the reflector material; and the pled core and heat conversion system physics is essential methods used to achieve thermal bonding between the to accurately designing and/or predicting the operating heat pipes and the reactor. The materials selected have envelope of self-regulating nuclear reactors. This means huge impacts on the overall performance of the reactor that design selections such as the type of uranium alloy system. and engineering details such as the fabrication meth- ods used to achieve thermal bonding between the heat In the self-regulation portion of the project, we inves- pipes and reactor core can have huge impacts on the tigated the physics of fast reactors in more detail. A overall performance of the reactor system. The purpose unique feature of highly reflected fast reactors is the on the first phase of this research was to investigate a large percentage (up to 50%) of fissions caused by matrix of materials and bonding techniques for the fuel/ neutrons that return from the reflector. An important heat pipe to optimize the choice of materials. feature of these reflected neutrons is their long average lifetime, which can be two to four orders of magnitude Another unique feature of highly reflected fast spec- longer than in-core neutrons. As a result, a substantial trum reactor is that very large percentage (over 50%) of fraction of the reactivity “worth” (neutron multiplica- the fissions is caused by neutrons that return from the tion) is delayed as the core waits for these very-long- reflector. The result is very long lifetimes of reflected lifetime reflected neutrons to return. These “geometric neutrons as compared to those that stay within the core. delayed” neutrons therefore can have a major influence The average lifetime of neutrons in the core is on the 738

Page 739

order of 10-8 s, while the average lifetime of neutrons re- • Uranium with 2.5% Vanadium turned from the reflector is on the order of 10-6 s to 10-4 • Uranium with 1.5%, 7% and 10% Molybdenum s. In effect, this lifetime increase creates additional de- • Addition of a can (stainless-steel) around the fuel layed neutron groups, referred to as “geometric delayed” neutron groups. Although the fraction of decay-based Material properties for pure Uranium were provided by delayed neutrons is <1%, they are absolutely essential to MST-6. Material properties with the additive materials the controllability of a nuclear reactor. “Geometric de- were provided courtesy of staff at the Y-12 plant. layed” neutron groups are on a shorter time scale but can be significant because they provide a much larger fraction There are only two candidate reflector materials that could of the flux. Previous studies of “geometric delayed” neu- meet the anticipated requirements: Beryllium (Be) and trons have shown that for large insertions of reactivity, the Beryllium Oxide (BeO). Other materials will not have suf- delayed neutron groups have a pronounced effect on the ficient neutron reflection to meet launch criticality safety ability of the reactor to handle transient events and could requirements. The material properties for Be and BeO potentially reduce peak power by several multiples lower were provided by MST-6. than when “geometric delayed” neutrons are ignored. The final issue was how to create an acceptable thermal- The purpose of the second phase of this project was to structural bond between the core and the heat pipes. The modify the LANL system code, FRINK to handle geometric first concept examined was to embed the heat pipes in the delayed neutrons automatically based on data derived reflector, because this simplifies system development. The from MCNP6. This would allow for the effect of geometric second option was to bond the heat pipes to the outside of delayed neutrons to be accounted for in the safety assess- the fuel because this lowers the temperature gradient. ment of these reactor types, where ability of the reactor to survive rapid insertions of reactivity during an accident are In order to measure of success from modeling, criteria of a concern. were needed to score each design configuration. The criteria are presented below. The criteria are all weighted Scientific Approach equally: The research was performed with the two principle LANL reactor design tools. The first design tool is the radiation • Lowest system mass transport code, MCNP61. MCNP6 (Monte Carlo Neutron • Smallest size Photon) is the LANL Monte Carlo code for radiation trans- port. The MCNP6 code is widely used in studies of ad- • Smallest delta temperature across reactor fuel and vanced reactor concepts. heat pipe • Least Impact on reactor neutronic performance The second design tool is the systems modeling code FRINK2. The systems model is a point kinetics neutronics • Least impact from off normal performance model coupled to a two-dimensional heat transfer model. • Least potential for long-term material interactions The model also includes a heat pipe model and a repre- • Least impact to launch safety accidents sentation of the energy removed by the operation of the Stirling engine. The choices of materials and bonding techniques formed a “matrix of calculations” to against the success criteria. The The process for the materials study included the selection results were used to down-select to the appropriate design of candidate fuel and reflector materials and a method for for the reactor concept. bonding the heat pipes to the reactor core. The process is described in more detail in the following paragraphs. For the second portion of this project, the systems code FRINK was modified to allow the prompt neutrons to be A high-uranium density fuel is needed to design a system parsed into geometrically delayed neutron groups. This that is attractive. The ideal fuel candidate from a mass involves multiple runs of MCNP to obtain the reactivity perspective is pure uranium HEU; however, pure uranium coefficients of each subdivision of the reactor (e.g. the metal becomes very soft at high temperature (>800°C). reactor is broken into fuel region, several regions for the The following materials were examined in this study as reflector, etc.) and the average lifetime of neutrons for that candidates for the fuel: component. Once the reactivity coefficients and neutron lifetimes are calculated, they and the geometry of the • Pure Uranium model are input to FRINK. The reactivity coefficients de- • U doped with small quantities of Fe or Si termine the reactivity worth of each component based on 739

Page 740

temperature. The neutron lifetime determines magnitude on small variations within this sub-matrix. Both Uranium of reactivity worth on neutron multiplication and hence fuel and a Molybdenum alloy had positives and nega- the power in a time-step. Once known, the power in a tives. U and doped U make for a smaller, lighter core that region is used to advance the thermal solution to the next has the best neutronics (they have the highest density time step. To provide an example of the impact of “geo- of uranium.) However, material properties are pushed metrically delayed” neutrons, transient calculations were to their limit and the probability of failure is higher than performed on the proposed reactor concept. desired, (this uncertainty in strength at high temperatures was described in the previous bullet.) Molybdenum added Accomplishments to Uranium was acceptable for size, weight and neutronics For the first phase of this project the materials study evalu- and eliminates most material limits issues. However, the ated the proposed combinations of materials and bonding size and mass was less than ideal for shipping and manu- techniques. The “matrix of calculations” consisted of a set facture. The final selection a 7% Molybdenum added to of 23 calculations done with MCNP and the systems model, Uranium because the material properties removed the FRINK. The results of these calculations gave the following issue of phase change, low melting temperature and issues insights: with creep strength at high temperature. Mass estimates for 7% Moly ranged from a low of 258 kg to over 300 kg, Vanadium performed slightly worse than Molybdenum in which was competitive to the pure Uranium systems with impacting neutronics for similar material qualities. Also a mass range of 226 kg to 255 kg. For this reason, the7% the data set for Vanadium was less robust than for Molyb- Moly alloy was chosen as the final material for the reactor denum (potential for adverse thermal issues.) Vanadium design. was then dropped from further consideration. The results of the material study produced a reference Heat pipes in the reflector did not perform as well as heat preliminary reactor design. This reactor concept is shown pipes at the edge of the fuel. The heat pipes in the reflec- in Figures 1 to 3. The core is a casting of HEU with 7% Mo- tor produced too large of a delta temperature for the lybdenum as an alloy, with an OD of 11 cm and a length of power level of the reactor. By placing the heat pipes at the 25 cm. There is a 3.7 cm B4C control rod that inserts into edge of the fuel the delta temperature was lower. Also, a a hole along the axial core centerline. The reference heat mechanical press fit of the heat pipes to the fuel is possible pipes have a super-alloy steel wick and shell. The peak fuel in this location. Finally, if a gap was introduced between temperature is limited to 1200 K, which limits the core the fuel and the reflector the reflector ran cooler and power to ~4 kWt. Each heat pipe is 1.27 cm in diameter thus less excess reactivity was needed in the system. For and ~4 m long. The baseline reflector material is BeO. The these reasons, the heat pipes at the edge of the fuel were reference power conversion system configuration uses selected for the design. multiple Stirling engines to produce 800 to 1000 Watts of electricity. Beryllium Oxide and Beryllium both performed well for the system neutronically. BeO has better high temperature For the second phase of this project, a series of calcula- properties and was selected as the reflector material. tions were performed to show the importance of “geomet- rically delayed” neutrons to safety calculations performed Configurations with a higher length to diameter ratio were for a highly reflected fast reactor. A series of reactivity favored over small ratios for launch safety considerations. insertion events ranging from $0.25to$3.00 were per- This was a discriminator for some of the Molybdenum plus formed for the aforementioned reactor over a range of re- U systems. activity insertion timing that ranged from instantaneous to Doped U systems behaved neutronically and thermally 500 seconds. The results of these calculations are shown identical to pure Uranium systems. It is believed that the in Figure 4 to 5. dopents will assist in grain refinement, yield strength and The modeling of “geometric delayed” neutron groups has ultimate strength. Creep strength was believed to be unaf- a significant impact on reactivity insertions of greater than fected. However, both U and doped U have a low creep $1($1 is a reactivity insertion that causes a prompt super- strength at the desired operating temperature. This issue critical condition and is a part of the accident set run for introduces uncertainty into the performance of the system, reactor safety calculations.) Figure 4 shows the difference as discussed in the next bullet. in energy inserted into the reactor accounting for geo- Uranium versus Uranium plus Molybdenum was a hard metrically delayed neutrons and without their inclusion. decision. So, difficult that a sub-matrix of these two op- Note the two orders of magnitude difference in the results. tions was done that included an additional 22 calculations For reactivity insertion accidents, correctly modeling these 740

Page 741

accidents is paramount to the development of the reactor concept. In most cases these reactor concepts will start with several dollars of excess reactivity at startup tempera- ture. Reactivity will be inserted slowly to allow the reac- tor core to establish thermal equilibrium during startup. Precise understanding of how a sudden reactivity insertion would impact reactor startup is a necessary licensing step. Figure 5 shows the results of reactivity insertions for a range of times and the predicted peak fuel temperatures. The results provide information of the necessary reactiv- ity insertion rate to prevent the melting of the reactor fuel during startup. Impact on National Missions Figure 2. Example Heat Pipe Bonding Technique NASA uses deep space science missions to explore the underlying fabric of the universe. Deep space missions have relied on radioisotope power systems for energy. The nuclear isotope that powers these systems is Plutonium 238. Currently, the United States has a dwindling supply of Plutonium 238 putting NASA deep space missions at risk. The small reactor system proposed by Los Alamos would have many advantages. The design is simple and relies on existing technologies. The use of uranium fuel and the low power of the reactor mean that safety and regulatory is- sues are greatly reduced or eliminated. The great potential of this reactor concept was recently recognized when it received a 2013 R&D 100 award. Los Alamos sees this small nuclear reactor as a stepping- Figure 3. Example Heat Pipe Bonding Technique stone to a reinvigoration of nuclear reactors for space ap- plications. Currently NASA is attempting to fund the tech- nology development of this reactor concept. Cut backs in !”#%!.' NASA funding are making this difficult, but all indications !"#%!!’ point to this technology going forward. !”#%!#' !"#%#-’ !”#%#,' !"#%#+’ !”#%#*' !"#%#)’ !”#%#(' !"#%#&’ #’ #”)’ !’ !”)’ .’ .”)’ &’ Figure 4. Peak Power After Reactivity Insertion Figure 5. Peak Fuel Temperature vs Reactivity Insertion Rate References

1. Pelowitz, D. B.. MCNP6 User’s Manual. 2005. Los Ala- mos National Laboratory Report LA-CP-05-0137 .
2. Poston, D. I., D. D. Dixon, T. F. Marcille, and B. W. Amiri. Figure 1. Reference Reactor Concept 741 %0/.%’”-,!%+(”)(#’&%#"! !"#%!,-”’%#9”‘%1()+#(+#(“,2)%3”#456+7%1()“‘5,(% 3”#4+,’%)%#%47:(;”’%,<%=>?,%.@AB”(‘0C%DED:7%‘“F”4+”;%G7%H”I% /01’231435678’ 93:;<‘261=>?’ @75A’231435678’ 93:;<‘261=>?’ 1()+#(+#(“,2)%3”#456+7%1()“‘5,(%.80%

Page 742

FRINK – A Code to Evaluate Small Reactor Transients. 2007. In Proceedings of Space Technology and Applica- tions International Forum (STAIF-2007). (Albuquerque, NM, 2007). Vol. #800, p. 50. Melville, New York: AIP Conf. Proc.. Publications McClrue, P. R., D. I. Poston, and D. D. Dixon. EXPERIMENTAL DEMONSTRATION OF A HEAT PIPE/STIRLING ENGINE NUCLEAR REACTOR. 2013. In American Nuclear Society Annual Meeting. (Atlanta, GA, 10-15 June 2013). , p. 1020. La Grange Park, Il: ANS. Poston, D. I.. CRITICALITY AND DYNAMIC BENCHMARKING OF THE DUFF REACTOR TEST. 2013. In American Nucle- ar Society Annual Meeting. (Atlanta, GA, 10-15 June 2013). , p. 1015. La Grange Park, Il: ANS. Poston, D. I., D. D. Dixon, P. R. McClure, and M. A. Gibson. A Simple, Low-Power Fission Reactor for Space Explo- ration Power Systems. 2013. In Proceedings of Nuclear and Emerging Technologies for Space 2013. (Albu- querque, NM, 25-28 Feb. 2013). , p. 15. Albuquerque: American Nuclear Society. Poston, D. I., P. R. McClure, D. D. Dixon, and M. A. Gibson. Experimental Demonstration of a Heat Pipe/Stirling En- gine Nuclear Reactor. To appear in Journal of Nuclear Technology. Poston, D. I., P. R. McClure, and D. D. Dixon. The DUFF Ex- periment – What Was Learned?. 2013. In Proceedings of Nuclear and Emerging Technologies for Space 2013. (Albuquerque, NM, 25-28, Feb. 2013). , p. 20. Albu- querque: ANS. 742

Page 743

Physics Exploratory Research Final Report pRad Measurements of Meteorites Chad T. Olinger 20130811ER Abstract 2) selecting features identified through non-destructive Nondestructively exploring three-dimensional density p-Rad tomography for subsequent LG SIMS measure- structure of meteorites establishes new signatures asso- ments. Collaborators from the University of New Mexico ciated with meteorite structure, providing new oppor- Institute of Meteoritics Carl Agee and Rhian Jones were tunities to investigate metamorphic relationships such instrumental in quickly providing samples for the prelim- as aqueous alteration veins, chondrule size and shape inary study in December as well as a sample of the 2013 distributions as well as identification of xenoliths for Chelyabinsk meteorite fall for p-Rad in this 3-month subsequent isotopic and elemental analyses. This LDRD reserve effort. In addition, we procured two samples of reserve proposal continues to develop capabilities using the Murchison meteorite, with the intent that measure- Proton Radiography (p-Rad) to nondestructively obtain ments of this relatively abundant sample will eventually meteorite structural information, a capability demon- motivate future P-Rad of the similar Sutter’s Mill meteor- strated for the first time in December 2012. In addition, ite, which, similar to the Chelyabinsk meteorite, repre- we use p-rad density signatures to guide sectioning for sents a bolide event equivalent to a 4kT explosion over complementary study using the Cameca IMS 1280 Large California on April 22, 2012. Geomoetry Secondary Ion Mass Spectrometer (LG SIMS) Scientific Approach and Accomplishments and scanning electron microscopy (SEM). In addition to these two activities, this project also took advantage of Due to activation of short-lived radionuclides from pro- a target of opportunity using the Lujan Center’s neutron ton radiography and the short (3-month) time frame of tomography capability. Very preliminary analysis sug- this study, we were unable to perform fresh P-Rad mea- gests the ability to nondestructively identify the size surements on the same samples we used for SEM and and location of specific mineral phases through neutron LG-SIMS measurements. Our strategy therefore evolved resonance spectroscopy using pixelated time-of-flight to performing SEM and LG SIMS measurements on the measurements. Abbott and Milton samples radiographed in December 2012, and obtaining new P-Rad measurements on one of Background and Research Objectives the Chelyabinsk meteorite samples provided by UNM. A one day target of opportunity was exploited in De- Programmatic efforts had priority on P-Rad beam time, cember of 2012 radiographing an H chondrite meteoritic and uncertainties associated with this encouraged us sample know as Abbott, which represents material that to also pursue a parallel path of neutron tomography at was processed on the surface of a planetessimal and the Lujan center. We had sufficient samples to provide a pallasite meteorite named Milton, which represents each of these three measurement pathways separate a rock formed near the core of a highly differentiated samples, with the goal of performing complementary parent body. These samples provided challenges for de- measurements where possible. The Chelyabinsk sample velopmental p-Rad tomography techniques. Significant that had p-Rad tomography performed is awaiting neu- internal structure could be identified in both of these tron tomography later this week. quick turn-around samples, demonstrating the power of p-Rad for geologic/planetary science studies and moti- To date, we have obtained bulk elemental compositional vating the current LDRD reserve proposal. maps on cross-sections obtained on the Milton (Figure 1) and Abbott (Figure 2) meteorites. The December 2012 We had the initial goals of 1) expanding on the number P-Rad tomographs of Abbott demonstrated a significant and types of meteorite samples analyzed with p-Rad and 743

Page 744

rind around the entire perimeter, which we conjectured to the analysis algorithm. These improvements prevented might represent fusion crust or desert varnish. However, the computational artifact of the thick “rind” observed in SEM elemental maps did not reflect any systematic change the original Abbott tomograph in December from being in elemental abundance across this zone. We now be- repeated in the more recent computed tomograph image lieve it is possible that this apparent rind was an artifact of the Chelyabinsk meteorite. Moreover, spatial resolu- of the computed tomographic inversion, where limbing tion has improved from ~100 microns in December 2012 (reduced energy attenuation where the proton path length to ~20 – 40 microns in the September 2013 radiograph of is reduced) can be an issue. These SEM images also reveal Chelyabinsk. Each of these tools individually establishes that what we believed to be chondrules within the mete- capabilities for nuclear forensics and potentially nuclear orite are, at least in many cases, silicates enclosed by metal archeology. Combined, they provide the opportunity for inclusions, but not necessarily individual chondrules with targeted destructive analyses, further advancing these rims. The SEM maps do reveal a highly heterogeneous fields while also advancing the state of the art for geologic structure that can now be correlated with the original and planetary science studies, all of which support the na- P-Rad tomographs. These SEM maps also transect veining tion’s leadership in science. features that show enhanced concentrations in Al and Fe. Plans were to obtain oxygen isotope transects with the LG SIMS on the Milton sample in mid-September at the time of the writing of this report, those values have not yet been reported. A ~1 cm diameter sample of the Chelyabinsk meteorite was radiographed at P-Rad under this effort, and can now be used to inform subsequent sampling. We have a movie of the 720 ½ degree steps of the P-Rad of this sample, as well as a computed tomograph movie of virtual slices mov- ing up through this sample (Figure 3 is a representative image from the raw radiograph). We also acquired spatially resolved neutron time-of-flight spectral transmission images of a ~ 3 cm sample of the Chelyabinsk meteorite (Figure 4). Spectral features in Fig- ure 4 reflect a clear neutron absorption feature at 129 eV. Figure 1. Milton SEM Elemental Map At the time of the writing of this report, the data are less than 24 hours old; the neutron resonance feature is consis- tent with 59Co absorption feature. This would identify the mineral phase to be most likely kamacite; having identified this feature, the entire sample can now be reanalyzed to nondestructively identify those phases in 3D. Our plan moving forward is to publish these results and develop LDRD and NASA cosmochemistry proposals that build on these preliminary results. Impact on National Missions Nuclear forensics is a significant direction for part of the global security mission of LANL. While these measure- ments are not specifically directed at nuclear forensics, they do exercise capabilities relevant to nuclear forensics and have generated results that are publishable in the open literature and demonstrate the value of coordinated nondestructive radiography techniques with sectioning Figure 2. Abbott SEM Elemental Map and subsequent elemental and isotopic mapping of sam- ples. These correlated measurements informed the p-Rad computed tomographic analysis resulting in improvements 744

Page 745

Figure 3. Proton Radiograph of Chelyabinsk Meteorite Figure 4. Neutron Radiograph of Chelyabinsk Meteorite with time-of-flight spectrum showing a Co-59 absorption feature. Publications Olinger, C. T.. Proton radiography of meteorite samples. 2013. Presentation to The Albuquerque Astronomical Society. 745

Page 746

Physics Postdoctoral Research and Development Continuing Project Quantum Simulations: From Superconductivity to Nanoscale Electronics Eddy M. Timmermans 20110711PRD2 Introduction We study cold atom systems and their use as quantum Benefit to National Security Missions simulators of superconductors and many-body effects This work may bring better understanding of high-Tc in nanoscale transport. This is a competitive field with superconductors: materials with zero electrical resis- broad implications, e.g., in condensed matter physics, tance. These materials have important applications like, nanoscale science, and quantum information. Quantum e.g., they can eliminate virtually all the loss of energy simulation uses the highly controllable nature of one that goes into resistive heating in the lines transmitting quantum system (here, a cold atom cloud) to create electricity. This research will also advance fundamental simplified analogs of complex materials and devices in understanding of materials on the basic quantum level. order to gain direct theoretical and experimental insight Both goals are becoming focus areas for numerous into the underlying physical phenomenon. We develop, sponsors including DOE and BES, for example, for both theoretically, a scheme for emulation of various super- applied and basic aspects of energy research. conductors in cold atom setups to address problems inaccessible in traditional condensed matter settings. In Progress particular, we develop a unified theory of spectroscopic The PD searched the literature on the response of measurements on cold atom and superconducting superconductors and superfluids to external perturba- systems in order to help understand the mechanisms tions. Surprisingly, most authors did not notice that their of high-Tc superconductivity as revealed by cold-atom results violate the important conservation law of par- simulations. From here, more complex systems will be ticle numbers. We found that this inconsistency is due studied to understand quantum and nonequilibrium to the neglect of collective excitations associated with effects in nanoscale electronic systems, as well as to the coherent state of superconductor and superfluids. provide additional measurement techniques capable of We successfully developed a consistent description of probing the cold atom systems. This will elucidate the the response of superconductors and superfluids that role of electron-electron interactions and dissipation in satisfies the conservation law. Also we addressed use- electronic transport – effects that are difficult to capture ful criteria for checking whether a theory respects the in traditional theoretical approaches to transport. conservation law and point out some pitfalls if one tries to fix the problem naively. Importantly, the PD’s theory The postdoc’s (PD) work characterizes the promising provides better estimations of the excitation spectrum methods for extracting information from cold atom and compressibility. systems that may reveal the main phenomena that underpin high-Tc superconductivity. Superconducting The PD has developed codes for studying quantum materials have the potential to revolutionize power dis- dynamics of atoms and applied them to new designs of tribution, i.e., create more efficient means of distributing devices using atoms and engineered optical trapping electricity that will be helpful in moving to green sources potentials. Since atoms are charge-neutral, connecting of energy among other things. Also, the PD investigates the system to a battery cannot drive an atomic current. nonequilibrium and inhomogeneous cold atom systems However, a recently developed technique let atoms in order to understand quantum effects in electronic exchange momentum and energy with photons and we transport, which is crucial in designing functional na- developed a method for driving atoms based on this noscale devices. scheme. We found that the current of a fully quantum system exhibits very different behavior from what one 746

Page 747

may expect by treating the atoms as classical objects. We phases and understand interplays among different species. also propose new designs of devices that allow one to con- When two systems exchange particles, correlations start to trol the direction and magnitude of the atomic current by build up. A quantity called entanglement entropy describes tuning the external photon source. This level of control had how correlated the two systems are and it is of broad inter- not been possible in the literature prior to the PD’s work. est ranging from black-hole physics to quantum informa- Experimentally one can mix two species of atoms together. tion. We will study the dynamics of the entanglement To elucidate the physics of such systems, the PD developed entropy when atoms are driven by a bias and flow across a theoretical framework for describing such a mixture. an artificial lattice generated by laser light. There have Since the inter-species repulsive interactions are tunable, been theoretical predictions based on modeling atoms as the system can either be a homogeneous mixture or it can billiards moving in the lattice. However, atoms should be separate into two parts with different species avoiding modeled as quantum waves and we can study them using each other (phase separation). At extremely low tempera- our quantum description. We run simulations and investi- ture, there has been an effective description of this system. gate how important quantum effects are in the entangle- However, at finite temperature there are many thermal ment entropy. excitations and a more sophisticated treatment is needed. We continue our study on energy transport in complex Using advanced quantum-mechanical methods the PD molecules using both analytical calculations and molecu- successfully included the interactions among those ther- lar-dynamics simulations to investigate how to control the mal excitations and found that the two possible states at thermal conductance of a given material. We found that low temperature continue to exist at higher temperatures. by increasing the binding energy on each atomic site of a Moreover, the inter-species interactions can be controlled molecular chain, the molecule conducts less heat. We aim by an external magnetic field. The PD thus generalized our to study how the thermal conductance changes if local theory to include dynamical effects when the interaction is structures of a molecule are altered as scientists do regu- changed suddenly. From the PD’s simulations we observed larly nowadays. Sorting out what kinds of local structures that the system develops interesting patterns reflecting can help increase or decrease the heat conductance could the stable state after the change. Our work provides more lead us to better designs of thermally controlled devices. insight into multi-species superconductors and superfluids and the results have been published. Conclusion In previous work the PD found that the heat transport The results from quantum simulations may provide more properties of DNA molecules can change dramatically insight into the parallels between cold atom and high- when DNA dissociates from its double-helix structure to Tc superconducting systems, which represents a unique two single strands due to high temperature. The PD suc- opportunity in the field of physics. The analogies may, cessfully designed a device utilizing this property of DNA. eventually, lead to the discovery of novel superconductors This device has a piece of DNA clamped by two metallic of higher transition temperatures. Superconductivity is one leads with different temperatures. By choosing differ- of the most significant discoveries in the 20th century and ent sequences or length of DNA, this device has thermal has important applications in dissipationless power trans- switching behavior at different temperatures. Immedi- mission and advanced medical instruments. The unified ately after his publication, a European and a US group theory, which we will develop for quantum transport in performed experiments and demonstrated novel devices nanoscale will lead to new design of nano devices. These inspired by our work. new nano devices will play a crucial role in next-generation electronics. Future Work This project addresses three major challenges: Publications Chern, G. W., C. C. Chien, and M. Di Ventra. Dynamically Recent experiments have demonstrated the ability to mix generated flat-band phases in optical kagome lattices. atoms in different quantum states/species. We plan to Physical Review Letters. extend our theoretical study of atomic systems to those Chien, C. C.. Spatially varying interactions induced in atom- multi-component systems. One major challenge is to ic gases by optical Feshbach resonance . 2012. Physics include inter-species interactions while respecting the Letters A. 376: 729. number conservation of each species. We derive the cor- responding equations describing physical quantities that Chien, C. C., D. Gruss, M. Di Ventra, and M. Zwolak. Inter- can be measured in experiments. By comparing our results action-induced conducting-nonconducting transition with experimental data, we may explore novel quantum of ultra-cold atoms in 1D optical lattices . 2013. New 747

Page 748

Journal of Physics. 15: 063026. Guo, H., C. C. Chien, and Y. He. Gauge invariant linear re- sponse theory of relativistic BCS superfluids . 2012. Chien, C. C., F. Cooper, and E. Timmermans. Large-N ap- Physical Review D. 85: 074025. proximation for single- and two-component dilute Bose gases . 2012. Physical Review A. 86: 023634. Guo, H., C. C. Chien, and Y. He. Theories of linear response in BCS superfluids and how they meet fundamental Chien, C. C., H. Guo, and K. Levin. Comment on “Density constraints. 2013. Journal of Low Temperature Physics. and spin response of a strongly-interacting Fermi gas in 172: 5. the attractive and quasi-repulsive regime”. 2012. Physi- cal Review Letters. 109: 118901. Guo, H., Y. He, C. C. Chien, and K. Levin. The compressibility in strongly correlated superconductors and superfluids: Chien, C. C., J. She, and F. Cooper. Mean-field description from BCS to BEC. To appear in Physical Review A. of pairing effects, BKT physics, and superfluidity in 2D Bose gases. Annals of Physics. Guo, H., Y. Li, Y. He, and C. C. Chien. Density and spin linear response of atomic Fermi superfluids with population Chien, C. C., K. A. Velizhenin, Y. Dubi, and M. Zwolak. Tun- imbalance in BCS-BEC crossover. Physical Review A. able thermal switching via DNA-based nano devices. 2013. Nanotechnology. 24: 095704. Mihaila, B., F. Cooper, J. F. Dawson, C. C. Chien, and E. Tim- mermans. Analytical limits for cold-atom Bose gases Chien, C. C., M. Zwolak, and M. Di Ventra. Bosonic and fer- with tunable interactions. 2011. Physical Review A. 84: mionic transport phenomena of ultra-cold atoms in 1D 023603. optical lattices . 2012. Physical Review A. 85: 041601. Mihaila, B., J. F. Dawson, F. Cooper, C. C. Chien, and E. Tim- Chien, C. C., and F. Cooper. Quench dynamics and emer- mermans. Auxiliary field formalism for dilute fermionic gence of phase separation in two-component atomic atom gases with tunable interactions. 2011. Physical Bose gases at zero temperature and above the BEC crit- Review A. 83: 053637. ical temperature. 2013. Physical Review A. 87: 045602. Velizhanin, K. A., C. C. Chien, Y. Dubi, and M. Zwolak. Driv- Chien, C. C., and M. Di Ventra. Dynamical crossover be- ing denaturation: Nanoscale thermal transport as a tween the infinite-volume and empty-lattice limits of probe of DNA melting. 2011. Physical Review E. 83: ultra-cold fermions in 1D optical lattices . 2012. Euro- 050906. physics Letters. 99: 40003. Wulin, D., B. M. Fregoso, H. Guo, C. C. Chien, and K. Levin. Chien, C. C., and M. Di Ventra. Controlling transport of Conductivity in Pseudogapped Superconductors: The ultra-cold atoms in 1D optical lattices with artificial Role of the Fermi Arcs. 2011. Physical Review B. 84: gauge fields. 2013. Physical Review A. 87: 023609. 140509. Cooper, F., B. Mihaila, J. F. Dawson, C. C. Chien, and E. Tim- Wulin, D., H. Guo, C. C. Chien, and K. Levin. Spin Transport mermans. Auxiliary field approach to dilute Bose gases in Cold Fermi gases: A Pseudogap Interpretation of with tunable interactions . 2011. Physical Review A. 83: Spin Diffusion Experiments at Unitarity . 2011. Physical 053622. Review A. 83: 061601. Cooper, F., C. C. Chien, B. Mihaila, J. F. Dawson, and E. Tim- Wulin, D., H. Guo, C. C. Chien, and K. Levin. Two-compo- mermans. Composite-field Goldstone states and Higgs nent optical conductivity in the cuprates: A necessary mechanism in dilute Bose gases . 2012. Physical Review consequence of preformed pairs . 2012. Physical Re- A. 85: 023631. view B. 86: 134518. Dawson, J. F., F. Cooper, C. C. Chien, and B. Mihaila. Lead- ing-order auxiliary field theory of the Bose-Hubbard model. 2013. Physical Review A. 88: 023607. Doi, K., M. Tsutsui, T. Ohshiro, C. C. Chien, M. Zwolak, M. Taniguchi, T. Kawai, S. Kawano, and M. Di Ventra. Elec- trochemical response of biased nanoelectrodes in solu- tion. Journal of Physical Chemistry C. Guo, H., C. C. Chien, Y. He, and K. Levin. Fundamental con- straints on linear response theories of Fermi superflu- ids above and below Tc. 2013. International Journal of Modern Physics B. 27: 1330010. 748

Page 749

Physics Postdoctoral Research and Development Continuing Project Shortcuts to Adiabaticity in Quantum Devices Wojciech H. Zurek 20110736PRD2 Introduction in Atomic, Molecular, and Optical Physics, an other re- The concept “Fast Good” pushed forward by the world- view accepted in Journal of Physics: Condensed Matter. leading chef F. Adria might be a motto born in modern This work has been devoted to the following accomplish- cuisine, but it actually applies to many areas of the ments. day-to-day life. When it comes to technology there is Prior to his arrival to LANL, the JRO fellow Adolfo del certainly no doubt that shortening production processes Campo proposed theoretically an experimental test of pays great dividends, and the quest for faster and more the Kibble-Zurek mechanism, the paradigmatic theory powerful devices such as computers has been running for defect formation across a phase transition. After for decades as dramatically illustrated by Moore’s law. starting a collaboration with the experimental group at But all quantum technologies crash with the Fast Good PTB led by Tanja Mehlstaubler, the first experimental trend at a fundamental level. At a microscopic scale, results on the universal defect formation dynamics in higher performance of a quantum device comes gener- inhomogeneous systems have been reported jointly. The ally at the price of slowing down its running time, as results are currently under consideration in Nature Com- dictated by text-book theorems in quantum theory. munications for publication. The focus of this research project is to spur quantum The understanding of inhomogeneous phase transitions technologies by changing this state of affairs and finding was reviewed and extended in collaboration with the PI ways out to fulfill both sides of the Fast-Good dichotomy. (Wojciech Zurek) and Tom Kibble, the founding fathers It will deepen the current understanding of dynamical of the paradigmatic model to describe the universal process, put forward new schemes in the state-of-the- dynamics of defect formation across phase transitions art experimental control of quantum devices, and accel- (the Kibble-Zurek mechanism). It has been shown that in erate the development of new benchmarks in quantum inhomogeneous systems defect formation can be con- metrology, simulation and computation. trolled, a possibility of great interest in material science. Benefit to National Security Missions The first review on the field of shortcuts to adiabaticity This project will allow for a better control of quantum was also coauthored by the Adolfo del Campo. systems. This has broad applications, that cut across many areas of interest to DOE and LANL. This includes New shortcuts to adiabaticity have been developed, nanoscience, quantum sensitive measurement, informa- generalizing the counter-diabatic driving technique tion processing, and related areas. While this is admit- which allows one to drive a system quickly mimicking tedly long-term research, the measurement tools that adiabaticity. The new approach is extremely powerful may develop could impact any national security mission since it applies to many-body interacting systems as well that depends on sensing technology. as non-linear systems where the old theory was not ap- plicable. Experimental tests of this technique have been Progress designed and proposed. In the last 12 months, Adolfo del Campo has submitted 7 publications to peer-review journals including two Physi- Further, shortcuts to adiabaticity have been applied to cal Review Letters, a third paper under consideration finite-time thermodynamics, leading to an experimen- in this journal, two manuscripts under consideration in tally realizable scheme to engineer with trapped ions Nature Communications, a review accepted in Advances a superadiabatic engine which operates at maximum 749

Page 750

efficiency and with tunable output power. Publications Campo, A. del. Shortcuts to Adiabaticity by Counterdiabatic Future Work Driving. 2013. PHYSICAL REVIEW LETTERS. 111 (10): -. In quantum technologies, when implementing in the labo- Campo, A. del, I. L. Egusquiza, M. B. Plenio, and S. F. Huel- ratory a given process or operation, one faces a trade-off ga. Quantum speed limits in open system dynamics . between the accuracy and speed of implementation. This 2013. Physical Review Letters. 110: 050403. is dictated by the quantum adiabatic theorem. Its break- down leads inevitably to formation of excitation, which is Campo, A. del, M. M. Rams, and W. H. Zurek. Assisted undesirable for a wide range of applications, such as the finite-rate adiabatic passage across a quantum criti- preparation of novel quantum phases in quantum simula- cal point: Exact solution for the quantum Ising model. tion, and adiabatic quantum computation. Suppressing 2012. Physical Review Letters, 109, 115703 (2012) . excitations is also of interest to a variety of operations in 109: 115703. the laboratory, like entangling strings of atom. Campo, A. del, T. W. B. Kibble, and W. H. Zurek. Causality and non-equilibrium second-order phase transitions in The goal is to propose methods to mimic adiabaticity, even inhomogeneous systems. 2013. J. Phys.: Condens. Mat- when the dynamics is ultrafast and adiabaticity breaks ter. 25: 404210. down. In other words, the project aims at preparing the same target state that would result in a truly adiabatic Campo, A. del, and M. G. Boshier. Shortcuts to adiabaticity (slow) process, but imposing simultaneously a fast imple- in a time-dependent box . 2012. Scientific Reports. 2: mentation. The JRO fellow will continue developing new 648. techniques in the field of shortcuts to adiabaticity and Campo, A. del, and W. H. Zurek. Universality of phase tran- explore their applications. sition dynamics: topological defects from symmetry breaking. Professor Tom Kibble’s 80th birthday’s Sym- This will include the design of of a superadiabatic engine posium. Edited by Gauntlett, J. P.. based on the counter-diabatic driving of particles undergo- ing Brownian dynamics, operating at maximum efficiency, Nigmatullin, R., A. del Campo, G. De Chiara, G. Morigi, and the design of a new experiment to test the Kibble- M. B. Plenio, and A. Retzker. Formation of helical ion Zurek mechanism in colloids. chains . 2011. LA-UR 11-06244. With respect to the trapped ions which were used in the Pyka, K., J. Keller, H. L. Partner, R. Nigmatullin, T. Burger- experimental results reported in FY13, colloids allow to meister, D. M. Meier, K. Kuhlmann, A. Retzker, M. B. access large system sizes, can be easily imaged, and are Plenio, W. H. Zurek, A. del Campo, and T. E. Mehl- greatly overdamped. All this features will allow us dto staubler. Topological defect formation and spontane- ous symmetry breaking in ion Coulomb crystals. 2013. overcome the main drawbacks of ion traps, and generate NATURE COMMUNICATIONS. 4: -. a new set of experiments on the universal dynamics of defect formation. Torrontegui, E., S. Ibanez, S. Martinez-Garaot, M. Modug- no, A. del Campo, D. Guery-Odelin, A. Ruschhaupt, X. Finally, Adolfo del Campo will study the fundamental limits Chen, and J. G. Muga. Shortcuts to Adiabaticity. 2013. to the speed of evolution which constitutes ultimate limits ADVANCES IN ATOMIC, MOLECULAR, AND OPTICAL to quantum computing and information processing. PHYSICS, VOL 62. 62: 117. Conclusion Though quantum technologies such as atomic clocks have been benefiting society since the fifties, it is clear that they are playing and ever increasing role. They have the potential to do better and completely outperform con- ventional electronics. The exploitation of quantum laws in metrology, shows how the race for ever better clocks have brought technological advances such as the GPS required for satellite navigation, ultimately transforming society in a global scale through new communication and business models. Other quantum technologies in a nascent stage, might revolutionize the fields of simulation, computation and communication. 750

Page 751

Physics Postdoctoral Research and Development Continuing Project Influence of Velocity Shear on Turbulent Magnetic Reconnection at the Earth’s Magnetopause William S. Daughton 20110750PRD3 Introduction forefront of high performance computing, both for scal- Magnetic reconnection is a basic process that occurs ing the calculations to large numbers of cores (~200,000) in hot ionized gases known as plasmas. As this process and in remote visualization and analysis of the results. develops, magnetic field energy is rapidly converted into We thus build capability for hydrodynamic simulations particle kinetic energy. This process plays an important of nuclear weapons. Scientific results from this research role in a wide variety of problems in space, laboratory will benefit NASA missions focused on understanding the and astrophysical plasmas. The focus of this project is near-Earth space environment. In particular, the upcom- on the dynamics of magnetic reconnection within the ing Magnetospheric MultiScale (MMS) mission will focus Earth’s magnetosphere – a protective magnetic bubble on the kinetic physics of magnetic reconnection with the that partially shields our planet from energetic particles. Earth’s magnetosphere. The same science is essential to In this environment, magnetic reconnection often oc- understanding the environment for satellites that sup- curs in the presence of intense velocity shear, which can port nuclear nonproliferation and other national security drive the formation of large-scale flow vortices, in much missions. the same way as in ordinary fluids. However, in high- Progress temperature plasmas, these vortices are closely coupled to the evolution of the magnetic field lines, which can Ionized gas (or plasma) is continually streaming outward then trigger the process of magnetic reconnection. Thus from the Sun. When this stream of plasma interacts in reality, these two processes can interact in a complex with the magnetic field of the Earth, it produces the manner, which can only be fully described with three- magnetosphere – a protective magnetic bubble that par- dimensional computer simulations. This project will lead tially shields our planet from energetic particles. The dy- to a new understanding regarding how velocity shear in- namics of the magnetosphere has long been of interest fluences reconnection by employing some of the world’s both scientifically and for practical impacts to modern fastest supercomputers such as the Kraken and Jaguar infrastructure, including damage to satellites, interrup- machines. The research will involve a combination of tion of GPS service, and power grid failures. The focus basic theory and large-scale kinetic simulations that de- of this Director’s fellowship project is to examine several scribe the temperature plasma at the most fundamental natural processes that can give rise to leaks in this shield. level. The project will utilize the plasma simulation code These leaks typically occur within a thin boundary layer VPIC, which has been carefully optimized to take full in which both the magnetic field and plasma velocity advantage of these new computers. In order to intelli- change rapidly. The rapid change in the magnetic field gently setup and interpret these simulations, some basic drives a process known as magnetic reconnection, which theoretical approaches will also be utilized. can change how field lines are connected in the plasma and release energy stored in these fields. On the other Benefit to National Security Missions hand, the strong velocity shear in these boundary lay- This research will exploit the newest generation of pet- ers can drive an instability known as Kelvin-Helmholtz, ascale supercomputers using a kinetic simulation code which leads to swirls and vortices in the flow. While both (VPIC) specially optimized for these machines. Some of these processes are interesting and important, they of these calculations will be performed locally at Los are usually studied separately. However, in the real mag- Alamos using NNSA computers, while other calculations netosphere they may often occur together and influence will be performed on flagship DOE supercomputers such each other in complex ways. We are addressing this as Jaguar at Oak Ridge. These efforts are at the very possibility by performing large-scale computer simula- 751

Page 752

tions using the LANL developed VPIC code which describes Publications the physics of these processes in a rigorous manner. Karimabadi, H., V. Roytershteyn, M. Wan, W. H. Matthaeus, During the past fiscal year, we have completed the first W. Daughton, P. Wu, M. Shay, B. Loring, J. Borovsky, 3D simulations of this process using 100,000 cores on the E. Leonardis, S. C. Chapman, and T. K. M. Nakamura. Jaguar computer at Oak Ridge National Laboratory. Many Coherent structures, intermittent turbulence, and dis- sipation in high-temperature plasmas. 2013. PHYSICS additional smaller scale simulations were also performed OF PLASMAS. 20 (1): -. on the Mustang machine at LANL. As described in a new paper, these 3D simulations feature the development of Nakamura, T. K. M., R. Nakamura, A. Alexandrova, Y. Kubo- magnetic flux rope structures around the periphery of the ta, and T. Nagai. Hall magnetohydrodynamic effects for vortex, leading to a complex tangled magnetic field and three-dimensional magnetic reconnection with finite fully turbulent layers. These 3D features lead to enhanced width along the direction of the current. 2012. JOUR- mixing of particles across the boundary layer, which may NAL OF GEOPHYSICAL RESEARCH-SPACE PHYSICS. 117: be important for modeling the low-latitude magnetopause -. boundary layer. While the paper describing these results is Nakamura, T., W. Daughton, H. Karimabadi, and S. Erikss- still in review, the preliminary results have attracted con- son. Three-Dimensional Dynamics of Vortex-Induced siderable attention at the Fall AGU meeting and Takuma Reconnection and Comparison with THEMIS Observa- Nakamura was asked to give an invited talk at the upcom- tions. 2013. Journal of Geophysical Research. 118: ing IAGA meeting, one of the largest international confer- doi:10.1002/jgra.50547. ences in geophysics. During the final months of this project, we will work to- wards analyzing existing simulations and performing a few additional simulations with more realistic initial conditions. We expect to finish several additional journal articles from this work. Future Work This Director’s funded fellowship will reach completion in November, 2013. As described in our new JGR paper, the results from this project are leading to a new under- standing of vortex-induced reconnection at the Earth’s magnetopause. These are the first fully kinetic 3D simula- tions of this process, and the results have raised a num- ber of new questions as well as testable predictions for future spacecraft observations. During the remainder of this project, we will perform a few additional simulations and work towards finishing the analysis of the existing 3D results. Our goal is to develop a better understanding of how the mixing across the layer is influenced by the initial conditions. We will document these results in additional journal articles, and present our findings at both domestic and international meetings. Conclusion Scientists believe that magnetic reconnection plays a central role in a wide variety of applications such as geo- magnetic substorms, solar flares and laboratory fusion machines. The ability to predict the complex nonlinear be- havior of high temperature plasma is of great practical im- portance for these applications. For example, the results of this project may have direct application to modeling the plasma environment surrounding the Earth where commu- nication satellites play a crucial role in modern society. 752

Page 753

Physics Postdoctoral Research and Development Continuing Project Non-equilibrium Transport in Nanoscale Thermal Energy Harvesting Jianxin Zhu 20120737PRD1 Introduction equilibrium thermoelectric transport is understood, the Energy harvesting and waste is a great bottleneck in the efficiency of thermoelectricity at the nanoscale can be supply of energy resources to a sustainable economy. optimized. In addition to developing carbon-free green energy Furthermore, by coupling these electronic and local sources, it is also beneficial to enhance the efficiency vibrational degrees of freedom to photons (light), the of energy utilization. On this aspect, control of energy thermoelectric transport out of equilibrium can be con- at the nanoscale creates unprecedented opportunities trolled. The research is aligned with the Energy Frontier for directing and conversion of energy. Among various Research Centers Program established by Department of forms of energy, heat, electricity, and light are three of Energy on the aspect of energy efficiency. the most familiar. Future generations of multimode en- ergy harvester rests with the integration of electronics, Progress photonics, and phononics (carrying heat). Therefore, it We have developed a thermal transport theory based is critical to gain a fundamental understanding of non- on the scattering wave approach for various interface equilibrium transport of electrons/phonons (vibrational systems, including normal metal or topological insu- thermal energy) at the nanoscale, the interplay between lator and superconductor interface. By applying the them as well as their coupling to photons. interfacial heat transfer formula, we have uncovered In this project, we will achieve this goal by consider- several anomalous thermal properties, such as thermal ing out-of-equilibrium transport through open quan- energy’s Klein tunneling, asymmetric Kapitza resistance tum nanoscale systems in contact with electrical and/ and negative differential thermal resistance. We have or thermal reservoirs. We will develop new theoretical also discovered a new class of Hall-like effects, that is, approaches to tackle the problem in a wide range of the asymmetric Andreev reflection is able to induce the parameter space, like the strength of electron-electron electric and thermal Hall-like effects in metal/anisotropic interaction and electron-phonon coupling. Scientifically, superconductor junctions. These findings could have we will address the effects of these interactions on ther- potential applications for the smart energy control in moelectricity, and pay special attention to the conditions various hybridized mesoscopic systems. for optimal thermoelectric figure of merit value. Further- We have uncovered the fundamental effects of geomet- more, we will consider the coupling of local vibration ric phase and topological bands in several scenarios of modes and electrons in these nanoscale systems to a energy transport. Our main results include: Geometric coherent optical field, and investigate the optical control phase effect can provide additional channel for energy of thermoelectric transport out of equilibrium. transfer, thus enhances the transfer efficiency; Topologi- cal bands of magnon can transfer the energy and spin Benefit to National Security Missions information robustly and non-dissipatively; The complex The development of a fundamental understanding of bands of nonequilibrium open systems can be clari- non-equilibrium transport of electrons/phonons (vibra- fied by the braid groups and process topological phase tional thermal energy) at nanoscale will have an im- transitions: the braid phase transition. These uncovered mediate impact to energy generation and harvesting. fundamental effects have potential applications for ef- It will support and guide experimental effort on energy ficient energy transport and harvesting. research at LANL. Once the effects of electron-electron and electron-lattice vibration mode coupling on the non- 753

Page 754

Spin Seebeck effect is a new phenomenon that tempera- Ren, J.. Predicted rectification and negative differential spin ture bias can produce a pure spin (without electric) current Seebeck effect at magnetic interfaces. 2013. arXiv.org. and an associated spin voltage. We have uncovered the in- Ren, J., J. Zhu, J. E. Gubernatis, C. Wang, and B. Li. Thermo- triguing rectification effect and negative different conduc- electric transport with electron-phonpn coupling and tance for spin Seebeck effect, so that spin Seebeck diode electron-electron interaction in molecular junctions. and transistor are now within our grasp. We expect our 2012. Physical Review B. 85: 155443. findings will act as new methods facilitating the functional use of heat and opens a new possibility of spintronics and Ren, J., S. Liu, and B. Li. Geometric heat flux for classical spin caloritronics. thermal transport in interacting open systems. 2012. Physical Review Letters. 108: 210603. Future Work Ren, J., and J. Zhu. Heat diode effect and negative differ- We will expand our present focus to include more many- ential thermal conductance across nanoscale metal- body degrees of freedom and extend the thermal energy dielectric interfaces. 2013. Physical Review B. 87: harvesting to the spintronics and spin caloritronics. By 241412(R) . integrating electron-phonon, electron-spin (magnon), magnon-phonon, electron-photon interactions in the non- Ren, J., and J. Zhu. Anomalous energy transport across equilibrium transport study of hybrid condensed matters topological insulator superconductor junctions. 2013. Physical Review B. 87: 165121. and nano-devices with interfaces, we expect to integrate the thermoelectricity, optomechanics, photovoltaics to Ren, J., and J. Zhu. Theory of asymmetric and negative dif- achieve the multi-mode efficient energy harvesting. ferential magnon tunneling under temperature bias: Towards a spin Seebeck diode and transistor. 2013. Upon gaining the knowledge of functional thermal devices, Physical Review B. 88: 094427. such as thermal diode, transistor, switch, memory, we are going to explore the design principle of “thermal meta- Ren, J., and J. Zhu. Asymmetric Andreev reflection induces materials”. We will spend time on developing the special electric and thermal Hall-like effects in anisotropic su- conformal field theory to the thermodynamics and heat perconductors. 2013. arXiv.org. transfer in various materials with possible hybrid inter- Ren, J., and N. A. Sinitsyn. Braid group and topological faces. Heat flux cloaking and concentrating in designed phase transitions in nonequilibrium stochastic dynam- solid states would be the first attempt to achieve the goal ics. 2013. Physical Review E. 87: 050101(R). of “thermal metamaterials.” Yan, G., J. Ren, Y. Lai, C. Lai, and B. Li. Controlling complex Conclusion networks: how much energy is needed?. 2012. Physical This project will develop a theoretical understanding of Review Letters. 108: 218703. non-equilibrium transport of electrons and phonons (carry- Zhang, L., J. Ren, J. S. Wang, and B. Li. Topological magnon ing thermal energy) at the nanoscale. It will investigate the insulator in insulating ferromagnet. 2013. Physical Re- effects of electron-electron interaction and electron-local view B. 87: 144101. vibration mode coupling on the non-equilibrium transport at the nanoscale. The study will address the optimization of the thermoelectric energy conversion. Finally, we will explore how thermoelectric transport out of equilibrium through the nanoscale systems will be affected by the cou- pling of electronic and local vibrational degrees of freedom to a coherent optical field, leading to the possibility of opti- cal control of energy conversion efficiency. Publications Chen, T., X. Wang, and J. Ren. Dynamic control of quantum geometric heat flux in a nonequilibrium spin-boson model. 2013. Physical Review B. 87: 144303. Li, N., J. Ren, L. Wang, G. Zhang, P. Hanggi, and B. Li. Pho- nonics: manipulating heat flow with electronic analogs and beyond. 2012. Reviews of Modern Physics. 84: 1045. 754

Page 755

Physics Postdoctoral Research and Development Continuing Project High Efficiency Upconversion of Infrared Radiation in Semiconductor-metal Nanostructures for Applications in Solar Energy Victor I. Klimov 20120746PRD1 Introduction this project can potentially help in the realization of such One factor that limits a power output of traditional solar breakthrough technologies that combine low fabrication cells is an inefficient use of the infrared (IR) portion costs with high power conversion efficiencies. of the solar spectrum leaking through a transparency Progress window of a semiconductor defined by its bandgap. This problem can potentially be resolved by “upconverting” During the past year, we have investigated PbSe/CdSe IR photons via the process of two-photon absorption core-shell colloidal quantum dots (QDs) prepared by the (2PA). In a 2PA material, an electron is promoted across cation-exchange method. First, large monocomponent the bandgap by absorbing simultaneously two low-ener- PbSe QDs are fabricated (radius R of ca. 4 nm), and then gy photons, which generates a higher energy exciton. For a CdSe shell of a controlled thickness (H) is prepared current state-of-the art 2PA materials the “threshold” by exchanging Pb for Cd in the near-surface layer of light intensity, above which 2PA becomes significant, is the dots. This procedure maintains the total size of the considerably higher than the solar flux. The situation can structure, therefore, the increase in the thickness of the be improved using light concentration. However, even CdSe shell is accompanied by the reduction of the core for a concentration factor of 1000, the solar light inten- size. In addition to core infrared (IR) emission, the thin- sity is still lower than the 2PA activation threshold. This shell QDs exhibit broad and structureless visible photo- implies that the utilization of 2PA ideas in solar energy luminescence (PL), originating, most probably, from the conversion requires new types of 2PA chromophores surface defect states. As the shell thickness increases, with cross-sections that are orders of magnitude higher the visible emission gets stronger and narrower. It also than in existing materials. The purpose of this project is progressively redshifts until it reaches ~650-680 nm for to explore two effects for a dramatic increase of 2PA of the 2.8-nm-thick shell samples. Performing an effective semiconductors. One is a “giant” field enhancement in mass modeling of the structures, we assign the IR band the presence of nanoscale metals and the other – 2PA to the 1Se-1Sh transition while the visible band to the through intra-gap states introduced by surface treat- 1Se-2Sh transition. The energy of the 2Sh state is above ment of nanoscale semiconductors. the valence band edge of CdSe, and therefore, the cor- responding hole wavefunction extends across the entire Benefit to National Security Missions volume of the core-shell QD. This is in contrast to the This project is relevant to LANL and DOE interests in 1Sh state, which is tightly confined to the core. both world-class fundamental science and strategic We have studied nonlinear absorption properties of the applications. Successful implementation of new strate- QDs described above by monitoring visible up-converted gies for capturing infrared light can yield solar-energy- emission, excited with the IR fs laser pulses that are conversion technologies relevant to DOE and LANL in the range of optical absorption of the PbSe core. missions in Energy Security. At the recent DOE workshop Quadratic dependence of the visible emission intensity on Basic Research Needs for Solar Energy Utilization, on excitation power confirms a two-photon excitation it was recognized that while solar photovoltaics is an mechanism. The measured two-photon absorption cross important clean energy source, the practical large-scale sections are about one order of magnitude larger than applications of solar cells would require new concepts those of core-only CdSe QDs and reach values up to and novel materials for solar energy conversion. Novel ~3×107 GM. We explain such an increase in the cross concepts of enhanced two-photon absorption in hybrid sections by resonance enhancement effect due to core semiconductor/metal structures that are explored in 755

Page 756

absorption, which is further enhanced by Auger-assisted to excitation of coherent oscillations of the electron gas upconversion. The latter effect is activated when two or known as surface plasmons (SPs). To study the effect of more holes are excited in the core. In this case, the energy SPs on 2PA, we will use well-defined semiconductor-metal released during Auger recombination of one of the holes systems recently developed by LANL researchers . These with a conduction-band electron is transferred to another structures comprise a metal gold or silver core over-coated core-localized hole. In our specially designed structures, with a thin layer of a dielectric (silica), followed by a layer this energy is sufficient to promote a hole to the shell of nanocrystal quantum dots (NQDs). These hybrid materi- region where it can recombine radiatively with an electron, als offer an unprecedented level of control of SP interac- producing visible PL. The quantum yield of this emission tions with NQD excitons that can be tuned by adjusting is directly proportional to the time of hole re-capture by the thickness of the silica spacer as well as the positions the core. To evaluate this time we have conducted time- of both excitonic and plasmonic resonances. Experimental resolved, streak-camera studies of visible emission using studies of the 2PA properties of NQDs combined with the- 400 nm excitation with 100 fs pulses. oretical calculations will provide directions for the synthe- sis of optimized structures with large 2PA cross sections. We observe that the dynamics of the initial IR peak does The experimental studies will be done using a state-of-the- not change with the excitation fluence, while the dynam- art 2PA spectrometer, similar to one recently developed by ics of the 600 nm peak is fluence-dependent. Both can be the applicant and implemented in studies of organic 2PA characterized by a biexponential decay with the time con- chromophores. A further increase in 2PA, if needed, will stants of ~20-30 ps and ~300 ps. As the excitation fluence be achieved by introducing “engineered” states within the is increased the ratio of the amplitudes of the two compo- band-gap region. As was recently demonstrated by LANL nents (fast-to-slow) remains constant (~10) for the 650-nm researchers, by treating PbS NQDs with thiol-based ligands peak and decreases from ~40 to ~5 for the 600-nm peak. it is possible to introduce mid-gap states that served as We attribute the initial fast component to hole capture by charge conduit in dark. Additionally, these states were the core, while the slower component to radiative decay optically active and participated in optical transition with of shell localized holes. Based on these measurements, the sub-gap energies. These observations indicate that the hole capture time is around 20-30 ps. This time constant is surface related mid-gap states can also serve as intermedi- significantly longer than characteristic times of intraband ate levels facilitating the 2PA process. relaxation in standard monocomponent PbSe QDs (~0.3 – 1 ps), which leads to a great enhancement of “hot” visible PL Conclusion in our core/shell structures. Utilization of infrared solar photons (IR) can significantly enhance the efficiency of photovoltaic cells. The goal of These preliminary studies of novel core-shell PbSe/CdSe this project is to explore the utility of plasmon-enhanced QDs indicate that they have a number of unusual and po- two-photon absorption (2PA) for conversion of IR photons tentially useful properties. One such property is dual-color into larger-energy excitations in semiconductor nanocrys- emission, which is characterized by two bands, one located tals. To accomplish this goal, we will conduct studies of in the IR and the other in the visible spectral range. These 2PA as a function of nanocrystal size, shape, and composi- new dots also demonstrate a more than ten-fold enhance- tion. We will further investigate the influence of plasmonic ment of a two-photon absorption cross section, which is field enhancement of 2PA cross-sections. Based on these contributed by both resonant core-mediated transitions studies we will identify optimal nanostructure designs and Auger-assisted upconversion. These QDs are promis- that can allow for capturing a significant fraction of solar IR ing for various applications, and can be used, for example, radiation at moderate concentration factors of ~1000. as dual-color probes in two-photon bioimaging or “up- converters” of low-energy solar photons in IR-enhanced Publications photovoltaics. Makarov, N. S., H. McDaniel, N. Fuke, I. Robel, and V. I. Kli- mov. Photocharging artifacts in measuremens of elec- Future Work tron transfer in quantum-dot-sensitized mesoporous As was demonstrated previously, enhancement of the non- titania films. J. Phys. Chem. Lett.. linear response can be achieved via utilization of plasmonic resonances in hybrid semiconductor-metal nanostructures. Makarov, N. S., Q. Lin, K. Velizhanin, and V. I. Klimov. PbSe/ CdSe core/shell colloidal quantum dots with enhanced For example, Raman scattering, which is a third-order optical nonlinearities and dual-band infrared/visible nonlinear process (same as 2PA) could be enhanced by emission. 013. In Conference on Nonlinear Optics. up to ~10^10 times in the presence of nanoscale metals. (Kohala Cost, HI, USA, July 21-26, 2013). , p. NTh1A.3. This effect has been explained by field enhancement due Washington: Optical Society of America. 756

Page 757

Physics Postdoctoral Research and Development Continuing Project Minority Carrier Devices Based on Concentric Nanowires: Theory and Experiment Quanxi Jia 20120747PRD1 Introduction Using high resolution transmission electron microscopy The goal of this work is to achieve a comprehensive un- (HRTEM), we have advanced in materials synthesis of Si/ derstanding on minority carrier dynamics at nanoscale, Ge NWs using vapor-liquid-solid (VLS) growth mecha- and identify novel physics intrinsically related to small nism. High quality Ge/Si NW with various designs have dimensions, in order to provide insights for nano-device been grown and characterized. The main research focus- design and optimization. This work will provide a robust es entail fabrication of novel NW devices based on Ge/ assessment of the fundamental limits and potential of Si heterostructures. This work is being conducted using nanowire devices operating with minority carriers, com- a large combination of tools and systems at CINT such plement current nanowire activities at CINT, and may as HRTEM and in-situ thermal annealing under TEM, inspire new research directions in the area of photovol- e-beam/photo lithography, Scanning electron micros- taics at LANL. The work will focus on Si photovoltaics and copy/Focused Ion Beam (SEM/FIB), metal deposition by SiGe heterostructured tandem solar cells to maximize e-beam evaporation and sputtering, plasma-enhanced potential solar cell performance from these devices. chemical vapor deposition (PECVD) and atomic layer de- A considerable effort will be paid on heterostructure position (ALD) of high-k dielectrics, inductively coupled transistor devices where the standby power and active plasma (ICP) dry etchings of Si, Ge and dielectrics, and power can be reduced with tailoring the composition thermal oxidation furnaces. The combination of high and contact engineering to these structures thereby ad- quality materials and advanced fabrication techniques dressing an increasingly important area of semiconduc- has resulted in great accomplishments, currently being tor technology, green-electronics. wrapped up for publication. Device physics of NW FETs Benefit to National Security Missions NW has been being considered as an attractive can- This work lies at the heart of solar cell technology and didate for new generation material due to its unique low power transistor devices, both of which tie well with configuration that allows effective electrostatic coupling the DOE and DOD energy agenda. These devices operate between gate control voltage and carrier conduction on minority carrier transport in semiconductors, and the channel. High performance Si/Ge NW FETs have been goal of this work is to achieve a comprehensive under- demonstrated, but so far, little work has been reported standing on minority carrier dynamics at nanoscale, and on the device physics and experimental proof for the identify novel physics intrinsically related to small dimen- advantages of the Ge/Si core/shell structure over stan- sions, in order to provide insights for nano-device design dard Si or Ge NWs. A large number of Ge/Si core/shell and optimization. This work will provide a robust assess- NWs with different diameters were processed into FETs ment of the fundamental limits and potential of nanow- and characterized to extract the size-dependence of FET ire devices operating with minority carriers, complement performance. Our findings are experimental evidence for current nanowire activities at CINT, and may inspire new the linear dependence of device’s on-current on diam- research directions in the area of photovoltaics at LANL. eter, which signify confinement of carriers near the NWs’ perimeter in the on-state. In the off-state, it is more dif- Progress ficult to turn off larger diameter device (higher required We continued to make excellent progress in all areas gate voltage, and slower current drop) due to larger relevant to this research project including growth, device volume of carriers to be depleted. Comparison between modeling and fabrication as well as characterization. Ge/Si core/shell and Ge NWs exhibited a factor of 3-fold 757

Page 758

increase of device on-current, which is shown to be due to core/shell NWFETs: standard NWFETs require large di- mobility enhancement in core/shell NWs. ameter to be able to drive large current, but such large diameters make it hard to turn off the device. We have Ultra-short metal/semiconductor/metal heterostructures developed a new device concept using Si/Ge/Si core/ for ultra-short channel length FETs multiple-shell structure that is theoretically proven to Aiming to create ultra-short (<10 nm) channel length FETs, overcome this fundamental drawback. We will address solid-state-reaction between Ni and Ge/Si core/shell NWs experimental challenges (growth, processing and mea- is utilized to create a NiGe( or Si) / Ge(Si)/ NiGe(Si) meta/ surement) to demonstrate a device prototype. semiconductor/metal with controlled size. Thermal an- nealing provides kinetic energy to Ni atoms to penetrate 3. Thermoelectric materials: Nanowires have been prov- into Ge/Si semiconductors and create metallic NiGe/NiSi en excellent thermoelectric materials with low thermal layers with high electrical conductivity, which can serve as conductivity due to their size in the nano range. With drain and source for the un-reacted Ge/Si semiconductor state-of-the-art device fabrication capability developed channel in FETs. Developments in device fabrication allow this year, we plan to further improve thermoelectric us to integrate FET device structures on a 50 nm thick performance of the NWs system by creating nanowire Si3N4 membrane in order to monitor and control reaction superlattice: a complex repetitive …NiGe(Si)/Ge(Si)/ mechanism in-situ with HRTEM. Reaction between Ni with NiGe(Si)/Ge(Si)/… heterostructure from the building Ge/Si heterostructure exhibited completely difference re- block NiGe(Si)/Ge(Si)/NiGe(Si) developed this year. action behaviors/mechanisms that haven’t been observed Multiple heterointerfaces between NiGe(Si) and Ge(Si) in Ni-Ge or Ni-Si reactions. In-situ HRTEM revealed atomi- is predicted to enhance phonon scattering for reduced cally smooth layer-by-layer propagation of Ni into the Ge/ thermal conductivity, while maintaining high electrical Si semiconductors, allowing for precise control of channel conductivity. length and junction abruptness. The achieved channel length of 5 nm is the shortest channel length in Ge/Si NW Conclusion FETs up to date. We expect to be able to model and realize Ge, Si, and Ge/Si heterostructured nanowire solar cell arrays with Advances in growth and processing developed this year fab-compatible processes that can outperform their bulk will enable novel materials and device configurations that counterparts. We also plan to reduce breakdown voltages can open new frontiers for NW research (as described in in avalanche photodiodes and in impact ionization field-ef- proposed work for next fiscal year). fect transistors by proper band-edge engineering available in our heterostructures wires. We will also work on tunnel Future Work field-effect transistors in the III-V material system where In the next fiscal year, we will concentrate on the materials steep turn-on characteristics, similar to impact ionization synthesis and device fabrication of Si/Ge heterostructured FETs, are feasible targeting low power operation. nanowires (NWs) for energy applications. The following specific tasks will target on developing low power con- Publications sumption high performance electronics (1 and 2) and high Dayeh, S. A., W. Tang, B. M. Nguyen, X. Dai, Y. Liu, Y. performance thermoelectric materials (3). Hwang, X. H. Liu, and R. Chen. Nanoscale Heteroge- neous Reactions and Interfaces in Ge/Si and for III-V

1. Ultra-short channel length field effect transistors on Si Integrated Devices. Invited presentation at ECS (FETs): We have successful realized a NiGe (or Si)/ Meeting. (San Francisco, CA, 27 Oct - 1 Nov. 2013). Ge(Si)/NiGe(Si) NW heterostructre with the Ge( or Si) semiconductor segment as short as 5 nm. The next Liu, Y., X. H. Liu, B. M. Nguyen, J. Yoo, J. P. Sullivan, S. T. Pic- raux, J. Y. Huang, and S. A. Dayeh. Tailoring Lithiation step is to fabricate FETs from such ultra-short channel Behavior by Interface and Bandgap Engineering at the lengths. Devices with <10 nm channel length are highly Nanoscale. To appear in Nano Letters. desirable as it requires much lower power consump- tion for the same performance as longer channel Nguyen, B. M., J. Yoo, S. A. Dayeh, P. Schuele, D. Evans, and length ones. Reducing the channel length is also a path S. T. Picraux. Design of Radial p–i–n Silicon Nanowires way for miniaturization, an industrial strategy to inte- for High-Performance Solar Cells. 013. In The Wonder grate more devices on the same area, hence increasing of Nanotechnology: Quantum Optoelectronic Devices performance and reducing fabrication cost. and Applications. Edited by Razegh, M., I. L. Esaki, and K. von Klitzing. , p. 823. Bellingham, WA: SPIE Press.
2. Novel design of NWFETs: On-going work has identified Nguyen, B. M., J. Yoo, S. A. Dayeh, and S. T. Picraux. Design a fundamental drawback in current design of Ge/Si 758

Page 759

of radial p-i-n Si nanowires for high performance solar cells. Presented at 11th international conference on in- frared Optoelectronics: Materials and Devices (MIOMD XI). (Evanston, IL, Sep. 2012). Nguyen, B. M., W. Tang, S. T. Picraux, and S. A. Dayeh. One dimensional transport in Ge/Si core/shell nanowires: physics and devices applications. Presented at CINT User Conference. (Albuquerque, 19-20 Sept.). Nguyen, B. M., W. Tang, and S. A. Dayeh. Ultra-short chan- nel field effect transistors based on Ge/Si core/shell nanowires. Invited presentation at SPIE Photonics West. (San Francisco, 2-7 Feb.). Nguyen, B. M., W. Tang, and S. A. Dayeh. Ultra-short chan- nel field effect transistors based on Ge/Si core/shell nanowires. Invited presentation at SPIE Photonic West. (San Francisco, CA, Feb. 2013). Nguyen, B. M., Y. Liu, W. Tang, S. T. Picraux, and S. Dayeh. Ultra-short channel field effect transistors based on Ge/Si core/shell nanowires. Invited presentation at SPIE Photonic West. (San Francisco, CA, Feb. 2013). Nguyen, B. M., Y. Liu, W. Tang, and S. Dayeh. Tailoring Ni-Ge/Si core/shell reaction rates through in-situ mi- croscopy for ultra-short channel nanowire field effect transistors. Presented at MRS Spring Meeting . (San Francisco, CA, April 2013). 759

Page 760

Physics Postdoctoral Research and Development Continuing Project Frustrated Materials Cristian D. Batista 20120751PRD2 Introduction semiconductors or intermetallic compounds. By model- One of the most spectacular manifestations of geometri- ing of bulk materials and films, we will establish guiding cal frustration is the presence of the novel states of mat- principles for finding novel states with optimized func- ter and associated phase transitions. These novel phases tionalities. This is a necessary step towards the dream of arise as a compromise between the opposite tendencies “materials by design.” that are inherent to frustrated geometries. A triangu- Progress lar plaquette is an archetypal building block for lattices with geometrical frustration. Many unusual magnetic The research of Dr. Gia-Wei Chern at LANL is focused on and orbital orderings in frustrated magnets are best the general area of frustrated materials. The main goal understood from the viewpoint that the system is in the is to understand and control the fundamental prop- vicinity of a metal-insulator transition. It is a challenging erties of these complex systems with applications to theoretical task to model these systems which lie in the magnetic storage, computing, and sensing. In particular, intermediate-coupling regime between deep Mott insu- his main effort in the past months has been devoted to lator and the metallic phase. To carry out such studies, understanding the novel transport properties in metal- we will employ a combination of analytical and numeri- lic frustrated magnets. By considering Kondo-lattice cal techniques. We will use mean field and semi-classical models in frustrated lattices, e.g. pyrochlore or kagome techniques for computing thermodynamic properties. lattices, Dr. Chern and his collaborators have uncovered For the numerical approach we will implement a Mote a novel mechanism for resistivity minimum, which does Carlo algorithm for solving the problem of itinerant elec- not originate from the conventional Kondo-screening, trons interacting with localized classical spins. Since the in some metallic pyrochlore magnets. By establishing action is quadratic in the fermion operators, the itinerant the connection between electronic transport and the electrons can be integrated out exactly, leading to an spin-spin correlations in the highly constrained coopera- effective non-local classical action for the local moments tive paramagnetic regime, the proposed mechanism that will be solved by applying the Metropolis algorithm. can explain similar resistivity minimum phenomena in All the thermodynamic properties will be computed frustrated magnets in general. under control because the statistical error can be made In addition, Dr Chern and his collaborators have also arbitrarily small by increasing the sample size. The tem- discovered a novel phase exhibiting a quantized Hall ef- perature dependence of the thermodynamic quantities fect in a highly degenerate ice manifold. A similar phase, will be directly compared against the experiments. dubbed chiral spin liquid, has recently been reported in pyrochlore irradiates. Their result thus provides a proof Benefit to National Security Missions of principle for the existence of such novel state of mat- Given the complexity of correlated materials and the ter. unlimited number of compounds, the discovery of new states of matter and functionalities must be organized Another main area of his research is the study of ar- around simple guiding principles. Our basic principle for tificially created frustrated systems, in particular the discovering complex and collective forms of matter that so-called artificial spin ice. One of the key advantages exhibit novel properties and respond in new ways to of these systems is that they allow direct experimental environmental conditions is to make, characterize and access to the microscopic degrees of freedom. With model classes of materials with controllable degree of other collaborators in LANL, Dr. Chern has proposed a electronic localization. These classes can either be Mott 760

Page 761

realization of the famous 3-coloring problem in systems solids, leading to band insulator with a nontrivialtopol- of colloidal particles interacting with an array of optical ogy. It has been shown recently that the Kitaev model traps. Experimental realization of such systems will allow can be realized in iridium oxides A2IrO3 that areMott us to study the dynamics of defects in this highly degener- insulators with strong spin-orbit coupling. We will in- ate critical phase. Artificial spin ice is a booming field in vestigate the general properties of topological Mottin- condensed matter physics. Among the most exciting recent sulators, their experimental characterization, and their developments is the emergence of magnetic monopoles connection with topological band insulators when the- in these materials. Emergent phenomena, which until now system undergoes a Mott transition.(5) Quasiparticles have only been accessible at low-T, will then be tailor- of orbital ice: recently, we have uncovered a nontrivial designed to manifest at room temperature. In particular, many-body ground-state solution forthis system. We the existence of mobile magnetic monopoles coupled to will study the elementary excitations in orbital ice and magnetic fields provides a promising approach to realize to pursue an algebraic approach forexactly solving the magnetricity, the control and manipulation of magnetic quantum orbital model. currents. With the rapid progress in nano-fabrications, it is now possible to realize artificial spin ice in various lattice Conclusion geometries. Working with collaborators both in LANL and • Find novel magnetic orderings in double-exchange other institutes, Dr. Chern has explored novel phenomena models and spin-density wave instability of Hubbard of monopoles in different geometries. One of the major model on frustrated lattices. findings along this direction is the phenomenon of mono- pole screening in lattices with mixed coordination num- • Understand the thermodynamic properties of metallic bers. spin ice, particularly the exotic chiral spin-liquid phase in pyrochlore compound Pr2Ir2O7. Future Work • Develop theoretical models for electron transport • Compute the magnetic phase diagram of double- in the emergent Coulomb phase in metallic spin ice exchange system on the frustrated lattices: The inter- systems. play of magnetic frustration and the nonlocal nature of electron-mediated spin interaction gives rise to • Develop numerical algorithms for studying the order- unusual magnetic orders and exotic phases. There are by-disorder phenomenon in frustrated magnets a growing number of compounds that are described by such models. We will develop sophisticated and effi- • Provide a theoretical modeling of magnetic and orbital cient Monte Carlo algorithms for computing the phase ordering in vanadium spinels. diagram of these models. Publications • Compute thermodynamic and transport properties Chern, G. W., A. Rahmani, and I. Martin. Quantum Hall ice. of metallic spin ices: A similar study for the metallic Physical Review Letters. spin-ice systems will shed some light on the strange chiral phase with disordered spins recently observed Chern, G. W., C. C. Chien, and M. Di Ventra. Dynamically generated flat band phases in optical kagome lattice. in pyrochlore Pr2Ir2O7. We will examine the dynamics Physical Review Letters. of monopoles induced by the electron-mediated RKKY interactions. Chern, G. W., C. Reichhardt, and C. J. O. Reichhardt. Frus- trated colloidal ordering and fully packed loops in ar- • Metal-insulator transitions in frustrated magnets: rays of optical traps. 2013. PHYSICAL REVIEW E. 87 (6): Many of the unusual magnetic and orbital orderingin -. pyrochlore magnets are best understood from the viewpoint that the system is close to a metal-insulator Chern, G. W., M. Morrison, and C. Nisoli. Degeneracy and transition. Indeed, some of them also exhibit a room- criticality from emergent frustration in artificial spin temperature metal-insulator transition that is accom- ice. To appear in Physical Review Letters. panied by magnetic ordering. We will explore the role Chern, G. W., R. M. Fernandes, R. Nandkishore, and A. V. of orbital degrees of freedom in such transitions for Chubukov. Broken translational symmetry in an emer- multiband Hubbard models. gent paramagnetic phase of graphene. 2012. PHYSICAL REVIEW B. 86 (11): -. • Mott transitions and physics in strongly coupled spin- orbital systems: Strong spin-orbit coupling candramati- Chern, G. W., S. Maiti, R. M. Fernandes, and P. Wolfle. cally change the band structure of weakly interacting Electronic Transport in the Coulomb Phase of the Pyro- 761

Page 762

chlore Spin Ice. 2013. PHYSICAL REVIEW LETTERS. 110 (14): -. Chern, G. W., and C. D. Batista. Spontaneous quantum Hall effect via a thermally induced quadratic Fermi point . 2012. Physical Review Letters. 109: 156801. Chern, G. W., and O. Tchernyshyov. Magnetic charge and ordering in kagome spin ice. 2012. PHILOSOPHICAL TRANSACTIONS OF THE ROYAL SOCIETY A-MATHE- MATICAL PHYSICAL AND ENGINEERING SCIENCES. 370 (1981): 5718. Chern, G. W., and P. Mellado. Magnetic monopole polarons in spin ices with mixed coordination numbers. Physical Review Letters. Chern, G. W., and R. Moessner. Dipolar Order by Disorder in the Classical Heisenberg Antiferromagnet on the Kagome Lattice. 2013. PHYSICAL REVIEW LETTERS. 110 (7): -. Choi, E. S., G. W. Chern, and N. B. Perkins. Chiral magne- tism and helimagnons in a pyrochlore antiferromagnet. 2013. PHYSICAL REVIEW B. 87 (5): -. Choi, E., G. W. Chern, and N. B. Perkins. Helimagnons in a chiral ground state of the pyrochlore antiferromagnets. 2013. EPL. 101 (3): -. Lin, S. Z., Y. Kamiya, G. W. Chern, and C. D. Batista. Reen- trant Berezinskii-Kosterlitz-Thouless Transition in a Tri- angular Ising Thin Film. Physical Review Letters. Mun, E. D., G. W. Chern, V. Pardo, F. Rivadulla, V. S. Zapf, and C. D. Batista. Magnetic field induced transition in multiferroic spinel CdV2O4. Physical Review Letters. Perkins, N. B., G. W. Chern, and W. Brenig. Raman scat- tering in a Heisenberg S=1/2 antiferromagnet on the anisotropic triangular lattice. 2013. PHYSICAL REVIEW B. 87 (17): -. Rahmani, A., and G. W. Chern. Universal Renyi mutual in- formation in classical systems: The case of kagome ice. 2013. PHYSICAL REVIEW B. 88 (5): -. Zhang, S., I. Gilbert, C. Nisoli, G. W. Chern, M. J. Erickson, L. O’Brien, C. Leighton, P. E. Lammert, V. H. Crespi, and P. Schiffer. Crystallites of magnetic charges in artificial spin ice. 2013. NATURE. 500 (7464): 553. 762

Page 763

Physics Postdoctoral Research and Development Continuing Project Atomic Coherence with X-Ray Free-electron Lasers Bruce E. Carlsten 20120754PRD2 Introduction Progress Atomic coherence with intense visible and infrared lasers During FY13 we have examined several experiments is an exciting emerging field, which allows exploration done at LCLS where atomic coherence was used to of strange atomic behavior though controlling recombi- modify transition rates of Neon. These experiments have nation rates and suppressing absorption. These effects shown that spectral brightness of self-amplified spon- include such as slowing down or speeding up light, taneous emission based X-ray free electron lasers is not efficient frequency up-conversion, and more efficient sufficient to provide noticeable coherent effects. It has photocells. Additional new phenomena may be possible also indicated that novel designs for future X-ray free by using intense X-ray lasers, such as the one proposed electron lasers are required in order to make them suit- for the LANL MaRIE facility. The purpose of this project is able for fundamental atomic and quantum research. to extend this new field to the X-ray regime. Currently, X- Search for the novel designs have required us to re-ex- ray lasers based on self-amplified spontaneous emission amine the basics of X-ray free electron laser theory. Our do not provide spectral brightness necessary to excite analysis showed that the quantum mechanical nature of atomic coherence properties and novel designs for X-ray electrons manifested as the spread of their wave packets free electron lasers are needed. We will first look at the becomes significant in hard X-ray free electron lasers generation of an optical comb with train of electron such as SACLA in Japan and LANL proposed MaRIE. This bunches prepared via laser modulation technique. As an is a 10% effect in the case of lasing at the fundamental alternative approach, we will study parametric interac- and increases to 30% for the third harmonic generation. tion of fundamental and sub-harmonic in bi-energetic It is a collective effect. electron beam in order to imprint longitudinal coher- ence of the sub-harmonic onto the fundamental. We We have developed a novel X-ray free electron laser will then consider how electron wave function spreading design that was inspired by the concept of optical fre- will affect the performance of a hard X-ray free electron quency combs. Starting with a train of electron bunches laser operating at lower electron energies for cost reduc- that generates a single coherent X-ray pulse per bunch, tion. We have three major focus areas: start-up noise, one generates a periodic train of X-ray pulses. In this gain reduction, and saturation process. We will model regime, the radiation spectrum becomes discrete with operation of these X-ray free electron lasers and apply spectral brightness being increased by at least an order the generated electric fields to modeling the excitation of magnitude. We did numerical simulations and con- of atomic transitions. This will allow us to quantify the firmed the predictions. It has been also discovered that amount of atomic coherence one could generate using such an electron beam is easier to prepare so that it has these new X-ray free electron lasers. improved gain. Furthermore, the propagation of a long train of short electron bunches in metal pipes is advanta- Benefit to National Security Missions geous since it does not suffer from resistive wall wake This research may lead to enhanced materials perfor- effects that introduce inhomogeneous energy loss on mance and devices relying on atomic technologies, such the beam. This work was presented at the LANL postdoc as higher efficient photocells and smaller and more competition and won an outstanding poster award. controllable devices based on quantum heat engines, quantum information storage and processing, atomic Future Work clocks, magnetometers, interferometers, and novel laser We expect progress will be made in three areas in FY14. systems. 763

Page 764

First, quantum effects of electron wave function on start- up noise, gain reduction, and saturation regime in the case of hard X-ray free electron laser operating at lower ener- gies will be evaluated. Second, novel X-ray free electron laser designs with higher spectral brightness will be stud- ied. Finally, quantum atomic effects that can be induced by this novel X-ray free electron lasers will be identified. In the first area, future hard X-ray free electron lasers, such as the one proposed for MaRIE, are proposed to operate at lower electron energies to reduce construction and opera- tion cost. In this new regime the quantum effects on the electron beam and the X-ray generation mechanism may become significant. Specific examples may include start-up noise modifications, gain reduction due to phase uncer- tainty, and quantum saturation regime where due to the spread of an electron wave function electrons are trapped and untrapped at the same time. In the second area, we will first look at optical comb generation in the X-ray domain where a train of electron bunches is used to excite a train of optical pulses; we will then look at parametric generation of coherent X-ray radiation in bi-energetic electron beam. We will finally use modeled electric field to drive atomic coherences in simple atomic systems in order to validate X-ray free electron laser applicability for funda- mental atomic and quantum research. Conclusion The goal of this project is to investigate novel designs of X-ray free electron lasers that offer improved spectral brightness needed to excite atomic coherence. The ex- pected results are: optical comb generation with train of electron bunches as well as the parametric generation in a bi-energetic electron beam; effects of an electron wave function on start-up noise, gain reduction, and saturation regime in the case of hard X-ray free electron laser oper- ating at lower energies; validation of X-ray free electron lasers for fundamental atomic and quantum research by apply generated electric fields to excite atomic coherence in simple atomic systems. 764

Page 765

Physics Postdoctoral Research and Development Continuing Project Precision Fission Cross Section Measurements with a Time Projection Chamber Fredrik K. Tovesson 20120755PRD2 Introduction The new technique for nuclear physics that is developed We intend to study nuclear reactions, particular fission, within this project will also provide new advanced capa- with a newly developed instrument (a Time Projection bilities for nuclear science research at LANCSE and other Chamber) at the Los Alamos Neutron Science Center facilities in the US and world wide. We will also advance (LANSCE). This goal is to measure these reactions with our understanding of the nuclear fission process within unprecedented accuracy in order to meet needs for the the scope of the project. nuclear energy and weapons programs. This is the first Progress time ever that a Time Projection Chamber (TPC) is ap- plied to this type of research, and if successful this could The main accomplishment this year was completing the have major impact on the field. The experiments will be analysis of the experimental data collected during the carried out at LANSCE, which is a world class facility for 2011-2012 LANSCE run cycle. This was a major achieve- nuclear research and one of only two places where this ment on the path to getting to the final nuclear data work could be performed. There are major difficulties measurement completed. in reaching the goal of this project, as a completely new There was also significant progress made in terms of technique is being developed. Particle tracking of fission coordinating the in-beam data collection efforts for the fragments, neutron beam characterization and radioac- TPC project, which involved: tive sample handling all poses challenging problems to solve. However, if these challenging problems are solve a • Supervising two undergraduate and two graduate major improvement in the accuracy that can be reached summer students in fission cross section measurements will be achieved. The data than can be measured with this approach will • Preparing and submitting safety documentation have far reaching implications for nuclear energy re- search, the weapons program, and basic science. • Setting up the experimental area, detector system, and auxiliary equipment Benefit to National Security Missions • Troubleshooting and maintaining experimental The results that will come out of this work build capabil- equipment during the run ity for high level milestones for the weapons program. The data will help better understand the performance A number of presentations were made through the year: of nuclear weapons, improve predictive capabilities, and support the science-based approach to the stockpile • Delivered an invited talk at an international confer- stewardship mission. ence - Collective Motion in Nuclei under Extreme Conditions (COMEX4), Hayama, Japan, October 2012 It will also support the nuclear data effort of the DOE nuclear energy programs such as the Advance Reactor • Cross-Section Evaluation Working Group Meeting – Campaign and Fuel Cycle Technologies. The cross sec- Brookhaven National Laboratory, November 2012 tions that will be measured were identified within the nuclear data effort as being in need of improvement, • International Conference on Nuclear Data for Sci- and will be used to improve simulations of new ad- ence and Technology – New York NY, March 2013 vanced reactor and fuel concepts. (Featured Talk, refereed conference proceedings to be published January 2014) 765

Page 766

• Gordon Research Conference on Nuclear Chemistry – New London NH, June 2013 • Submitted and defended a proposal to the LANSCE Program Advisory Committee to measure the U- 238/U-235 (n,f) cross-section ratio using the TPC Future Work One task for FY14 is to collect more beam data at the Los Alamos Neutron Science Center (LANSCE) on nuclear fission of Uranium-238 and Uranium-235 using a Time Projection Chamber (TPC). This measurement will be performed in the 2013 LANSCE run cycle, which ends in December 2013. The second task is to analyze the data collected with the TPC in order to determine the probability for nuclear fission to occur in Uranium-238 when bombarded with neutrons of different energies. This is a complex analysis that involves the study of tracks made by fission fragments in the detector, as well as determining the kinetic energy of the incident neutrons that induced the reaction. The overall goal for this fiscal year will be to produce a preliminary fission probability (a so-called reaction cross section) as a function on incident neutron energy for Ura- nium-238 relative to Uranium-235. Conclusion Within this project we expect to demonstrate that the TPC can reach the expected accuracy of approximately 1% in fission cross section measurements. We will deliver the U-238/U-235 fission ratio, and develop all the necessary procedures to run beam experiments for Pu-239. We will also develop a an expanded research program for the TPC, which will include more detailed studies of the nuclear fis- sion process. Publications Meharchand, R., F. Tovesson, A. B. Laptev, and K. Meier- bachtol. TPC measurement support. 2012. LANL inter- nal report LA-UR-12-25409. Meharchand, R., T. Hill, and L. Snyder. Analysis of TPC Single Sextant U-238/U-235 Engineering In-Beam Data. 2012. LANL Internal Report LA-UR-12-25452. 766

Page 767

Physics Postdoctoral Research and Development Continuing Project The Assembly of Primeval Galaxies and the Birth of the First Quasars Hui Li 20120756PRD2 Introduction Progress Soon after the Big Bang, gravity caused dark matter and Primordial supernova explosions are key to the forma- primordial gas to collapse into a vast cosmic network of tion of the first galaxies (and hence the science objec- filaments and halos (roughly spheroidal clumps of dark tives of my Director’s Fellowship) because they chemi- matter and gas). These halos congregated into the first cally enrich them with the first heavy elements and may primitive galaxies 400-600 million years after the Big trigger second-generation star formation in them. They Bang. Protogalaxies were also the first great nucleosyn- will also be important to detecting such galaxies, since thetic engines in the universe, polluting the intergalactic some of these primitive structures would otherwise be medium (IGM) with heavy elements from intense winds too dim to be detected in wide-field surveys. In the first and radically transforming generations of stars after year of my Fellowship we have modeled the light curves them and making planets and life possible. Such galax- and spectra of the first SNe in the universe with the ies are also believed to be the birthplaces of the seeds radiation hydrodynamics code RAGE: core-collapse, pair- of the supermassive black holes (SMBH) residing at the instability, Type IIn and supermassive thermonuclear centers of most massive galaxies today. Because they events. We have discovered that these ancient explo- lie at the edge of the observable universe, these galax- sions will be detectable by the James Webb Space Tele- ies are beyond the realm of current instruments but will scope (JWST) and the Wide-Field Infrared Survey Tele- soon be detected by the James Webb Space Telescope scope (WFIRST) at the earliest epochs in the universe. (JWST) and the Thirty-Meter Telescope (TMT). State-of- These simulations are having a real science impact on the-art protogalaxy simulations fail to resolve the cycle the astronomy community. David Spergel and Jason Ka- of primordial star formation and explosion in each halo liari from Princeton and STScI have consulted us on filter as they are assembled into the first galaxies. We will de- designs for WFIRST (which is now in its planning stages) velop a new generation of protogalaxy simulations that to detect these ancient events in the near infrared will bridge all spatial scales relevant to their formation. (NIR) and have solicited contributions from LANL on the We will also build the first numerical models ever of the science case for WFIRST to be submitted to NASA. We birth of the first quasars. have also been contacted by several independent groups Benefit to National Security Missions of astronomers for assistance in devising searches for Pop III SNe. Reporters have also contacted us about our This research addresses the fundamental question on work, which now appears in the June 12, 2013 issue of how the primeval galaxies and black holes in the early New Scientist (The Supernova that Blew up a Galaxy). universe. The results from this study will provide first- My numerical models will culminate in the publication of principle physical understanding of the formation of first 6 Astrophysical Journal publications. objects under extreme conditions. They will also prepare us for the upcoming new observations by the nation’s We have also begun to simulate the formation of the leading telescopes, both ground-based and space-borne. first quasars in the universe 500 Myr after the big bang This research contributes to the basic science capabili- and their evolution down through cosmic time in large ties at the Laboratory. It also provides the necessary simulation boxes with the cosmology code Enzo. The physics basis for many challenging problems associated births of these quasars will soon be discovered by suc- with the Beyond the Standard Model Grand Challenge at cessors to the Swift mission such as JANUS or LOBSTER. LANL. If we find that such black holes can reach a billion solar 767

Page 768

masses by redshift z = 7, we will have explained one of the outstanding problems in cosmology today: the appearance of supermassive quasars less than a billion years after the big bang. Future Work We will perform cosmological simulations to study the formation of primeval galaxies and investigate how they will contribute to future structure formation. Our simula- tion campaign will unveil not only the nature of primeval galaxies but also how they began early chemical enrich- ment and reionization of the universe. Our models will also determine the initial conditions of massive galaxy formation down through cosmic time. These simulations will leverage LANL’s expertise in radiation transport, explo- sions and turbulent subgrid mixing and have the potential to transform their subject, putting LANL at the forefront of high-redshift cosmological structure formation. Conclusion Our simulation campaign will unveil not only the nature of primeval galaxies but also how they began early chemical enrichment and reionization of the universe. Our models will also determine the initial conditions of massive galaxy formation down through cosmic time. These simulations will leverage LANL’s expertise in radiation transport, explo- sions and turbulent subgrid mixing and have the potential to transform their subject, putting LANL at the forefront of high-redshift cosmological structure formation. Publications Johnson, J., D. Whalen, W. Even, C. Fryer, A. Heger, J. Smidt, and K. Chen. The biggest explosions in the universe. 2013. The Astrophysical Journal. 775 (2): 8. Whalen, D., A. Heger, K. Chen, W. Even, J. Smidt, C. Fryer, M. Stiavelli, and H. Xu. Supermassive population III supernovae and the birth of the first quasars. To ap- pear in The Astrophysical Journal. Whalen, D., J. Johnson, J. Smidt, A. Heger, W. Even, and C. Fryer. The supernova that destroyed a protogalaxy: prompt chemical enrichment and supermassive black hole growth. 2013. The Astrophysical Journal. 774 (1): 10. Whalen, D., J. Johnson, W. Even, C. Fryer, A. Heger, J. Smidt, and K. Chen. The biggest explosions in the universe. II. To appear in The Astrophysical Journal. Whalen, D., L. Frey, W. Even, J. Smidt, C. Fryer, M. Stia- velli, D. Holz, A. Heger, S. Woosley, C. Lovekin, and A. Hungerford. Finding the first cosmic explosions I: pair- instability supernovae. To appear in The Astrophysical Journal. 768

Page 769

Physics Postdoctoral Research and Development Continuing Project Searching for Sterile Neutrinos with MircoBooNE Richard G. Van De Water 20120757PRD2 Introduction the data acquisition system (DAQ). This complex system The project is to work on the next generation large scale is now reading out test electronics and is operating at neutrino experiment (MircoBoone) at Fermi National Ac- the required speed to run the entire system. celerator Lab (FNAL contact person: Sam Zeller, 630-840- He has also begun development of neutrino reconstruc- 6879, [email protected]). The experiment plans to follow tion algorithms and has made significant progress in up on the recent tantalizing neutrino oscillation mea- developing techniques to determine the vertex position surements from LSND and MiniBooNE that are hinting of neutrino events. at the possibility of a new fundamental particle called sterile neutrinos. Such a discovery would have profound Lastly, Wesley has been asked by the collaboration (over effects ranging from fundamental properties of the Stan- 100 scientists) to lead the reconstruction algorithm dard Model, to the origin and evolution of the Universe group. This is in recognition of his outstanding scientific and Cosmology. MicroBoone will search for neutrino talents. oscillations using a new neutrino detection technology involving a Time Projection Chamber in liquid Argon. Future Work This spatially high resolution detector can separate many The goal in the next year is to complete and commission classes of events, allowing a rejection of background work on the Data Acquisition System (DAQ) that reads photons from signal electrons. This capability will allow out the time projection chamber in the liquid Argon. a definite determination of the origin of the LSND/Mini- Computers are used to read out the 10,000 channels BooNE signals. There will be many technical challenges of electrons that record the signals. The DAQ code in making this new technology work as planned. must constantly monitor the electronics and read out the channels when data is available. The data are then Benefit to National Security Missions assembled into events to be permanently recorded on MicroBooNE will be testing a new technology of time disk for subsequent analysis. The trigger information projection chamber in liquid Argon. This new technol- and detector status most constantly be monitored and ogy will require much research to understand and make recorded by the DAQ. The goal in the first year was to work as expected. We will be contributing directly to the develop the low level drivers that read out individual work to readout the detector and also reconstruct the channels. This forms the basic building block of the DAQ. events that will allow us to make physics measurements Also, minimal DAQ code will be developed to help debug of Argon neutrino nuclear cross sections and neutrino and characterize the electronics during the assembly and oscillations. Also, if the oscillation measurements reveal installation of the electronics in the detector hall. the existence of sterile neutrinos, then a re-evaluation of the standard model of particle physics and evolution Conclusion of the Universe is required to accommodate this new In the next two year period we expect to build the particle/state of matter. Fundamental work of this sort detector, commission, and begin taking data. If the new builds capability for sensitive nuclear detection, of inter- technology works as planned, then we can determine est to nuclear security programs. if the MiniBooNE signal is due to electrons, as opposed to photons, which is expect if the origin of the signal Progress is neutrino oscillations. A paper will be published on Wesley has been a major player in the development of these results. 769

Page 770

Physics Postdoctoral Research and Development Continuing Project 3D Turbulent Magnetic Reconnection Experiments and Simulations Thomas P. Intrator 20120768PRD3 Introduction kinetic scales, and display sporadic, bursty, and turbulent Background: The plasma universe including our solar dynamics. system is threaded with magnetic fields. These magnetic Benefit to National Security Missions fields are typically “frozen” into the magnetohydrody- namic (MHD) plasma fluid frame, and in many cases of One of the missions undertaken by DOE Office of Sci- interest play an essential role in organizing the plasma. ence is Fusion Energy Sciences. Plasma science forms the The twisting, folding, and annihilation of magnetic field basis for research that is needed to establish our ability can energize particles. Our knowledge of fundamental to harness the power of the stars in order to generate principles underlying local magnetic structure allows us fusion energy on earth. The research required for fu- to extend understanding to remote stars and galaxies, sion energy’s success is intimately tied to rich scientific solar system space weather, and magnetic fusion energy questions about some of nature’s most extreme envi- devices. ronments, inside and outside of stars, and has practi- cal implications to industry beyond energy as well. The Magnetic reconnection is a general category of impor- impact of a fusion energy source will be huge, as energy tant plasma physics processes in which magnetic field drives much of our domestic and foreign policies. A ma- lines reorganize, and usually change their topology while jor fusion energy scenario requires magnetized plasmas, annihilating the associated magnetic flux. The typi- and the situation in real life is fully 3D. Our proposed cal two-dimensional (2D) Sweet-Parker physics picture research speaks directly to the fundamental science of invokes oppositely directed magnetic fields that are magnetized plasmas. advected by plasma fluid motion towards each other, where they collide and form a thin discontinuity and an NASA also has a mission of Science, as we explore the induced current sheet. Magnetic flux diffuses into the Earth, solar system and universe beyond, to reap the thin “diffusion” region, and plasma jets are flung out in benefits of Earth and space exploration for society. Our slingshot fashion sideways as reconnecting magnetic research project aims to unravel some fundamental fields first bend, reconnect to other field lines, and fi- questions about how magnetic structure “works” in nally straighten out. However it is becoming increasingly three dimensions. apparent that these thin current sheets are inherently Progress unstable to the formation of islands or plasmoids, each of which has a current rope threading the O-point of The reconnection scaling experiment (RSX) at LANL each island. The necessity of including the magnetic flux provides diagnostics to observe reconnection in the full rope dynamics forces us to confront the 3D nature of the three dimensions from macro to micro length scales. problem. Any useful reconnection theory should account We will measure 3D interactions of reconnecting cur- for the 3D features. rent sheets, as they shred into islands (which are current ropes in 3D) that cascade down to kinetic scales, and Proposed work: The reconnection scaling experiment display sporadic, bursty, and turbulent dynamics. (RSX) system at LANL provides diagnostics to observe reconnection in the full three dimensions from macro to During the past 7 months (2012Nov-2013May) since micro length scales. We will measure 3D interactions of Dr. Feng arrived at LANL, we have been characterizing reconnecting current sheets, as they shred into islands the 3D behavior of a single flux rope. The behavior of a (which are current ropes in 3D) that cascade down to single flux rope allows us to understand key and simpli- 770

Page 771

fied features of this complicated 3D, two flux rope prob- lem. Two post docs (J. Sears and Y. Feng) have measured the non linear dynamics of plasma magnetic fields, plasma pressure, and flows. We have amassed a huge 3D dataset (> 2000 shots, 25000 probe locations) for a single Recon- nection Scaling Experiment (RSX) flux rope. The 3D analysis shows unexpected behaviors. We are in the process of intpreting the experimental results, and already have a story that seems destined for Physical Review Letters. Future Work A large array of multiple plasma guns will be fabricated and installed on the Reconnection Scaling Experiment. This will allow us to create flux ropes of varying sizes and shapes by merging, 1,2,3,4 or more plasma gun jets into single streams of plasma. Current sheets will then be created from rows of plasma guns and we will experimentally char- acterize their interactions. In parallel, we will start to set up the input parameters for the VPIC particle in cell code, so that we can simulate the RSX experiment computation- ally. Conclusion Magnetic reconnection plays a crucial role in many dif- ferent astrophysical and laboratory plasma phenomena, which are all intrinsically 3D. Recent theories and simu- lations implicate generic current sheet instabilities that generate islands. Cascades of plasmoids down to kinetic spatial scales naturally link reconnection, turbulence, cas- cades, and dissipation. We propose the first experiments to verify the 3D dynamics of magnetic field annihilation and reconnection. We will carry out 3D computational simulations and benchmark them to experimental data to better understand the formation and interaction of flux ropes within the Earth’s magnetosphere, and the basic physics of laboratory, space and astrophysical plasmas. Publications Feng, Y., J. Goree, and B. Liu. Longitudinal viscosity of two- dimensional Yukawa liquids. 2013. PHYSICAL REVIEW E. 87 (1): -. Goree, J., B. Liu, and Y. Feng. Diagnostics for transport phe- nomena in strongly coupled dusty plasmas. To appear in Plasma Physics and Controlled Fusion. Intrator, T. P., X. Sun, L. Dorf, J. A. Sears, Y. Feng, T. E. We- ber, and H. O. Swan. Flux ropes and 3D dynamics in the Relaxation Scaling Experiment. To appear in Plasma Physics and Controlled Fusion. 771

Page 772

Physics Postdoctoral Research and Development Continuing Project Measurement of Orthogonal Velocity Components Using Photon Doppler Velocimetry Matthew E. Briggs 20120771PRD4 Introduction could indeed observe the phenomenon as proposed. The primary research will be experiments using PDV in The second is a much longer paper laying out the details laboratory and gun shots and the addition of the refer- of how to apply our analysis and comparing to data. In ence electric field to the scattered electric field in the addition, Erik has laid out a program for the Los Alamos theoretical analysis. This could demonstrate the extent Dynamics Summer School that will engage a team of to which PDV with dynamic speckle analysis can simulta- undergraduates in exploring aspects of his research. The neously measure axial and lateral target surface veloci- students are arriving next week. He has also prepared a ties, using existing test setups. This capability would paper for the post-doc poster day. provide a detailed view of a structure’s response to Future Work excitation, thereby lifting a fundamental limitation on currently deployed velocimetry techniques. During the first year we will complete a literature survey on the topic, and present a research plan at the annual The theoretical development of dynamic speckle mea- Photon Doppler Velocimetry Workshop. We will decide surements for extracting lateral motion from PDV data which of the various analysis tools in the literature for records, along with any corresponding experimental extracting lateral velocity from dynamic speckle to use investigations are unique and appropriate for publishing first, test it with existing laboratory experimental ve- in a journal that covers optical sensor research. locimetry apparatus, and then develop a plan for the second year. Benefit to National Security Missions Velocimetry is used widely across many applications. By We have accomplished the above as planned, and Erik giving insight into the lateral velocity (new) simultane- is now preparing a 2nd paper on the topic (his first has ously to longitudinal velocity (existing), our understand- already been published.) Our plan for next year is to ex- ing of material responses to driving forces will be quali- pand the testing of the speckle analysis on experiments tatively advanced. This will provide significant additional involving straining and non-straining surfaces in order to constraints on the physics and modelling of systems see if we can distinguish between the two. important to a broad range of federal missions. In par- Conclusion ticular, velocity measurements are primary or secondary on many of our most important nuclear weapon safety, To demonstrate the extent to which PDV with dynamic security, reliability & non-proliferation measurements. speckle analysis can simultaneously measure axial These measurements are also primary on many material and lateral target surface velocities, using existing test response to shock experiments. setups. This research will be of broad interest to the velocimetry community, and to the larger community Progress of material and weapons scientists. Even qualitative We have accomplished the work as planned: We have information about lateral motion fills a limitation on completed 3 rounds of laboratory tests that have dem- velocimetry measurements that has been present since onstrated that we can indeed measure the phenom- the start of the field. If quantitative analysis of lateral enon as intended, and elucidated the physical principles motion turns out to be robust, then the applications will underlying the phenomenon. Erik has published his first become much more wide ranging, as PDV is now being paper on the topic and is now preparing a 2nd paper used around the world. on the topic. The first paper was a note that showed we 772

Page 773

Publications Briggs. Laser speckle dynamics sensing applications. To Briggs, M., D. Knierim, E. Moro, and S. McGrane. Optical appear in IMAC-XXXII. (Orlando, 2014). distance measurements to recover the material ap- proach missed by optical velocimetry. To appear in Shock Compression of Condensed Matter. (Seattle, July 2013). Hull, L. M., M. E. Briggs, and E. A. Moro. The effect of tan- gential velocity components and surface evolution on photon Doppler velocimetry measurements. 2013. Los Alamos National Laboratory Report. MOro, E. A.. New Developments in photon Doppler ve- locimetry. To appear in Shock Compression of Con- densed Matter. (Seattle, July, 2013). Moro, E. A.. Extracting transverse motion from speckle dynamics in photon Doppler Velocimetry. Presented at Postdoc Research Day at Los Alamos National Labora- tory. (Los Alamos, 2013). Moro, E. A., M. E. Briggs, and L. H. Hull. Point-measure- ment of strain-induced speckle dynamics, observed with photon Doppler velocimetry. To appear in 15th International Detonation Symposium. (San Fransisco, 2014). Moro, E. A., M. E. Briggs, and L. M. Hull. A comparison of techniques for extracting transverse speed from pho- ton Doppler velocimetry signal content. To appear in IEEE Sensors. (Baltimore, 2013). Moro, E. A., M. E. Briggs, and L. M. Hull. “H4442, H4442 Test Report: Surface evolution of radially expanding cylinders and the measured response in photon Dop- pler velocimetry. 2013. Los Alamos National Labora- tory. Moro, E. A., and M. E. Briggs. Note: Simultaneous mea- surement of transverse speed and axial velocity from a single optical beam. 2013. Review of Scientific Instru- ments. 84: 016110. Moro, E. A., and M. E. Briggs. Defining parametric-depen- dencies for the correct interpretation of speckle dy- namics in photon Doppler Velocimetry. Applied Optics. Moro, E. A., and M. E. Briggs. Extracting transverse motion from speckle dynamics in photon Doppler velocimetry. Presented at 2013 Nuclear Explosives Design Physics Conference. (Livermore, January, 2014). Moro, E. A., and M. E. Briggs. Extracting lateral motion from the dynamic amplitude modulations-the dynamic speckle-of a PDV signal. 2012. In 2012 PDV Workshop at Sandia National Laboratory. (Albuquerque, 2012). , p. 1. Columbus: Ohio State University. Warren, W. J., L. Ott, E. Elmore, E. A. Moro, and M. E. 773

Page 774

Physics Postdoctoral Research and Development Continuing Project The Role of Microscopic Eye Movements in Visual Perception John S. George 20120773PRD4 Introduction Such efforts have been supported by recent and ongo- Eye movements are fundamental for processing and per- ing LDRD projects and are highly relevant for missions in ceiving visual information about the surrounding world. DOE, DoD, and the intelligence community. Large-scale eye movements, called saccades, facilitate Progress exploration of the visual environment by successively positioning the fovea (the region of retina that provides During the first 8 months of this project, Dr. Schei and greatest visual acuity) on areas of interest in the visual collaborators have made significant progress on all of field. During periods of eye “fixation,” microscopic eye the key scientific and technological objectives outlined in movements–shifts on the order of a photoreceptor cell– the research proposal and initial datasheet. are observed in all vertebrate animals. Due to complex Electrophysiological studies of isolated retina from tiger experimental requirements, the role of these move- salamander have produced robust recordings of the ments, called ocular tremor, is not well understood. We population electroretinogram (an integrated response hypothesize that ocular tremor contributes to low-level over a large population of retinal neurons) as well as visual processing by generating patterns of synchronous reliable recordings of neural spiking (action potentials firing between neighboring retinal neurons, and that produced by individual neurons) generated by controlled correlated firing encodes unique spatiotemporal visual light stimuli. A series of experiments were conducted to information, improving signal-to-noise ratios and visual simulate biological eye movements by mechanically jit- acuity. I will study neural firing in response to visual stim- tering a small light source, which was projected through ulation of isolated retina, and characterize the perceptu- the microscope optics onto the retina. We observed a al consequences of eye movements in human subjects. strong correlation between neural oscillations apparent These experiments are inspired by LANL computational in the electrophysiological response (a time histogram of models that simulate retinal and neural consequences of recorded spikes) and the frequency of stimulus jitter. ocular tremor. The work exploits laboratory technologies for high spatial/temporal resolution measurement of Experimental results were compared to the results and neuronal function in order to provide further insight into predictions of neural simulations, and a poster was visual signal encoding and processing. presented at the Vision Sciences Society meeting in May of 2013. Benefit to National Security Missions The objective of the work is to better understand the Combined electrophysiological and optical recordings of mechanism that supports the encoding and processing retinal responses have been obtained. This should help of visual information by the retina and cortical visual identify the specific cellular sources of dynamic electrical systems of the brain. This work will enable better responses. interpretation of neural signals for diagnostics of pathol- ogy, and facilitate treatment of retinal degeneration or A new video-based stimulation system is being assem- neural injury via neural prosthetic devices such as those bled which can be used for both in vitro retinal physiol- developed with DOE support in the Artificial Retina ogy and for human perceptual and physiological experi- project and by NIH. The work will also enhance our ments. understanding of neural mechanisms of visual cognition, Basic setup of a fundus camera suitable for noninvasive supporting the development of neural-inspired compu- imaging in situ of the human retina has been completed, tation systems for image processing and interpretation. 774

Page 775

except for coupling the optical system to an electronic im- specialized “analog” neurons of retina, as well as the spik- ager, and integrating the video stimulation system. (These ing activity of the retinal ganglion cells that constitute the components are presently being used for tissue physiology optic nerve. I will combine high- speed video microscopy experiments.) with high-density electrophysiological measurements to investigate the relationship between neural firing patterns Basic setup has been completed of an adaptive optics and optical responses imaged in retina. system intended to allow imaging of individual photore- ceptors in the human retina through fundus camera optics. Specific Aim 3: Characterize and simulate ocular tremor This should allow us to track microscopic eye movements in the human eye. I will use high-speed optical imaging noninvasively with human subjects in perceptual experi- techniques with an ophthalmoscope camera employing ments, as well as to characterize intrinsic eye movements adaptive optics to correct aberrations in eye’s optics. This in salamanders. instrument will allow us to characterize optical tremor in an intact human eye, and to measure neural responses Enhancements to the high-speed camera system (firmware to tremor or to synthetic motion of a visual stimulus. The upgrades to the camera, and replacement of the computer objective is to assess the role of physiological encoding in based data acquisition system) have been completed. perceptual judgment. Schei is mentoring a Los Alamos Scholarship Fund student Conclusion intern during the summer of 2013. This student is design- Physiological measures of natural or simulated ocular ing and implementing human perceptual experiments tremor will enhance our understanding about retinal pro- (employing fast video presentations with or without simu- cessing and encoding visual information, by testing and im- lated eye movements) that will be used to characterize proving current computational models of retinal function. perceptual performance in preparation for future experi- A deeper understanding of neural encoding and processing ments that will additionally collect noninvasive optical and will enable better interpretation of neural signals for diag- electrophysiological measures. nostics of pathology, prosthetics to treat neural injury, and neural-inspired computation systems. Furthermore, these Schei has continued to contribute to an ongoing LDRD experiments will provide tools to define the role of ocular ER project on Advanced Neural Interfaces, assisting with tremor in human visual perception by combining physi- measurements of retinal electrophysiology used to validate ological and psychophysical measurement techniques. and characterize performance of novel devices produced by the project. Publications In addition to the VSS presentation, this work was present- George, J., J. L. Schei, P. Schultz, and G. Kenyon. Implica- ed at Los Alamos Postdoc Research Day and was awarded tions of microscopic eye movements for retinal encod- honorable mention. Another publication based on Schei’s ing. 2013. In Vision Sciences Society. (Naples, Fl, 10-15 May 2013). Vol. 13, 9 Edition, p. 1341. New York: Jour- graduate work has recently been accepted for publication nal of Vision. in the disciplinary journal Sleep. Schei, J. L., G. Kenyon, and J. George. Contributions of Future Work microscopic eye movements to retinal encoding. To Specific Aim 1: Simulate ocular tremor in an isolated am- appear in Optical Society of America Vision Meeting. phibian retina by jittering a visual stimulus, while measur- (Houston, 4-6 Oct. 2013). ing electrophysiological activity of retinal neural popula- tions. I will use commercial microelectrode arrays, as well as high-density nano-pillar arrays developed and fabricated at CINT, for high fidelity neural measurements of neural ensemble firing patterns. The objective is to characterize neural firing patterns in response to moving stimulation patterns, and to confirm the proposed physiological foun- dation of perceptual enhancement. Specific Aim 2: Correlate neural spiking activity to intrinsic optical responses of isolated retina. Electrophysiological activity alters the optical properties of neurons and can be used to measure subthreshold activity and responses of 775

Page 776

Physics Postdoctoral Research and Development Continuing Project Using Jet Production to Investigate quark-Gluon Plasma at RHIC Xiaodong Jiang 20120775PRD4 Introduction and Particle physics research. The project, using a new With the PHENIX experiment at Relativistic Heavy Ion experimental probe of particle jet production, is aimed Collider (RHIC) of DOE’s Brookhaven National Labora- at answering the question of “what are the properties tory (BNL), this research is in part a study of two-particle and the behaviors of the extremely hot an dense Quark- correlations demonstrating how the path-length through Gluon Plasma formed within the first few seconds right the Quark-Gluon Plasma (QGP) affects the fast moving after BigBang, the start of our Universe ? “. This project particles to a degree beyond that expected within many is aligned with the Nation’s Nuclear Physics Long Range theoretical expectations. Since then he have been help- Plan (DOE Office of Science), advancing our basic knowl- ing the PHENIX experiment, in which LANL participates edge of fundamental nuclear physics. The hardware heavily, extend the current suite of measurements by and software tools developed through this project and separating out fast moving charm and bottom quarks. the technical expertise gained through this project will One major goal of this project is to use the fact that also benefit LANL’s mission on energy security, national heavy quarks traverse the QGP differently depending on security and nuclear non-proliferation. the mass of the quasi-particle excitations to study the This project expand the study of physics of Quark Gluon properties of Quark-Gluon Plasma. Experimentally, this Plasma into using particle-jets to investigate the proper- approach requires excellent particle tracking near the ties of quark gluon plasma. This topic and the research collision point and to that end PHENIX has installed a method proposed is a brand new direction in the field new set of vertex tracking detectors, the Vertex Detector of High Energy Nuclear Physics. Therefore, this project (VTX) and the Forward Vertex Detector (FVTX), the latter is not improperly augmenting the existing P25 research a $5 million DOE project led by the LANL P-25 group. program currently funded by DOE. Specifically the postdoc fellow will continue to contrib- ute to the data acquisition system of the PHENIX experi- With the PHENIX experiment at Relativistic Heavy Ion ment, of which he is already a well-respected expert, Collider (RHIC) of DOE’s Brookhaven National Laboratory and working on the reconstruction algorithms of the VTX (BNL), LANL team at BNL contact: Dr. Xiadong Jiang and detector. He is also expected to participate deeply in the Dr. Ming Liu 505 606 0437. next decadal plan to upgrade the PHENIX detector (the sPHENIX proposal, and the forward-sPHENIX proposal) to Progress fully reconstruct jets and thereby better characterize the Over the last six months significant progresses and ac- quasi-particle description of the Quark-Gluon Plasma. complishments have been made on the following fronts: This project expands the study of physics of Quark Gluon • RHIC PHENIX Heavy Ion physics analysis using Run- Plasma into using particle-jets to investigate the proper- 2012 data and information from Vertex Detector ties of quark gluon plasma. This topic and the research (VTX) and forward Vertex Detector (FVTX). Particle method proposed is a brand new direction in the field tracking codes, with a brand new tracking algorithm of High Energy Nuclear Physics. Therefore, this project has been developed and tested. Raw data processing is not improperly augmenting the existing P25 research of U+U, Cu+Au, Au+Au collisions have been complet- program current funded by DOE. ed with the new package of tracking code, physics analysis have made good progresses, and various Benefit to National Security Missions PHENIX “preliminary-results” have been reported in This project is at the very frontier of High-Energy Nuclear 776

Page 777

international conferences. Mike continue sto lead the key measurements and design parameters of the cen- Heavy Flavor Physics working group within the PHENIX tral barrel and the forward sPHENIX detectors. Collaboration. • Physics analysis of jet observables in heavy-ion and • A new run (RHIC Run-2013, with the FVTX detector p+p collisions. fully functional) was completed Jan-June 2013, with This project expands the study of physics of Quark Gluon 510 GeV proton+proton collisions. Mike’s efforts on Plasma into using particle-jets to investigate the proper- supporting FVTX day-to-day operation, calibrations and ties of quark gluon plasma. This topic and the research online detector performance checks have earned him a method proposed is a brand new direction in the field of high respect within the Collaboration. Physics analysis High Energy Nuclear Physics. Therefore, this project is not process is well underway, with Mike leading the efforts improperly augmenting the existing P25 research program of FVTX tracking. Mike has been selected to give invit- current funded by DOE. ed PHENIX presentations at international conferences such as HardProb2013, Physics-In-Collision-2013. Conclusion • On the future PHENIX upgrades, Mike has the lead in With the PHENIX experiment at Relativistic Heavy Ion jet production simulations for sPHENIX and Forward- Collider (RHIC) of DOE’s Brookhaven National Laboratory sPHENIX, defined physics requirements for building (BNL), LANL team at BNL contact: Dr. Xiadong Jiang and Dr. new spectrometer/detectors. Several progress reports Ming Liu 505 606 0437. have been made by Mike to the PHENIX Collaboration on jet shape, jet-identification and jet-reconstruction • Combined tracking analysis code of Phenix VTX and during Collaboration meetings, leading to major tech- FVTX detector. nical decisions by the Collaboration towards the future • Data quality control during the operation in run-2013 construction of sPHENIX and forward-sPHENIX detec- and run-2013 for VTX and FVTX detectors, tors. • Preliminary analysis results of PHENIX run-2012 data Future Work on heavy flavor quark, and “jet-like” observables in With the PHENIX experiment at Relativistic Heavy Ion Heavy-Ion collisions. Collider (RHIC) of DOE’s Brookhaven National Laboratory (BNL), LANL team at BNL contact: Dr. Xiadong Jiang and Dr. • Preliminary design report of sPHENIX and forward- Ming Liu 505 606 0437. sPHENIX upgrades, using jet observable to probe QGP in Heavy Ion Collision. The Postdoc Fellow is expected to work very closely with P-25 group members on the PHENIX experiment focusing Publications on heavy ion physics with the Vertex Tracker (VTX) and Adare, A., C. Aidala, N. N. Ajitanand, Y. Akiba, H. Al-Ba- the Forward Vertex Tracker (FVTX ) as well as preparing for taineh, et.al. Cold-Nuclear-Matter Effects on Heavy- future upgrades. As such, the technical responsibilities of Quark Production in d+Au Collisions at root S-NN=200 his research include: GeV. 2012. PHYSICAL REVIEW LETTERS. 109 (24): -. • PHENIX experiment at RHIC: heavy quark energy loss Adare, A., C. Aidala, N. N. Ajitanand, Y. Akiba, R. Akimoto, and flow, large -gap two-particle correlations H. Al-Ta’ani, et.al. J/psi suppression at forward rapidity in Au plus Au collisions at root s(NN)=39 and 62.4 GeV. • sPHENIX upgrade at RHIC: jet physics in both heavy ion 2012. PHYSICAL REVIEW C. 86 (6): -. collisions and polarized protons Adare, A., S. Afanasiev, C. Aidala, N. N. Ajitanand, Y. Akiba, H. Al-Bataineh, et.al. Spectra and ratios of identified Specific items include: particles in Au plus Au and d plus Au collisions at root • Hardware responsibilities for the FVTX project: DAQ s(NN)=200 GeV. 2013. PHYSICAL REVIEW C. 88 (2): -. and detector operations in Run13 and Run14. Adare, A., S. Afanasiev, C. Aidala, N. N. Ajitanand, Y. Akiba, H. Al-Bataineh, et.al. Transverse-momentum depen- • Software responsibilities for the FVTX and VTX proj- dence of the J/psi nuclear modification in d+Au colli- ects: FVTX track reconstruction, integration of VTX and sions at root s(NN)=200 GeV. 2013. PHYSICAL REVIEW FVTX algorithms, physics analysis C. 87 (3): -. • sPHENIX physics and detector design: Simulations of Adare, A., S. Afanasiev, C. Aidala, N. N. Ajitanand, Y. Akiba, 777

Page 778

H. Al-Bataineh, et.al. Measurement of transverse sin- gle-spin asymmetries for J/psi production in polarized p + p collisions at root s = 200 GeV (vol 82, 112008, 2010). 2012. PHYSICAL REVIEW D. 86 (9): -. Adare, A., S. Afanasiev, C. Aidala, N. N. Ajitanand, Y. Akiba, H. Al-Bataineh, et.al. Inclusive cross section and single transverse spin asymmetry for very forward neutron production in polarized p plus p collisions at root s=200 GeV. 2013. PHYSICAL REVIEW D. 88 (3): -. Adare, A., S. Afanasiev, C. Aidala, N. N. Ajitanand, Y. Akiba, H. Al-Bataineh, et.al. Neutral pion production with re- spect to centrality and reaction plane in Au plus Au col- lisions at root S-NN=200 GeV. 2013. PHYSICAL REVIEW C. 87 (3): -. Adare, A., S. Afanasiev, C. Aidala, N. N. Ajitanand, Y. Akiba, et.al. Medium Modification of Jet Fragmentation in Au plus Au Collisions at root S-NN=200 GeV Measured in Direct Photon-Hadron Correlations. 2013. PHYSICAL REVIEW LETTERS. 111 (3): -. Adare, A., S. Afanasiev, C. Aidala, N. N. Ajitanand, Y. Akiba, et.al. gamma (1S+2S+3S) production in d plus Au and p plus p collisions at root s(NN)=200 GeV and cold-nucle- ar-matter effects. 2013. PHYSICAL REVIEW C. 87 (4): -. Adare, A., S. Afanasiev, C. Aidala, N. N. Ajitanand, Y. Akiba, et.al. Double-spin asymmetry of electrons from heavy- flavor decays in p plus p collisions at root s=200 GeV. 2013. PHYSICAL REVIEW D. 87 (1): -. 778

Page 779

Physics Postdoctoral Research and Development Continuing Project Extending the Time-scale of Protein Simulations Using Accelerated Molecular Dynamics Techniques Arthur F. Voter 20130777PRD1 Introduction Office. This work could also lead to improved under- The goal of this project is to apply the parallel replica standing of bio-chemical reactions inside the cell that dynamics (ParRep) method to the dynamics of protein could be used for energy-related applications, also an folding. ParRep advances an infrequent-event system area of interest to the CSGB. rapidly from state to state by parallelizing time over Progress many replicas of the system in each state. A recent ad- vance in the understanding of the generality of ParRep, This project began just a few months ago. Exploratory allowing for more general state definitions, opens the simulations have been performed using the GROMACS possibility of effective application to complex systems code on the alanine dipeptide system. Tens of thou- such as proteins, something that was previously thought sands of trajectories were initiated from a single geom- to be impossible in most cases. This new approach etry, and the escape time distribution was determined will be developed and tested on a small model peptide for various definitions of the state boundary, as defined system such as alanine dipeptide, for which the exact by rectangular regions in the phi-psi (Ramachandran) dynamics are known, and then extended to more com- plot. As predicted by the quasi-stationary distribution plicated systems such as Met-enkephalin, verifying the theory, after a suitable dephasing time, the decay of folding characteristics against existing data. If this new the population from the defined state becomes cleanly methodology can be developed to be effective for ac- exponential. Thus, as anticipated, this system and this curate protein folding, the potential payoff is substantial, type of state-boundary definition would allow treatment as most protein folding and function takes place on time with parallel replica dynamics to accelerated the rate of scales longer than are accessible with direct molecular state-to-state transitions. However, one aspect that is dynamics. Moreover, this methodology should transfer less than satisfactory for using this as a model system is well for study of dynamics in other complex and soft- that the average escape time from the deepest state is matter systems important in nanotechnology, energy not long enough to get significant parallelization gain. conversion and alternative energy generation. Thus, we are now in the process of choosing a second system to continue our study, one that has deeper states Benefit to National Security Missions that give less frequent events. The accurate description of processes such as protein Future Work folding, protein function, drug function, etc., on biologi- cally relevant timescales is a long-standing problem. In the first year, the parallel-replica dynamics method This project aims to develop a methodology to extend will be implemented into the GROMACS bio-simulation the accessible simulation time scale for these types of package. We will then begin developing the new, gen- processes. If this project succeeds, it will impact not eralized methodology on a small model peptide system only bioscience, but other areas where complex or soft- such as alanine dipeptide, where the dynamics will be matter dynamics are important. Such a capability could compared to the known exact dynamics for that same be extremely useful to understand active site protein model system. We will investigate definitions for the chemistry, which could lead to the design of bio-inspired states of the system, with the goal of finding the largest catalysts with high efficiency and selectivity. This spe- separation of time scales between the dephasing time cifically addresses a key mission of the Physical Biosci- (the time to settle into the equilibrium or steady-state ences program of the Chemical Sciences, Geosciences, & distribution) and the average escape time from the state. Biosciences (CSGB) Division of the Basic Energy Science We will also begin investigating how to determine, in 779

Page 780

general, the length of this dephasing time. This is the most challenging part of the project. Conclusion The overall goal of this project is to develop a viable methodology for application of the parallel-replica dynam- ics method to protein folding. Success in this project will impact biochemistry, as time scales for important protein dynamics, DNA dynamics, membrane function, etc., are beyond the reach of standard molecular dynamics. Suc- cess will also impact other fields where soft-matter or com- plex dynamics are important, as accurate atomistic evolu- tion of these systems on long time scales beyond about one microsecond is currently impossible. Examples include heterogeneous catalysis and processing steps in surface coatings for solar cells or semiconductors. 780

Page 781

Physics Postdoctoral Research and Development Continuing Project Electronic and Photonic Transport in Chiral Materials and Nanostructures Diego A. Dalvit 20130781PRD1 Introduction Faraday rotation experiments ongoing at LANL. Graphene and topological insulators are the most recent • Study of how chiral materials, including graphene discoveries in material science, and represent the cutting and topological insulators and superconductors, edge of condensed matter physics. Among many applica- influence Casimir interactions at the nanoscale. tions, these materials are the most promising candidates to replace semiconductors in commercial electronics. Conclusion They will make possible, for example, to create a micro- The expected results of this project is a comprehensive processor with orders of magnitude higher speed and study of the transport properties of strongly spin-orbit almost no energy losses. Spin-orbit coupling in these coupled materials including topological insulators and systems represent a form of chirality, which is a property superconductors, as well as nanostructures built from prevalent in Nature. The study of the electronic and chiral materials such as graphene superlattices. Specifi- photonic chiral materials offers exciting opportunities cally, we will study how density-density and current- for discovering interesting new physics and technologi- current correlation functions depends on the spin-orbit cally important applications, including spintronics-based coupling, external magnetic fields, and electron-electron information processing, new material probe techniques, interactions. We will also perform research on the opti- and Casimir force repulsion. cal properties of chiral materials with the intention to elucidate how electromagnetic fluctuation-induced (Ca- Benefit to National Security Missions simir) forces depend on the physical parameters of such This project will impact LANL/DOE mission on basic en- systems, with the ultimate goal of determining optimal ergy sciences, materials, and nanotechnology, as it will strategies for controlling the Casimir effect. develop new theoretical and numerical tools to describe some of the most exciting materials in condensed matter physics, including graphene, topological insulators, and topological superconductors. It also has the potential of attracting interest from external agencies, such as DARPA, that have shown recent interest in controlling Casimir forces and in energy harvesting. Progress The postdoc arrived July 2013; therefore, there is no progress to date. Future Work The tasks to be performed are: • Study of charge/spin density response functions in chiral materials, that give important information about the charge/spin quantum fluctuations and their collective modes that can be measured by electron energy loss spectroscopy and spin-noise 781

Page 782

Physics Postdoctoral Research and Development Continuing Project Exploring Doubly Parasitic Radioisotope Production Via Secondary Neutron Fluence From the 100 MeV IPF Irradiations Eva R. Birnbaum 20130782PRD2 Introduction scientific purposes. The Isotope Production Facility (IPF) at Los Alamos Benefit to National Security Missions National Laboratory (LANL) produces the radioactive isotopes strontium-82 and germanium-68 for use in A comprehensive understanding of the secondary neu- medical imaging. The IPF’s high proton beam current and tron flux at the Isotope Production Facility will expand lengthy irradiations produce a secondary neutron field the capabilities of the Isotope Program to study and with a utilitarian scale that is beyond the reach of medi- produce a broader suite of isotopes for medical, national cal cyclotrons and energetically distinct from reactor security, and basic science applications. Studies will also neutron fluences. There are currently no large facilities improve nuclear models and provide an avenue for study in the United States with access to such a significant, of materials under extreme environments. high-energy neutron flux, yet IPF secondary neutrons’ Progress potential for research in novel methods of isotope production and materials science remains unexploited. Dr. Engle has made good progress in his primary propos- These radioisotopes are necessary to a variety of scien- al area to fully characterize the secondary neutron flux tific and medical fields, and in many cases the stability at the Isotope Production Facility. Based on simulations of future uninterrupted supply is uncertain. Additionally, calculated with MCNPX code, a series of materials were LANL’s globally recognized expertise in the characteriza- chosen to experimentally probe this neutron flux. A set tion and handling of radioisotopically and chemically of special activation foils were introduced into the neu- diverse products of fast neutron irradiations is uniquely tron field during the last IPF run cycle and brought to the capable of facilitating such an exploration. TA-48 count room for measurement. The radioisotope products in these foils yield information on neutron- Current two week irradiations produce several tera- initiated reactions. Measurements using these foils have neutrons per second, approaching the scale of medium- characterized these neutrons’ flux energy distribution, sized research reactors. The Monte Carlo N-Particle and efforts are now underway to investigate the produc- eXtended (MCNPX) code, developed at LANL, has been tion of 47Sc for medical radiotherapy, and to character- used to simulate the secondary neutron field and op- ize the products from high energy neutron irradiation timize the design of research targets that can confirm of thorium targets for the production of alpha emitting predictions about the size and charter of this neutron actinide isotopes. A second generation neutron target is flux. Using these simulations, special activation foils being developed which will enable longer-term, produc- have been introduced into the neutron field and the tion scale irradiations of chosen targets in a safe environ- radioisotope products of neutron-initiated reactions. ment tailored for downstream radiochemical isolation of Once computational predictions have been confirmed by desired radioactive products. experiment, MCNPX codes will be further used to design future experiments in materials science development Concurrently, Dr. Engle has benefitted from the proton and to produce radioisotopes in the service of the larger irradiation capabilities of the LANSCE and IPF facilities, scientific community. Delivery of these and other iso- which have been used to initiate nuclear formation cross topes whose study is proposed makes possible targeted section measurements for several radioisotopes of inter- radiotherapy for cancer patients, experiments in solid- est to medical and basic science research. Cross section state physics, diagnosis of a wide range of pathologies, measurements are in progress for a number of isotopes and the development of improved materials for civil and with potential medical application. Isotopes under 782

Page 783

investigation include: the lanthanide isotope 153Gd, which measurements of activation experiments using test is useful as a flood source for instrument calibration and foils and quantification of radioisotopes produced in imaging; the 165Tm/165Er parent/daughter pair, consid- the foils by neutron-initiated reactions. ered as novel approaches to targeted radioimmunotherapy • Use the validated codes to design a series of test ir- of malignant and viral diseases; and a comprehensive un- radiations and radiochemical separations to produce derstanding of radioactive impurities co-produced during scandium-47 using enriched titanium (Ti) and vana- proton irradiations of thorium to produce 225Ac for alpha dium (V) targets. therapy of disease. Two publications have been submitted on these works and an additional two manuscripts are in • Irradiate these test materials (Ti and V) at IPF and use preparation. gamma spectroscopy to quantify radioactive products and possible impurities. Dr. Engle has also been engaged by the ongoing LDRD-ER funded effort to measure neutrino mass via calorimetric • Conduct 200 and 100 MeV incident proton energy ir- electron capture spectroscopy of the rare-earth isotope radiations of thulium targets to complement 800 MeV 163Ho. The collaborative effort within C-IIAC, ISR-2, and irradiations completed in December 2012 and oversee NEN-5 to investigate potential routes to production of large assay of these irradiated samples using the C-NR coun- quantities of this isotope has resulted in an invited talk at troom and TA-48 Hot Cell Wing facilities. the NuMass 2013 meeting in Milan and in the submission of an additional manuscript since the Reines fellowship • Analyze spectroscopy of 800 MeV p + Tm irradiation was awarded. foils to calculate nuclear formation cross sections. Ap- ply these cross sections to the evaluation of spallation- The Isotope Program has recently been directed by the regime isotope production schemes. Office of Science to pursue the development of metal- lic rubidium targets, to replace RbCl as the material of • Finalize analysis of 40 – 200 MeV p + Tb residual cross choice for irradiations to produce 82Sr for cardiac positron sections to evaluate the potential for Gd-153 produc- emission tomography. Dr. Engle contributes to this project tion at IPF. Make determination regarding securing IP in collaboration with several LANL experts in accelerator or publication of results. targetry and separations chemistry. • Complete publication of 40 – 800 MeV p + Th target Since receiving the Reines award, Dr. Engle has also pre- residual cross sections. sented nuclear data at the 2013 International Nuclear Data Meeting in New York and at the 8th Targeted Alpha Ther- • Continue contributions Rb metal targetry and produc- apy Meeting at ORNL, and has consented to serve on the tion of Ho-163 for neutrino mass measurement. scientific committee of the 15th semi-annual Workshop on Targetry and Target Chemistry in Prague, scheduled for the Conclusion summer of 2014. The goal of this project is to elucidate a more complete understanding of the secondary neutron flux at the Iso- Future Work tope Production Facility. These experiments will define • Finish computational modeling of the Isotope Produc- production methodologies that will supply radioisotopes tion Facility neutron flux using LANL-developed state- on the cutting edge of research in several fields. Produc- of-the-art computational tools like MCNPX/MCNP6 and tion of isotopes including scandium-47, actinium-225, CINDER90. Work with Mike James (NEN-5), Stepan and radium-225 for medical radioimmunotherapy will be Mashnik (XCP-3), and others to investigate accuracy investigated. The neutron flux can also be used to study of evaluated data libraries and MCNP-implemented the effects of high radiation environments and thus fuel physics models, which are applied to the problem of development of materials and devices that can withstand computational predictions of isotope production using the extreme conditions in space or power production de- the IPF secondary neutron flux. vices such as fusion reactors. • Continue investigation of IPF’s potential for contribu- Publications tion to fusion energy research program through testing Engle, J. W., E. R. Birnbaum, F. M. Nortier, J. A. Rau, K. D. of tritium breeding, chemical speciation and migration John, and H. R. Trellue. Purification of (PU)-P-242 by through test materials. irradiation with thermal neutrons. 2013. NUCLEAR IN- STRUMENTS & METHODS IN PHYSICS RESEARCH SEC- • Tailor the predictions of these models to experimental TION B-BEAM INTERACTIONS WITH MATERIALS AND 783

Page 784

ATOMS. 298: 70. Engle, J. W., E. R. Birnbaum, H. R. Trellue, K. D. John, M. W. Rabin, and F. M. Nortier. Evaluation of Ho-163 produc- tion options for neutrino mass measurements with microcalorimeter detectors. 2013. NUCLEAR INSTRU- MENTS & METHODS IN PHYSICS RESEARCH SECTION B-BEAM INTERACTIONS WITH MATERIALS AND ATOMS. 311: 131. Engle, J. W., S. G. Mashnik, J. W. Weidner, L. E. Wolfsberg, M. E. Fassbender, K. Jackman, A. Couture, L. J. Bitteker, J. L. Ullmann, M. S. Gulley, C. Pillai, K. D. John, E. R. Birnbaum, and F. M. Nortier. Cross sections from pro- ton irradiation of thorium at 800 MeV. 2013. PHYSICAL REVIEW C. 88 (1): -. 784

Page 785

Physics Postdoctoral Research and Development Continuing Project Theoretical Investigation of Nucleon and Nuclear Structure at Very High Energies Ivan M. Vitev 20130783PRD2 Introduction extract such information from data collected in the US The theoretical study and experimental exploration of and abroad. It also provides much-needed theoretical the internal structure of nucleons (protons and neu- guidance for and interpretation of the experimental trons) and nuclei are of fundamental importance to sci- results from the flagship nuclear physics program in US ence and have recently entered a new exciting phase. In and abroad. This project will also help firm the scientific the past decades an understanding of nucleons in terms case for a future Electron Ion Collider. of quarks and gluons (partons) has successfully emerged. Progress Progress has been made in constructing a “one-dimen- sional” picture of the nucleon, in the sense that we Since the beginning of fiscal year 2013, we have made “only” know about the longitudinal motion of partons in significant progress toward the ultimate goal of this fast moving protons and neutrons. However, the par- project. We have published 4 papers: two in Physical ton’s motion that is perpendicular (transverse) to the Review D, one in Physics Letters B, and one in Physical proton momentum is still largely unknown and urgently Review Letters. We are also actively involved in prepar- needed in order to construct the 3-dimensional image ing a white paper on the future experimental direction of the proton. Since the transverse momentum compo- on spin physics. This white paper discusses future critical nent is usually much smaller than parton’s longitudinal measurements and theoretical advances in spin physics component, it is critical to develop accurate theoretical for the upgraded Relativistic Heavy Ion Collider (RHIC) tools to extract this information. We propose two unique at Brookhaven National Laboratory. In particular, our ways to attach this problem. One is through the trans- proposed observables and simulations are essential for verse spin dependence of experimental observables, as understanding the transverse spin dependent observ- transverse spin can correlate with the parton transverse ables. momentum. The other way is through a discovery of Two highlights of our work concern the evolution and a novel nuclear dependence, as the small transverse universality properties of the so-called Sivers effect. The motion can be amplified by the nuclear size. We will Sivers function represents the probability to find unpo- develop the first solid theoretical framework that can larized patrons inside a transversely polarized proton. utilize both methods to pinpoint the partons’ transverse It contains both the longitudinal and transverse motion motion in both the proton and the big nucleus. of the partons. Understanding how the Sivers function changes with energy and in different processes is critical Benefit to National Security Missions to extract the information on the Sivers function consis- This project ties directly into the Laboratory mission tently from the world-wide experimental data. in scientific innovation and discovery. Understanding the structure of the nucleon is central to the nuclear In a recent paper published in Physical Review D, we physics component of the National Science Foundation computed the next-to-leading order contribution to (NSF) and Department of Energy (DOE) Office of Sci- the transverse momentum weighted Sivers asymmetry ence portfolios. Specifically, this project will provide new in semi-inclusive hadron production in deep inelastic and unique 3 dimensional imaging information for the scattering (SIDIS). From this calculation, we were able to nucleon structure at the ~20% level. It will also give es- extract the momentum scale evolution of the relevant sential insights into how the quark and gluon dynamics parton correlation function. This parton correlation are modified in a large nucleus. It will produce theoreti- function allows us to pin down the averaged parton cal tools for the community to accurately and reliably 785

Page 786

transverse motion inside the transversely polarized pro- clear dependence of experimental spin observables in both ton. Our calculation also demonstrates how this averaged electron-nucleus and proton-nucleus collisions. We will use transverse motion changes when one probes it at different these theoretical formalisms to interpret the experimental momentum scales. data collected at the major experimental facilities in the US and abroad and extract the valuable information on the In another paper, published in Physical Review Letters, we parton’s transverse motion. This will provide unique 3-di- studied how the Sivers effect changes when going from mensional imaging information for the proton and nuclear SIDIS to single inclusive jet production in proton-proton structure at high energies. collisions. In particular, we evaluated the initial-state and final-state interactions between the active parton and the Publications remnants of the proton, which are responsible for the dif- Aschenauer, E. C., A. Bazilevsky, K. Boyle , R. Fatemi , C. ference of the observed Sivers effect in different processes. Gagliardi, M. Grosse-Perdekamp, Z. B. Kang, Y. Kov- By doing so, we found the relation between the Sivers chegov, and J. Lajoie. The RHIC Spin Program: Achieve- function in these two different processes. We first per- ments and Future Opportunities . 2013. White Paper formed a global fitting of the existing SIDIS experimental on RHIC spin physics to the Tribble Panel and Nuclear data to extract the Sivers function. We then started with Science Advisory Committee. this Sivers function, properly took into account the differ- Gamberg, L., Z. B. Kang, and A. Prokudin. Indication on the ences in the Sivers function in single inclusive jet produc- process-dependence of the Sivers effect. 2013. Physical tion, and made a prediction for the spin asymmetry for Review Letters. 110 (23): 232301. the jet production. We compared our prediction to the recent available experimental data from AnDY collabora- Gamberg, L., and Z. B. Kang. Single transverse spin asym- tion at RHIC at Brookhaven National Laboratory and found metry of prompt photon production. 2012. Physics Let- reasonable agreement. This provides a first indication on ters B. 718 (1): 181. the process-dependence of the Sivers effect in different Kang, Z. B.. QCD and RHIC spin physics. Invited presenta- processes. tion at 2012 Fall Meeting of the APS Division of Nucle- ar Physics. (Newport Beach, CA, 24-27 Oct. 2012). Future Work As emphasized in our project description, we will take Kang, Z. B.. Polarized p+A, single spin asymmetries. Invited two complementary approaches to pin down the parton’s presentation at BNL-LANL-RBRC Joint Workshop on The Physics of p+A Collisions at RHIC. (Upton, NY, 7-9 transverse motion in nucleons and nuclei. Focusing on Jan. 2013). the spin dependence, in the next fiscal year we will derive energy evolution equations for the parton momentum Kang, Z. B.. Single transverse spin asymmetries in polarized distribution inside the proton. This will allow for a direct SIDIS and pp scattering. Invited presentation at The 5th comparison between theory and experimental data taken Workshop of the APS Topical Group on Hadronic Phys- at very different center of mass energies. We will then ics. (Denver, CO, 10-12 Apr. 2013). implement these evolution equations in the description of the experimental data on semi-inclusive hadron produc- Kang, Z. B.. Forward physics from a theoretical perspective. Invited presentation at STAR Meeting on eSTAR Letter tion in deep inelastic scattering (SIDIS) processes. Using a of Intent, Forward-Upgrades and Results from U+U global fitting procedure, we will extract for the first time Collisions. (Los Angeles, CA, 28-30 Aug. 2013). the parton transverse momentum distributions with much better accuracy. On the nuclear dependence side, we will Kang, Z. B., I. Vitev, and H. Xing. Transverse momentum- focus on the multiple quark and gluon scattering in a big weighted Sivers asymmetry in semi-inclusive deep in- nucleus that generates transverse motion. In particular, elastic scattering at next-to-leading order. 2013. Physi- we will perform the first calculation of the multi-parton in- cal Review D. 87 (3): 034024. teraction contributions to the same SIDIS process with full Kang, Z. B., and B. Xiao. Sivers asymmetry of Drell-Yan pro- next-to-leading order accuracy. This will enable us to gain duction in the small-x regime. 2013. Physical Review D. experience and apply the same method to other experi- 87 (3): 034038. mental processes. Vitev, I.. Transv. Invited presentation at Transverse momen- Conclusion tum-weighted Sivers asymmetry in semi-inclusive deep We will develop a new consistent theoretical formalism inelastic scattering at next-to-leading order. (Newport for evaluating the cross sections in polarized reactions with News, VA, 6-10 May 2013). protons and nuclei. We will also investigate the novel nu- 786

Page 787

Physics Postdoctoral Research and Development Continuing Project The Intersection of the Energy, the Intensity, and the Cosmic Frontiers in the Era of the Large Hadron Collider and XENON experiments Michael L. Graesser 20130789PRD2 Introduction Benefit to National Security Missions The overall goal of the project is to develop new meth- One of the central goals of particle physics and the DOE ods and theories that may deepen our understanding Office of Science High Energy Physics is to understand of particle physics in two areas. One area is at distances the physics of matter at very short distances - 1000 100-1000 times smaller than a proton, where we think times smaller than the size of a proton - and at ener- the physics “responsible” for giving mass to fundamental gies 100-1000 times higher than the energy in a proton. particles is occurring. The other is during epochs of the Currently the main experimental facility to achieve that very early universe, when the abundance of dark matter goal is the Large Hadron Collider (LHC) located outside of was being set and when the excess of matter over anti- Geneva, Switzerland. This experiment recently discov- matter occurred. ered a new particle that has the physical features of the so-called “Higgs boson”. To do that, the project will develop new models for dark matter interacting with the Higgs boson, and simultane- Another important thrust of particle physics and the ously, develop new experimental tools that will improve DOE Office of Science High Energy Physics is to under- the ongoing searches at the Large Hadron Collider (LHC) stand the physical properties of dark matter. for new particles. The discovery of the Higgs boson brings into sharp focus The proposed research to look for connections between two important scientific questions: how does matter - dark matter and the Higgs boson may provide us with a such as quarks and the electron - gain inertial mass has new framework for understanding why the Universe has become acute; and why does the Universe appear to more particles than anti-particles. The proposed work have more particles than anti-particles? will either discover such a new framework, or provide us The discovery of additional new particles at the LHC will with a deeper understanding of the limitations of new provide more information that may help address the interactions between the Higgs boson and dark matter. first question. To that end, while at Los Alamos Tuhin An important research tool for understanding the impli- Roy will develop new search strategies to improve the cations of particle theories in a collider environment is experimental reach of the LHC. the use of jets. Jets are energetic sprays of particles that Tuhin will also develop new models of dark matter are detected at collider experiments. Jet substructure interacting with the Higgs boson, which may address the refers to the physical properties of the particles inside second question. The properties of such theories may be the jet. The substructure of a jet depends on the physics tested at the LHC and/or future dark matter direct detec- that produced it. tion experiments. The proposed research in jet substructure physics will generate important tools that will: improve the poten- Progress tial for the LHC to discover new particles; and improve This project started September 2013; therefore, there is the ability of the LHC to perform more accurate mea- no progress to date. surements. Applied to the Higgs boson, these methods may be quite rewarding, since they may enable experi- Future Work mentalists at the LHC to discover and measure various In the first year the proposed work will: decay modes of this particle. 787

Page 788

Develop tools for separating jets produced by the strong force, from those produced by the decay of a new particle As described under “Project Description”, “jets” are an important research tool for understanding the implications of particle theories in a collider environment. Jets are energetic sprays of particles, such as of protons, neutrons, pions, kaons,…, that are produced from the collision of two protons at a collider, and then detected inside the experiment. Jets are produced from Standard Model (SM) processes - specifically the strong force - but can also be produced in the decay of a new particle. The first step in achieving this goal is to identify specific classes of models of Beyond-the-Standard Model (BSM) physics that will be the target for discovery. Then a new idea for jet substructure will be applied to these classes of BSM models and SM backgrounds. The significance of the new idea(s) will be evaluated through computer simula- tions. Propose new interactions between dark matter and Higgs boson; then identify regions of the parameter space that allows for the Higgs boson to undergo a “first order” phase transition A phase transition refers to the transition of matter from one phase to another, such as water boiling to form water vapor or water cooling to form ice. A “first order” phase transition refers to a phase transition in which the transi- tion occurs through the formation of bubbles. An example of a first order phase transition is water boiling. A first order phase transition is one of the conditions that more or less must occur in the early universe in order for more baryons to be produced than anti-baryons. (Protons and neutrons are baryons.) Conclusion The overall goal is to advance our understanding of particle physics at distances much smaller than a proton and dur- ing epochs in the early universe. New methods are needed to improve the potential for the Large Hadron Collider (LHC) to discover new particles and to improve accuracy. The results of this proposal may be quite rewarding, since, for example, they may enable the LHC to discover and measure various decay modes of the Higgs boson. Formulating new interactions between dark matter and the Higgs boson may lead to a new understanding for why the universe has more particles than anti-particles. 788

Page 789

Physics Postdoctoral Research and Development Continuing Project Boosting New Physics Discoveries with Jet Substructure Christopher Lee 20130794PRD2 Introduction Benefit to National Security Missions The Standard Model (SM) of particle physics is the most Two key missions of the DOE Office of Science in High- precise and successful theory in the history of science, Energy and Nuclear Physics are 1) to improve our but it still cannot explain dark matter or the huge hier- understanding of Quantum Chromodynamics (QCD), the archy of energy scales (from nuclear forces to gravity) theory of the strong interactions between quarks and found in nature. Much excitement today comes from gluons, constituents of all ordinary matter; and 2) to predictions of new models of elementary particles that search for new particles and forces beyond the Standard could help solve these mysteries and be detected at the Model that will explain the origin of matter, dark matter, Large Hadron Collider (LHC). and masses of elementary particles. The LHC collides protons at high energies and records Understanding properties of jets of hadrons produced by the resulting particles that are produced. The most com- energetic quarks and gluons are key to achieving these mon type of SM final state contains collimated streams objectives. They reveal the behavior of QCD itself and of hadrons called jets, produced by radiation from high- contain evidence of new particles that decay to jets. energy quarks or gluons. On occasion, new, non-SM par- This project tackles problems at the forefront of QCD ticles may be produced and also decay into jets. Finding perturbation theory and strategies to search for signa- them is a very difficult needle-in-the-haystack problem tures of new physics in the substructure of jets. Physi- to distinguish jets produced by new particles from the cally accurate predictions of jet cross sections require vastly larger number of jets produced by ordinary SM resumming arbitrarily many soft and collinear quark and processes. This makes high-precision predictions of the gluon emissions in and from jets. We will develop and SM jet background essential. Unfortunately, jet cross apply tools for resummation to achieve unprecedented sections are among the least precisely understood due precision in predicting jet cross sections for probing QCD to the often prohibitively difficult nature of the calcula- and new physics at the Large Hadron Collider (LHC). In tions. concert we will invent and develop new strategies for This research aims to vastly increase the precision of finding and characterizing jet substructure and interpret- theoretical jet cross section predictions. We will develop ing those characteristics as signatures for new physics. strategies to characterize and probe the substructure of Historically, individuals hired for their expertise in this jets, which can serve as a discriminant between signal area have gone on to address a range of Laboratory chal- and background jets. We will apply our expertise in lenges, including bio and energy security, thanks to their theoretical methods to resum the effects of infinitely broad analytical skills. many quarks and gluons radiated from jets to predict jet cross sections to high precision and accuracy. We will Progress develop new methods to predict cross sections depen- dent on subjet measures. We will thereby improve the This project started September 2013; therefore, there is theoretical input into Monte Carlo event generators that no progress to date. simulate large numbers of particles in high-energy colli- Future Work sions. All of these tools together will greatly advance our ability to find evidence of new physics at the LHC inside In the first year of this project, we will: hadronic jets. 789

Page 790

• Evaluate previously proposed measures of jet substruc- ture for their amenability to precise theoretical calcu- lation, and if necessary improve them or invent new measures that are better suited to precise predictions. • Develop theoretical tools to predict subjet cross sec- tions, by extending the tools of soft collinear effective theory (SCET) to handle multiple jet directions con- tained within a single, larger jet. • Apply the new theoretical tools to predict signal and background cross sections for searches for heavy par- ticles such as top quarks, Higgs bosons, or beyond-the- Standard-Model particles that decay to hadronic jets. Conclusion In this new, exciting era in particle physics, this project will play a crucial role in guiding what observables to measure at the Large Hadron Collider and how to interpret such measurements for signals of new physics. We focus on probing the substructure of jets of hadrons to which new physics particles decay and which can be used to distin- guish jets produced by known and by new physics. We will make simultaneous advances in search algorithms, theo- retical predictions, and computational programs for jet cross sections that are all part of a comprehensive strategy for finding new physics using hadronic jets. 790

Page 791

Physics Postdoctoral Research and Development Final Report Physics of Cosmic Ray Shocks and the High Energy Universe Hui Li 20100611PRD2 Abstract galactic dynamo or star formation could only be fully un- Astrophysical plasmas are magnetized and well-de- derstood if cosmic rays are included in the dynamics the scribed on large scales by Magnetohydrodynamics or interstellar medium (ISM) and molecular clouds. Another MHD. Their dynamics are affected by turbulence, which classic example of the importance of the mutual interac- is observed in the Sun, solar and stellar winds, accretion tion between cosmic rays and MHD fluid is the problem disks and spiral galaxies, violent outflows from active ga- of cosmic ray acceleration in supernova remnants, i.e. lactic nuclei, supernova explosions and so on. Another the problem of the origin of the Galactic cosmic rays basic component of magnetized astrophysical plasmas themselves. We know that cosmic rays are present in su- is so-called cosmic rays (CRs), charged relativistic par- pernova remnants in large amounts but we are not sure ticles, mostly protons, with energies from 10^8 to 10^21 why their acceleration is efficient. It has become clear eV. The mysterious property of CRs is that they have that ambient scattering of cosmic rays is grossly insuf- an energy density comparable to that of kinetic mo- ficient to provide sizable acceleration; to get around this, tions and magnetic fields, e.g., around 1 eV/cm^3 in our it was proposed that astrophysical shocks are strongly Galaxy, which is likely related to the interaction between modified by cosmic rays and become turbulent. The CRs and MHD plasmas, the process which is still poorly reason why this problem has been around for more than understood. The physics of CR acceleration in shocks, thirty years is that in order to be efficient, diffusive shock where CRs have a pressure comparable to gas pressure acceleration must involve interaction of cosmic rays and strongly modify the shocks, is dominated by the streaming in front of the shock with the undisturbed ISM mutual interaction of CRs and turbulence. plasma, creating a precursor turbulence, which in turn reduces the cosmic ray mean free path, making accelera- Background and Research Objectives tion more efficient. Without this back-reaction the ac- Magnetohydrodynamic turbulence is a complex nonlin- celeration in supernova remnants should be negligible, ear problem. It is multidisciplinary in nature, as highly contrary to observations. Despite this problem having conductive plasmas are found not only in astrophysical attracted a lot of attention, most existing approaches to objects, but in our own solar system, in the solar wind, it are as yet heuristic. Full understanding is complicated and also in laboratory plasma experiments. MHD turbu- by the strongly nonlinear supersonic dynamics of the lence is ubiquitous in these environments. Turbulence precursor MHD turbulence interacting with collisionless must be considered to understand large-scale flows. This particles via magnetic field. was realized a long time ago, making turbulence a major A number of widely different approaches have been theme in hydrodynamics for the past century. MHD used to tackle the cosmic ray acceleration problem. For turbulence and its applications for astrophysics are much example, PIC simulations are the most first-principle- younger, but even now it is clear that there is a wealth of based and used to solve particle dynamics in self-consis- phenomena to study. The next degree of complexity in tently evolved electromagnetic fields. Unfortunately, PIC astrophysical plasmas arises when one considers cosmic simulations in three dimensions will always be limited in rays – highly energetic particles, which interact with the range of scales they cover. Only a few hundred, or at MHD fluid and which are dynamically important in many best, a few thousand plasma skin depths can be reached astrophysical environments, such as galactic disks and by three-dimensional PIC. Although this is tiny compared supernova remnants. MHD turbulence and its interac- to astrophysical scales, some ideas in diffusive shock tion with cosmic rays is an interesting problem, but hard. acceleration (DSA) can be checked by PIC. Monte Carlo Cosmic rays are dynamically important: processes as 791

Page 792

simulations assume certain diffusive properties and can, deemed universal by some researchers, actually saturates indeed, cover a wide range in cosmic ray energies but, un- at high Reynolds numbers. Our claim on alignment and fortunately, require a priori knowledge of particle scatter- spectral slope was based on a rigorous numerical argu- ing. Important advances in the particle acceleration theory ment called resolution study, in contrast with previous were made by realizing that the precursor magnetic fields claims based on subjective criteria. We also measured the have to be significantly amplified. In this respect impor- Kolmogorov constant of MHD turbulence and anisotropy tant progress has been made in the studies of cosmic ray constant of MHD turbulence. This was the first and till now current instability. Another mechanism is the small-scale the only conclusion of this kind. Spectral and anisotropy dynamo excited by the baroclinic term in strongly modified properties that we obtained have applications in all areas shocks. Although we would like to know the outcome of of astrophysics, plasma physics and solar and heliospheric the full kinetic streaming instability in the precursor, this is physics. In our view, our most important achievement to- hard due to the tangled nature of the magnetic field gener- date is a model for so-called imbalanced MHD turbulence. ated by small-scale dynamo. While the Goldreich-Sridhar model has become a standard description of MHD turbulence, it only considered purely Scientific Approach and Accomplishments balanced (zero cross-helicity) case, which is rarely seen in MHD turbulence is essentially nonlinear and cannot be nature. Imbalanced turbulence appears whenever there treated perturbatively. In this situation it is especially is a strong source of perturbations such as the Sun in the important to use direct numerical simulations as a guide solar wind or the central engine in AGN. Also, the stochas- to theoretical understanding. We used numerics to un- tic nature of turbulence requires knowledge of the general derstand basic physical processes, rather than trying to imbalanced case to fully understand globally balanced simulate a real astrophysical object and obtain a final turbulence. Our model has been successful in describ- answer. We studied supersonic, weakly compressible and ing the anisotropy difference observed in the imbalanced incompressible MHD turbulence and developed novel case. Driven by the interest in ISM turbulence, which is methods of data analysis, and in so doing executed some both strongly magnetized and supersonic, we and our of the highest resolution numerical experiments ever collaborators discovered anisotropy in density perturba- performed. In 2013, we were awarded DOE INCITE alloca- tions. It turns out that despite density perturbations being tion of 35 million CPU hours to study MHD turbulence. Our initially created by random shocks, the structure of these versatile MHD code that solves MHD, hydro and reduced perturbations, which are revealed by second-order struc- MHD equations are currently the fastest available. Not only ture function of log-density, is primarily created by Alfvenic do we run MHD simulations, but also as a step towards shearing and has a scale-dependent anisotropy similar to understanding MHD-particle interaction we developed a Goldreich-Sridhar anisotropy. As a first step to the cosmic particle tracing code that was used for calculating scatter- ray acceleration problem, we proposed a model of diffu- ing and diffusion of fast particles in magnetic fields, gener- sive shock acceleration (DSA) where the magnetic field in ated in MHD simulations. Later, we started using PIC and the precursor is amplified due to turbulence excited by the relativistic MHD codes. Careful application of numerics to baroclinic term. It turns out that this mechanism is rather theory, and, likewise, clear formulation of the theoretical efficient in generating large-scale magnetic fields. We also problem to resolve it with numerics, have resulted in sev- studied cosmic ray scattering in MHD turbulence by both eral key findings in the MHD theory and cosmic ray scat- numerical and analytical means. We showed that for low tering outlined below. We have proposed that the process energy cosmic rays existing in turbulent MHD environment of magnetic energy generation in turbulent conductive particle instabilities play a key role, dominating scattering, medium is universal in large Reynolds number flows, and as long as compressive MHD modes are present. More supported this by analytical arguments and high resolu- recently, we used numerical simulations with particle tion numerical simulations. We measured the efficiency of tracing to directly evaluate CR scattering frequencies and small-scale dynamo, which is the fraction of the supplied diffusion tensor in MHD turbulence. An important recent energy that goes to magnetic energy. Our study addressed development is the discovery of asymmetric diffusion of a long-standing puzzle: why almost all media in astrophysi- magnetic field lines. Stochasticity of magnetic field lines is cal environments are magnetized, even in situations where important for particle transport properties. Magnetic field turbulence exists for only a few dynamical times or where lines separate faster than diffusively in turbulent plasma, the large-scale dynamo is absent. We were able to resolve which is called superdiffusion. We discovered that this a long-standing debate between a -5/3 and -3/2 spectral superdiffusion is pronouncedly asymmetric, so that the slope of MHD turbulence. By running the highest resolu- separation of field lines along the magnetic field direction tion MHD simulations ever performed, we found that the is different from the separation in the opposite direction. process of so-called scale dependent dynamic alignment, While the symmetry of the flow is broken by the so-called 792

Page 793

imbalance or cross-helicity, the difference between for- mately better predict space weather. ward and backward diffusion is not directly due to imbal- ance, but a non-trivial consequence of both imbalance and Publications non-reversibility of turbulence. The asymmetric diffusion Beresnyak, A.. Spectral Slope and Kolmogorov Constant of perpendicular to the mean magnetic field entails a variety MHD Turbulence. 2011. Physical Review Letter. 106: of new physical phenomena, such as the production of par- 5001. allel particle streaming in the presence of perpendicular Beresnyak, A.. Universal Nonlinear Small-Scale Dynamo. particle gradients. Such streaming and associated instabili- 2012. Physical Review Letters. 108 (3): 035002 (4 pp.). ties could be significant for particle transport in laboratory, space, and astrophysical plasmas. We were also interested Beresnyak, A.. Asymmetric Diffusion of Magnetic Field in reconnection and associated release of energy. Magnetic Lines. 2013. Astrophysical Journal Letter. 767: 39. reconnection is a topological rearrangement of the mag- Beresnyak, A.. On the Rate of Spontaneous Magnetic Re- netic field lines, leading to the release of magnetic energy, connection. Nature. which is associated with solar X-ray flares. Magnetic field lines are supposed to be frozen into the well-conducting Beresnyak, A.. Basic Properties of MHD Turbulence in the plasma; so quiet laminar reconnection is extremely slow Inertial Range . 2012. MNRAS. 422 (4): 3495. and cannot explain observed phenomena. This prompted research into collisionless reconnection. The stochasticity Beresnyak, A., H. Xu, H. Li, and R. Schlickeiser. MHD Turbu- lence and cosmic ray reacceleration in Galaxy Cluster. of magnetic field lines due to ambient turbulence leads to To appear in Astrophysical Journal. fast reconnection, also the tearing instability of the thin current sheet was proposed as a driver of resistivity-inde- Beresnyak, A., H. Yan, and A. Lazarian. Numerical Study of pendent reconnection, which was shown to be consistent Cosmic Ray Diffusion in Magnetohydrodynamic Turbu- with two-dimensional simulations. By doing three-dimen- lence. 2011. Astrophysical Journal. 728: 60. sional high-resolution simulations of nonlinear evolution of the thin current sheet we found that it spontaneously Beresnyak, A., and A. Lazarian. Scaling laws and diffuse locality of balanced and imbalanced magnetohydrody- evolves into a turbulent current layer and shows a constant namic turbulence. 2010. Astrophysical Journal Letters. reconnection rate of 0.015 of Alfven speed and the dissipa- 722 (1): L110. tion rate per unit area which is independent of resistivity. This demonstrates that nearly ideally conductive fluids can be fairly resistive in presence of magnetic discontinuities. The significant fraction, 40%, of magnetic energy is dissi- pated inside of the current layer, which, at the early stages of reconnection process, can result in huge dissipation per unit volume and explain electron acceleration and X-ray flares. Impact on National Missions One of the missions of NNSA is to achieve inertial confine- ment fusion (ICF). This requires not only technical work, but also theoretical understanding of fundamentals of plasma physics. Recently, several authors have brought up the spontaneous generation of magnetic fields in ICF and its importance for plasma and fluid instabilities in the tar- get. There have been several proposals to impose external magnetic fields to the target to study its influence on the instabilities and confinement. Our work on universal non- linear small-scale dynamo places a lower limit on magnetic energy growth in well-conducting plasmas, and is, there- fore, important for ICF. Another challenge on a national level is predicting space weather. My work on MHD turbulence, applied to the solar wind is helping to better understand heliosphere and ulti- 793

Page 794

Physics Postdoctoral Research and Development Final Report Non-Condon and State Interaction Effects in Carbon Nanotubes Stephen K. Doorn 20100629PRD3 Abstract Scientific Approach and Accomplishments This project focused on probing electron-phonon 1) Studies of non-Condon behavior: Earlier work at coupling behaviors in carbon nanotubes by employ- LANL demonstrated that the non-Condon behaviors ing resonance Raman spectroscopy of pure chirality manifest themselves as a strong asymmetry in plots (ex- samples. The results have overturned a number of long citation profiles) of how Raman intensity changes with held assumptions in nanotube science and have gen- excitation energy for a specific vibrational mode. One of erated several important first-time observations. The the project goals was to determine how this asymmetry results appear or will appear in a number of high-profile depends on vibrational mode and on specific optical publications. Briefly, we have demonstrated for the first transition that is probed. time quantum interference effects in the nanotube Ra- Asymmetry has already been demonstrated for high fre- man response. We have also shown the breakdown of quency (so-called G-modes near 1600 cm-1in frequency) the Condon approximation in metallic and semiconduct- modes [1]. Its presence for the important low-frequency ing nanotubes. We have for the first time also probed (near 200-300 cm-1) radial breathing mode (RBM) had in detail the double resonance Raman process of the not yet been established. To resolve the asymmetry for 2D mode and shown novel behaviors of the LO and TO the RBM required optical excitation into the lowest-en- phonon modes in both metallic and semiconducting ergy nanotube transitions (E11). The result showed the nanotubes. RBM displays no non-Condon effect (Figure 1a). Impact Background and Research Objectives of excitation into different Eii was probed using the higher frequency G-modes. Excitation into the lowest Our goal is to understand the interaction of carbon energy E11 transitions were found to give smaller asym- nanotubes and light. Recent results [1] from the car- metries than displayed for excitation into the next higher bon nanotube effort at LANL suggest that molecular energy E22 transition (see Figures 2 and 3). Higher vibrations and electronic states are connected. More energy excitations (E33 and E44) were also probed, with technically, the transition moment dipole, which deter- similar results found as for E22. The results of these Eii mines strength of electronic excitations, depends on the and vibrational mode dependence studies are currently vibrational coordinate, indicating a breakdown in the being written up for submission as an article to Physical fundamental Condon approximation for understanding Review B. spectroscopic quantum behavior. The approximation states that electronic excitations occur on a fast times- Results on the semiconducting structures described cale relative to vibrational motion. This project aims to above suggested that specific symmetry tubes (called probe this violation of the Condon approximation using near-armchair structures) will have more pronounced Raman spectroscopy. The goal is to probe the structural non-Condon effects. To test this, excitation profiles and energy dependence of this behavior. To do this were also obtained on (6,6) and (7,7) species of nano- requires access to high purity single-structure samples, tubes—true armchair structures. The result showed which are only available at LANL through a collabora- an increased asymmetry compared to other structures tion with colleagues at NIST [2]. Measurements on such studied to date. The armchair results are currently being samples are also allowing us to probe quantum interfer- written up as a paper to be submitted to Physical Review ence phenomena and fundamental electron-phonon Letters. The results are translating into follow on theory coupling behaviors in carbon nanotubes, all of which work aimed at understanding the symmetry impacts on ultimately relate back to optical behaviors of interest for the non-Condon behavior. applications. 794

Page 795

2. Studies of Quantum Interference Behavior: We have will settle existing debate on how dopants interact with successfully demonstrated for the first time unambiguous metallic nanotube electronic structure. Specifically, the signatures of quantum interference in the Raman response Kohn anomaly and its effect on the LO phonon in carbon of carbon nanotubes. The result also led to an understand- nanotubes is a well known and studied phenomenon in ing of the relative behavior of the signs of matrix elements nanotube research. The Kohn anomaly induces a strong defining how electrons and phonons interact in nanotubes, electron-phonon coupling of the LO phonon to electrons at which underlie a number of other important optical behav- the Fermi-level which leads to a broadening and down shift iors. Non-Condon effects were demonstrated to also occur of the phonon frequency. At the same time the TO pho- for high energy transitions and interference effects were non is predicted to up-shift. It is possible to “switch off” demonstrated to be a sensitive probe of energy separation the Kohn anomaly by changing the Fermi energy, which of electronic states that are near each other in energy. As can be achieved by gating or doping. The top of Figure 4 one specific example, in the (10,5) structure the G-plus and shows the G-mode (LO and TO phonons) Raman spectra of G-minus phonon intensities were found to reverse (Figure armchair carbon nanotubes [(8,8), (7,7), and (6,6)], which
3. from what is normally. This behavior is a direct conse- we electro-chemically doped, from -1.2 to 1.2 V (bottom quence of strong destructive interference occurring for the to top). The bottom spectra nicely show the down shifted G-plus mode while the G-minus experiences strong con- and broadened features, which are related to the LO-pho- structive interference. This behavior only occurs as a result non. The two narrow features at ~1580 and ~1590 cm^-1 of the E33 and E44 interfering transitions having zero are related to the TO phonon. As predicted by theory they energy separation. The results were published in Physical have higher frequencies compared to semiconducting Review Letters, 108, 117404, (2012). nanotubes, which are not effected by the Kohn anomaly. When we dope the sample by increasing the potential we
4. Analysis of the 2D Mode in Pure Chirality Samples: are able to “switch off” the Kohn anomaly, as is apparent Probing of the 2D (near 2600 cm-1) vibrational mode from the up-shift and narrowing of the LO-phonon related behavior is only possible in pure chirality samples and is peaks. However, the peaks which are related to the TO important for probing phonon structure in tubes and how phonons are not showing any significant shift (bottom it couples to various momentum regions of the electronic plot in Figure 4). If the unusually high frequencies of the structure. Analysis of the 2D data requires modeling of TO-phonon would be related to the Kohn anomaly the electronic and phonon densities of states and how they “switch off” should result in a down-shift of up to 60 cm-1. interact under particular optical excitations. The software Therefore we conclude that the increase of the TO-phonon to do this analysis has been completed and evaluated frequencies is related to a screening of the phonon by free with several test case systems. The program has also now carriers in metallic tubes rather than the Kohn-anomaly. been applied to particular nanotube structures for which These results overturn a long-held assumption in nanotube we have good experimental data. The 2D phonon band electron-phonon coupling behavior and are currently being is related to a double resonant Raman process. The two written up for submission to Physical Review Letters. phonons involved in the process scatter an excited electron in a way that the intermediate states are real, which leads 5) Chirality dependence of the LO and TO phonons in to an enhancement of the Raman signal. The exciting as- Semiconducting Carbon Nanotubes: Raman spectroscopy pect about this process is that the involved phonons have of the G-band frequencies in pure chirality semiconducting non-zero momentum and that the momentum varies with nanotube samples has allowed us to accurately determine the initial electronic state thus with the excitation energy. the structural dependence of both the G-plus and G-minus This mode is therefore suitable to experimentally map the (LO and TO) phonon modes. The result opens up the TO phonon dispersion of the particular carbon based macro mode as a new route to structural identification of tubes. molecule. As an example of our modeling capability, iIn Additionally, the results allow us to determine the origins Figure 3 we simulated the 2D mode Raman spectrum as a of structural dependences based on rehybridization, con- function of excitation energy for a (9,1) nanotube. Initial finement, and electron-electron interaction effects. The experimental results show a good agreement with this work is published in ACS Nano, 6, 904 (2012). simulation. The results are currently being written up for submission to Physical Review B. Impact on National Missions Results of this project will have a direct bearing on devel-
5. Electronic doping of “armchair” metallic nanotubes: oping optical and electronic properties of carbon nano- Electrostatic doping of films of armchair metallic tubes tubes for photonic and energy harvesting applications and have been carried out with an eye towards evaluating may also contribute to sensing and spectral tagging appli- the doping effect on phonon frequencies. The result cations. As a result this work will have direct relevance to 795

Page 796

the mission of the DOE-BES funded Center for Integrated Nanotechnologies and the potential applications will be of interest to agencies including NIH, DOE, DHS, and DOD with potential impact on threat reduction and renewable energy missions. a Figure 2. Resonance Raman G-plus (green circles) and G-minus (orange circles) band resonance Raman excitation profiles for the (10,5) tube in E33/E44 excitation. b c Figure 3. Plot of calculated 2D double resonance Raman intensity as a function of excitation energy and phonon frequency as mod- eled for the (9,1) tube. Figure 1. a) RBM resonance Raman excitation profile for (9,1) tube in E11 excitation. b) G-band resonance Raman excitation profile for (9,1) tube in E11 excitation. c) G-band resonance Ra- man excitation profile for (9,1) tube in E22 excitation. 796

Page 797

2. Tu, X., S. Manohar, A. Jagota, and M. Zheng. DNA sequence motifs for structure-specific recognition and separation of carbon nanotubes. 2009. Nature. 460:

Publications Duque, J. G., H. Telg, H. Chen, A. K. Swan, A. P. Shreve, X. Tu, M. Zheng, and S. K. Doorn. Quantum interference between the third and fourth excitonic states in semi- conducting carbon nanotubes. 2012. Physical Review Letters. 108: 117404. Telg, H., J. G. Duque, M. Steiger, X. Tu, F. Hennrich, M. M. Kappes, M. Zheng, J. Maultzsch, C. Thomsen, and S. K. Doorn. Chiral index deendence of the G+ and G- Ra- man modes in semiconducting carbon nanotubes. 2012. ACS Nano. 6 (1): 904. Figure 4. Top Panel: Plot of LO and TO Raman spectrum for an armchair-enriched sample of carbon nanotubes as doping potential is varied from +1.2 V (top spectrum) to -1.2 V (bottom spectrum. Bottom panel: Plot of LO and TO frequencies as dop- ing potential is changed. Figure 5. Left Panel: Plot of LO and TO Raman spectra for a range of nanotube chiralities as labeled. Right Panel: Plot of TO frequency as a function of nanotube transition energy and tube diameter for E11 through E44 excitations and demonstrat- ing the ability to now use the TO band as a means for identifying nanotube structure. References

1. Duque, J. G., H. Chen, A. K. Swan, A. P. Shreve, S. Kilina, S. Tretiak, X. Tu, M. Zheng, and S. K. Doorn. Violation of the Condon approximation in semiconducting carbon nanotubes. 2011. ACS Nano. 5 (6): 5233. 797

Page 798

Physics Postdoctoral Research and Development Final Report A New Regime of Carrier Multiplication Using Intraband Re-excitation of Nanocrystals Victor I. Klimov 20100632PRD4 Abstract gies. For example, a single-junction PV with an ideal CM This project’s goal has been to answer the question: yield can produce a power conversion efficiency exceed- What are the dominant decay mechanisms that compete ing 40%, which is a considerable improvement over the with carrier multiplication in semiconductor nanocrys- traditional Shockley-Queisser limit of ~31%; the ideal tals? Carrier multiplication (CM) is a process whereby ab- CM yield is described by a staircase function in which sorption of a single photon results in multiple electron- each increment of the incident photon energy by the hole (e-h) pairs (excitons). This process could benefit a band-gap results in a new e-h pair, corresponding to an number of solar-energy conversion technologies, most increase of quantum efficiency (QE) by 100%. notably photocatalysis and photovoltaics (PVs). The work In bulk semiconductors, CM has been observed since the in this project was primarily motivated by an outstanding 1950’s and the underlying mechanism is typically attrib- challenge in this field, which has been the lack of capa- uted to impact ionization. Two important parameters of bility to quickly discern between candidate materials for this process are the activation threshold and the e-h pair enhanced CM in nanocrystals (NCs). By studying materi- creation energy, which is the energy lost by the ionizing als of a variety of compounds and shapes we have been particle in a single impact ionization event. In the ideal able to rationalize relative changes in CM yields of these case, energy conservation dictates that the minimal materials and ascribe these changes to known physical values of the threshold and e-h pair creation energy are processes. In fact, as our understanding has grown, we Eg and 2Eg, respectively. However, in bulk semiconduc- have developed an ever more sophisticated phenome- tors, because of the additional restrictions imposed by nological model which accurately describes the observed momentum conservation and fast intraband energy trends. This has resulted in the first ever prediction and losses due to phonon emission, both of these quantities realization of increased CM efficiency in a material, sug- are considerably higher than the ideal limit. In wide-gap gesting our approach will be useful for a more rational semiconductors, typical values are ~4Eg for the thresh- search of future materials with higher CM. old and ~3Eg for the e-h pair creation energy, becoming even greater for narrow-gap materials. As a result, the Background and Research Objectives benefits of CM in PVs based on traditional bulk solids are The realization of low-threshold carrier multiplication negligible; specifically, the CM-induced enhancement in (CM) in semiconductor nanocrystals (NCs) is a promising the power conversion efficiency expected for an opti- step toward realization of next generation solar cells – mized SiGe alloy is less than 1%. low cost photovoltaics (PVs) with efficiencies beyond the Shockley-Queisser limit. CM or multiexciton generation Quantum confined semiconductor particles or quantum (MEG) is a process whereby multiple electron-hole (e-h) dots (QDs), on the other hand, have been expected to pairs (excitons) are generated via absorption of a single increase CM yields due to their discrete structure of photon. In bulk semiconductors, CM occurs via impact electronic states. Firstly, a wide separation between NC ionization where a valence band electron is promoted to discrete states has been expected to suppress phonon the conduction band via a collision with a high-energy emission due to a phonon bottleneck and thus favor charge carrier. In this traditional picture, the efficiency CM, as was first pointed out by Nozik in 2002. Secondly, of CM is set by the competition between impact ioniza- three-dimensional (3D) spatial confinement leads to tion and phonon emission. Understanding what sets the relaxation of translational momentum conservation, limits of CM efficiency is an important challenge due to which should reduce the threshold for CM as well as the its direct impact on solar-energy conversion technolo- 798

Page 799

e-h pair creation energy. These considerations motivated a us to come to this formulation came from considering the renewed interest in CM with a focus on quantum-confined decay processes available to a photogenerated exciton as nanomaterials. an effective time window in which CM can occur. Then, ex- pressing the CM yield for excitation energies just above the The first experimental observation of enhanced CM in threshold energy for CM to occur, we were able to express quantum dots as compared to bulk material was reported the CM yield directly as the product of the two timescales for PbSe NCs by the Klimov group in 2004, where CM above [1]. The demonstration that this is more generally was detected using the fast Auger decay signatures of true requires significantly more calculation but neverthe- multiexcitons in transient absorption (TA). Following this less arrives at exactly the same formulation, as reported in initial report, CM in NCs has become a subject of intense [2]. What is most significant, however, is the experimental experimental and theoretical investigations. By now, proof which validates this approach. measurements show excellent agreement between data sets obtained by different groups and/or different spec- Naively, based on the approach outlined above, one needs troscopic techniques. Furthermore, the results of spectro- only to measure the CM timescale and cooling rate (kcool) scopic studies of CM have recently been validated by direct of the competing process to predict the CM performance. photocurrent measurements in photovoltaic devices, Unfortunately, while kcool can be at least inferred from demonstrating that CM can be useful in real devices. analyzing intraband relaxation below the threshold for CM, direct measurements of the CM timescale are not The outstanding challenge, however, has been the lack of so straightforward and have not yet been carried out. To an established theory which would allow for direction on overcome this problem, we infer the CM timescale from how to further increase CM efficiencies. This research has measurements of the inverse process – biexciton Auger re- successfully addressed this lack of direction in the form of combination. During this process, one exciton recombines well thought out experiments which delineate between by transferring the recombination energy to an electron or the potential physical properties that set CM yields. a hole of the second exciton, producing a hot exciton. This is precisely the inverse of CM, where a hot exciton relaxes Scientific Approach and Accomplishments by creating a new e-h pair. Thus, CM and Auger decay are To address the decay mechanisms available to a photo- described by the same interaction Hamiltonian and are generated charge and how these mechanisms set the CM therefore interconnected. yield, we have studied the influence that size, shape and composition have on the CM process in semiconductor Another important simplifying feature of CM rates is the nanocrystals. Using measured biexciton Auger lifetimes observation of universal scaling of Auger lifetimes with the and intraband relaxation rates, we have rationalized rela- NC volume, termed volume-scaling. Volume-scaling was tive changes in CM yields as a function of composition. originally discovered in studies of CdSe NCs, and subse- Indeed, by studying PbS, PbSe, and PbTe NCs for a variety quently the universality of this feature was established for of sizes we determine that the significant difference in the many other compositions, including the lead chalcogen- CM yields in these compounds comes from the dissimilari- ides studied here. While the exact physical mechanism ties on their non-CM relaxation channels, i.e., the process- underlying these universal trends is not fully understood, es that compete with CM. We further explored the role it likely relates to the relaxation of momentum conserva- of nanostructure shape in the CM process. We observe in tion, which diminishes the role of the exact band structure PbSe that via a moderate NC elongation (aspect ratio of in Auger recombination. The consequence is that universal ~7) we can obtain a 50-to-80% increase in the multiexciton volume-scaling of Auger lifetimes suggests that CM rates in yield compared to spherical nanoparticles. Together, these NCs are also likely to be dominated by NC size and not by studies demonstrate that CM can be understood within composition. This would further suggest that the observed the framework of competing mechanisms and provide a differences in e-h pair creation energy and CM yield for prescription for how to increase CM yields. similarly sized NCs of different compositions is mostly due to the differences in intraband cooling rates. To test the va- The important result of this project has been the formula- lidity of these conclusions we have conducted side-by-side tion and then experimental proof that the e-h pair creation studies of CM yields, intraband cooling, and Auger rates, in energy can be represented in terms of two timescales: the NCs of our three different lead chalcogenides: PbSe, PbS, timescale which characterizes CM; and, the competing and PbTe. cooling rate, which we assume to be dominated by phonon emission. As we will show below, this resulted in the first In Figure 1 we show all of the data required to make a ever predictive prescription on how to choose a material qualitative prediction of the relative CM yields for these with increased CM yields. The original insight that allowed three materials based on Auger timescales and phonon 799

Page 800

cooling rates. In Figure 1a, we show biexciton Auger and increases approximately linearly with the NR volume lifetimes as a function of confinement energy (Ec): Ec = (Figure 3a). On the other hand, CM measurements shown Eg - Eg0, where Eg0 is the bulk energy gap. Consistent with in Figure 3b (excitation at 3.1 eV), show that NRs are either volume scaling, NCs of all three compositions show a simi- similar or more efficient than quantum dots. We see, how- lar dependence of Auger lifetime on Ec and similar density ever, that for a given excitation energy, the NR data show a of states further imply similar CM time constants. significant spread in CM values indicating that some other parameter(s) play an important role in determining the In Figure 1b, we show a representative data set which overall CM yield. A significant discovery that we made [3] we use to extract the intraband cooling rate near the is that CM performance depends strongly on the degree of band edge in PbSe NCs. Specifically, we measure the time elongation, i.e. the aspect ratio (L/d), as shown in Figure required to populate the ground state after excitation. 3c. These data show a progressive increase in the multi- We calculate the energy loss rate from the ratio of the exciton yield NRs up to an aspect ratio of ~5-7 and then a 1P-1S energy spacing to the time required to populate the decrease for larger aspect ratios. Repeating this measure- 1S state. Conducting these measurements for all three ment for various diameters indicates that this aspect-ratio materials we find the energy loss rates are related by dependence is band-gap independent. The universality of (PbTe):(PbSe):(PbS) ≈ 1:2:4. Then, we assume that a similar this “optimal aspect ratio” may greatly help ongoing theory ratio between cooling rates also holds at energies relevant work to understand the underlying physics of CM in nano- to CM. Given this comparison of relaxation rates and the crystals. Importantly for this work, however, is the clear similarity in CM time constants expected for these three demonstration that CM rates play just as vital a role in the compounds, we would expect CM yields to increase going CM process as competing mechanisms such as phonon from PbS to PbSe and then to PbTe. emission. The test of our prediction is shown in Figure 2 for which we The content of this report has been borrowed heavily from see that we clearly confirm the predicted trend. CM yields the project publications on these subjects. The references in PbSe NCs are systematically higher than in PbS NCs and include [1-5]. then further increased in PbTe NCs. For example, for at excitation of 4.3Eg, the CM yield is 15%, 35%, and 65% in Impact on National Missions PbS, PbSe, and PbTe NCs, respectively. Interestingly, the ra- This work has directly addressed many FY13 grand chal- tio between these yields (1:2.3:4.3) is close to the relative lenges related to energy generation; these include: “Ma- ratios of the cooling rates. This effort represents the first terials Discovery Science to Strategic Applications” and ever confirmation of a predicted increase in CM yields as “Energy and Earth Systems.” This work has provided vital a function of nanocrystal composition and should provide direction toward the practical use of carrier multiplication for a more rational search of NC materials with even higher to enhance photon-to-charge conversion in real devices. CM efficiencies. For an example, the results in [1] directly influenced the decision to use PbSe as the nanocrystal material composi- One potential critique of the above approach is that the tion in a recent paper published in Science Magazine which CM timescales were essentially the same for all three for the first time observed external quantum efficiency compounds. One could argue that this approach only es- greater than unity within the solar spectrum. Finally, the tablishes the importance of competing decay mechanisms results here provided first ever confirmed prediction of en- such as phonon emission and in fact does not establish the hanced carrier multiplication in a material (PbTe) and have importance of the CM time constant in setting the CM effi- provided a prescription for how to increase CM efficiency ciency. To address this issue, we designed and completed a in future materials. series of experiments on the same compound but of differ- ent shapes [3, 4]. Because all shapes are composed of the same material we ensure that the competing mechanism (phonon cooling) is the same for all samples; this allows us to attribute any changes in the Auger rate and CM yield to changes in the carrier-carrier interaction responsible for these two processes. We examine CM yields and Auger recombination in PbSe for both spherical NCs (quantum dots) and NRs. The measurements indicate that the Auger time constant tends to be longer in NRs than in spherical nanoparticles 800

Page 801

malized photon energy; measurements were made using a fixed excitation energy (3.1 eV) and differently sized NCs. Despite their similar Auger time constants (see Figure 1a), these materials show strikingly different CM yields, indicative of different e-h pair creation energies. These trends are consistent with the trends in competing cooling rates (kcoolPbS>kcoolPbSe>kcoolPbTe) Figure 1. Quantitative studies of Auger lifetimes and 1P-1S energy loss rate used as “surrogates” for CM time constants and kcool, respectively. (a) Biexciton Auger lifetimes for PbS, PbSe, and PbTe NCs plotted as a function of confinement energy dem- onstrate a universal trend seen previously for NCs of a variety of compositions. (b) The buildup of 1S bleaching for PbSe NCs mea- sured by TA is used to estimate the intraband cooling rate. The red data points represent the instantaneous Coulomb shift due to hot carriers; it is followed by slower growth due to population transfer from the 1P to the 1S state. The 1S-1P energy separa- tion is derived from the absorption spectrum (inset). Figure 3. Auger time constants and carrier multiplication (CM) quantum efficiency for PbSe nanorods (NRs). (a) NRs (red circles) appear to demonstrate similar, although systematically higher, size dependence of Auger time constants as spherical particles (black squares). (b) Quantum efficiency of CM measured for vari- ous diameters and aspect ratios plotted as function of normal- Figure 2. Quantum efficiencies (QEs) for PbTe (blue triangles), ized photon energy; the measurements were conducted at 3.1 eV PbSe (black circles) and PbS (red squares) as a function of nor- excitation (red and green symbols) and 4.65 eV excitation (blue 801

Page 802

triangles). The NR results are compared to our measurements Pandey, J. M. Pietryga, and V. I. Klimov. Carrier multipli- on PbSe quantum dots (QDs) (black solid squares), sometimes cation in PbTe, PbSe, and PbS quantum dots within the showing enhancement but not always. (c) Upon plotting the framework of two competing energy relaxation mecha- enhancement factor of NRs to QDs we see a clear aspect ratio nisms. Nature Physics. dependence that does not depend on the initial band gap. This feature hints at universal shape dependence of CM in NRs. Stewart, J., L. Padilha, M. Mumtaz. Qazilbash, J. Pietryga, A. Midgett, J. Luther, M. Beard, A. Nozik, and V. Klimov. References Comparison of Carrier Multiplication Yields in PbS and

1. Stewart, J., L. Padilha, M. Mumtaz. Qazilbash, J. Pi- PbSe Nanocrystals: The Role of Competing Energy-Loss etryga, A. Midgett, J. Luther, M. Beard, A. Nozik, and Processes. 2012. NANO LETTERS. 12 (2): 622. V. Klimov. Comparison of Carrier Multiplication Yields in PbS and PbSe Nanocrystals: The Role of Competing Energy-Loss Processes. 2012. NANO LETTERS. 12 (2):
3. Stewart, J. T., L. A. Padilha, W. K. Bae, W. K. Koh, L. Li, A. Pandey, J. M. Pietryga, and V. I. Klimov. Carrier multipli- cation in PbTe, PbSe, and PbS quantum dots within the framework of two competing energy relaxation mecha- nisms. Nature Physics.
4. Padilha, L. A., J. T. Stewart, R. L. Sandberg, W. K. Bae, W. K. Koh, J. M. Pietryga, and V. I. Klimov. Aspect ratio dependence of Auger recombination and carrier multi- plication in PbSe nanorods. Nano Letters.
5. Padilha, L. A., J. T. Stewart, R. L. Sandberg, W. K. Bae, W. K. Koh, J. M. Pietryga, and V. I. Klimov. Carrier mul- tiplication in semiconductor nanocrystals: Influence of size, shape and composition. To appear in Accounts of Chemical Research.
6. Koh, W. K., A. Y. Koposov, J. T. Stewart, I. Robel, J. M. Pi- etryga, and V. I. Klimov. Heavily doped n-type PbSe and PbS nanocrystals using ground-state charge transfer from cobaltocene. Nature Materials. Publications Koh, W. K., A. Y. Koposov, J. T. Stewart, I. Robel, J. M. Pietryga, and V. I. Klimov. Heavily doped n-type PbSe and PbS nanocrystals using ground-state charge transfer from cobaltocene. Nature Materials. Padilha, L. A., J. T. Stewart, R. L. Sandberg, W. K. Bae, W. K. Koh, J. M. Pietryga, and V. I. Klimov. Aspect ratio depen- dence of Auger recombination and carrier multiplication in PbSe nanorods. Nano Letters. Padilha, L. A., J. T. Stewart, R. L. Sandberg, W. K. Bae, W. K. Koh, J. M. Pietryga, and V. I. Klimov. Carrier multiplication in semiconductor nanocrystals: Influence of size, shape and composition. To appear in Accounts of Chemical Re- search. Stewart, J. T., L. A. Padilha, W. K. Bae, W. K. Koh, L. Li, A. 802

Page 803

Physics Postdoctoral Research and Development Final Report Heavy quarks in Cold Nuclear Matter and in the Quark Gluon Plasma Ming X. Liu 20100635PRD4 Abstract as they normally are in nature. The effects on particle In this project, we study the properties of a new state production of this de-confined medium can reveal of matter that may have been absent from the Universe how this novel medium behaves. The two large experi- since a first few milliseconds after the Big Bang. To probe ments at RHIC have both measured electron decay of this hot dense state we use heavy quarks, with whimsi- D mesons containing a charm quark with mass around cal names like “charm” and “beauty”. These quarks are 1.8 GeV/c2 and of B mesons containing a beauty quark distinct from the up and down quarks of normal mat- with mass around 5.3 GeV/c2. The high transverse ter. This hot dense matter, called a Quark Gluon Plasma momentum electrons from these heavy-quark mesons (QGP), is created in high-energy heavy-ion collisions at are suppressed by the same amount as mesons contain- the Relativistic Heavy Ion Collider (RHIC) at Brookhaven ing light quarks when compared to their production on National Laboratory. A new silicon vertex detector free nucleons. This surprising observation suggests that (FVTX), developed under LANL leadership, identifies the effect of the medium on meson production is more heavy quarks. These quarks travel only a short distance complex than first thought. To gain a full understanding before decaying into muons. Although the distance is of the phenomena, more precise measurements of the short, it can be measured by this new FVTX detector, heavy quark mesons for a variety of different collision thereby detecting the ephemeral life of a heavy quark. systems and over a broad kinematic range are needed. Bound states of two heavy quarks are also detectable. Another interesting related effect is the dissociation of The quarks experience the effect of the QGP as they heavy quarkonia (pairs of charm or beauty quarks; J/ψ, travel through it on the way to our detectors. ψ’ and ϒs) because of color screening in the deconfined medium. This dissociation is supposed to happen at a The new FVTX detector was built and added to the exist- certain density or temperature of the medium for each ing PHENIX detector at RHIC at the end of 2011. The quarkonium state. The observation of heavy quarkonia new detector has been commissioned early this year such J/ψ, ψ’ and ϒ dissociation in heavy ion collisions, 2012 when it saw its first colliding beams. Collisions with along with lattice QCD estimates of their dissociation various types of high-energy beams including gold and temperatures, can reduce the current uncertainty in the cooper nuclei, and polarized protons were collected in temperature achieved in the hot-dense medium. 2012 and delivered the data required to obtain a com- plete picture of the heavy quark physics; and to extract Initial-state effects also influence the production of these properties of the QGP. Analysis for physics results is in heavy-quark objects, such as the modification of the ini- progress now and preliminary results show that we can tial parton distribution in nuclei compared to those in a clearly identify muon tracks with the new FVTX detec- free nucleon. These effects are not well understood and tors. Our final data analysis will deliver scientific results are largest in the forward direction. that are not possible within the baseline project. The goal of this project was to identify the heavy quark (charm or bottom) and study the quark-mass and mo- Background and Research Objectives mentum dependency of the heavy-quark suppression Collisions of heavy ions at center-of-mass energies of as well as to search for angular asymmetries relative to 200 GeV performed at the Relativistic Heavy Ion Collider the reaction plane of the collision. These measurements (RHIC) at Brookhaven National Lab have shown that a have been made for particles scattered in angles around new form of matter has been formed where quarks and 90 degrees relative to the beam (mid-rapidity). The FVTX gluons (partons) are not confined in individual nucleons 803

Page 804

will allow the study of heavy quarks in smaller angles (charm and beauty) within the PHENIX experiment and (forward rapidity) exploring regions where the initial state in the world, he represented all latest results from the effects are more pronounced. This information will be Relativistic Heavy Ion Collider (RHIC) at the Brookhaven important as an input to the models, which describe the National Laboratory. hydrodinamical behavior of the QGP. The LANL group has been leading the construction and Scientific Approach and Accomplishments installation of the FVTX upgrade detectors in the PHENIX experiment at RHIC. This new detector was successfully in- The PHENIX forward detectors can track and identify stalled at the end of 2011 and participated in the data tak- muons with a set of trackers and absorbers (where only ing during 2012 run. Cesar has played a major role during muons can penetrate). The heavy quarks we want to mea- this commissioning preparing the integration of the FVTX sure are included in the sample of muons detected by the electronics with the complicated data acquisition system of forward detectors. Despite the large rejection of other par- PHENIX and debugging it during installation and during the ticles that are not muons in the absorbers, there are too U+U run. many pions produced in the collisions and many of then can penetrate them and be detected as a muon. Besides, With the arrival of 2012 physics data, Cesar worked on pions also decay in muons and if that happen before they physics analysis of the new FVTX muon data, including reach the absorber they can also contribute to the back- developing of the tracking code to match FVTX track with ground in our muon sample from heavy-quarks. Muon track in p+p and A+A collisions, analysis of high mass dimuon pairs from J/Psi decays and simulations that pro- We have built and installed a new Forward Silicon Vertex vide direct preparation for using the new FVTX detector. Detector (FVTX), which consists of four silicon mini-strip Improvements on the J/psi mass peak resolution were al- stations in front of each of the two PHENIX muon detec- ready achieved when using FVTX hit information to recon- tors. The two FVTX detectors can perform tracking of struct dimuons in p+p collisions. With additional dimuon particles going in the direction of the forward detectors counts we hope to start to see a second peak from ψ’ with before reach any absorber material. These tracks indicate a mass slightly larger than the J/ψ. He is well positioned to how far the particle was produced from the collision point. lead the first analysis using the new FVTX from this year’s Pions that could cross the absorber material usually have data as soon as the data reconstruction is finish by the end their direction modified inside the absorbers, so a track of this year. from these pions seen before (in FVTX) and after (in MuTr) the absorber will not match. Muons from pion decays are Pre-FVTX analysis are also being completed by Cesar. A list mostly created far from the collision point allowing their of final and preliminary results are: rejection using FVTX. Muons from heavy quarkonia (J/ψ, ψ’ and ϒ1S+2S+3S), Drell-Yan and W decays are produced at Final analysis of quarkonia (J/psi, Psi’, Chi_c) in 200 the collision point. Open heavy-quark mesons D (B), con- GeV p+p collisions, which result in, a long paper taining a charm(bottom) valence quark, typically travel a [PRD85,092004(2012)], which is the reference for all quar- distance of 123-312 (460-500) m before decay. Therefore, konia measurements, made by PHENIX. Preliminary result the displacement-vertex measurement of the particles of the asymmetry of J/ψ angle detection relative to the crossing the FVTX will help in the identification of muons collision reaction plane. This measurement can indicate if and they sources. the J/ψ suppression observed in Au+Au collisions depends on the path length in Au+Au or if J/ψ yields also contain We successfully commissioned and took p+p, U+U and coalescence or regeneration of open charms in the QGP. Cu+Au collisions data in 2012 using FVTX. Through this The result are 5 times more precise than the original one project, Cesar has continued to establish himself as a lead- obtained in 2007 due to enhanced statistics and care- ing international expert in the field of heavy-ion physics ful handling of the backgrounds. However, even with the and has positioned himself to play a lead role in the new precision we have we can’t still claim any asymmetry or physics that will be enabled by the new LANL-led Forward distinguish for different theoretical scenarios. The addition Silicon Vertex Detector (FVTX). of the data set taken in 2011 can further reduce the uncer- tainties on this measurement. He was nominated to represent the PHENIX collabora- tion and gave plenary invited talks about heavy quark and Work on determining the nuclear modification of Upsilons quarkonia at the Quark Matter Conference in 2011, this in Au+Au collisions. A clear Upsilon peak in the invariant year’s Nuclear Dynamics Winter Workshop and two talks in mass distribution of dimuons was found for the first time the International Workshop on Heavy Quark Production in in such collisions in the forward rapidity at RHIC. Heavy-Ion Collisions. As a leading expert on heavy-quarks 804

Page 805

5. Cesar is the leader of the Quarkonia Topical Group, chair lisions at sqrt(s_NN) = 200 GeV\}. 2011. Phys.Rev.. C84: of the Muon Software meetings, first contact person for 024904. all issues involving the forward detectors software and Adare, A., and others. \{Event Structure and Double Helic- extended member of the PHENIX Speakers Bureau. ity Asymmetry in Jet Production from Polarized $p+p$ Collisions at $$\begin{gathered} sqrt \\ \{s \\ \}=200 \end{gathered}$$~GeV\}. 2011. Phys.Rev.. Impact on National Missions D84: 012006. This work addressed the fundamental study of de-confined quark-gluon matter created in very high-energy heavy- Adare, A., and others. \{Identified charged hadron produc- ion collisions, which mimic the first few milliseconds of tion in $p+p$ collisions at $$\begin{gathered} sqrt \\ \{s \\ \}=200 \end{gathered}$$ and 62.4 the universe. These studies are the primary goal of the GeV\}. 2011. Phys.Rev.. C83: 064903. forefront Relativistic Heavy Ion Collider facility that is the Adare, A., and others. \{Azimuthal correlations of electrons leading U.S. Nuclear Physics facility. It also addressed a from heavy-flavor decay with hadrons in p+p and U.S. Nuclear Science Advisory Committee (NSAC) Goal that Au+Au collisions at sqrt(s_NN)=200 GeV\}. 2011. Phys. deals with the structure of nuclei at a very small momen- Rev.. C83: 044912. tum scale where gluons in those nuclei become saturated and alter their structure, compared with their normal Adare, A., and others. \{Measurement of neutral mesons in structure in free protons. p+p collisions at $sqrt(s)$= 200 GeV and scaling prop- erties of hadron production\}. 2011. Phys.Rev.. D83: Publications 052004. Adare, A., S. Afanasiev, C. Aidala, Ajitanand, Adare, A., and others. \{Nuclear modification factors of phi N.N., Y. Akiba, and others. \{Production of omega me- mesons in d+Au, Cu+Cu and Au+Au collisions at sqrt(S_ sons in p+p, d+Au, Cu+Cu, and Au+Au collisions at NN)=200 GeV\}. 2011. Phys.Rev.. C83: 024909. sqrt(s_NN)=200 GeV\}. 2011. Adare, A., and others. \{Cross Section and Parity Violating Adare, A., S. Afanasiev, C. Aidala, Ajitanand, Spin Asymmetries of W+/- Boson Production in Polar- N.N., Y. Akiba, and others. \{J/psi suppression at for- ized p+p Collisions at sqrt\{s\}=500 GeV\}. 2011. Phys. ward rapidity in Au+Au collisions at sqrt(s_NN)=200 Rev.Lett.. 106: 062001. GeV\}. 2011. Adare, A., and others. \{Cross section and double helicity Adare, A., S. Afanasiev, C. Aidala, Ajitanand, asymmetry for Eta mesons and their comparison to N.N., Y. Akiba, and others. \{Cold Nuclear Matter Effects neutral pion production in p+p collisions at sqrt(s)=200 on J/psi Yields as a Function of Rapidity and Nuclear GeV\}. 2011. Phys.Rev.. D83: 032001. Geometry in Deuteron-Gold Collisions at sqrt(s_NN) = 200 GeV\}. 2010. Phys.Rev.Lett.. Adare, A., and others. \{Measurement of Transverse Single- Spin Asymmetries for J/psi Production in Polarized p+p Adare, A., and others. \{Suppression of back-to-back hadron Collisions at sqrt(s) = 200 GeV\}. 2010. Phys.Rev.. D82: pairs at forward rapidity in d+Au Collisions at sqrt(s_

NN)=200 GeV\}. 2011. Adare, A., and others. \{High pT direct photon and pi0 trig- Adare, A., and others. \{Observation of direct-photon col- gered azimuthal jet correlations and measurement lective flow in sqrt(s_NN)=200 GeV Au+Au collisions\}. of kT for isolated direct photons in p+p collisions at 2011. sqrt\{s\}=200 GeV\}. 2010. Phys.Rev.. D82: 072001. Adare, A., and others. \{Measurements of Higher-Order Adare, A., and others. \{Azimuthal anisotropy of neutral Flow Harmonics in Au+Au Collisions at sqrt(s_NN) = pion production in Au+Au collisions at sqrt(s_NN) = 200 GeV\}. 2011. 200 GeV: Path-length dependence of jet quenching and the role of initial geometry\}. 2010. Phys.Rev.Lett.. Adare, A., and others. \{Ground and excited charmonium 105: 142301. state production in $p+p$ collisions at $$\begin{gathered} sqrt \\ \{s \\ \}=200 \end{gathered}$$ GeV\}. 2011. Adare, A., and others. \{Elliptic and hexadecapole flow of charged hadrons in Au+Au collisions at sqrt(s_NN) = Adare, A., and others. \{Heavy Quark Production in p+p and 200 GeV\}. 2010. Phys.Rev.Lett.. 105: 062301. Energy Loss and Flow of Heavy Quarks in Au+Au Colli- sions at sqrt(s_NN)=200 GeV\}. 2010. Adare, A., and others. \{Transverse momentum dependence of meson suppression Eta suppression in Au+Au col- Adare, A., and others. \{Suppression of away-side jet frag- lisions at sqrt\{sNN\} = 200 GeV\}. 2010. Phys.Rev.. C82: ments with respect to the reaction plane in Au+Au col- 011902. 805

Page 806

Adare, A., and others. \{Transverse momentum dependence of J/psi polarization at midrapidity in p+p collisions at sqrt(s) = 200-GeV\}. 2010. Phys.Rev.. D82: 012001. Adare, A., and others. \{Trends in Yield and Azimuthal Shape Modification in Dihadron Correlations in Rela- tivistic Heavy Ion Collisions\}. 2010. Phys.Rev.Lett.. 104: 252301. Adare, A., and others. \{Detailed measurement of the e+ e- pair continuum in p+p and Au+Au collisions at sqrt\{sNN\} = 200 GeV and implications for direct photon production\}. 2010. Phys.Rev.. C81: 034911. Adare, A., and others. \{Enhanced production of direct pho- tons in Au+Au collisions at sqrt(sNN)=200 GeV and im- plications for the initial temperature\}. 2010. Phys.Rev. Lett.. 104: 132301. Adare, A., and others. \{Double-Helicity Dependence of Jet Properties from Dihadrons in Longitudinally Polarized p+p Collisions at sqrt\{s\}=200 GeV\}. 2010. Phys.Rev.. D81: 012002. Afanasiev, S., and others. \{Enhancement of the dielectron continuum in s(NN)**(1/2) = 200-geV Au+Au colli- sions\}. 2007. Phys.Rev.Lett.. Silva, C. da, and others. \{Recent Heavy Flavor and Quar- konia Measurements from the PHENIX Experiment\}. 2011. Silva, C. da, and others. \{Decay Angular Distributions for the J/Psi Measured by PHENIX\}. 2011. 806

Page 807

Physics Postdoctoral Research and Development Final Report The Formation and Evolution of Black Holes in the Universe Christopher L. Fryer 20100636PRD4 Abstract 10,000 times that of our sun. Such stars are generally Many questions at the forefront of modern astrophysics thought to collapse to black holes. Since they start out and cosmology pertain to the formation and growth of very massive, it is possible for them to grow to become the first black holes in the universe. Here we describe the supermassive black holes, which have been detected our work studying these questions over the past months, at very early cosmic times. with particular attention to one of the main models for One of the main aims of our research was to gain a early black hole formation via the collapse of supermas- better understanding of how such black holes formed, sive stars. We made considerable progress in under- and we were largely focused on how and where their standing the formation and fate of these objects, and supermassive star progenitors formed. The sites of their our research into this and related topics has resulted in formation are the cores of the earliest nascent galax- 13 successful publications. Our work has shed light not ies, and the formation and evolution of each depends only on the origin of the first black holes, but also on the in crucial ways on the other. Therefore, we studied a first planets and powerful early supernovae, and overall number of aspects of galaxy and black hole formation, on how the first galaxies in the universe came to be. using a variety of techniques, in hopes of understanding the whole picture. Background and Research Objectives It is now well established that most massive galaxies, We carried many projects from conception to publica- including our Milky Way, harbor supermassive black tion, resulting in a total of 13 publications during the holes at their centers. There are a variety of correlations duration of the project. These included the first analyti- between the observed properties of galaxies and the cal modeling of the growth of radiating supermassive masses of their central black holes, and this point to an stars; some of the most realistic large-scale cosmologi- interwoven evolution of galaxies and black holes going cal simulations to date, which allowed to pin down the back to the formation of the first such objects in the formation sites of supermassive black holes in the early early universe. One of the principal questions challeng- universe; and utilization of the available observational ing astrophysicists and cosmologists today is how the data on the nature of black hole growth, in order to first supermassive black holes formed and evolved. In show that supermassive stars, while exotic, are indeed particular, there have now been detected numerous the strongest candidates for the seeds of the largest black holes with masses exceeding a billion times that supermassive black holes. These works have helped to of our Sun, already in place less than one billion years propel forward the study of supermassive star and black after the Big Bang, when the universe was less than ten hole formation, with numerous international groups percent of its present age. now engaging in related research that builds upon, and is motivated by, our own. One of the most promising and fascinating explanations for these early behemoths is that they initially formed In addition, we have pursued other questions with bear- from the collapse of supermassive stars that formed ing on the formation of the first galaxies and which help from the primordial gas. In this scenario, the primor- to get at the question of how we can verify our theories dial gas is prevented from forming hydrogen molecules for how and where the first black holes formed. Among that allow it to cool and form typical primordial stars. the other projects that we have carried through publica- Instead, the gas remains hot until it collapses under its tion are studies of the explosions of supermassive stars, own gravity and forms a star with a mass in excess of 807

Page 808

which were among the most energetic thermonuclear emitted form stars in the first galaxies, from early to later events in the history of the universe; studies and simula- times (left to right). The intensity of the molecule-dissoci- tions of bright supernovae and gamma-ray bursts which ating radiation field increases steadily, and this eventually could have originated in the first galaxies; and two inves- leads to the assembly of hot primordial gas from which tigations into how the first planets in the universe formed supermassive stars form (from Johnson, Dalla Vecchia & around the stars inhabiting the first galaxies. Khochfar 2013). Scientific Approach and Accomplishments Another example of a cosmological simulation is shown in Figure 4, which shows the blast wave created by the We adopted various technical approaches to the questions explosion of a supermassive star in one of the most ener- we addressed. For some of the more exploratory projects, getic explosions in the universe. This work showed how many addressing questions that had not previously been such enormous explosions, expected to be as powerful as considered in much detail, we used analytical techniques. 10,000 typical supernova, impacted their host galaxies and In this way, we estimated the masses to which primordial highlighted how the heavy elements they produced may stars can grow, in the face of the high-energy radiation be found in ancient stars still inhabiting our galaxy today. that they emit, as shown in Figure 1. This work helped to While the cosmological simulation was needed to capture support the idea that supermassive stars could actually be the large-scale evolution of the blast wave, we simulated assembled in the early universe. The maximum mass to the small-scale evolution is great detail as well, using the which stars can grow by accreting gas at a constant rate. Los Alamos RAGE code. The combination of these two cal- As shown by the blue line, the radiation emitted by the star culations made this most detailed cosmological supernova will stop accretion only after it has grown to a high mass, calculation ever carried out, Figure 4. if it accretes gas at a sufficiently high rate, as expected for supermassive stars in the early universe. Impact on National Missions We also made the first estimate of the amount of heavy Among the simulations we carried out in the course of our elements required to form the first planets in the universe, research were some of supernovae run with the Los Ala- as shown in Figure 2. We found that our theory, which was mos RAGE radiation hydrodynamics code. We were able built from the main, overarching theory for planet forma- to uncover problems with the implementation of gravity tion, was consistent with the growing amount of data on in this code, and since then significant progress has been exoplanetary systems. This work received a surprising made in correcting these problems. Indeed, our supernova amount of coverage in the popular press, being covered on simulations have been an influential utilization of radiation msnbc.com, space.com and Astrobiology magazine. transport codes developed at Los Alamos. The colored lines show the minimum heavy element In addition, as regards new staff, the postdoc funded for abundance, here shown as the iron abundance ([Fe/H]) this research, Jarrett Johnson, has subsequently joined the as a proxy, for planet formation. The black points show scientific staff of the Lab in XTD-6. Wes Even, a postdoc col- the heavy element abundances of known planets, none of laborator on our research, has also joined the Lab as staff which exhibit heavy element abundances below the mini- in CCS-2. mum we estimated (from Johnson & Li 2012). Other projects demanded much more complex approach- es, and for some of these we carried out sophisticated, large-scale cosmological simulations in which we modeled in detail the formation of galaxies and the emergence supermassive stars. In one such simulation, we included the build-up of the early stellar radiation field, which is expected to lead to supermassive star formation (due to the photodissociation of hydrogen molecules), as shown in Figure 3. This simulation, which included star formation and heavy element enrichment, in addition to radiation from stars, was arguably the most realistic and complete simulation of the formation of the first galaxies ever done, Figure 1. The maximum mass to which stars can grow by ac- Figure 3. creting gas at a constant rate. As shown by the blue line, the radiation emitted by the star will stop accretion only after it has The build-up of the radiation field produced by the light grown to a high mass, if it accretes gas at a sufficiently high rate, 808

Page 809

as expected for supermassive stars in the early universe (from Publications Johnson et al. 2012). Agarwal, B., S. Khochfar, J. L. Johnson, E. Neistein, C. Dalla Vechia, and M. Livio. Ubiquitous seeding of supermas- sive black holes by direct collapse. 2012. Monthly No- tices of the Royal Astronomical Society. 425 (4): 2854. Elliott, J., J. Greiner, S. Khochfar, P. Schady, J. L. Johnson, and A. Rau. The long gamma-ray burst rate and the correlation with host galaxy properties. 2012. ASTRON- OMY & ASTROPHYSICS. 539: A113. Johnson, J. L.. Angular momentum transport by thermal emission in black hole accretion disks. 2011. Astrono- mische Nachrichten. 332 (8): 841. Johnson, J. L.. The First Galaxies: Theory and Simulations. 2013. In The First Galaxies: Theoretical Predictions and Observational Clues. Edited by Wiklind, T., B. Mobash- Figure 2. The colored lines show the minimum heavy element er, and V. Bromm. Vol. 396, p. 177. Berlin: Springer- abundance, here shown as the iron abundance ([Fe/H]) as a Verlag. proxy, for planet formation. The black points show the heavy ele- ment abundances of known planets, none of which exhibit heavy Johnson, J. L., C. Dalla Vecchia, and S. Khochfar. The First element abundances below the minimum we estimated (from Billion Years project: the impact of stellar radiation Johnson & Li 2012). on the co-evolution of Populations II and III. 2013. Monthly Notices of the Royal Astronomical Society. 428: 1857. Johnson, J. L., D. J. Whalen, H. Li, and D. Holz. Supermas- sive Seeds for Supermassive Black Holes. To appear in The Astrophysical Journal. Johnson, J. L., D. J. Whalen, W. Even, C. L. Fryer, A. Heger, J. Smidt, and K. Chen. The Biggest Explosions in the Uni- verse. The Astrophysical Journal. Figure 3. The build-up of the radiation field produced by the light Johnson, J. L., and H. Li. Constraints on planet formation emitted form stars in the first galaxies, from early to later times via gravitaional instability across cosmic time. 2013. (left to right). The intensity of the molecule-dissociating radia- Monthly Notices of the Royal Astronomical Society. tion field increases steadily, and this eventually leads to the as- 431: 972. sembly of hot primordial gas from which supermassive stars form (from Johnson, Dalla Vecchia & Khochfar 2013). Johnson, J. L., and S. Khochfar. The Contribution of Su- pernovae to Cosmic Reionization. 2011. Astrophysical Journal. 743: 126. Johnson, J., B. Agarwal, D. Whalen, C. Dalla Vecchia, C. Fry- er, S. Khochfar, Hui Li, and M. Livio. The growth of the stellar seeds of supermassive black holes. 2012. In First Stars IV - from Hayashi to the Future ; 21-25 May 2012 ; Kyoto, Japan. Vol. 1480, p. 313. Johnson, J., D. Whalen, C. Fryer, and Hui Li. The growth of the stellar seeds of supermassive black holes. 2012. Astrophysical Journal. 750 (1): 66 (12 pp.). Figure 4. Two views of the blast wave produced by the explo- Johnson, J., and Hui Li. The First Planets: the Critical Metal- sion of a supermassive star in the early universe, as seen on the licity for Planet Formation. 2012. Astrophysical Journal. sky (in units of seconds of degrees). Such supernova explosions 751 (2): 81 (11 pp.). would have been among the most energetic thermonuclear events in the history of the universe (from Johnson, Whalen, Whalen, D. J., W. Even, L. H. Frey, J. L. Johnson, C. C. Love- Even, Fryer, et al. 2013). kin, C. L. Fryer, M. Stiavelli, D. E. Holz, A. Heger, S. E. 809

Page 810

Woosley, and A. L. Hungerford. Finding the First Cosmic Explosions I: Pair-Instability Supernovae. To appear in The Astrophysical Journal. 810

Page 811

Physics Postdoctoral Research and Development Final Report Auger-recombination-free Nanocrystals by Rational Design of Confinement Potential Victor I. Klimov 20110544PRD1 Abstract to simple fabrication processes such as spin-coating, The goal of this project has been to achieve fundamental ink-jet and roll-to-roll printing. Over the last few years, understanding of the interplay between structural char- several examples of LEDs based on nanostructured ma- acteristics and photophysical properties of colloidal core/ terials have been reported by different groups, and have shell nanocrystals (NCs) with designed interfacial compo- shown a remarkable tunability across the entire visible sition. The main aim of these studies has been the devel- spectrum. opment of NCs, which would allow us to achieve lasing The performance of NC-LEDs is greatly limited by fast in the regime of electrical injection. New light-emitting nonradiative AR. In conventional ‘single-component’ and especially lasing structures based on low-cost, NCs, Auger lifetimes are proportional to NC volume, a chemically synthesized materials would benefit numer- seemingly universal trend that has been observed for ous technologies from solid-state lighting and displays to many different semiconductors. The development of optical telecommunication and quantum cryptography. approaches for reducing AR rates, while still preserving NCs are particularly attractive for these applications due a significant degree of spatial confinement, is a long- to their efficient and size-tunable emission. A significant standing goal in the field of NCs. Recently, core-shell impediment to applications of NCs in lasers is fast non- nanostructures of CdTe/CdSe, CdZnSe/ZnSe, and CdSe/ radiative Auger recombination (AR), which limits optical CdS (thick-shell samples) as well as CdSe/CdS quantum- gain lifetimes to sub-ns timescales. Recent studies of dot/quantum-rod structures have been shown to exhibit LANL researchers have demonstrated that AR is signifi- a significant suppression of AR as assessed from single- cantly reduced in so-called “giant” NCs (g-NCs) compris- NC data or time-resolved ensemble studies. The reasons ing a small CdSe core overcoated with a thick CdS shell for this dramatic suppression of AR in thick-shell CdSe/ (Garcia-Santamaria et al., Nano Lett. 2009). The goal of CdS NCs (as well as in other reported nanostructures) this project has been to elucidate the reasons underlying are still not understood. Recent theoretical calculation the suppression of AR in these NCs by conducting direct by Cragg and Efros demonstrated that AR can be reduced measurements of multiexciton dynamics in conjunction by several orders of magnitude by replacing the sharp with studies of NC structure/composition via analytical step-like potential with smooth profile. In core-shell spectroscopic methods such as Raman scattering and flu- NCs, this is achievable by promoting a gradual transition orescence line narrowing (FNL). The insights gained from of the compositional profile between the core and the these studies have been used to design novel structures shell domains, which yields an alloyed region of mixed in which AR is efficiently suppressed which have allowed composition. The goal of this project has been to achieve us to extend the optical gain lifetime and increase its optimized NCs with suppressed Auger recombination magnitude and also boost the efficiency of NC-based through engineering of the confinement potential. light emitting diodes (LEDs). Scientific Approach and Accomplishments Background and Research Objectives The first part of this Project consisted of an experimental Colloidal semiconductor nanocrystals (NCs) have been investigation of the fine structure of the excitonic states considered as promising materials for optoelectronic in CdSe/CdS heterostructures where spatial separation applications because they exhibit particle-size-tunable between electrons and holes can be tuned by control- emission color and high (>90%) intrinsic quantum yields ling shell thickness (Fig. 1). This study has allowed us to in emission. Furthermore, NCs are produced through gain deeper understanding of the photophysics of this inexpensive, scalable solution-based synthesis amenable 811

Page 812

class of nanomaterials and to formulate a new paradigm meV). Remarkably, with increasing H, we also observe the for controlling exciton dynamics via engineered electron- emergence of a new FLN feature at LO12 = 60 meV (see, hole exchange interaction. Specifically, we showed that e.g., the spectrum of the 7 ML-shell sample), which is due the splitting energy between dark and bright excitonic to the combination of LO1 and LO2 modes. Such a “com- states in core/shell CdSe/CdS NCs can be tuned externally binatory” mode is a signature of the CdSeS alloy and is by controlling the degree of delocalization of the electron indicative of the development of an interfacial “graded” re- wavefunction in the CdS shell (Fig. 2). By applying this new gion where the mismatch between CdSe and CdS lattice is strategy, we were able to fabricate NCs with light-emission gradually recovered. The increases in the amplitude of the properties, which are independent of magnetic field up to LO12 feature correlates with the reduction in the intensity 7 Tesla and of temperature from 300 K to 1.5 K. Because of of the CdSe-related phonon replica, indicating progressive its fundamental and technological importance, this study, consumption of the core material, which accompanies the published Nature Communications, has been highly cited formation of the alloy layer. (23 citations in one year) and has already resulted in sev- We quantify these observations by plotting relative intensi- eral follow-up publications from other groups. ties of the signals related to three different components Based on this newly acquired knowledge of excitonic fine of the NCs as a function of R (Fig. 5): the interfacial alloy structure, we have able to use CdSe/CdS NCs as a model region [Aalloy; triangles in (a)], the CdSe core [ACdSe; system to investigate the role of interfacial alloying on AR solid circles in (b)], and the CdS shell [ACdS; open circles in the light of a recent theoretical paper by Cragg & Efros in (b)]. In these plots, the signal related to a certain NC (Nano Letters, 2009) that predicted orders of magnitude region (e.g., the integral intensity of the LO1 and the 2LO1 suppression of AR in NCs where the abrupt potential step features in the case of the CdSe core) is normalized by the is substituted by a smooth potential gradient. Studying total area of all phonon replicas in the FLN spectrum. The thick shell CdSe/CdS NCs through ultrafast optical spec- data presented in this way clearly indicate that most of the troscopy we observed a very significant suppression of AR interfacial alloying occurs until the shell thickness reaches that greatly exceeded expectations based on established ~9 MLs, while further shell growth is not accompanied by arguments, such as volume scaling (Fig. 3). any significant expansion of the alloy region. Interestingly, the largest change in the Auger decay time also occurs in This observation suggested a dominant role of the interfa- the same range of shell thicknesses for which we observe cial composition, which defines the shape of the confine- the formation of the alloyed interface. Specifically, τ2A ment potential, in AR suppression. Probing elusive sub- changes by more than two orders of magnitude (from 250 nanometer variations in the NC composition constituted ps to 31 ns; the latter value corresponds to β = 4), when H a significant experimental challenge. To this end, we have increases from 0 to 11 ML. This increase is about ten times developed a new experimental method based on fluores- larger than that predicted based on changes in the effec- cence line narrowing spectroscopy (FLN, Fig. 4). In this tive exciton volume and e-h spatial overlap. method, NCs are excited with a cw laser on the red side of the BE absorption feature, which effectively selects a At larger values of H, the Auger time increases more slowly sub-ensemble of the largest NCs in the sample. As a result, and can be ascribed to the reduction in the e-h overlap the low-temperature FLN spectra typically exhibit narrow integral. For example, when H grows from 9 to 14 ML, emission peaks due to zero-phonon and phonon-assisted which corresponds to the range where the alloy thickness excitonic transitions. As a result of this spectroscopic is essentially constant, we observe a three-fold increase selection, using FLN we can clearly resolve features due to in τ2A (from 31 ns to 90 ns). This correlates well with an phonon-assisted transitions involving various longitudinal approximately three-fold reduction in Θeh (from 0.24 to optical (LO) modes; based on the intensity of these modes 0.09) as derived from changes in single exciton dynamics we can analyze the composition of the NC within the vol- (τx increases from 45 ns to 115 ns, Fig. 3a). These results ume sampled by a photoexcited exciton. strongly suggest that the major contribution to the ob- served suppression of AR arises from the softening or Figure 4a shows the 1.55 K FLN spectra of CdSe core-only smoothing of the confining potential due to formation of NCs and a few CdSe/CdS NC samples with increasing shell an alloyed layer at the CdSe/CdS interface. Through this thicknesses. The spectrum of core-only NCs exhibits a novel approach, we provided the first experimental veri- feature due to exciton recombination assisted by the LO fication of Efros’s theory, which we published in a highly phonon of CdSe (LO1; 25 meV), with overtones at 2LO1 (50 cited article in Nano Letters (34 citations; highlighted in meV) and 3LO1 (75 meV). The growth of the CdS shell re- Chemical & Engineering News). sults in additional peaks due to phonon-assisted transitions involving a CdS LO phonon (LO2 at 35 meV and 2LO2 at 70 One important consequence of suppressed AR is a sig- 812

Page 813

nificant increase in emission efficiencies of multiexciton The EL response of a representative LED comprising an states. For example, in the core-only samples studied in active layer made from 4-nm diameter core, 16-mono- this work, the emission quantum yield of biexcitons is layer shell CdSe/CdS NCs is reported in Fig. 5a. A plot of ~1%. However, it reaches 40 – 50% in CdSe/CdS structures luminance versus voltage reveals a clear EL ‘turn-on’ at a with shell thickness of 9 ML and is as high as 60 – 80% in bias of 3.0 V. The LED reaches a maximum luminance of samples with a 19 ML shell (Fig. 5b, inset). This increase 2000 Cd/m2 at a current density (J) of 1200 mA/cm2 and in PL yields of multiexciton states can greatly benefit ap- voltage of 11 V (Fig. 5a). The EL spectrum overlaps almost plications of NCs in technologies involving light emission precisely with the PL spectrum (Fig. 5b), where the latter such as lasing, optical amplification and solid-state light- was obtained from a solution of the NCs prior to device ing. To explore the applicability of CdSe/CdS NCs to LEDs, fabrication. Interestingly, at higher current densities, we do we conducted a detailed systematic study of the effect of observe a weak, high-energy band that we tentatively attri- shell thickness on device performances, showing a tenfold bute to emission from higher order excited states (inset of increase of LED efficiency for thick shell NCs. We explained Fig. 5b). Such emission from charged or multiexciton states this increase by greatly reduced rates of AR and strongly becomes possible in the case of efficient suppression of suppressed NC-to-NC energy transfer with respect to con- nonradiative Auger recombination, as shown in the previ- ventional systems for which these parasitic non-radiative ous section. Significantly, the luminance level for ‘standard processes severely reduce the electroluminescence ef- video brightness’ (200 Cd/m2) arises in this device at 6.5 V ficiency. This study has been published in Nano Letters. and 180 mA/cm2. EQE reaches a maximum of 0.17% at 100 mA/cm2 (7.0 V) with a peak luminance efficiency of 0.36 As demonstrated above, core/shell NC in which a CdSe Cd/A. A photograph of the device (Fig. 5c, inset) reveals core is overcoated by a thick CdS shell show dramatically uniform pixel brightness over the relatively large area of suppressed AR and are therefore interesting candidates the device (0.03 cm2) at a bias of 9 V. The device shows no for efficient NC-LEDs. In addition, optical characteristics sign of degrading when tested repeatedly after storage in of these systems (including PL QY and long-term optical air for more than a month. Importantly, earlier reports us- stability) were observed to be largely independent of the ing similar architectures and conventional CdSe/ZnS core/ ligand coating. However, there are several potential issues shell NCs resulted in an EL efficiency less than ~6x10-3 that could make these systems unsuitable for LED appli- Cd/A, which is nearly two orders of magnitude lower than cations. Currently, solution-phase QYs are at most ~50%, that of our 16-ML-shell CdSe/CdS device. compared to near-unity in the case of optimized thin-shell NCs. In addition, the density of NCs in the LED active layer Impact on National Missions is effectively reduced due to the much larger size of a thick As mentioned above, NCs are appealing candidates for effi- shell NC compared to a conventional system. Together, cient solid state lighting technologies. Given the size of the these features could render an active layer less emissive already existing market, applied research in this field and compared to a standard NC film. Furthermore, the thick efficient transfer of knowledge to the industrial entities shell comprising the relatively wide band-gap semiconduc- opens new opportunities for economic growth. In addition, tor, CdS, could inhibit charge injection and raise the LED there is a chance that new governmental regulations will turn-on voltage. To investigate these issues, we utilize a further push for this paradigm shift thereby providing an simple LED device architecture for comparing g-NCs to additional stimulus or the development of the solid state conventional thinner-shell systems. lighting industry, with obvious increase of its economic relevance. Clearly, companies with privileged access to In our experiments (Fig. 5), we have used a simple device technical expertise will be in a leading position within the architecture, which comprises a NC active layer in between industry and will benefit of an important time advantage a PEDOT:PSS-coated indium-tin oxide (ITO) anode and a that could allow them to establish themselves on the mar- LiF/Al cathode. The LEDs were fabricated on ITO-coated ket before other players become competitive. Although glass substrates. The PEDOT:PSS layer (~30 nm) was spin- the fundamental knowledge of the physical processes coated onto the substrate and annealed at 250 °C for underpinning the operation of efficient NC-based devices 20 min inside an inert-atmosphere (UHP Ar) glovebox. is not trivial, their production, once properly understood The high-temperature annealing yields a compact hole- and optimized, is relatively inexpensive compared to other conducting layer. The NC active layer was then spin-coated materials such as epitaxial semiconductor structures, (~2500 rpm) from a 10 mg/ml NC octane solution. Finally, which instead require expensive facilities typical of micro- 2 nm of LiF and 100 nm of aluminum were thermally electronic industry. In addition, the technical entry barrier evaporated onto the NC layer through a shadow mask to is low, which makes it particularly easy to be transferred to form the top electrode. companies. 813

Page 814

Figure 1. Spatial probability distribution of the hole (red area) and electron (green areas) for R = 1.5 nm and H = 1.6 nm, H = 2.8 nm and H = 7.6 nm. Figure 4. Compositional analysis of CdSe/CdS NCs based on the intensities of phonon replica in FLN spectra. (a) The 1.55 K FLN spectra (black lines) of core-only CdSe NCs (bottom curve) and four CdSe/CdS NC samples with R0 = 1.5 nm and different shell thicknesses (H = 1.6, 2.8, 5.6, and 7.6 nm, corresponding values in ML are reported next to each curve); the horizontal-axis zero corresponds to the excitation energy. Gray lines show Gaussian fits of the FLN spectra that account for contributions of phonon replicas associated with LO modes of CdSe (LO1), CdS (LO2), and CdSeS (LO12); surface phonons were not included in the fit- ting procedure. (b, c) Spectrally integrated intensity of the FLN features due to the alloy mode (triangles in (b)), the CdSe mode (solid circles in (c)) and the CdS mode (open circles in (c)) plotted Figure 2. Dark-bright splitting energy as a function of the CdS as a function the nanoparticle radius. Crosses are the calcula- shell thickness (in monolayers) for core/shell CdSe/CdS NCs. tions. Figure 3. (a) Biexciton AR lifetimes. Lines show AR lifetimes expected for the cases of linear scaling with effective exciton vol- Figure 5. a) Top panel: Current density (full black circles) and ume and the volume-scaling corrected for e-h spatial separation luminance (empty circles) characteristics of a NC-LED incorpo- (dashed line). (b) “Intrinsic” biexciton emission quantum yield; rating a monolayer of CdSe/16CdS NCs. Lower panel: PL of the this expression does not account for “extrinsic” carrier losses due, active layer compared to the EL spectrum of the same NC-LED as e.g., to trapping at surface defects. in the top panel at increasing working voltage (namely, 5, 6, 7, 8 V). The inset shows a detail of the high-energy portion of the EL spectra at increasing driving voltage (as indicated by the arrow). b) External quantum efficiency of the same device as a function of current density. Inset: Photograph of a working LED at driving voltage of 9V (J = 450 mA/cm2) showing uniform EL emission from several LEDs on the same substrate. 814

Page 815

Publications Brovelli, S., C. Galland, R. Viswanatha, and V. Klimov. Tun- ing radiative recombination in cu-doped nanocrystals via electrochemical control of surface trapping. 2012. Nano Letters. 12 (8): 4372. Brovelli, S., R. D. Schaller, S. A. Crooker, Garci&acute, Y. Chen, R. Viswanatha, J. A. Hollingsworth, H. Htoon, and V. I. Klimov. Nano-engineered electron-hole exchange interaction controls exciton dynamics in core-shell semiconductor nanocrystals. 2011. Nature Communi- cations. 2: 280 (8 pp.). Galland, C., S. Brovelli, W. K. Bae, L. A. Padilha, F. Mein- ardi, and V. I. Klimov. Dynamic hole blockade yields two-color quantum and classical light from dot-in-bulk nanocrystals. 2013. Nano Letters. 13: 321. GarciÃÅa-SantamariÃÅa, F., S. Brovelli, R. Viswanatha, J. Hollingsworth, H. Htoon, S. Crooker, and V. Klimov. Breakdown of volume scaling in auger recombination in CdSe/CdS heteronanocrystals: The role of the core- shell interface. 2011. Nano Letters. 11 (2): 687. Pal, B., Y. Ghosh, S. Brovelli, R. Laocharoensuk, V. Klimov, J. Hollingsworth, and H. Htoon. ‘Giant’ CdSe/CdS core/ shell nanocrystal quantum dots as efficient electrolu- minescent materials: Strong influence of shell thick- ness on light-emitting diode performance. 2012. Nano Letters. 12 (1): 331. Pandey, A., S. Brovelli, R. Viswanatha, J. M. Petryga, V. I. Klimov, and S. A. Crooker. Long-lived photoinduced magnetization in copperdoped ZnSe–CdSe core–shell nanocrystals. 2012. Nature Nanotechnology. 7: 792. Viswanatha, R., S. Brovelli, A. Pandey, S. A. Crooker, and V. I. Klimov. Copper-doped inverted core/shell nanocrys- tals with “permanent” optically active holes . 2011. Nano Letters. 11: 4753. 815

Page 816

Physics Postdoctoral Research and Development Final Report Universal Physics with Ultracold Atoms Joseph A. Carlson 20110633PRD2 Abstract ‘phases’ of atoms, dimers, and trimers connected by This project was directed at studies of universal physics crossovers and phase transitions. The phase diagram is in cold atom systems. The goal is to elucidate the prop- shown in Figure 1. erties of these systems, allowing insight from one field The weak coupling regime is shown on the left of the to strengthen our understanding across several fields. diagram. In this regime two species form a standard Topics that were studied included new types of cross- superfluid with the remaining species unpaired. The over physics in multicomponent Fermi gases, Efimov coupling increases toward the right of the diagram. At effects in quantum magnets and in two-dimensional the bottom of the figure the effective range is nearly systems. These topics have applications in cold atoms, zero and three atoms of each species can collapse to a nuclear physics, and condensed matter. deeply bound state. The three-component system has a transition from a system with an unpaired species to one Background and Research Objectives without – an ‘atom-trimer continuity’ transition analo- Strongly-interacting Fermi systems appear in many gous to the quark-hadron duality that exists in nuclear subfields in physics and their understanding is important matter at zero temperature with increasing density. This but often challenging. Ultracold atoms are ideal to tackle work was published in Physical Review Letters and is the such problems because of the prominent advantage subject of much new theoretical and experimental inves- of the ability to control the atom-atom interaction. In tigation. Extensions of this work are likely to be impor- particular, the limit of strongest interaction (unitarity tant in nuclear physics where alpha particle clustering is limit) has attracted considerable attention because of important. the universality meaning that its properties are indepen- dent of details of the interaction potential. Physics in the Dr. Nishida then worked with Christian Batista (T-4) and unitarity limit can shed light on various strongly inter- Yasuku Kato on the Efimov effect in quantum magnets. acting systems ranged from high-Tc superconductors to [2] The Efimov effect is a discrete scaling symmetry in neutron stars and is thus an important interdisciplinary few-body systems that relate few-body states at definite field. The goal of my research is to take the advantage of ratios of binding energies. It is typically thought of as ultracold atoms to deepen and enlarge our understand- a continuum few-body effect. Nishida and collabora- ing of the universal few-body and many-body physics tors discovered that the Efimov also exists in quantum that are relevant beyond ultracold atoms and to explore magnets, many-body systems typically defined on a new frontiers across traditional disciplines in physics. lattice. This novel understanding has led to new searches for Efimov effects in a variety of systems. This paper was Scientific Approach and Accomplishments published in Physical Review Letters. Dr. Nishida obtained several important physics results during his Oppenheimer Fellowship at Los Alamos. Another important area of investigation Dr. Nishida Dr. Nishida first studied a new type of crossover phys- conducted was developing a ‘super’-Efimov effect in ics in three-component Fermi gases.[1] The standard two-dimensional systems.[3] The Efimov effect has two component system (e.g. spin up and spin down) traditionally been studied in normal three-dimensional has a smooth crossover from the BCS (weak pairing) systems. Many condensed matter systems have reduced regime to the BEC (strong pairing into bosons) regime dimesionality induced by anisotropies in the background as the strength of the coupling is increased. The three- structure of the material. Dr Nishida and collaborators component system has a richer structure, with different 816

Page 817

found that spinless fermions in two-dimensions fine-tuned yond. 2012. PHYSICAL REVIEW X. 2 (4): -. to a p-wave resonance can form an infinite tower of bound Nishida, Y.. Impossibility of the Efimov effect for p-wave in- states with a univeral doubly-exponential scaling pattern. teractions. 2012. PHYSICAL REVIEW A. 86 (1): 012710. This states could be observed in cold-atom experiments. This paper was published in Physical Review Letters. Nishida, Y.. Probing strongly interacting atomic gases with energetic atoms. 2012. PHYSICAL REVIEW A. 85 (5): Dr. Nishida published several other papers as part of this 053643. project, including studies of electron spin resonance ina dilute magnon gas,[4] and magnetic and nematic orderings Nishida, Y.. Probing strongly interacting atomic gases with in spin-1 antiferromagnets with single-ion anisotropy. The energetic atoms. 2012. PHYSICAL REVIEW A. 85 (5): latter was published in collaboration with scientists in T-4. 053643. Nishida, Y.. New Type of Crossover Physics in Three-Com- Impact on National Missions ponent Fermi Gases. 2012. PHYSICAL REVIEW LETTERS. This project made a significant impact on our understand- 109 (24): -. ing of strongly-correlated systems in condensed matter and nuclear physics. In particular the work on multicomponent Nishida, Y.. Impossibility of the Efimov effect for p-wave in- Fermi systems is influencing studies of nuclear physics, teractions. 2012. PHYSICAL REVIEW A. 86 (1): -. where four species (spin up and spin down neutrons and Nishida, Y.. Probing strongly interacting atomic gases with protons) can cluster into alpha-particle like states. Similar energetic atoms. 2012. PHYSICAL REVIEW A. 85 (5): -. investigations are being pursued in condensed matter and cold atoms. This work will help break the standard two- Nishida, Y., S. Moroz, and D. T. Son. Super Efimov Effect of component paradigm (neutrons or protons paired) used to Resonantly Interacting Fermions in Two Dimensions. study large nuclei. These examinations are critical to the 2013. PHYSICAL REVIEW LETTERS. 110 (23): -. study of neutron-rich matter and nuclei at FRIB (the Facility Nishida, Y., Y. Kato, and C. D. Batista. Efimov effect in quan- for Rare Isotope Beams) at MSU. tum magnets. 2013. NATURE PHYSICS. 9 (2): 93. Dr. Nishida’s work will also be critical for understanding Nishida, Y., and D. Lee. Weakly bound molecules trapped novel states of matter theoretically and experimentally, with discrete scaling symmetries. 2012. PHYSICAL RE- including studies conducted at LANL. VIEW A. 86 (3): 032706. Figure 1. Phase Diagram of Three Species of Fermions ver- Nishida, Y., and D. Lee. Weakly bound molecules trapped sus coupling strength (horizontal axis) and effective range with discrete scaling symmetries. 2012. PHYSICAL RE- (vertical axis). VIEW A. 86 (3): -. References Wierschem, K., Y. Kato, Y. Nishida, C. D. Batista, and P. Sen- gupta. Magnetic and nematic orderings in spin-1 anti-

1. Nishida, Y.. New Type of Crossover Physics in Three- ferromagnets with single-ion anisotropy. 2012. PHYSI- Component Fermi Gases. 2012. PHYSICAL REVIEW CAL REVIEW B. 86 LETTERS. 109 (24): -.
2. Nishida, Y., Y. Kato, and C. D. Batista. Efimov effect in quantum magnets. 2013. NATURE PHYSICS. 9 (2): 93.
3. Nishida, Y., S. Moroz, and D. T. Son. Super Efimov Effect of Resonantly Interacting Fermions in Two Dimensions.
4. PHYSICAL REVIEW LETTERS. 110 (23): -.
5. Wierschem, K., Y. Kato, Y. Nishida, C. D. Batista, and P. Sengupta. Magnetic and nematic orderings in spin-1 antiferromagnets with single-ion anisotropy. 2012. PHYSICAL REVIEW B. 86 (20): -. Publications Knap, M., A. Shashi, Y. Nishida, A. Imambekov, D. A. Abanin, and E. Demler. Time-Dependent Impurity in Ultracold Fermions: Orthogonality Catastrophe and Be- 817

Page 818

Physics Postdoctoral Research and Development Final Report A New Drift Shell Integration Technique for Inner Magnetospheric Space Weather Models Josef Koller 20110735PRD2 Abstract This project aims to develop a drift shell model using a Energetic particles within the magnetosphere undergo novel technique to efficiently calculate the drift shell L* three periodic motions including the gyration, bounce, while preserving a high fidelity of accuracy as the stan- and drift with associated adiabatic time scales and dard method. Instead of using the common empirical invariants. The particle drift shell L*, the third adia- magnetic field model, which can reasonably represent batic invariant, is computationally demanding from the a statistical, average magnetosphere, a magnetically standard method because the global magnetic field lines self-consistent inner magnetosphere model LANL RAM- need to be traced and integrated. This project developed SCB, which solves the magnetic field from fundamental a new tool using the novel neural network technique equations of plasma physics, is employed for the de- to efficiently calculate the drift shell L* in microseconds termination of the drift shell. This project will not only instead of seconds without losing accuracy. Further- evaluate the applicability of the newly developed neural more, since a global magnetic field configuration is the network in predicting the L* and its accuracy, but also essential element in obtaining the drift shell, this project demonstrate the application in addressing the physical uses the magnetospheric configuration form a first- processes in the radiation belt dynamics. The L* neu- principle inner magnetosphere model LANL RAM-SCB, ral network approach will remove the computational rather than the commonly employed empirical magnetic bottleneck in the accurate radiation belt modeling or field model. The newly developed L* neural network is data assimilation, therefore allowing for real-time space capable of removing the computational bottleneck and weather forecasting. hence allows for real-time radiation belt modeling and Scientific Approach and Accomplishments space weather forecasting. A modern machine learning technique (i.e., feed-for- Background and Research Objectives ward artificial neural network) allows one to establish During geomagnetic storms triggered by abrupt solar the complicated relationship between the external solar activities, the severely enhanced energetic electron wind driving and responses within a complex nonlinear flux and radiations in the Van Allen radiation belts will system like the terrestrial magnetosphere system. The pose threats to both spacecraft and astronauts. The artificial neural network, inspired from the biological understanding and modeling of the radiation belt are neuron system that is typically composed of the stimula- therefore of great importance both practically and scien- tion and the response, consists of one input layer, one tifically. The overarching goal of this project is to inves- hidden layer, and one output layer (Figure 1). In this tigate the physical processes that are responsible for study, the input parameters on the neurons, include the radiation belt dynamics, such as transport, loss, and time, location, solar wind and interplanetary magnetic energization. The dynamics of energetic particles can be field conditions, and magnetospheric disturbance index, conveniently expressed and visualized by three adiabatic and the output layer is the particle drift shell L*. The invariants. While the first two are relatively easy to be input parameters are multiplied by weights w and then computed provided with the magnetic field information modified by an activation function f, producing the locally or along a field line, the third invariant, i.e., the output at each neuron to be inter-connected to the particle drift shell L*, is computational expensive since neurons in the next layer. The output L* therefore can be numerical tracing of the global magnetic field lines is described by: needed in the standard approach described in Roederer Training a neural network refers to a process that (1970). 818

Page 819

generalize the relationship between the input and the CODIF instrument and found reasonably good agreements prescribed target. The training begins with an initial guess during storm times. of the weights in the network and then iteratively adjusts L* neural network with Tsyganenko empirical magnetic them through an optimization algorithm until the error field models and the physics-based LANL RAM-SCB model between the target and the network output is minimized We developed the L* neural networks trained from seven or reaches a tolerance. Once the training is completed (i.e., different empirical magnetospheric field models and one the weights and bias are determined), given any input, the first-principle inner magnetosphere model RAM-SCB using drift shell L* can be easily computed. the traditional integration technique. Their applicability One important component in the training data is the under different geomagnetic activity conditions is also desired output, that is, the drift shell L*, which is used for evaluated. These neural networks are found to be highly supervising the learning/training process. This prescribed reliable with the prediction efficiency (PE) about 0.98, L* is obtained from numerical tracing of global field lines, expect for extreme conditions when the magnetosphere the traditional standard approach, within a global mag- is largely disturbed. These L* neural networks have been netospheric magnetic field model by means of the IRBEM released on the website: http://lanlstar.net, implemented Fortran library (http://sourceforge.net/projects/irbem). to the python-based tool SpacePy developed at LANL, and The commonly used global magnetic field model is Tsy- delivered to the Coordinated Community Modeling Center ganenko empirical model, such as T89, T96, T01Storm, (CCMC) at NASA for public application. T01Quiet, and TS05. While these empirical models repre- Quantify the effect of the magnetopause shadowing on sent a statistical, average magnetosphere, a first-principle radiation belts dropouts magnetic field model RAM-SCB self-consistently solves the Radiation belts changes dramatically during geomagnetic global magnetic field configuration based on plasma phys- storm times. Sudden dropouts of electron flux are com- ics. In this study, the RAM-SCB magnetic field configura- mon phenomena. However, the relative contribution of dif- tion is employed to produce the L* for the learning of the ferent loss mechanisms remains far from well understood. neural network. We use the 1D radiation belt diffusion code to study the In order to achieve a trustworthy neural network, sufficient effect of the magnetopause shadowing on the dropouts. training data is required to avoid issues like over-fitting and The initial condition is obtained by averaging over quiet memorization of the input-output pattern. The training time results from DREAM for different (u, K) combinations. patterns are generated with random input. The location, The magnetopause in the model is represented by the last one of the input parameters, is randomly selected within closed drift shell L* which is K-dependent. The loss process the domain (r: [1.03, 11.5] RE, θ: [-90°, 90°], ϕ: [0, 360°] in in the simulation includes the drift loss out of the magne- spherical coordinates). Each location corresponds to a ran- topause that is coupled with outward radial diffusion. The dom time from 01/01/1994 to 01/01/2005 (for empirical resulting PSD is converting to omni-directional flux, which magnetic field models) or from 01/01/2001 to 01/01/2007 is compared to GPS flux observations to quantify the effect (for the RAM-SCB model) and a random pitch angel from 0 of the loss process on the main phase electron dropout. to 90°. The results indicate that the magnetopause shadowing together with the outward radial diffusion can account for Validation of the self-consistent inner magnetosphere 67-88% of the total electron loss in the high-speed solar model RAM-SCB wind stream (HSS) events. This physics-based model couples the ring current-atmo- sphere interaction model (RAM), which solves the bounce- On the L* accuracy and its effect on the PSD radial profile averaged ring current particle phase space distribution We utilizes the phase space density (PSD) matching tech- function [Jordanova et al. 2006] in a non-dipole magnetic nique developed by Chen, et al. [2005] to estimate the L* field, with a 3-D Euler potential based plasma equilibrium accuracy during quiet time. This technique compares two magnetic field code (SCB) [Zaharia et al. 2004; 2008] that PSD values at the same Phase Space Coordinates (i.e., the calculates the magnetic field self-consistently in force bal- same adiabatic invariants u, K, L*)), using a matching ratio ance with the anisotropic ring current distributions. We which is usually deviates from unity due to errors mainly carried out validation study of this self-consistent inner from the L* calculation from the underlying global mag- magnetosphere model by simulating several geomagnetic netospheric model. The PSD data from three LANL-GEO storms with different activity levels and by comparing satellites is used to measure the L* error. Results indicate against a variety of in-situ satellite measurements includ- that the neural network L* with different underlying mag- ing magnetic fields from POLAR and Cluster spacecraft, netospheric models has uncertainties statistically below energy differential ion (H+, O+, and He+) flux from Cluster/ 0.4, implying that these neural networks can be reliably 819

Page 820

used to calculate the drift shell L* during geomagnetically Center and DoD Air Force Research Lab and Space Com- quiet time. However, the neural network with the underly- mand. Understanding the space environment is critical to ing T01Storm model is found to have the most accurate L* fielding the satellite systems that are needed to monitor among all the Tsyganenko empirical models. nuclear nonproliferation. The underlying space physics builds capability to understand the effects of nuclear deto- To investigate how the L* uncertainty affects the PSD esti- nations in space, supporting stockpile stewardship. mation, 100 magnetosphere configurations is created us- ing solar wind input parameters randomly disturbed based on their measurement uncertainties. These 100 global magnetospheric configurations are then used for the L* calculation at a fixed location. These L* values and the corresponding PSD consequently form a pseudo-Gaussian distribution. A change of 0.2 in the L* can vary the PSD by 30%. By taking L* and PSD at different locations from the above magnetospheres, the PSD radial profile is found to be shift but with its shape retained. Radiation belt data assimilation using RAM-SB Figure 1. Left: The architecture of a multi-layer neural network. magnetospheric field configuration Right: The learning/training procedure of a neural network. The LANL-GEO and POLAR spacecraft flux measurements are converted to PSD data for assimilation of the radiation References belt electron distribution during a moderate geomagnetic

1. Yu, Y., J. Koller, S. Zaharia, and V. Jordanova. L* Neural storm event in Dec 19-22, 2002. While he “flux-to-PSD’ Network from Different Magnetospheric Field Models. conversion is highly dependent on interpolation method
2. Space Weather. : S02014. along the process, the cubic-spline interpolation method is found to be sufficiently more accurate than linear interpo-
3. Yu, Y., V. Jordanova, S. Zaharia, J. Koller, J. Zhang, and lation method in producing reliable PSD input data. Using L. Kistler. Validation Study of the Magnetically Self- localized adaptive inflation (LAI) technique in the ensemble Consistent Inner Magnetosphere Model RAMS-SCB. Kalman filter enables successful data assimilation by ap-
4. Journal of Geophysical Research - Space Physics. propriately accounting for model errors. Since the L* is 117 (A03222): 1. highly dependent on the magnetic field model utilized, we compared above assimilated results with that using an 3. Yu, Y., and A. Ridley. Exploring the influence of iono- empirical model Tsyganenko 2001 magnetic field configura- spheric heavy ion outflow on magnetospheric dynam- tion in converting the PSD data. The two assimilated radia- ics: dependence on the source location. To appear in tion belts show similarities in terms of capturing dynamic Journal of Geophysical Reserach. features in the storm, including the peak location, dropout, and recovery of the phase space density. By comparing the 4. Yu, Y., J. Koller, and S. Morley. Quantifying the effect of predicted flux from the data assimilation with flux observa- magnetopause shadowing on radiation belt electron tion from an independent satellite, we found that using the dropout. To appear in Journal of Geophysical Research. RAM-SCB magnetic field configuration for obtaining PSD
5. Rastaetter, L., and Y. Yu. Geospace Environment Model- data for the data assimilation provides better estimation ing 2008-2009 Challenge: Dst Index. To appear in Space of the radiation environment, implying that the RAM-SCB Weather. models a more realistic representation of the global mag- netospheric configuration.
6. Yu, Y., J. Koller, V. Jordanova, and S. Zaharia. On the L* neural network and its accuracy. Journal of Geophysi- Impact on National Missions cal Research. This program has built a tool to calculate particle drift shells in the inner magnetosphere. The computational 7. Yu, Y., J. Koller, V. Jordanova, S. Zaharia, and H. Godi- speed is at least six orders of magnitudes faster than nez. Radiation Belt Data Assimilation using RAM-SCB standard integration techniques. The tool has enabled us, Magnetic Field Configuration. Journal of Geophysical for the first time, to develop high precision space weather Research. data assimilation and forecasting models, which are an important component to NOAA Space Weather Prediction 820

Page 821

Publications Rastaetter, L., and Y. Yu. Geospace Environment Modeling 2008-2009 Challenge: Dst index. To appear in Journal of Geophysical Research. Yu, Y., J. Koller, S. Zaharia, and V. Jordanova. L* Neural Network from Different Magnetospheric Field Models. 2012. Space Weather. : S02014. Yu, Y., J. Koller, V. Jordanova, S. Zaharia, and H. Godinez. Ra- diation belt data assimilation using RAM-SCB magnetic field configuration. Journal of Geophysical Research. Yu, Y., J. Koller, V. Jordanova, and S. Zaharia. On the L* neu- ral network and its accuracy. Journal of Geophysical Research. Yu, Y., J. Koller, and S. Morley. Quantifying the Effect of Magnetopause Shadowing on Radiation Belt Electron Dropout. Journal of Geophysical Research. Yu, Y., V. Jordanova, S. Zaharia, J. Koller, J. Zhang, and L. Kistler. Validation Study of the Magnetically Self-Con- sistent Inner Magnetosphere Model RAMS-SCB. 2012. Journal of Geophysical Research - Space Physics. 117 (A03222): 1. Yu, Y., and A. Ridley. Exploring the Influence of Ionospheric Heavy Ion Outflow on Magnetospheric Dynamics: De- pendence on the Source Location. To appear in Journal of Geophysical Research. 821

Page 822

Physics Postdoctoral Research and Development Final Report “Listening” to the Noise of a Single Electron Spin: Ultra-sensitive, Non- perturbative Spin Noise Spectroscopy Scott A. Crooker 20110747PRD3 Abstract spin resonance is in marked contrast to conventional Measurement of electron and nuclear spins forms the nuclear- or electron-spin resonance measurements, basis of modern magnetic resonance technologies such which necessarily perturb the spins. Nevertheless, this as NMR, MRI, and electron spin resonance. Driven by approach is guaranteed by the Fluctuation-Dissipation the desire and need for ever-increasing sensitivity and Theorem. Importantly, noise-based measurements resolution (e.g., for MRI), the number of detected spins become viable alternatives to conventional (dissipative) necessarily decreases. The goal of this project is to reach methods as the few-particle limit is approached (N→1), the limit of single quantum-mechanical spins – that is, since noise signals fall only as sqrt(N) rather than as N. a single electron — and to measure the intrinsic fluc- Our scientific goal was therefore to develop an ultra-sen- tuations of that single spin. At present, however, the sitive magnetometer capable of “listening” to the tiny direct magnetization measurement of a single spin is a spin fluctuations of truly quantum-mechanical objects: tremendously difficult proposition, due to the extreme single electron spins trapped in semiconductor quantum sensitivity required (conventional NMR or MRI lacks the dots. New instrumentation for efficient digital signal sensitivity by about a factor of a trillion!). Nonetheless, processing based on configurable hardware (FPGAs) are from viewpoints of both fundamental measurement and essential to this work. basic science, single-spin detection represents a truly landmark achievement in measurement science. Recent Scientific Approach and Accomplishments advances in optical detection and efficient digital signal Yan Li, the postdoc supported by the project, has suc- processing, coupled with alternative measurement ap- cessfully demonstrated that our scheme for using lasers proaches based on passive spin noise detection (rather to sensitively measure spin fluctuations works — and than conventional perturbative studies), suggest a viable works very well — for electrons trapped in semicon- route that forms the basis of this work. ductor quantum dots. We completed our first set of measurements, wherein we showed that not only are Background and Research Objectives spin fluctuation signals measurable using the optical Not all noise in experimental measurements is unwel- techniques that we proposed, but that the data are of come. Certain fundamental noise sources contain ex- sufficient quality to achieve a detailed understanding of tremely valuable information about the system itself – a spin relaxation and coherence for the case of electron classic example being the inherent voltage fluctuations spins in semiconductor quantum dots. This initial work that exist across any resistor (Johnson noise), from which was written up and published in the high-profile journal temperature can be determined. In magnetic systems, Physical Review Letters (Yan Li et al), and the paper was fundamental noise exists in the form of small, random especially highlighted by the journal Editors. The refer- spin fluctuations. For example, statistical fluctuations of ence can be found in the reference section below. N paramagnetic spins should generate “spin noise” of order sqrt(N), even in zero magnetic fields. Crucially, the As a result of her data, our theoretical component has frequency spectra of these tiny fluctuations - if measur- made significant advances in our understanding of how able - reveal the important dynamic properties of the single electron spins couple to a bath of nuclear spins spins (such as spin precession frequencies, decoherence (directly addressing the “central spin problem” — a topic times, and interparticle couplings), without ever having of hot theoretical interest). This work was also submitted to excite, pump, or drive the spin ensemble away from to and was accepted at Physical Review Letters. thermal equilibrium. Such a noise-based approach to 822

Page 823

Again, as a result of Yan’s data, our close collaboration with in silicon nanowires through oxide tunnel barriers . the Dortmund and Bochum University groups of Professor 2013. Nano Letters. 13: 430. Manfred Bayer and Dirk Reuter continues. We have just published a joint theoretical/experimental paper on the use of these new experimental techniques as a high-reso- lution spectroscopic tool. Again, this paper was accepted at the prestigious journal Physical Review Letters (Zapasskii et al). In November 2012 Yan finished construction of the new ultra-stable single-quantum dot microscope, which works very well. Yan has shown that we can measure the absorp- tion from a single quantum dot, which is a remarkable achievement. She is currently measuring noise signals from this quantum dot – thereby achieving our goals. We are currently writing up these results for publication. Impact on National Missions This project will build Laboratory capabilities in the devel- opment of novel measurement methods that enable new scientific discovery at Los Alamos National Laboratory. It is based on recent technological advancements that enable the use of FPGAs and on the very large data storage and manipulation capabilities that have now become avail- able. Our basic approach to noise spectroscopy goes well beyond the specific example of electron spin noise that will be investigated here and the advantages of noise spectros- copy as an experimental probe are very general. The proj- ect will advance new Laboratory capabilities that underpin a wide variety of Laboratory missions. Demonstration of single spin noise spectroscopy measurements will gener- ate worldwide attention and will form a robust platform for nanoscale spintronics at LANL. The team will work with DOE program managers to develop a BES Materials Science funded program and to obtain DARPA funding for extend- ing this work. Publications Li, Y., N. Sinitsyn, D. L. Smith, D. Reuter, A. D. Wieck, D. R. Yakovlev, M. Bayer, and S. A. Crooker. Intrinsic spin fluctuations reveal the dynamical response function of holes coupled to nuclear spin baths in (In,Ga)As quan- tum dots . 2012. Physical Review Letters. 108: 186603. Sinitsyn, N. A., Y. Li, S. A. Crooker, A. Saxena, and D. L. Smith. Role of nuclear quadrupole coupling on deco- herence and relaxation of central spins in quantum dots . 2012. Physical Review Letters. 109: 166605. Zapasskii, V. S., A. Greilich, S. A. Crooker, Y. Li, G. G. Kozlov, D. R. Yakovlev, and M. Bayer. Optical spectroscopy of spin noise . 2013. Physical Review Letters. 110: 176601. Zhang, S. X., S. A. Dayeh, Y. Li, S. A. Crooker, D. L. Smith, and S. T. Picraux. Electrical spin injection and detection 823