Stellar-Relevant Emission-Based Opacity Experiments at the Orion Laser Facility HEDS Seminar Series Madison E. Martin on behalf of Orion team October 21, 2021 LLNL-PRES-828262 Lawrence Livermore National Laboratory LLNL: — Greg Brown (PI, experiment lead) — Mark Foord (theory lead) — Madison E. Martin (design lead) — Ronnie Shepherd — Mike MacDonald — Klaus Widmann — Antonia Hubbard — Carlos Iglesias — Daniel Aberg — Michael Kruse — Duane Liedahl — Paul Grabowski — Rich London — Joe Nilsen — Mehul Patel — Dylan Cliche — Howard Scott — Bob Heeter — Heather Whitley AWE: — Dave Hoarty (PI) — J. Morton — S. Richardson — Colin Brown — B. Foote LLNL-PRES-828262 • Motivation • Experimental Platform • Recent Fe Campaign • Modeling Approaches — Radiation Hydrodynamics modeling — Atomic kinetics and radiation transfer calculations • Ongoing & Future Work • Summary LLNL-PRES-828262 • Radiation hydrodynamic design calculations rely on energy transport models that include material opacities as either theoretically based formulas or tables • High energy density (HED) systems can include a large range of ionization stages which may require statistical methods • Experimental validation of opacity theory is needed • Disagreements between solar observations and models may be mitigated by increased opacity [1] - Bailey et al. [2]: measured iron opacity values 30-400% larger than predicted We are interested in opacities at plasma conditions equivalent to the upper radiative region of the Sun (plasma conditions in standard solar model, Bahcall, et al., 2005): region of interest T = 200 - 400 eV, n_e = 10^23 - 10^24 cm^-3 (10^-1 to solid density). Figure adapted from London et al., APS, DPP, 2013. [1] Bahcall et al. The Astro. J., 2005 [2] Bailey et al., Nature, 2015 • Laser -> hot electrons -> heat target -> x-ray emission • High T at solid density • Measure specific intensity (I_nu) • Dopant (eg. Mg) used to infer T and rho from emission • Infer opacity (kappa_nu): kappa_nu = - ln(1 - I_nu / B_nu(T)) / (rho * Delta_l) • We are currently focused on directly comparing simulated emission spectra with measured emission spectra (I_nu) - explore sensitivities to modeling assumptions - understand plasma evolution Target setup: ~ 1 ps laser pulse, I ~ 10^18 W/cm^2 incident on low-Z Tamper (CH) surrounding buried layer (e.g. Fe with K, Cl, S dopant) observed by x-ray spectrometer. Solution to Rad. Transport Equation in 1D for thin slab: I_nu = (eta_nu / kappa_nu) * (1 - e^(-kappa_nu * rho * Delta_t)) LTE: B_nu(T) = eta_nu / kappa_nu Solve for opacity: kappa_nu = - ln(1 - I_nu / B_nu(T)) / (rho * Delta_l) Figure adapted from R.A. London and J. I. Castor, HEDP, 2013. Related papers: Hoarty et. al, HEDP, 2007 & 2013 & 2017 Orion provides access to a portion of phase space that is complementary to other facilities. • Orion is a combined long and short pulse facility delivering: — 10 long-pulse beams (500J each, 0.1-10ns, 335nm) — 2 short-pulse beams (500J each, 500fs, 1054nm) • 1 short-pulse beam has been converted to green operation to increase pulse contrast (200J, 500fs, 527nm) • 200 J is achieved using two independent doubling crystals LLNL-PRES-815223 Characterized using: • PIXE (proton induced X-ray emission) • EDX (electron beam) • Profilometry Metrology results for FeS layer: • layer thickness ( 174 ± 12 ) nm • composition: Fe 60.2 at%, S 39.8 at% • areal density: o Fe : 61 ± 4 µg cm^-2 o S : 23 ± 2 µg cm^-2 Target Parameters: • Parylene-N: 3 µm • FeS: 0.16 µm (50 µm diameter) • KCl: 0.06 µm • C: 0.015 µm • Parylene-N: 3 µm • Substrate: 2 mm x 2 mm Laser Parameters: • Delta_t: 1 ps • energy: ~ 165 J • Focus: 100 µm • Irradiance: ~ 2 x10^18 W/cm^2 LLNL-PRES-815223 For these shots, we fielded: — 2 time-resolved spectrometers: • Temperature inference — 4 time-integrated spectrometers: • Emission of interest (Fe L-shell) • Temperature & Density — 1 Pinhole Camera (GXD): • Uniformity Diagnostic Overview Table: • AXIS 608 | Crystal: RAP | Filters: 8 µm Be | Sees: Fe L | View Angle: 25° | Front/Back: F • AXIS 609 | Crystal: CsAP | Filters: 16 µm Be | Sees: S K | View Angle: 65° | Front/Back: B • OHREX-I | Crystal: Quartz | Filters: 51 µm Be | Sees: Fe K | View Angle: 19.3° | Front/Back: F • OHREX-I | Crystal: KAP | Filters: 51 µm Be | Sees: Fe L | View Angle: 19.3° | Front/Back: F • HBS | Crystal: CsAP | Filters: 8 µm Be | Sees: Fe L | View Angle: 70° | Front/Back: F • MK II | Crystal: CsAP | Filters: 16 µm Be | Sees: S, Cl He_alpha, Ly_alpha, S beta-lines | View Angle: 76° | Front/Back: F • Titan | Crystal: PET | Filters: 16 µm Be | Sees: S, Cl, K k-shell | View Angle: 76° | Front/Back: B • GXD | Crystal: ... | Filters: 8 µm Be | Sees: Imaging | View Angle: 20° | Front/Back: B LLNL-PRES-815223 • Crystals: — AWE: X-ray tube/Excaliber w/ double crystal spectrometers — LLNL: EBIT w/ quantum calorimeter (ECS) • Filters: — LLNL: EBIT with ECS — AWE & LLNL: Profilometry • Image Plate: — AWE: X-ray Tube/Excaliber sources LLNL-PRES-815223 HYDRA [1]: • Parametric heating source • Target and laser parameters • Opacity model • Radiation transport methods • Electron conduction model -> Output: evolution of plasma conditions, post-process for simulated emission Cretin [2]: • Single pair of plasma conditions • Time dependent atomic kinetics • Optical depth effects • Line shapes • Radiation transfer effects -> Output: Simulated emission, Ionization, charge state distributions [1] M. Marinak et al., POP, 1996, 1998, & 2001 [2] H. A. Scott, JQSRT 71, 689 (2001) Design parameters: • laser energy • laser pulse shape • target dimension HYDRA [1]: • Opacity model (LTE versus non-LTE) • Radiation transport model • Electron conduction model YORICK [2] Analysis: • Calculate radiation transport along a specified ray Observe effect on: • Temperature • Density • Emission • Opacity Martin et al., POP 24, 022705 (2017) Martin et al., HEDP 26, 26-37 (2018) • Assume 1-Dimensional geometry • Opacity is modeled using HYDRA DCA package — Relies on super-configuration-based atomic models • Energy source — Do not model laser rays or hot electrons — Deposit internal energy (J/g) uniformly in target and proportional to laser pulse (conversion efficiency) — 1 ps Gaussian pulse: t_bar >= 1.5 * FWHM Martin et al., POP 24, 022705 (2017) Plot comparing electron temperature vs time (0 to 20 ps): • AXIS 609 fit w/FLYCHK (AWE) • AXIS 609 fit w/SCRAM • HYDRA-DCA 1.4 keV peak • HYDRA-DCA 1.1 keV peak Internal FY21 Milestone Report: LLNL-TR-820896 • Nonlocality parameter (Kn) is used to assess whether non-local transport is important: Kn = lambda_c / L_T L_T = T / |grad T| (electron temperature gradient length scale) lambda_c (thermal collisional mean free path) • HYDRA includes a nonlocal electron transport package to model electron thermal conduction for strongly driven plasmas — Extension of Schurtz, Nicolai, and Busquet (SNB) non-local thermal transport model [1] • Comparison of Baseline (Lee & More conduction) vs SNB across time (0-40 ps) against threshold for potential importance. [1] Schurtz et al. POP 7, 4238 (2000) M. Patel (2020) Comparison of Te (keV) vs Time (ps) for FeS and KCl layers (1.5 to 3.5 ps): • Baseline vs SNB • Average (solid lines) and Max/Min (dashed lines) plotted • Shows non-local transport predicts slightly lower peak temperature and wider range of conditions across layers. M. Patel (2020) • Average Te in FeS Layer: time-dependent NLTE atomic kinetics vs steady-state NLTE atomic kinetics (0 to 20 ps) • Emission at time of peak Te in FeS Layer (1 degree from target normal, 2.40 - 2.70 keV), showing S He-alpha and S Ly-alpha lines. • Average S Charge State: 15 (time dependent) vs 15.1 (steady state) • Takeaway: Assuming steady-state NLTE atomic kinetics has a small impact on average Te in FeS Layer but slightly increases S ionization. • Average Fe Charge State in FeS Layer vs time (0 to 30 ps): peak average Te ~1.4 keV at 2 ps comparing non-LTE and LTE. • Shot S11219 Data and Models (1.000 - 1.200 keV): time-integrated Fe L-shell spectrum comparing non-LTE, LTE, and HBS data. • Takeaway: Both non-LTE and LTE models have reasonable agreement with the time-integrated Fe L-shell data. Comparison of experimental temperature inference (AXIS 609 fit w/ FLYCHK, AXIS 609 fit w/ SCRAM) to HYDRA-DCA 1.4 keV and 1.1 keV simulations over 0-20 ps. Internal FY21 Milestone Report: LLNL-TR-820896 • Time-integrated Fe L-shell (1.000 - 1.200 keV): Comparison of peak ~1.1 keV model, peak ~1.4 keV model, and HBS data. • Time-integrated Fe K-shell (6.63 - 6.72 keV): Fe j, Fe r,q,t, Fe y, Fe m,t, and Fe w lines compared between peak ~1.1 keV, peak ~1.4 keV, and OHREX data. • Takeaway: Both models have reasonable agreement with Fe L-shell but overpredict the Fe K-shell emission. • Time-integrated K-shell (2.4 - 3.2 keV): MKII data vs models showing S He-alpha, S Ly-alpha, Cl He-alpha, Cl Ly-alpha, S He-beta, S Ly-beta. • Time-integrated K-shell (2.8 - 3.6 keV): Titan data vs models showing Cl He-alpha, Cl Ly-alpha, S He-beta, S Ly-beta, Cl He-beta, K He-beta. • Takeaway: Both models underpredict S He-alpha and S Ly-alpha, but the higher temperature model overpredicts Cl He-alpha, Cl Ly-alpha, and K He-beta. Plasma Conditions: • Fixed density, Te = 0.8 - 1.4 keV • 1D slab of FeS (areal density conserved) • Less expensive and not dependent on assumed heating in rad-hydro Cretin [1]: • No radiation transfer • With radiation transfer • Optical depth effects • Ti effects Output: • Ionization • Charge state • Emission Compare with Data and other Spectral Codes: • Measurement • SCRAM (used for Te inference) • ENRICO (non-LTE code) [1] H. A. Scott, JQSRT 71, 689 (2001) • Average ionization <Z> vs Te (0.8 - 1.4 keV) for FeS rho=2 g/cc: Fe (Be-like, Li-like, He-like) and S (He-like, H-like, bare), comparing with RT and without RT. • S ion fractions vs Te (0.8 - 1.4 keV): Fractions of Li-like, He-like, H-like, and bare S comparing with RT and without RT. • Takeaway: RT = radiation transfer. Radiation transfer slightly affects average ionization but noticeably changes individual charge state concentrations. R. London (2020) • Fe He-alpha region for Te = 1.2 to 1.8 keV (E - 6600 eV, 40 to 120 eV): Shows strong temperature sensitivity of Li-like satellites, He-like y, and He-like w lines (optical depth at line center ~10). • OHREX spectrum of Orion shot 11219 OHREX showing measured Li-like sat., He-like y, and He-like w lines. R. London (2020) • Effect of layer thickness on Fe He_alpha lines: Comparing data to Cretin calculations with nominal thickness multipliers (1.0, 0.15, 0.1) at Te = 1.8 keV. • Effect of Ti on y line profile: Comparing data to Ti = 1.2 keV and 1.8 keV at Te = 1.8 keV and areal density 15% of nominal. R. London (2020) Best Cretin simulation vs S11219 OHREX data (E - 6600 eV): • Parameters of Cretin fit: Te = 1.8 keV, Ti = 1.2 keV, rho*Delta_z = 15% of nominal. • Simulated line positions shifted to match data (showing matched Y and W lines). R. London (2020) Closed-loop modeling workflow: • Assumed energy deposition (parameterized in Time and Spatial dimension(s)) • -> HYDRA [1]: Predict plasma conditions • -> Cretin [2]: Simulate multiple diagnostic signals • -> Compare Measured and Synthetic Data (Emission, Opacity) • -> Refine assumed energy deposition. [1] M. Marinak et al., POP, 1996, 1998, & 2001 [2] H. A. Scott, JQSRT 71, 689 (2001) • Modeling and Design: — Complete predictions and analysis of FY22 campaign using current HYDRA and Cretin models [Martin et al.] — Extend HYDRA models to include 2D effects [R. London, et al.] — Replace ray trace with post-processing using Cretin for simultaneously constraining plasma evolution and atomic kinetics by comparing simulated and measured spectra [D. Cliche, et al.] — Improve understanding of sensitivities of the Orion platform using our best available models [Martin, et al.] • Experimental: — Continue to improve calibration of spectrometers [Brown, MacDonald, Shepherd] — Complete software development to more efficiently reduce data and aid with preliminary analysis during future FY22 campaign [MacDonald] — Continue to develop multiple-temperature inference algorithm [MacDonald and Liedahl] — Complete installation and testing of STOHREX to provide time-resolved density diagnostics [Brown, et al.] • Theory: — Complete predictions and analysis of FY22 campaign using NLTE models (eg. SCRAM, ENRICO, Cretin) [Foord, et al.] — Improve accuracy, convergence, and speed of ENRICO [Foord, et al.] — Generate NLTE tables as part of Autonomous Multiscale SI [Gaffney et al.] • Experimental validation of opacity theory is needed for HED conditions • Short pulse lasers can be used to conduct opacity measurements • We have conducted Fe emission experiments at AWE's Orion Laser Facility • A previously developed 1D HYDRA methodology was used to study sensitivities to radiation transport, non-local electron transport, and assumed atomic model • Cretin was used to investigate radiation transfer and plasma conditions on ionization and emission • We have only discussed a small fraction of the ongoing work associated with this campaign and milestone LLNL-PRES-828262 This document was prepared as an account of work sponsored by an agency of the United States government. Neither the United States government nor Lawrence Livermore National Security, LLC, nor any of their employees makes any warranty, expressed or implied, 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 product, 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 Lawrence Livermore National Security, LLC. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States government or Lawrence Livermore National Security, LLC, and shall not be used for advertising or product endorsement purposes.