This week, a company called Casimir Inc. emerged from “stealth mode” to announce that it had raised significant funding from venture capitalists willing to roll the dice on free energy. That’s right: a startup has gotten serious backing to develop sources of perpetual free energy. The people behind this fantastic new energy generator also brought us the wildly successful WTF thruster EM-drive that could supposedly directly convert electricity into a propulsive force. (Its one practical application was in the show Salvation, where it was treated with the same detailed attention to physical laws as Galaxy Quest’s Omega-13.) With that success, who are we to be skeptical? Casimir Inc. is convinced it can squeeze energy from the vacuum via the Casimir force (hence the subtle reference in the name). The Casimir force is a real thing, arising from the fact that a vacuum is not actually nothing. Instead, it is filled with a froth of virtual particles becoming real in pairs, waving to us, annihilating each other, and sinking back into the soup of virtual particles. The Casimir force emerges when we create an imbalance in the spatial distribution of these virtual particles, leading to a pressure as the Universe seeks to equalize the distribution. To get to the heart of the Casimir force, we need to talk about modes—one of my favorite concepts (and my chosen instrument of torture for undergraduate students). Put simply, a mode describes how a photon can spread out and occupy a space. The larger the space, the more ways a photon can occupy that space. The Universe is big, so it has an enormous number of modes. But if you create a confined space within the Universe, such as the gap between two closely spaced metal plates, there are only a few modes available, and they are less likely to be occupied by a particle. Between the plates, we have no particles; outside the plates, we have particles. This excess of particles, as they bounce off the plates, drives them together. This is the Casimir force. And, yes, a force means energy is involved, so in principle, you could extract a small amount of energy from the plates moving together. You would immediately lose all that energy to move the plates apart again if you want to repeat the process. To get a continuous flow of energy, according to Casimir Inc., the setup needs to be slightly different. The plates are fixed so they can’t move together, and a row of pillars is placed between the plates. The plates and pillars are then connected via the device that will be powered (a load). From here, things get hazy because the details are missing—or at least highly obfuscated. The idea is that electrons will tunnel from the plates to the electrodes but not tunnel in the reverse direction. Tunneling is the quantum process by which a particle can pass through a barrier to another region of space. Again, modes are involved: The electron is in a mode in one place and will occupy a mode in its new location. All else being equal, the chance of tunneling forward is the same as tunneling backward because the overlap between the modes is symmetrical. This will not generate energy. But there are a couple of ways to make tunneling asymmetric (and hence might work). The one proposed in the paper by Casimir Inc. is that the structure it has designed somehow has modes between the plates that follow the same rules as a hydrogen atom, and the tunneling involves an electron going from a high-energy state to a low-energy state (at least in hydrogen, these would represent different energetic states). If this is the case, one would expect a net flow of electrons from high energy to low energy via tunneling. Unfortunately, although the math in the paper looks fine (I’ve not checked it in detail), the underlying assumption—that the modes are identical to those in a hydrogen atom—seems quite shaky. The more likely route is the same as that used in quantum cascade lasers. In these systems, electrons tunnel from one location to another with almost the same energy. But in their new location, they quickly lose energy (via an acoustic wave generated by the crystalline material that holds the electrons), which traps them there. This is an actual mechanism based on well-established physics. It does, however, rely on very specific material properties and precise structural engineering. Given that, it’s probably not what would happen in the proposed device, either. Nevertheless, the company claims to have made a device and measured a drop in voltage between the plates and pillars. The company also claims this voltage is predicted in a paper that doesn’t appear to have any predictions, which is necessary for success. I would be shocked if Casimir, Inc. had not measured a potential difference. For a decade, surfaces of materials were the bane of my existence. Surfaces are not simple and can exhibit all sorts of weird properties due to missing atoms, crystalline boundaries, and impurities from fabrication techniques. If Casimir, Inc.’s team chose the right metal and the pillars were thin enough, they might even have fully oxidized on exposure to air, making them very different from the plates next to them. All of these material properties play a role in generating the potential measured by a probe, independent of any special Casimir forces or vacuum fluctuations. But let’s give the company the benefit of the doubt and assume it will observe (or already has observed) an electron flow from the plates to the pillars due to the Casimir force—it’s not impossible. These electrons still need to be coaxed through a load where they can give up their energy. That means connecting the pillars and plates to wires, each of which will introduce a potential difference due to the point of contact between different metals. To overcome that potential difference, charge will have to accumulate in the pillars. This will reduce the potential difference between the pillars and the plates and slow the flow of the tunneling current. Eventually, the whole charge pump will grind to a halt, leaving no current flow. In other words, I expect that no useful energy will be extracted. But I do value the company’s service in burning through a bunch of VC money. Supposedly related paper: Physical Review Research, 2026: 10.1103/l8y7-r3rm