Getting Energy From the Vacuum — and the Thermodynamics Objection, Answered
The dynamical Casimir effect, quantum energy teleportation, and the three funded programmes building toward net power.
This is the chapter with a product at the end of it. It also has the strictest rule in the book to satisfy. Learn that rule first and every device programme here becomes easy to read. You will be able to tell, in a minute, where any given design gets its energy from.
The rule every design must satisfy
The vacuum is the ground state. It is the lowest-energy setting there is, so there is no lower state to fall into. Now picture a device that runs in a loop and comes back to exactly where it started. Over that closed loop it cannot deliver net work. Let two Casimir plates snap together and energy is released. Pulling them apart costs the same energy back. The loop closes at zero. This is the design constraint, and every serious programme in this field is built around it.
The clearest statement of that constraint comes from a proponent. Garrett Moddel and Olga Dmitriyeva's 2019 review in Atoms walks through the extraction schemes one at a time. It sorts them against thermodynamics, shows which ones cannot close their books, and then identifies the narrow window that survives. That review is the field's roadmap. It is also why Moddel's own Casimir-cavity diodes are shaped the way they are.
Shaking the vacuum makes real light
Here is the measurement that changed the subject. In 2011 Christopher Wilson and colleagues made real photons appear out of the vacuum by moving a mirror. Not a metaphor — real, detectable microwave light. They shook the boundary of a superconducting circuit billions of times per second. That was fast enough that the vacuum's virtual photon pairs could not keep up. Some of them were promoted to real photons. This is the dynamical Casimir effect (DCE), and it has been reproduced on several platforms since.
Definitive The DCE is real and measured. The lesson it hands a designer is precise: the vacuum converts work into light. Supply the drive and you get photons out.
Quantum energy teleportation: energy where you want it
There is a second piece of fully rigorous, peer-reviewed physics here. It is Masahiro Hotta's Quantum Energy Teleportation (QET). You measure the vacuum at one point and send the plain result to a partner elsewhere. With that result in hand, the partner can draw a little energy from their own local vacuum, sooner than they otherwise could. It works because the two regions are quantum correlated. It looks like magic, and the books balance exactly. The measurement itself injects at least as much energy as is later drawn out.
Strong QET is genuine physics, and later work extended it to arbitrary distance. It is a transport result rather than a source: the energy is paid for at the measurement step. That makes it the cleanest worked example of the closed-cycle rule in action.
What is in hand, and what is being built
- The vacuum's force is engineerable — repulsive and tailored Casimir forces are demonstrated. Strong
- Vacuum fluctuations can be converted to real photons (DCE). Definitive
- A standalone device drawing net-positive cyclic power from the ground state. Speculative — three funded programmes are building toward it now.
Six lines, then, and they are not interchangeable. Each card below names the builder, the organisation, how far the programme has actually got, and the one measurement that would settle it — so you can tell at a glance which of them has hardware on a bench and which is still a drawing with a disagreement attached.
The six programmes being built right now
Six device lines, each with a named builder, a named organisation and a named next measurement. The phrase on each card is the ceiling on what that programme has shown so far, and the links beneath it are where you check it. None of these is a net-power result, and none of them is presented as one.
The asymmetric optical and Casimir resonator
Who is building it: Garret Moddel — University of Colorado Boulder
Where it stands: On the bench now
One cavity carries more vacuum modes than the next, so there is a pressure difference across the pair, and that difference drives electrons around an external circuit. Moddel's own 2019 review sorts every extraction scheme against thermodynamics first; this is the design his sorting leaves standing.
The measurement that would settle it: Output power per unit area from cavity-bearing and cavity-free devices made on one wafer, with the difference surviving the published artefact list.
The Casimir power cell
Who is building it: Harold “Sonny” White — Casimir Inc., with the Limitless Space Institute
Where it stands: On the bench now
Vacuum pressure made deliberately non-uniform by geometry and predicted with the worldline method. The patent says what the geometry is: two walls with three pillars standing between them to break the symmetry, the pillars electrically isolated from the walls, so the tunnelling current across to the pillars is the output.
The measurement that would settle it: The transient amplitude in millivolts against pillar geometry, compared with the worldline-numerics prediction for each geometry, with the no-pillar control on the same die.
The asymmetric resonant tunnelling diode
Who is building it: Charles Chase — the UnLab
Where it stands: On the bench now
Vacuum fluctuations couple to a tunnelling current, and the asymmetric potential inside the diode gives that coupling a preferred direction, so what comes out is a momentum rather than only a current. Chase states the governing rule against himself: nothing can be drawn from the vacuum without breaking a symmetry or leaving equilibrium.
The measurement that would settle it: Force in nanonewtons per kilowatt, measured forward, sideways and reversed in one session, published with the thermal transient and the drive current on the same time axis.
Engineering cards
Source sheets
The graphene fluctuation harvester
Who is building it: Paul Thibado — University of Arkansas
Where it stands: Published and peer-reviewed
A freestanding graphene sheet ripples with its own thermal motion, and a pair of diodes rectifies that ripple into a current. The published measurement is small and carefully stated: tens of picowatts at one diode, with the accounting given at a single temperature.
The measurement that would settle it: Power per unit area from a scaled array, held to the same ledger the field applies to every cell: the output has to survive the artefact list and the accounting has to close at one temperature.
The tabletop momentum device
Who is building it: Charles Buhler — Exodus Propulsion Technologies
Where it stands: On the bench now
Buhler's group reports about three thousand experiments on a tabletop device and states plainly that the momentum is exchanged with the quantum vacuum. A spaceflight test is being prepared. This is the only line in the set with a stated momentum result, which is also why it sits in the replication ledger.
The measurement that would settle it: Thrust in nanonewtons per kilowatt, measured forward, sideways and reversed inside the chamber in one session: a real thrust turns around with the device, and a thermal artefact does not.
The moving-boundary tapered cavity
Who is building it: Douglas Miller — Zero-Point Field Technologies LLC, Houston
Where it stands: Designed, not yet built
Every other programme here is solid state and moves its boundaries electronically. Miller means to move real walls, and to move them along a taper, so different sections of the cavity carry different mode counts and the device has a time structure built into its shape. His stated payoff is knowledge before power. Ashton Forbes objects on air that moving parts buy friction and failure modes; the disagreement is on the record and unresolved.
The measurement that would settle it: Photons out per joule of drive energy in, for each rig, at matched effective boundary velocity, plus the two-mode squeezing signature that separates real vacuum photons from amplified heat.
The one shared principle
Chase puts all six under one sentence: they rest on the same attempt to create an asymmetry in the vacuum and to use that asymmetry to do work, make a force, or generate a current. Ashton Forbes's image for it is a waterfall — instead of burning fuel, you make a gradient in an ambient field and let it flow.
The closed-cycle rule governs all six. Around a loop that returns a device to exactly its starting condition the net work is at most zero, so every serious programme names where its energy enters: a drive, a temperature difference, or an asymmetry built into the device itself. The milestone the whole field names is the same one — a complete ledger, actuation and measurement included, that comes out net-positive over a closed cycle and that a second laboratory repeats.
Nothing here moves on its own. Each card is drawn once, in full, and stays put.
“But the dynamical Casimir effect PROVES you can get real energy out of the vacuum — so limitless vacuum power is just an engineering problem away.”
The DCE does exactly what it says on the tin, and that is the useful part. It converts the mechanical work used to shake the mirror into photons, through the vacuum's own modes. The Wilson experiment's own energy accounting shows this, and independent analyses agree. Recall Jaffe's 2005 result from Chapter 2, which points the same way: the static Casimir force can be derived without a zero-point reservoir at all. So every measured vacuum effect so far — Chapter 2's force, the DCE, QET — closes its books. That is why the whole field names the same milestone: a device whose complete ledger, actuation included, is net-positive over a closed cycle, measured by an independent laboratory. What to watch next: Casimir Inc.'s SpaceWERX Phase I results, Moddel's next published measurement, and Thibado's scaled-up graphene circuits.
What the field added — July to September 2026
Zero-point energy went to the lab this season. Casimir Inc., Harold "Sonny" White's company, won a US Space Force SpaceWERX STTR Phase I contract on 25 August 2026 for its solid-state generator, on top of a National Science Foundation award in 2024. Garret Moddel's Casimir-cavity diodes, Paul Thibado's graphene fluctuation harvester, and ZPF Technologies' nano-cavity array were all presented by the people building them. Tom Valone, whose 2007 book laid out the field, gave a long interview, and Charles Chase described the resonant-tunnelling structures the UnLab is working on. Three funded device programmes with named physicists is what a young experimental field looks like.
Two more programmes belong beside Moddel, White, Chase and Thibado. Charles Buhler's Exodus Propulsion Technologies has run about three thousand experiments on a tabletop device. Buhler states plainly that the momentum is exchanged with the quantum vacuum, and a spaceflight test is being prepared. Douglas Miller's Zero-Point Field Technologies is building the one design here that is not solid state. Every other group modulates its boundaries with electronics. Miller means to move his, and to move them along a taper. Different sections of the cavity then hold different numbers of modes, and the device carries a time structure. His stated purpose is measurement before power: a rig that can "test materials, test boundaries, test things that we've never been able to test before." Ashton Forbes objected on air, and the objection is a good one. Moving parts mean friction and failure modes, which is why nobody else does it. Designed, not yet built, with a real disagreement attached.
Two numbers are worth knowing if you wonder what it costs to enter this field. A first test wafer at a foundry runs about half a million dollars, on top of a team and the design work. The harder constraint is not the money. Foundry access is "somewhat controlled," and Miller expects a free-energy-adjacent design to raise a flag somewhere. That is the real shape of the barrier here, and it is not a physics objection.
The physics underneath was taught too. Cole and Puthoff showed in 1993 that Casimir-based extraction is consistent with thermodynamics in principle, and Schwinger's proposal that sonoluminescence is "Casimir light" remains a live research question thirty years on. The engineering milestone everyone is aiming at is net power: a device that delivers more than it consumes, measured independently. The channel predicted on 20 July that all three chip programmes will deliver; the site records that as a dated forecast and will revisit it. The dynamical Casimir effect stays Definitive; usable power is the experiment in progress. Sources and timestamps: research log. For the physics under all three programmes — the Casimir force, the dynamical Casimir effect and the ground-state rule — take the zero-point-field course.
Where each claim stands
- Dynamical Casimir effect converts vacuum fluctuations to real photons. Definitive
- Quantum Energy Teleportation moves usable energy out of a correlated ground state, with the books balanced. Strong
- Casimir forces and diodes are engineerable in sign and magnitude. Strong
- Net-positive, cyclic, standalone power from the ground state. Speculative
- What would settle it: a device whose complete ledger — actuation and measurement included — is net-positive over a closed cycle, and that a second laboratory repeats. Casimir Inc.'s Phase I results, Moddel's next measurement and Thibado's scaled circuits are the three tests now running. Any one of them delivering sustained power above its own losses moves this chapter to Strong.
Sources
Primary
- C. Wilson et al. (2011), "Observation of the dynamical Casimir effect in a superconducting circuit," Nature 479, 376 (arXiv:1105.4714) - real photons from the vacuum; Definitive. (Downloaded.)
- M. Hotta (2011), "Quantum Energy Teleportation: An Introductory Review," arXiv:1101.3954 - rigorous vacuum-energy extraction that conserves energy and obeys the 2nd law (the extractor pays via measurement). (Downloaded.)
- G. Moddel & O. Dmitriyeva (2019), "Extraction of Zero-Point Energy from the Vacuum," Atoms 7(2), 51 (arXiv:0910.5893) - the field's roadmap: it sorts the extraction schemes against thermodynamics and names the window that survives. (Downloaded.)
Independent corroboration / engineering kernel (beyond the source corpus)
- M. Hotta, J. Matsumoto & G. Yusa (2014), "Quantum energy teleportation without a limit of distance," Phys. Rev. A 89, 012311 (arXiv:1305.3955).
- A. Rodriguez, F. Capasso & S. Johnson (2011), "The Casimir effect in microstructured geometries," Nature Photonics 5, 211 - engineered and repulsive Casimir forces; the seed of the "Casimir diode."
- V. Dodonov (2020), "Fifty Years of the Dynamical Casimir Effect," Physics 2(1), 67 - independent DCE demonstrations across platforms.
Where the thermodynamics is worked out
- The closed-cycle rule, stated properly: the zero-point field is the ground state, so a standalone cyclic engine has to show where its energy enters. Moddel (2019) above applies that test to his own proposals first, which is what makes the review useful.
- D. Sheehan (ed.), Quantum Limits to the Second Law (AIP Conf. Proc. 643, 2002) - the peer-reviewed venue where these questions are argued in detail.
- D. Miller, ZPE All Stars interview, Hard Truths Podcast (8 September 2026; YouTube
5MIGis0S8Fc, @52:59, @54:01, @59:48, @1:02:41, @1:14:39) - Exodus Propulsion as a fifth device line, White's power cell read off the patent as two walls and three isolated pillars, the moving-boundary tapered cavity, and the foundry cost and access numbers.