The second law of thermodynamics is the law that says usable energy runs down: heat spreads, gradients fade, and no cyclic machine can turn ambient heat entirely back into work. Prof. Daniel P. Sheehan, Professor of Physics at the University of San Diego, has held a long-running challenge to that law does not begin with energy from nowhere, but with a more surgical question: are there special systems — surfaces, membranes, superconductors, junctions, cavities, plasmas, and statistical ladders — that can organize ordinary thermal motion into work in ways standard thermodynamics forbids? The following list is a guided tour of twenty-five experiments, thought experiments, device proposals, and research programs that Sheehan has discussed as part of the modern challenge to the second law.

From Maxwell’s Demon to Thermodynamic Zombies

Professor Daniel Sheehan’s work belongs to a larger tradition that began with Maxwell’s demon, the famous 19th-century thought experiment in which a tiny being sorts molecules to create a pressure or temperature difference. Most physicists believe the traditional demon is defeated by information costs: the demon must observe, remember, and erase information, and that erasure restores the entropy balance. Sheehan’s preferred descendants are different. He calls them “Maxwell zombies” because they do not think, remember, or decide. They are passive physical systems that sort through their material properties.

The key idea is boundary asymmetry. In many ordinary thermodynamic treatments, surfaces and interfaces are treated as secondary. Sheehan argues that this can be misleading. Real physical systems have surfaces, membranes, depletion regions, Debye sheaths, catalysts, optical cavities, and chemical gradients. These boundaries can store free energy, impose asymmetries, and sometimes create macroscopic gradients from microscopic thermal motion.

None of the items below should be read as a settled overthrow of the second law. Some are thought experiments. Some are proposed tests. Some are reported laboratory anomalies. Some are theoretical device concepts. Some have been criticized or remain unresolved. Their importance is that they define the experimental frontier: what would it take to show that the second law is not absolute, but instead a powerful approximation with limits?

In the list below, the title of each item links to the best public resource I could find: a journal article, open repository page, public explainer, patent-related page, or video/event source. The paragraph below each title explains what the experiment or concept is and why Sheehan treats it as relevant to second-law challenges.

The 25 Experiments That Defy Thermodynamics

1. Maxwell’s Demon — the original molecular-sorting challenge. Maxwell’s demon is the ancestor of nearly every later second-law challenge. In the classic thought experiment, a tiny being opens and closes a door between two gas chambers, letting faster or slower molecules pass in selected directions until a temperature or pressure gradient appears. It matters because it shows the heart of the problem: if microscopic motion can be sorted, thermal disorder can become useful work. Modern physics usually rescues the second law by charging the demon an information-erasure cost, but Sheehan argues that this answer does not automatically apply to passive “zombies” that do no measurement or memory processing.
2. Duncan’s Radical Radiometer — a chemical pressure zombie. Todd Duncan’s comment on Sheehan’s pressure-gradient work imagined a radiometer-like device inside a sealed blackbody cavity, with opposing vane faces that interact differently with a dissociating gas. One surface would split molecules; the other would recombine atoms. This could create a persistent pressure difference capable of turning a vane and doing work. It matters because it turns Maxwell’s demon into a passive surface-chemistry problem: no tiny intelligence, no trapdoor decisions, just surfaces that sort molecules by their normal chemical behavior.
3. Duncan’s Temperature Zombie — a heat-engine version of the paradox. The temperature version of Duncan’s paradox replaces the pressure imbalance with a temperature imbalance. One surface cools because it dissociates molecules and absorbs bond energy; the other heats because it recombines atoms and releases bond energy. If that temperature difference persists inside a single-temperature environment, it can drive an ordinary heat engine. It matters because it aims directly at the Kelvin–Planck statement of the second law, which forbids a cyclic device whose only net effect is converting heat from one reservoir into work.
4. Surface-Ionized Q-Plasma Experiments. In surface-ionized plasmas, atoms or molecules interact with hot surfaces and emerge ionized in ways that depend on the surface’s work function. Sheehan has pointed to Q-plasmas as systems where ion populations and velocity distributions can be surface-specific and non-Maxwellian under conditions that resemble blackbody environments. It matters because these experiments support one of his central claims: surfaces can maintain stationary nonequilibrium features that bulk equilibrium thermodynamics would tend to erase.
5. High-Temperature Plasma Differential-Current Experiments. In Sheehan’s early plasma-based challenges, a blackbody cavity contains a low-density plasma and a probe or boundary arrangement that permits a differential current between materials at the same nominal thermal environment. The proposed paradox is that a plasma potential or surface-specific ionization behavior can generate an asymmetric current or momentum flux that can be routed through a load. It matters because it was one of the early experimentally grounded examples of Sheehan’s broader template: a boundary asymmetry creates an exploitable reservoir of free energy.
6. Chemically Maintained Steady-State Pressure Gradients. Sheehan’s 1998 Physical Review E paper proposed that low-density gases in sealed blackbody cavities could sustain pressure gradients if chemically active surfaces desorb different species at different rates. Standard equilibrium reasoning says a gas in a closed isothermal cavity should not maintain such gradients, except for familiar cases like gravity. It matters because this paper provided the theoretical basis for Duncan’s later radiometer objection and for the whole epicatalytic line of second-law experiments.
7. Gas-Filament Hydrogen Dissociation Tests with Tungsten and Rhenium. In these tests, tungsten and rhenium filaments were heated in hydrogen, helium, and vacuum while the required electrical power was measured. Helium served as an inert control, while hydrogen allowed dissociation and recombination chemistry. Rhenium reportedly required more power to maintain the same temperature, indicating stronger hydrogen dissociation. It matters because this experiment established the chemical asymmetry used to predict the later tungsten–rhenium blackbody-cavity temperature difference.
8. Tungsten–Rhenium Blackbody-Cavity Duncan Paradox Experiment. This is the central laboratory experiment in Sheehan’s public case. Tungsten- and rhenium-coated thermocouples were placed inside a high-temperature blackbody cavity made mostly of tungsten or rhenium and exposed to vacuum, helium, and hydrogen. In vacuum and helium, the thermocouples agreed; in hydrogen, they reportedly diverged by more than 120 kelvin under some conditions. It matters because it was presented as an experimental realization of Duncan’s temperature paradox: two surfaces in a nearly single-temperature cavity maintaining different temperatures.
9. Complementary Rhenium-Cavity Sign-Reversal Test. The sign-reversal test repeated the blackbody-cavity experiment with the dominant cavity material reversed. In a tungsten-dominated cavity, the rhenium-coated thermocouple reportedly cooled relative to tungsten; in a rhenium-dominated cavity, tungsten reportedly heated relative to rhenium. It matters because sign reversal is harder to dismiss as a fixed thermocouple offset or wiring artifact. The result follows the proposed chemical logic: the dominant cavity surface sets the hydrogen/atomic-hydrogen environment, and the minority surface reacts against it.
10. Epicatalytic Thermal Diode. The epicatalytic thermal diode is the device concept that grows out of Duncan’s temperature zombie. Two chemically distinct surfaces interact with a gas so that one side cools through dissociation and the other warms through recombination, creating a sustained temperature difference. In principle, that gradient could power a thermoelectric generator or heat engine. It matters because it is the engineering bridge from “we measured a temperature anomaly” to “we might build a heat recycler.”
11. Room-Temperature Epicatalysis with Formic Acid, Methanol, Teflon, and Kapton. High-temperature hydrogen-metal experiments are difficult to commercialize, so Sheehan’s group also looked for room-temperature epicatalysis. In screening experiments, polymer surfaces such as teflon and kapton were tested with gases including formic acid and methanol, and small but reproducible differences in dimer desorption were reported. It matters because room-temperature epicatalysis would be far more relevant to practical energy devices than refractory metals glowing near 2,000 kelvin.
12. Iodine–Metal Epicatalysis Tests. Sheehan’s American Scientist article briefly mentions work by David Miller suggesting results consistent with epicatalysis in iodine–metal systems. The public description is limited, so this should be treated as a reported supporting line rather than a fully documented experiment. It matters because it suggests that epicatalysis, if real, may not be limited to hydrogen-metal or formic-acid/polymer systems; it may be a broader gas-surface phenomenon.
13. Solid-State Maxwell Demon / Open-Gap p–n Junction. The solid-state Maxwell demon uses an open-gap p–n junction. A conventional p–n junction has a built-in potential in its depletion region; Sheehan and collaborators proposed a geometry in which that potential appears across a vacuum gap, creating a charged-capacitor-like electric field. It matters because the energy stored in that field is argued to originate from thermal carrier motion. If the field can be cyclically discharged and restored, the device becomes a solid-state second-law challenge.
14. Hammer-and-Anvil NEMS/MEMS Oscillator. The hammer-and-anvil oscillator is the proposed work-extraction stage for the open-gap p–n junction. A movable silicon element is pulled toward an opposing plate by an intrinsic electric field, discharges on contact or near-contact, springs back, and recharges. If the mechanical and electrical time constants are matched, the device could oscillate. It matters because a static field alone is not enough; the second-law challenge requires a repeatable cycle that extracts mechanical work.
15. Supradegeneracy and Equilibrium “Population Inversion”. Supradegeneracy is a statistical-mechanics proposal by Sheehan and Larry Schulman. In ordinary equilibrium, high-energy states are less populated because the Boltzmann factor suppresses them. But if the number of available states increases rapidly enough with energy, degeneracy can overcome that suppression, creating something like population inversion at equilibrium. It matters because it challenges the intuition that equilibrium always favors lower-energy occupation in the way simple textbook examples suggest.
16. Superdegenerate Thermophotovoltaic. The superdegenerate thermophotovoltaic applies supradegeneracy to photovoltaic and thermophotovoltaic devices. Instead of requiring a single high-energy photon to lift an electron across a band gap, a ladder of dopant states could allow many small thermal steps upward. It matters because, if successful, the device would turn low-grade thermal radiation into electrical work in a way that appears to challenge the usual thermodynamic limits on converting ambient heat into useful electricity.
17. Asymmetric Membrane Concentration Cell / Thermal Battery. The thermal battery uses a chemically asymmetric membrane to separate ions into different concentrations, then uses an ordinary concentration cell to generate electricity from that gradient. The concentration cell itself is conventional; the contested element is the membrane’s claimed ability to regenerate the gradient through anisotropic diffusion driven by ambient thermal motion. It matters because it is one of the most concrete, room-temperature, device-like examples in Sheehan’s recent work.
18. Nikulov’s Asymmetric Superconducting Rings. Alexey Nikulov’s superconducting-loop work concerns persistent currents, resistance, and voltage effects in mesoscopic superconducting rings, often near the superconducting transition. The proposed challenge is that thermal fluctuations and quantum constraints may produce directed electrical behavior in systems that appear to be at equilibrium. It matters because it represents a low-temperature quantum route to second-law challenges, distinct from Sheehan’s chemical, plasma, and membrane systems.
19. Peter Keefe’s Coherent Magnetocaloric Superconducting Engine. Keefe proposed a second-law challenge involving Type I superconducting particles undergoing magnetic-field-induced transitions between normal and superconducting states. The Meissner effect expels magnetic fields during the transition, and Keefe argued that certain adiabatic magnetocaloric processes in small superconductive particles could produce entropy behavior forbidden in bulk thermodynamics. It matters because it tests whether mesoscopic superconducting phase transitions obey the same entropy accounting as macroscopic systems.
20. Graphene Trampoline / Fluctuation-Induced Current. Paul Thibado’s graphene work studies freestanding graphene sheets that ripple at room temperature. Coupled to a circuit with diodes, the motion is reported to produce a fluctuation-induced current or load power associated with thermal motion. It matters because it revisits one of the oldest Brownian-motion questions in thermodynamics: can random microscopic motion be rectified into useful macroscopic electrical output without an external temperature gradient?
21. Thermo-Charged Capacitors and Thermionic Emitters. Germano D’Abramo’s thermo-charged capacitor proposal uses electrodes with different work functions so that thermionic emission at room temperature creates a macroscopic voltage between them. The device is proposed to draw on a single thermal reservoir. It matters because it is a direct Kelvin–Planck-style challenge: if the capacitor can charge and deliver current using only ambient thermal energy, the second law’s prohibition on single-reservoir work extraction is in play.
22. Moddel’s Optical-Cavity-Induced Current / Casimir-Cavity Device. Garret Moddel’s optical-cavity work uses metal–insulator–metal tunneling structures adjoining submicron optical cavities. The reported devices produce zero-bias current, interpreted by the authors as related to cavity-modified optical or vacuum modes and hot-carrier injection. It matters because, depending on interpretation, the device could touch the boundary between zero-point energy, thermal equilibrium, and second-law accounting.
23. James W. Lee’s Transmembrane Proton-Localization Proposal. James W. Lee’s membrane work concerns transmembrane electrostatically localized protons and the possibility of thermotrophic features in mitochondrial energetics. The idea is that asymmetric biological membranes may localize protons at liquid–membrane interfaces in ways that alter how environmental heat participates in bioenergetics. It matters because it applies the boundary-asymmetry template to biology, where membranes, gradients, and energy conversion are already central.
24. Proposed Gas-Cycle Challenges Mentioned by Sheehan. Sheehan has mentioned additional gas-cycle concepts as possible high-power-density second-law challenges, but I could not verify a dedicated public, non-transcript source for the specific Ken Rowan proposal mentioned in the earlier transcript. The linked paper is the best public contextual source because it lays out Sheehan’s general template for second-law devices. It matters as a placeholder for an important category: gas-cycle devices that would need to show boundary asymmetry, thermal regeneration, independent work extraction, and cyclic reset.
25. Broader Gravitational, Ideal-Gas, and Other Second-Law Challenges. Sheehan and Čápek’s monograph surveys a wider world of proposed second-law challenges beyond the better-known epicatalytic, plasma, semiconductor, and membrane examples. These include theoretical challenges involving ideal gases, gravitational systems, quantum effects, and other nonstandard thermodynamic scenarios. It matters because Sheehan’s work is not a single anomaly claim; it is part of a larger research program asking whether the second law has a domain of applicability rather than unlimited reach.

The Beginning of a New Scientific Frontier?

The value of this list is not that it definitely proves the second law has been broken. It does not. The second law remains one of the most successful principles in science, and extraordinary claims against it demand extraordinary experimental discipline. Hidden reservoirs, incomplete cycles, surface contamination, measurement artifacts, ordinary chemical energy, and uncounted entropy costs are the natural first suspects.

What makes the list useful is that the challenges are specific. They are not vague appeals to “free energy.” They name materials, mechanisms, surfaces, membranes, cavities, junctions, fields, and proposed work-extraction cycles. A reader can move from Maxwell’s demon to Duncan’s radiometer, from tungsten and rhenium to graphene and superconductors, from formic acid on polymers to chemically asymmetric membranes.

The recurring pattern is boundary physics. Sheehan’s argument is that the thermodynamic limit teaches scientists to ignore boundaries, while real devices live by them. Surfaces dissociate molecules. Membranes bind ions. Junctions maintain fields. Cavities alter modes. Superconducting loops impose quantum constraints. If any loophole exists, it is likely to be found in these structured places where bulk equilibrium intuition is least reliable.

The honest ending is unresolved. Most or all of these challenges may eventually be absorbed into conventional thermodynamics. But the test is empirical, not rhetorical. A true second-law exception would need independent replication, complete energy and entropy accounting, and a closed cycle that repeatedly extracts work while resetting itself. Until then, these twenty-five ideas form a map of the frontier: the places where Sheehan and others think nature’s most famous “no” might still be worth testing.

References

Each item title above links to a journal article, public article, patent page, or other online resource focused on that specific experiment or concept. The list was curated from Daniel P. Sheehan’s published work, public talks, and related second-law challenge literature.

For a broad overview of Sheehan’s framework, see his 2022 article “Beyond the Thermodynamic Limit: Template for Second Law Violators,” which explains the boundary-asymmetry template behind many of the examples in this list.

For a more accessible introduction to Maxwell zombies and epicatalysis, see Sheehan’s American Scientist article “Maxwell Zombies: Conjuring the Thermodynamic Undead.”

For the central tungsten–rhenium blackbody-cavity experiment, see “Experimental Test of a Thermodynamic Paradox” in Foundations of Physics.

For a book-length treatment of the field, see Vladislav Čápek and Daniel P. Sheehan’s Challenges to the Second Law of Thermodynamics: Theory and Experiment.