From Random Noise to Useful Work: A Superconducting Feynman Ratchet

A superconducting diode can favour current in one direction. That does not automatically make it an engine: something still has to supply the energy. A new October 2026 preprint asks a more demanding question. Can a circuit turn the random electrical fluctuations associated with a temperature difference into useful work, without an alternating-current drive doing the pumping?^[1]

Konstantinos Kontogeorgiou, İnanç Adagideli, Géraldine Haack and Fabian Hassler propose a superconducting implementation of Feynman’s ratchet: a microscopic engine whose directional motion emerges from unequal thermal environments. Their study, posted on October 12, presents a circuit model and numerical simulations, not a newly measured device.^[1]

The distinction matters for quantum-energy research. The proposal connects engineered superconducting components with heat-to-work conversion, but the authors explicitly describe the phenomena studied as essentially classical. It is neither an experimental quantum advantage nor a beyond-Carnot result. Its contribution is a specific architecture for making thermal noise do electrical work under well-defined nonequilibrium conditions.^[1]

The central insight: a one-way element supplies direction, not energy. In this proposal, the energy resource is the temperature difference between two reservoirs.^[1]

Why a tiny ratchet needs two temperatures

Imagine a paddle wheel connected to a toothed wheel with a spring-loaded catch. Random molecular impacts shake the paddles. The teeth appear to permit rotation only in one direction, suggesting that ordinary thermal agitation could steadily lift a weight. Feynman used this setup to expose the missing ingredient in that appealing argument: the catch also fluctuates.^[3]

At equal temperatures, the catch sometimes jumps out of the way, allowing backward motion. The apparent one-way machine cannot indefinitely extract work from a single equilibrium environment. Cooling the ratchet relative to the paddles changes the balance, making directed operation possible because the machine now has access to a temperature difference.^[3]

The 2026 paper translates that logic into circuit variables. Instead of a mechanical axle and catch, it uses coupled superconducting phases and an asymmetric Josephson potential. Its equilibrium calculation returns zero average voltage when the two reservoirs have the same temperature and no external load current is imposed.^[1]

The circuit: a cold branch, a hot branch and a directional landscape

The proposed device has two main circuit branches. Each contains a resistor associated with a thermal reservoir, and the branches interact through inductive coupling. The upper branch also contains the superconducting diode. In the engine configuration discussed by the authors, the upper reservoir is colder than the lower one.^[1]

A Josephson junction links superconductors through a weak connection. Its current depends on the relative superconducting phase, a variable that can be pictured as the angular difference between two coordinated oscillations. In this model, designing the current–phase relationship creates an energy landscape that is asymmetric: moving through it in one direction differs from moving in the other.^[1]

For a concrete implementation, the authors consider eight Josephson tunnel junctions arranged into four parallel branches inside the diode element. This internal network should not be confused with the two larger branches connected to the thermal reservoirs. The junction network supplies the engineered asymmetry; the reservoir branches supply the nonequilibrium fluctuations.^[1]

The equations track two phase variables pushed by random currents from the resistors. A changing superconducting phase corresponds to a voltage through the Josephson relation. A preferred average direction of phase motion therefore produces an average electrical voltage, rather than the rotation of a miniature physical wheel.^[1]

Directionality is necessary, but not sufficient

The numerical design has a diode rectification coefficient of approximately 0.6. That number describes the asymmetry between critical currents in opposite directions. It is not a 60% heat-engine efficiency. The paper separately calculates electrical work output and the heat extracted from the hot reservoir to assess thermodynamic efficiency.^[1]

This is a useful discipline across superconducting electronics. A device can have impressive directional selectivity while producing little power, or require substantial resources to maintain its operating conditions. Directional response, energy conversion and whole-system efficiency are different measurements.

What the simulations actually predict

The authors first calculate the voltage with no external load current. A temperature difference produces a finite voltage, and the voltage changes sign when the thermal ordering is reversed in the regime studied. With equal reservoir temperatures, the equilibrium result remains zero.^[1]

They then examine the Seebeck coefficient: how sensitively the open-circuit voltage responds to a small temperature difference near equilibrium. The reported maximum is approximately 0.38 microvolts per kelvin for a resistance of 1 kilohm, at the specified dimensionless coupling and temperature. This is a model prediction for a particular parameter choice, not a measured sensitivity.^[1]

Open-circuit voltage alone does not establish useful output power. The authors therefore introduce an imposed current that models a load. Within the engine-operating range, increasing the load current reduces the voltage. For the approximately linear response shown, maximum output power occurs when the voltage has fallen to half its open-circuit value.^[1]

That headline requires its conditions. The approximately 12% result occurs at dimensionless coupling κ = 0.01, with the cold-reservoir temperature set to zero and the hot reservoir near 17 times a characteristic temperature T₀. The paper defines T₀ through the Josephson energy scale; 17T₀ does not mean 17 kelvin. At κ = 0.1, the corresponding maximum reported efficiency is approximately 2%.^[1]

The researchers select temperatures to optimize the open-circuit voltage and evaluate efficiency at maximum power. These are results within a particular parameter scan, not a proof of the globally best possible engine. The idealized cold-bath limit also should not be read as a practical instruction for reaching absolute zero.

Why this differs from earlier superconducting ratchets

Superconducting ratchet devices are not new. A 2024 preprint by Christoph Schmid and colleagues reported fabricated YBa₂Cu₃O₇ Josephson diodes that rectified an applied alternating current into a direct voltage. The study reported loaded operation, an output power up to 0.2 nanowatts and conversion efficiency up to 75% in its driven-ratchet setting.^[2]

Those results provide an important experimental precedent for directional superconducting circuits. But they do not show the same energy-conversion process as the October 2026 proposal. In the earlier experiment, an applied electrical drive supplies the energy being rectified. In the new model, the intended resource is heat flowing between reservoirs at different temperatures.^[1][2]

It would therefore be misleading to rank the new 12% prediction against the older 75% figure as though they were competing versions of the same heat engine. The energy inputs differ, as do the operating conditions and the evidence level: numerical proposal versus fabricated driven device.

The right comparison starts with the input: converting an applied AC signal into DC output is not the same task as converting heat from a hot reservoir into work.

Can the same circuit work as a refrigerator?

The paper also examines reversed operation. With a suitable externally supplied current in the opposite direction, the circuit can consume work and move heat from the colder reservoir toward the hotter one. The authors identify a cooling regime close to the parameters that maximize the Seebeck response.^[1]

This is the proposed Peltier-cooler mode. It does not mean that the circuit spontaneously cools its own heat source while simultaneously producing free power. Engine operation extracts work while heat flows downhill in temperature; refrigeration spends work to push heat uphill. Measuring the signs of the reservoir heat currents is essential to telling those regimes apart.^[1]

A practical follow-up would need to establish how much heat can be removed from a real target, how the effect changes under load, and whether unwanted heat leaks overwhelm it. A voltage signal is an encouraging diagnostic, but a refrigerator must demonstrate net cooling of the intended component.

The Floquet connection—and where it stops

Floquet engineering uses periodic driving to reshape a system’s effective dynamics. This proposal takes a different route: its central engine mechanism does not require an externally imposed periodic AC drive. The authors explicitly distinguish their two-temperature engine from superconducting ratchets operated by external AC forcing or engineered noise sources.^[1]

For Floquet researchers, that contrast is valuable. It creates a useful comparison between two strategies: actively driving a device to control its directional response, and building a directional element into a circuit powered by thermal imbalance. The present paper supports the second strategy; it does not demonstrate a Floquet-enhanced version of the engine.

One possible future research question is whether periodic control could improve the useful operating range of a related circuit. That is an editorial extension, not a result reported here. Any such experiment would need to include the control field’s energy input rather than attribute all output to heat harvesting.

What would make this useful rather than merely elegant?

The architecture’s appeal lies in translating an influential thermodynamic thought experiment into a superconducting circuit model compatible in principle with cryogenic electronics. The authors emphasize miniaturization and circuit integration as motivations, but the work remains a numerical proposal.^[1]

For evaluating future experimental claims, we would prioritize four tests:

• Verify the energy source. Demonstrate that the relevant output disappears at thermal equilibrium, and distinguish temperature-driven output from pickup or unintended external forcing.
• Measure power under load. Establish an operating range with positive delivered electrical power, rather than reporting only open-circuit voltage.
• Close the heat balance. Measure hot-side heat extraction and cold-side heat rejection alongside work output, including parasitic pathways.
• Report the operating overhead. Separate the efficiency of the small converter from the power needed for refrigeration, thermal bias, instrumentation and any added control drive.

Those are proposed evaluation criteria, not completed milestones in the preprint. They also clarify the likely near-term value: a controllable experimental platform for studying nanoscale energy conversion, rather than an immediate source of household or grid electricity.

The takeaway: engineer the fluctuations, keep the accounting

The October 2026 study offers a specific way to connect superconducting directionality with a genuine two-temperature heat-engine model. Its distinctive feature is not a record-breaking efficiency claim. It is the attempt to make a Josephson circuit play the role of Feynman’s ratchet, using thermal fluctuations to produce electrical work and reversing the process to obtain cooling.^[1]

For the wider quantum-energy community, that is a useful development precisely because its boundaries are clear. The evidence is theoretical and numerical, the reported mechanism is essentially classical, and the energy comes from a thermal gradient. A future demonstration would advance miniature thermal machines without needing to overturn thermodynamics.^[1]