Quantum thermodynamics is often presented through headline limits: Carnot efficiency, maximum power, quantum advantage, and the tantalizing possibility that coherence might help thermal machines do something classical engines cannot. But a quieter set of limits is just as important for real devices: how noisy is the useful output? If a nanoscale heat engine produces power only in rare, erratic bursts, its average efficiency tells only part of the story.

A new May 2026 preprint by Sergi Vidal, Alba Mayor-Fernandez, and Rosa Lopez, titled “Quantum Coherence Reshapes Thermodynamic Bounds for Thermal Machines” (arXiv:2605.04648), goes directly at that reliability question. The paper studies a coherent two-terminal quantum conductor operating as a heat engine, refrigerator, or heat pump, and asks how quantum coherence changes the thermodynamic uncertainty relations that connect fluctuations, useful currents, and entropy production.

The next generation of quantum energy devices will not be judged only by peak efficiency. They will be judged by whether coherent transport can deliver useful power with controlled fluctuations and honest entropy accounting.

Why Thermodynamic Uncertainty Matters

Thermodynamic uncertainty relations, usually called TURs, are among the most practical ideas to emerge from modern nonequilibrium thermodynamics. In plain language, a TUR says that precision is not free. If a tiny machine produces a current — charge, heat, particles, or work — then making that current more reliable generally costs entropy production. Classical stochastic engines face a trade-off among power, efficiency, and noise.

That trade-off is not an abstract nuisance. At nanoscale dimensions, fluctuations can be comparable to the average signal. A molecular motor, single-electron pump, quantum dot heat engine, or superconducting thermal machine can look impressive on average while being too noisy to serve as a reliable component. TURs give researchers a way to ask: is the machine not only efficient, but precise?

3

Operating modes analyzed in the new work: quantum heat engine, refrigerator, and heat pump, all within the same coherent two-terminal conductor framework.

The Vidal–Mayor-Fernandez–Lopez paper focuses on the multidimensional version of the TUR, or MTUR. Instead of considering one current at a time, the MTUR treats a matrix of currents and covariances. That matters because real thermal devices usually move several quantities together. A thermoelectric conductor, for example, may carry charge and heat simultaneously, and the cross-correlations between those currents can be as important as their separate fluctuations.

Where Quantum Coherence Enters

In a classical stochastic picture, transport is often modeled as hops among states. Quantum conductors are different. Electrons, photons, phonons, or quasiparticles can propagate as waves, interfere with themselves, and maintain phase coherence across a device. That coherence can reshape conductance, suppress or enhance noise, and produce correlations that have no simple classical counterpart.

The new paper asks a careful question: does coherent transport let thermal machines evade classical performance bounds? The answer is nuanced. The authors report that classical performance limits on efficiency and coefficient of performance remain constrained by the TUR when finite power or finite heat pumping is maintained, even in regimes dominated by coherent transport. At the same time, they identify conditions where TUR and MTUR violations are optimized, especially through cross-correlations near linear response.

What Is a Thermodynamic Uncertainty Relation?

A TUR links the relative noise of a current to entropy production. It is a speed-limit-style statement for small thermal machines: producing a precise, steady output usually requires dissipation. Quantum coherence can alter the relation, but it does not make thermodynamic bookkeeping disappear.

This is exactly the kind of result the beyond-Carnot community needs. It neither dismisses quantum advantage nor accepts every advantage claim at face value. Instead, it separates several questions that are too often blended together: Can coherence improve correlations? Can it reduce fluctuations? Can it maintain finite useful output? And after all resource costs are counted, what bound actually applies?

Why This Connects to Floquet Engineering

The 2026 preprint is not itself a Floquet-drive paper in the narrow sense. Its relevance to Floquet energy science is broader and important. Floquet engineering uses periodic driving to create effective Hamiltonians, open gaps, pump currents, and control nonequilibrium steady states. Those same driven systems are natural candidates for nanoscale engines, refrigerators, and quantum energy converters. But once a drive is introduced, the device typically has multiple coupled flows: heat from reservoirs, work from the drive, particles through contacts, and entropy into the environment.

That is precisely where multidimensional uncertainty bounds become valuable. A periodically driven thermoelectric engine may show an appealing average current, but a practical designer must also track current noise, heat leakage, drive-induced heating, and correlations among outputs. MTUR-style analysis gives a language for doing that accounting in a way that is not fooled by a single impressive number.

Floquet control can sculpt the pathways of energy flow. Thermodynamic uncertainty relations ask whether those sculpted pathways are stable, precise, and worth the entropy they cost.

There is also a direct conceptual bridge through coherent transport. Floquet devices often rely on interference in quasienergy space: electrons or photons absorb and emit drive quanta, opening sidebands that act like additional transport channels. Those sidebands can enhance performance, but they can also introduce extra noise. The new coherence-bound work helps frame the design problem: a useful driven energy converter should engineer not only the mean current, but the full covariance structure of charge and heat.

Heat Engines, Refrigerators, and Heat Pumps

One strength of the paper is that it treats several machine modes within a unified framework. In heat-engine mode, the device converts a heat flow into useful output. In refrigerator mode, it uses work or bias to pull heat from a cold reservoir. In heat-pump mode, it moves heat from cold to hot, with performance measured by a coefficient of performance rather than by engine efficiency.

These distinctions matter because “better” means different things in different modes. A heat engine wants high efficiency and finite power. A refrigerator wants high coefficient of performance while maintaining heat flow out of the cold side. A heat pump wants useful heating with minimal energetic cost. In each case, fluctuations can undermine usefulness. A cryogenic quantum chip, for example, does not benefit from a cooler that works only intermittently or injects uncontrolled heat noise into sensitive qubits.

The authors’ emphasis on finite power and finite cold-to-hot heat flow is especially important. Some theoretical loopholes appear only as output goes to zero, where a machine can approach a formal limit while becoming practically useless. By keeping finite currents in view, the work stays close to engineering reality.

2026

The year quantum-coherent thermal machines are being analyzed not just for average efficiency, but for precision, covariance, and fluctuation bounds.

Cross-Correlations: The Hidden Resource

The most interesting technical message for non-specialists may be the role of cross-correlations. If two currents fluctuate together in a useful way, the joint device can be more predictable than either current appears alone. The multidimensional TUR captures this because it treats the covariance matrix rather than a single variance.

For Floquet and quantum-energy researchers, this suggests a design principle: do not optimize every channel independently. In coherent devices, energy, charge, and entropy flows can be correlated by interference. Periodic driving may strengthen, weaken, or redirect those correlations. A future Floquet heat engine might deliberately tune drive frequency, phase, and coupling geometry to create favorable current correlations, much as photonic engineers tune resonators to shape bandwidth and noise.

There is a caution, however. Correlation is not magic. A device can violate a simplified classical-looking TUR while still obeying a more complete bound once all relevant currents and entropy production are included. That is why multidimensional formulations are so important. They reduce the chance that a claimed quantum advantage is merely a bookkeeping advantage.

Beyond-Carnot Without Confusion

Floquet.ca often covers “beyond-Carnot” ideas because driven quantum systems can appear to sidestep familiar textbook boundaries. The key word is appear. Carnot’s theorem applies to a specific idealized conversion of heat between two reservoirs. Real quantum devices may include coherent work reservoirs, measurement, feedback, nonthermal baths, squeezed reservoirs, time-dependent Hamiltonians, and strong coupling. Those ingredients can change the relevant bound, but they do not abolish thermodynamics.

The new coherence paper contributes to this more mature conversation. It shows that quantum coherence can reshape uncertainty relations and enhance joint precision through correlations, while still leaving robust constraints on performance when finite useful output is required. That is not a disappointing result. It is a roadmap for credible quantum energy claims.

Why “Beyond Carnot” Requires Extra Accounting

If a device uses coherence, measurement, periodic driving, or engineered reservoirs, those resources must be included in the thermodynamic balance sheet. A real advantage is one that survives after the drive, noise, information, and entropy costs are counted.

What Experimentalists Should Watch

The experimental frontier is moving quickly. Just this month, Tuomas Uusnäkki, Timm Mörstedt, Wallace Teixeira, Miika Rasola, and Mikko Möttönen reported an initial demonstration of a quantum heat engine based on dissipation-engineered superconducting circuits in Nature Communications. Their platform used a quantum-circuit refrigerator as a tunable thermal reservoir and a flux-tunable transmon qubit as the working medium, measuring positive output powers and efficiencies over a few quantum Otto cycles.

That superconducting result and the new TUR analysis point toward the same next step: quantum thermal machines need fluctuation metrology. Average power and efficiency are necessary, but not sufficient. Future demonstrations should report heat and work distributions, covariance between relevant currents, entropy production estimates, and how those quantities change under coherent control or periodic modulation.

Floquet platforms are well positioned for this because drive parameters are highly tunable. Frequency, amplitude, phase, pulse shape, and reservoir coupling can be swept systematically. The challenge is measurement: extracting full counting statistics for heat and work in quantum systems is difficult. But without those statistics, it is hard to know whether a device is a useful machine or merely an elegant average.

The Practical Energy Takeaway

No one should expect quantum-coherent heat engines to replace gas turbines or grid-scale batteries. Their plausible near-term role is more specialized: cooling, powering, and stabilizing quantum technologies themselves. Quantum processors, cryogenic sensors, microwave photonic circuits, and nanoscale detectors all need precise energy management at the level where classical thermodynamic intuition becomes incomplete.

In that context, the paper’s message is highly practical. Coherence can be a resource, but reliability is also a resource. A quantum thermal machine that produces slightly less peak power but far lower noise may be more valuable than one with a spectacular average and poor precision. Similarly, a Floquet energy converter whose sidebands produce favorable heat-charge correlations may outperform a device optimized only for maximum current.

The future of quantum energy engineering is not a single race to beat Carnot. It is a multidimensional optimization over power, efficiency, precision, coherence, and entropy production.

Selected Research Cited

The headline is simple: coherence changes the design space for quantum thermal machines, but it does not eliminate the need for disciplined thermodynamic accounting. For Floquet engineering, that is good news. Periodic driving offers a powerful way to shape coherent transport; uncertainty relations help reveal which shaped flows are precise enough, efficient enough, and honest enough to matter.

Explore Beyond-Carnot Thermodynamics

See how Floquet engineering, coherence, and engineered reservoirs are changing the way researchers think about quantum energy conversion.

Explore the Science →