Beyond the Fermion Limit: A Bosonic Route to More Quantum Engine Power

How much power can a quantum heat engine deliver? The answer depends not only on its temperature difference and internal design, but also on what carries the energy. A recent theoretical study argues that replacing fermionic carriers with bosonic ones changes a widely used nanoscale power benchmark—without changing the Carnot limit.

In Exclusion Statistics as a Thermodynamic Resource in Quantum Heat Engines, Sampurna Karmakar, Aziz Hasan and Sourin Das derive a bosonic maximum-power bound approximately 1.52 times the corresponding free-fermion bound. Their June 17, 2026 arXiv preprint also proposes a magnetic spin-chain implementation and investigates a mathematical bridge between bosons and fermions.[1]

For quantum energy research, the interesting move is conceptual. Engineers usually improve a device by reshaping its structure or its control signals. This work asks what changes when the carriers themselves obey a different rule for occupying quantum states.

Beating a fermionic power benchmark is not beating Carnot. One concerns how much power a particular transport model can deliver; the other concerns the efficiency of converting heat into work.

First distinguish power from efficiency

Power is useful energy delivered per unit time. Efficiency is the useful output divided by the heat taken from the hot reservoir. A device can improve one while worsening the other. A highly selective energy filter, for example, might let only especially productive carriers through, while admitting too few of them to generate much power.

Robert S. Whitney's quantum-thermoelectric work established bounds on maximum power and on efficiency at specified power within a nonlinear scattering description of electronic transport. In that framework, maximum efficiency approaches Carnot at zero output power and falls as the required output increases. These are microscopic, transport-based constraints, not simply a restatement of classical thermodynamics.[3]

The new study revisits the carrier assumption behind that benchmark. Rather than asking whether a more elaborate electronic filter can outdo the optimized electronic filter, it asks whether changing the occupation statistics changes the optimization problem itself.[2]

That distinction keeps the claim appropriately narrow. A bound can be universal within a specified class of systems without applying to every conceivable quantum engine. Channel count, transport assumptions and the type of carriers all belong in an honest comparison.

Fermions, bosons and the occupancy rule

Electrons are fermions. The Pauli exclusion principle prevents identical fermions from occupying the same single-particle quantum state. Bosons do not face that particular restriction: multiple bosons can occupy a state. Those different rules produce different reservoir occupation distributions, even before one chooses the structure connecting the reservoirs.

A helpful analogy is a transport terminal with different boarding rules. The routes and gates matter, but so does how many travellers can occupy each available slot. Changing the boarding rule changes the stream arriving at the gates. This is only an analogy: quantum statistics is a physical property, not a software setting that turns ordinary electrons into bosons.

In the paper's Landauer–Büttiker description, carriers travel through energy-resolved channels between hot and cold reservoirs. A transmission function specifies how readily a carrier of a given energy crosses the connecting structure. The reservoir distributions specify how populated those incoming states are. Both ingredients determine the current and useful power.[2]

The authors replace the fermionic distributions with bosonic ones, optimize the allowed transmission function, and then optimize the bias. They report that the different occupancy rule permits a larger ideal maximum power under the matched assumptions of their comparison.[2]

What the new bound actually says

For the channel normalization used in the paper, the fermionic benchmark is approximately 0.0321π² k_B²(ΔT)²/h. The proposed bosonic counterpart is (ln 2)² k_B²(ΔT)²/h, where ΔT is the reservoir temperature difference, k_B is Boltzmann's constant and h is Planck's constant.[1]

Readers need not calculate those expressions to understand the result. Both share the same temperature-difference dependence and basic quantum scale. The change is in the numerical coefficient associated with the optimized carrier statistics. That is why the comparison can reveal a statistics-driven advantage rather than merely the benefit of making a device larger.

The optimization still requires energy selection. The paper finds that maximum power uses a threshold-like transmission: block energies below a crossover point and transmit those above it. Below that crossover, transport works against the desired useful output. The crossover is set by where the two reservoir occupation distributions intersect.[2]

For maximum efficiency at a chosen finite power, the answer changes to a finite transmission window, often called a boxcar filter. Its upper cutoff excludes carriers that bring too much heat relative to their useful contribution. Narrowing the window improves selectivity but reduces throughput. The bosonic result does not eliminate this trade-off.[2]

A possible testbed: a chain carrying spin waves

Abstract bosonic reservoirs are not yet a laboratory engine. To make the idea more concrete, the authors model a one-dimensional ferromagnetic spin chain between two leads. Its relevant excitations are magnons: quantized spin waves that can carry spin angular momentum without being an ordinary electrical charge current.[2]

In the low-temperature, small-fluctuation regime used by the calculation, the magnetic system can be approximated by a quadratic bosonic model. This approximation is essential. It is not permission to apply the same ideal-carrier result unchanged to a hot, strongly interacting magnet.

The proposal uses magnetic fields to set spectral band edges and a magnetochemical bias. The central region's field helps establish the energy-filter threshold. The study calculates transmission through the chain and its contacts, rather than assuming the physical structure produces a mathematically perfect step.[2]

The numerical example—and its scale

The paper considers a hot reservoir at 300 millikelvin and a cold reservoir at 10 millikelvin, with a baseline field of about 2 tesla. For its chosen parameters, the calculated maximum output is 10.58 picowatts, compared with an ideal bosonic bound of 11.58 picowatts and a fermionic benchmark of 7.64 picowatts at the same temperature difference.[2]

The authors attribute the remaining gap from the ideal bosonic value to the physical transmission threshold being smoother than an abrupt step. They report that the example retains over 91% of the ideal bosonic maximum.[2]

These figures are scientifically useful precisely because they are modest and specific. Picowatt-scale output at cryogenic temperatures is not a replacement for a room-temperature generator. Converting the model's magnetic work into useful output, measuring it accurately and accounting for the supporting apparatus are separate engineering challenges.

Fractional exclusion: a bridge, not a magic dial

The study then introduces Haldane fractional exclusion statistics. A parameter called g describes how adding particles changes the number of available single-particle states. The model connects the bosonic case, g = 0, with the fermionic case, g = 1.[2]

Within that generalized transport calculation, reducing exclusion below the fermionic value increases the optimized maximum power and reduces the required bias. This supplies a continuous theoretical comparison instead of only two isolated endpoints.[2]

But a continuously adjustable parameter in an equation is not automatically an experimentally adjustable control. The authors mention spinon excitations in an antiferromagnetic Heisenberg chain as a possible connection to fractional statistics, while explicitly leaving a rigorous transport analysis of that realization outside the study.[2]

It is also worth distinguishing this idea from loosely saying that “anyons improve engines.” This paper's specific subject is exclusion statistics and transport power bounds. It does not establish that every system with unusual exchange statistics will reproduce the same engine advantage.

Why this is not beyond-Carnot efficiency

The new preprint itself retains the Carnot endpoint. In its optimized efficiency–power curves, the different statistical cases approach Carnot as output power tends to zero. The distinction appears in their finite-power performance and maximum-power values—not in an efficiency above Carnot for the stated thermal setup.[2]

For any eventual experiment, a useful energy ledger would identify the hot-reservoir heat input, the output delivered against the bias, losses and any additional driving. A device-level assessment would separately include refrigeration and control overhead. Otherwise, a gain inside the microscopic working medium can be confused with a net gain for the whole installation.

Where Floquet engineering could enter

This is not a Floquet-engine experiment. The reported enhancement comes from carrier statistics in a transport calculation, not from periodically modulating the device. That boundary matters on a site devoted to driven quantum systems: a relevant quantum-energy result should not acquire a Floquet label it has not earned.

There is nevertheless a useful research connection. Floquet engineering offers ways to reshape dynamical spectra and transition pathways; the present work highlights occupancy statistics as another ingredient in transport performance. A future project could ask whether periodic control helps implement a suitable energy filter for bosonic carriers. That is an editorial research question, not a result established by this preprint.

Such a project would have to count the drive's energy contribution and reassess the transport model. It could not simply import a stationary bound while ignoring drive-assisted channels. Better filtering is promising only when its benefits survive that accounting.

What would make the next result convincing?

A follow-up should test the claimed mechanism, not just show a larger current. Our assessment checklist is:

• Match the comparison: keep reservoir temperatures and channel normalization explicit when invoking the fermionic benchmark.
• Characterize the filter: determine whether the actual transmission has the required threshold and useful bandwidth.
• Verify the carrier regime: establish whether the small-fluctuation, effectively noninteracting magnon approximation remains appropriate.
• Measure useful output: distinguish spin or heat flow from work delivered to a load.
• Report losses and overhead: separate microscopic conversion performance from total apparatus consumption.

The strongest conclusion today is not that a new energy technology has arrived. It is that a familiar quantum-engine benchmark depends on a physical assumption worth testing. If the proposal survives detailed modelling and experimental scrutiny, carrier statistics could join energy filtering and reservoir design as a practical variable in quantum thermodynamics.

That would be an advance without requiring a broken thermodynamic law: more room for useful performance inside a carefully defined operating regime, with a clearer explanation of where the extra room comes from.

Sources and further reading

1. Karmakar, Hasan and Das, “Exclusion Statistics as a Thermodynamic Resource in Quantum Heat Engines,” arXiv:2606.19310, submitted June 17, 2026. Preprint metadata and abstract. Read source [1]
2. Full text of arXiv:2606.19310v1. Primary theoretical derivation, magnetic spin-chain calculation and efficiency–power analysis. Read source [2]
3. Robert S. Whitney, “Finding the quantum thermoelectric with maximal efficiency and minimal entropy production at given power output,” Physical Review B 91, 115425 (2015). Open arXiv version. Read source [3]