Floquet Heat Transport Needs a Better Benchmark

Quantum energy devices are often described as if their job were simple: absorb heat here, dump heat there, and use a precisely timed drive to turn the difference into useful work. The real physics is messier. Once the working system is periodically driven, energy can move through Floquet sidebands, multiphoton resonances and coherences that have no static equivalent. If the theoretical model averages away the wrong pieces, a quantum heat engine, refrigerator or thermal router can look better, worse or cleaner than it really is.

A new August 2026 arXiv paper by Luca Magazzù, Christoforus Dimas Satrya, Aleksandr S. Strelnikov, Bayan Karimi and Jukka P. Pekola, titled Heat transport in driven quantum systems: Comparison between the Floquet-Redfield equation and the master equation in the instantaneous eigenbasis, tackles that modelling problem directly. The team studies heat transport in a periodically driven spin-boson model and compares a weak-coupling Floquet-Redfield benchmark with more approximate master-equation descriptions, including one written in the system’s instantaneous energy basis.

The headline result is practical: driven quantum heat transport is not only about the average energy gap. It is about which sidebands, coherences and multiphoton pathways survive the approximation.

For Floquet.ca’s core themes, this is not a minor bookkeeping issue. Floquet engineering, quantum thermodynamics and beyond-Carnot claims all depend on honest heat-current calculations. Before a driven qubit can be treated as the working element of a thermal machine, researchers must know whether the equations they use capture the heat current created by the drive, the bath and the system’s quantum coherence together.

The open-system problem behind Floquet energy devices

Floquet theory is the mathematics of periodic motion. If a Hamiltonian repeats in time, the system can be described by quasienergies and time-periodic Floquet modes. In plain language, the drive dresses the system: a qubit, molecule or material no longer has only its bare transition frequency, but a ladder of drive-shifted transitions separated by integer multiples of the modulation frequency.

That picture is powerful for quantum energy. A periodic drive can open transport channels that a static device does not have. It can make an otherwise forbidden heat transition possible, tune a resonance without rebuilding hardware, and create fractional-frequency peaks where several drive photons combine to bridge one energy gap. But real devices are open systems. They exchange energy with baths, substrates, resonators, control lines or engineered environments.

The standard tool for this regime is a master equation. It tracks the reduced density matrix of the system while replacing the bath with correlation functions and transition rates. The challenge is deciding which master equation is accurate enough. A fully microscopic model of system plus bath is usually too expensive. A strongly simplified equation may be easy to solve but may discard the coherence or sideband structure that carries the heat current.

Floquet-Redfield as the benchmark

Magazzù and colleagues start from a Caldeira-Leggett description of the bath and derive heat-current formulas for a periodically driven open system. Their central benchmark is a Floquet-Redfield treatment in the weak system-bath coupling limit. Importantly, for the steady-state heat current, this benchmark does not require adding the Markovian or secular approximations that often simplify the problem further.

That matters because two common simplifications can change the physics. A Markov approximation assumes the bath memory dies quickly enough that the system’s present evolution does not depend on its detailed past. A secular approximation drops rapidly rotating terms, often decoupling populations from coherences. Both can be justified in the right regime. Both can mislead if applied too aggressively.

The paper’s Floquet-Redfield equation keeps the coupling between different Fourier components of the periodic steady state, as well as the coupling between populations and coherences in the Floquet basis. For heat transport, that is not an aesthetic choice. The heat current is itself a dynamical quantity, and its average over one drive period can receive contributions from terms that a simpler treatment suppresses.

Fractional-frequency heat peaks

The most accessible result comes from the driven qubit. The authors calculate the steady-state heat current into the bath as a function of the drive frequency. They find peaks not only at the main resonance, where the drive frequency matches the qubit frequency, but also at fractional resonances: drive frequencies that match a fraction of the qubit’s frequency. These are multiphoton processes. Several drive quanta work together to activate a transition.

For a smart non-physicist, an analogy helps. Imagine a staircase where one big step is too high to climb with a single small push. If the pushes are timed correctly, two, three or four smaller pushes can combine to move the system up or down the same energy step. In a quantum heat-transport setting, those timed pushes are drive photons, and the staircase is the qubit’s dressed energy structure.

The paper also highlights a symmetry effect. In the absence of applied static bias, the peaks at even fractional frequencies are suppressed. When a finite bias is introduced, the selection rule is relaxed and those even peaks can appear. That is a design-relevant detail: a tiny static asymmetry can change which multiphoton heat channels are visible.

Floquet heat transport is sideband engineering. The useful question is not only “how much heat flows?” but “which photon-assisted channel carried it?”

Where the full secular approximation falls short

The paper compares the Floquet-Redfield benchmark with the full secular approximation. The full secular treatment gives a clean Pauli-type master equation for averaged populations in the Floquet basis, with heat currents built from summed transition rates. This is appealing because it is compact and physically readable. However, the comparison shows that it can distort peak heights.

At low-frequency multiphoton peaks, the full secular approach overestimates the heat current relative to the nonsecular Floquet-Redfield calculation. At the broad main resonance, it can underestimate the peak, in the paper’s example giving roughly half of the value predicted by the nonsecular treatment. The reason is not mysterious: the discarded coherences and Fourier components are not just decorative. They affect the steady-state density matrix and the current formula.

This is an important warning for quantum heat-engine studies. If an approximation gets the populations approximately right but misses the current carried by coherence or sidebands, it may still produce the wrong power prediction. For a device paper, a factor-of-two change near resonance is not a philosophical disagreement; it can decide whether a proposed operating point looks viable.

The instantaneous-basis approach is not useless

The authors do not simply reject simpler master equations. They also derive and test a master equation in the instantaneous eigenbasis of the driven qubit. Instead of describing the system in the Floquet basis, this approach follows the energy eigenstates of the Hamiltonian at each instant. That can be especially natural in the adiabatic regime, where the drive changes slowly compared with the system’s internal motion.

This regime matters for thermal machines. Many quantum engines and refrigerators operate through cyclic modulation of energy gaps, couplings or fields. If the cycle is slow enough, instantaneous energy levels provide an intuitive thermodynamic picture: the system is being compressed, expanded, heated and cooled along a driven path. The paper finds that the instantaneous-basis master equation agrees well with the Floquet-Redfield benchmark especially at low drive frequency, and an analytical solution reproduces the numerical behaviour over broad parameter ranges when the drive amplitude is not too large.

The useful message is therefore conditional, not dogmatic. Floquet-Redfield gives a strong steady-state benchmark. The instantaneous-basis approach can be a practical modelling tool in the adiabatic thermal-machine regime. The full secular Floquet picture can still be useful when its assumptions are valid. The researcher’s job is to match the equation to the operating regime rather than choosing the tidiest formula by habit.

Why this matters for beyond-Carnot language

Floquet engineering is often discussed alongside beyond-Carnot thermodynamics because periodic driving supplies a resource that classical textbook engines do not include in the same way. But this is exactly why heat-current modelling must be conservative. A drive is not free. If a device appears to extract unusual power, redirect heat, or cool below an expected limit, the first question should be whether heat, work and drive-assisted transitions have been separated consistently.

The new paper is valuable because it improves that diagnostic layer. It gives theorists and experimentalists a clearer way to ask: are the observed heat-current peaks genuine multiphoton transport, an artifact of a secular approximation, or a feature that only appears when coherences are retained? That question applies directly to superconducting qubits, driven quantum dots, nanocaloritronic circuits and other platforms where Pekola’s broader research community has long measured heat at the single-device scale.

• For quantum heat engines: power curves near resonance may need nonsecular checks before being treated as design targets.
• For thermal diodes and routers: symmetry, bias and sideband selection can determine which direction or channel dominates heat flow.
• For quantum batteries: periodic charging protocols need the same caution when identifying injected work versus dissipated heat.
• For Floquet materials: bath structure and relaxation pathways decide whether a driven state is useful or merely heated.

A better map for quantum thermal hardware

The paper does not claim a new engine efficiency record, a direct route to energy harvesting, or a finished device. Its contribution is more foundational: it shows how to compare heat-transport theories in a setting simple enough to solve but rich enough to contain the sidebands, coherences and multiphoton resonances that matter in real driven devices.

That is exactly the kind of progress the field needs. Quantum energy research is moving from ideal cycles to hardware-constrained operation. In that world, the best theory is not always the most elegant one. It is the one that tells engineers when a resonance is real, when an approximation is safe, and when the energy ledger has quietly lost a term.