What Survives the Bath? Topology in Open Floquet Quantum Systems

Designing a quantum material with attractive mathematical properties is only half the challenge. The other half is finding out what survives when that material exchanges particles and energy with its surroundings. A periodically driven system may look topological on paper, yet settle into a state that tells a different story.

A preprint by Görkem D. Dinc, Alexander Schnell and Andy M. Martin, revised on October 21, 2026, tackles that distinction in a driven quantum chain connected to thermal reservoirs. Its central result is not a new battery or heat engine. It is a way to identify topology in the mixed, periodically repeating state that the open system actually reaches.^[1][2]

The study finds two symmetry-dependent regimes, including one where a topological marker switches during each driving cycle. For quantum-energy research, the lesson is practical even though the work is theoretical: a convincing design must characterize the operating state, not just the equations that generate its motion.^[2]

Why a repeating drive changes the question

Floquet engineering uses a repeating external influence to alter a quantum system's behavior. Rather than manufacture a different material, researchers can change the timing, strength or frequency of a drive. The resulting dynamics can support properties that have no direct counterpart in the undriven system.^[2]

One example is topology: a global structure that can remain unchanged under small, appropriate deformations. In certain one-dimensional chains, this structure is associated with states localized near the ends. A useful analogy is a knotted loop: modest reshaping need not undo the knot. But the analogy has limits, because quantum protection depends on specific symmetries and gaps.^[2]

For an isolated, chiral-symmetric Floquet chain, the classification can involve two independent integers, associated with the zero and π gaps of its quasienergy spectrum. Quasienergy is the energy-like quantity used for periodically driven motion; its repeating structure creates possibilities absent from an ordinary static spectrum.^[2]

The new work asks whether that familiar classification describes the state left behind after reservoirs have continuously acted on the system. Within the authors' ensemble-geometric-phase framework, it generally does not simply carry over.^[2]

A chain connected to the outside world

The model is a driven Su–Schrieffer–Heeger, or SSH, chain. Picture a row of paired sites, labeled A and B. Particles can hop within each pair and between neighboring pairs. The drive rhythmically changes those two hopping strengths in opposite directions.^[2]

The A and B sites also couple to fermionic thermal reservoirs. These environments can inject and remove particles, and their temperatures and chemical potentials help determine the state of the chain. Chemical potential describes the reservoir's tendency to exchange particles; it is an additional control beyond temperature.^[2]

This is an open system, so a single perfectly specified quantum wavefunction is generally insufficient. The researchers instead describe a mixed state: a statistical quantum description that retains information about populations and correlations. After initial transients, the state repeats with the driving period rather than becoming completely time-independent.^[2]

Key insight: A repeating steady state is not a frozen state. Its topology can change within a cycle even after the long-time operating pattern has settled.^[2]

That distinction matters whenever measurements take place at a particular phase of the drive. Looking once per cycle can miss changes that happen between snapshots.

Classify the state, not only the generator

A generator specifies how the system evolves. Its spectrum can reveal oscillation frequencies, decay behavior and topological structure. But the state reached under that evolution is a different mathematical object, and it need not inherit every feature of the generator.^[2]

In the particular bath model studied here, with the stated approximations, the matrix controlling homogeneous covariance evolution is insensitive to the bath temperatures and chemical potentials. The actual steady state is not. A classification based only on that matrix would therefore miss environmental changes that reshape the operating state.^[2]

The authors examine the state's purity spectrum instead. For their Gaussian fermionic system, correlations can be organized into a covariance matrix, from which a spectrum analogous to a band structure can be constructed. Its gaps and boundary-localized modes help diagnose mixed-state topology.^[2]

“Analogous” is important. These purity eigenvalues are not ordinary energy levels, and an edge mode in the purity spectrum is not automatically a channel delivering electrical power. The spectrum describes structure in the quantum state, not a ready-made energy-conversion measurement.^[2]

A geometric phase for a mixed quantum state

The central diagnostic is the ensemble geometric phase, abbreviated EGP. It generalizes geometric-phase ideas to mixed states by using a many-body expectation value. Bardyn and colleagues introduced this approach in a 2018 Physical Review X paper, showing how certain topological information can remain quantized even when the state is mixed.^[4]

Molignini and Cooper subsequently explored finite-temperature topological transitions in an open SSH setting. Their 2023 study showed that temperature can participate in changing topology, rather than merely washing out every interesting feature.^[3]

The October 2026 revision brings periodic driving and a microscopic treatment of the reservoirs into the same analysis. It tracks a complex-valued expectation value, called Z, whose angle is the EGP. An intuitive picture is an arrow in a plane: the arrow's angle and its trajectory through a full cycle provide different possible diagnostics.^[2]

Regime one: topology switches during the cycle

When the relevant steady-state symmetry holds, the expectation-value arrow is confined to the imaginary axis. Its angle is pinned to one of the two values above. In the model's convention, one value identifies the nontrivial regime and the other the trivial regime.^[2]

The striking result is that the state can move between these regimes within a single driving period. In the calculations, the switches track purity-gap closings and the appearance or disappearance of edge-localized modes in the open-boundary purity spectrum. This establishes a bulk-boundary correspondence for this symmetry-preserving case.^[2]

The authors define the fraction of each period spent in the nontrivial regime. That fraction varies continuously with driving frequency. Think of it as a duty cycle for a diagnosed topological property, not as a duty cycle for experimentally demonstrated useful output.^[2]

This distinction prevents a tempting overinterpretation. A longer nontrivial interval might become a useful design variable in a future protocol, but the present study does not show that it stores more energy, transfers more heat or produces more work. Those quantities require their own calculations.

Why the reservoir model cannot stay on autopilot

The drive does not merely change motion inside the chain. It also changes the transition channels through which the system exchanges particles and energy with its environment. In Floquet language, those transitions can involve different numbers of drive quanta.^[2]

The researchers account for this using Floquet–Born–Markov theory, without imposing the rotating-wave approximation. Their treatment is still approximate: it assumes weak coupling and a Markovian reservoir description, and it neglects Lamb-shift terms. It is not a claim to solve every strongly coupled or memory-rich environment.^[2]

An important comparison holds the dissipative part to a static construction while applying the drive to the coherent dynamics. In that simplified calculation, the within-cycle topological switching disappears. For this model, neglecting how the drive modifies the dissipator removes the phenomenon under investigation.^[2]

Regime two: count the winding, not the edge modes

Breaking the relevant symmetry, for example through a chemical-potential bias between the baths, frees the expectation-value arrow from the imaginary axis. It can trace a closed loop over the drive period. If that loop avoids the origin, its signed number of encirclements defines an integer winding number.^[2]

The calculations reveal regions with different winding numbers as temperature or driving frequency changes. In selected hot-bath parameter regimes, the transitions are consistent across the tested system sizes; some cold-bath plateaus instead move as the chain length changes. The paper therefore distinguishes finite-size observations from evidence of size-stable behavior.^[2]

Here, the important protective condition is that the trajectory stay away from the origin, where its angle becomes undefined. It is not the same purity-gap criterion used in the symmetry-preserving regime. Nor do the authors find a correspondence between this winding number and instantaneous purity-spectrum edge modes.^[2]

Most importantly for an energy audience, they report no signature of net transport across the chain associated with the nonzero winding phases. A winding geometric marker is therefore not evidence of quantized charge pumping in this open, driven setting.^[2]

The measurement challenge: a robust angle can have a tiny signal

There is another reason not to jump from a clean theoretical classification to an engineered component. The magnitude of the expectation value carrying the phase becomes exponentially suppressed with increasing chain length in the studied scaling analysis. The angle can remain mathematically meaningful while the signal becomes difficult to resolve.^[2]

This separates two forms of robustness. A topological integer may remain stable against a parameter change, yet extracting that integer experimentally may demand exceptional sensitivity. Larger systems are not automatically easier demonstrations, and a stable numerical invariant does not guarantee a convenient laboratory readout.

The 2018 foundational work proposed an interferometric measurement concept for mesoscopic ultracold-atom ensembles. However, the new paper explicitly leaves measurement procedures for its fast, nonadiabatic protocol as an open question. It also asks whether more experimentally practical diagnostics could reproduce the same classification without the strongly suppressed signal.^[2][4]

What this contributes to quantum-energy engineering

The immediate contribution is a better way to ask design questions. Before describing a Floquet structure as a protected energy pathway, researchers should establish which state exists under driving and dissipation, what property is protected, and how that property is measured.

For a future energy-device study, a sensible follow-up checklist would be:

• Track the entire cycle. Compare phase-resolved behavior with once-per-period measurements.
• Validate the environment model. Check whether drive-modified exchange channels change the predicted result.
• Measure transport separately. Establish particle currents, heat currents and useful output instead of inferring them from a topological label.
• Include operating costs. Account for drive power, reservoir preparation and any readout or feedback resources.
• Test accessibility and scale. Show that the relevant signal survives finite precision, realistic noise and changes in system size.

These are proposed engineering next steps, not achievements claimed by the preprint. In particular, nothing in the reported topology establishes beyond-Carnot efficiency. Applying thermodynamic efficiency bounds requires an explicit accounting of heat, work and any additional nonequilibrium resources; a geometric phase is not a substitute for that accounting.

The bottom line

This revised study gives Floquet research a useful distinction: the topology of the dynamics and the topology of the resulting mixed state are not interchangeable. Under one symmetry condition, the state can switch topological character within a cycle. Without that symmetry, a different winding diagnostic becomes relevant, with different protection and different limitations.^[2]

For practical quantum-energy research, that is progress in diagnosis rather than a claim of new power generation. Understanding what survives contact with a thermal environment—and what can actually be measured—is necessary groundwork for deciding whether a driven quantum material can do useful work.