The Signal
At Intermediate Frequencies, Resonance Alone May Not Stabilize Floquet States
A theoretical study finds that cavity-assisted stabilization of periodically driven quantum states depends not only on resonance but also on the strengths of competing transitions. The authors’ simulations show that those channels can coope
Quantum.world Team · Written with AI · fact-checked ·
“At intermediate driving frequencies, this resonance condition alone is not sufficient for stabilization.”
Zhuocheng Ma, Hongzheng Zhao and Liang-You Peng, the paper’s authors, write.
The three Peking University researchers submitted their paper to arXiv on 29 October 2026, challenging a simple expectation in cavity-assisted control of periodically driven quantum systems: matching a cavity’s frequency to a transition does not always make the desired state stable. Their calculations show that the outcome also depends on the relative strengths of competing transitions. In some frequency ranges, the cavity helps drive the system toward a mixture rather than a single Floquet state.
A cavity can open two paths
The study considers a periodically driven two-level spin coupled to a lossy cavity. The cavity can absorb energy from the spin and then shed it as a photon, creating a route for the driven system to relax into a selected Floquet eigenstate. Floquet states describe a system whose Hamiltonian repeats in time; their quasienergies play a role analogous to energy levels, but are defined modulo the drive frequency.
The intended route is not necessarily the only one. Ma, Zhao and Peng identify two competing transition-and-photon-loss pathways, which they call Blue and Red. Each tends to stabilize a different Floquet state, so the steady state depends on which pathway wins rather than simply whether one is resonant.
In the high-frequency regime, the resonance condition usually does select the dominant route: the paper estimates that a resonant transition can be enhanced over an off-resonant one by a factor of order (ω/κ)², where ω is the drive frequency and κ is the cavity’s photon-loss rate. At intermediate frequencies, however, off-resonant transitions can become strong enough to challenge that advantage. The paper’s transition-rate calculation includes both the energy detuning and the relevant transition matrix elements.
Resonance lines do not guarantee a pure state
The authors numerically solve the open-system dynamics across drive and cavity frequencies, tracking both the spin’s entropy and its fidelity with individual Floquet states. In their two-level example, the simulations use A = 60 and Ω = 100 for the drive and spin parameters, with cavity coupling g = 0.05A, photon loss κ = 0.05A and spin relaxation γ = 0.05κ. These are model parameters, not reported measurements from a physical device.
The calculations reveal two distinct failure patterns. Along some resonance lines, the resonant route loses to a competing transition with a larger matrix element, leaving a high-entropy mixed state. In other narrow frequency windows, two pathways become nearly resonant and their strengths converge; there, changing the cavity frequency cannot recover stabilization in the model.
The study also distinguishes ordinary crossings of quasienergy labels from Floquet-zone hybridization. Near hybridization, states from neighboring Floquet zones mix and the competing transition strengths become equal, removing a preferred direction for relaxation. That removes the one-way flow of population needed to settle into a single Floquet eigenstate.
The same test extends to spin ladders
Ma, Zhao and Peng extend their analysis to multilevel spin-S systems, where multiple Floquet levels create a larger set of possible transitions. Their argument uses the spin system’s su(2) structure: transition strengths favor the same direction between neighboring levels, while a single cavity frequency can satisfy resonance conditions across those neighboring pairs. Stabilization succeeds when the resonant pathway and the preferred transition direction agree; competition or hybridization can still defeat it.
For a five-level example, the authors calculate transition strengths and steady-state fidelities at representative drive frequencies. The results reproduce the three regimes identified in the two-level model: cooperative stabilization, competition between resonance and transition strength, and breakdown near degeneracy or hybridization.
A model-based diagnostic, not an experimental result
The paper’s practical contribution is a way to assess when cavity-assisted stabilization may work: check both the resonance condition and the transition matrix elements that set the relative rates of competing channels. The authors stress that competition depends on the Hamiltonian; it is not guaranteed to occur in every Floquet model. Their work is theoretical and numerical, and does not demonstrate the effect in an experiment.
Extending the analysis to driven many-body systems is the next question identified in the paper. Such systems may have denser quasienergy spectra and more competing pathways, complicating the same comparison. For the models studied here, the central result is specific: a frequency match can identify an available transition, but its strength relative to rival transitions determines whether the cavity actually selects a Floquet state.