Coherent Phonons Make Floquet-Bloch States Last Longer

Floquet engineering usually begins with light: shine an intense, periodic electromagnetic field on a crystal, and the electrons can temporarily behave as if the material has a new band structure. A new preprint points to a quieter driver. Instead of using the laser field itself as the clock, Yu-Chan Tai, Chih-Wei Luo, Noriaki Takagi, Hiroshi Ishida, Chun-Liang Lin and Ryuichi Arafune report evidence that a coherent lattice vibration—a phonon with a well-defined phase—can dress electronic states into Floquet-Bloch states on graphene-covered Ir(111).

The paper, “Phonon-driven Floquet-Bloch states probed by quantum beat spectroscopy”, posted to arXiv on June 29, 2026, is not a claim of a new power plant or room-temperature energy device. Its importance is more foundational: it demonstrates a time-domain method for seeing Floquet sidebands when their spacing is too small to resolve cleanly in ordinary energy spectra, and it shows that material vibrations can sustain the Floquet drive beyond the lifetime of a femtosecond laser pulse.

If light-driven Floquet matter is a camera flash, phonon-driven Floquet matter may be more like a ringing bell: the excitation is launched quickly, but the periodic motion can keep time after the optical flash is gone.

Why Floquet engineers care about the clock

Floquet theory describes systems whose Hamiltonian repeats in time. The mathematical foundations go back to Joseph Shirley’s 1965 solution of the periodically driven Schrödinger equation and Hideo Sambe’s 1973 quasienergy formulation. In solids, this becomes the Floquet-Bloch picture: ordinary Bloch bands, which arise from spatial periodicity in a crystal, acquire a temporal periodicity as well. The result is a ladder of quasienergy copies, or sidebands, separated by the energy quantum of the drive.

For the last decade, the most visible path has been optical Floquet engineering. Experiments have reported Floquet-Bloch states on topological-insulator surfaces, light-induced anomalous Hall signatures in graphene, and subcycle build-up and dephasing of driven bands. These are beautiful demonstrations, but they face a practical constraint. To generate a strong effect without overheating or damaging the sample, researchers often use intense ultrafast pulses. The dressed state may therefore exist only during the pulse—roughly a hundred femtoseconds in the new paper’s comparison.

That is enough time to prove a concept. It is much less attractive for eventual devices that need controlled energy flow, switching, sensing, or thermodynamic cycles. A Floquet energy technology needs not just an exotic state, but a usable state: long enough lived, phase coherent, and measurable without destroying the very order being engineered.

The new ingredient: a coherent phonon

A phonon is a quantized vibration of a lattice. In everyday language, it is a tiny collective motion of atoms in a solid. Most phonons in a warm material are thermally populated and phase-random; they jiggle, but they do not provide the clean periodicity that Floquet theory needs. A coherent phonon, by contrast, is launched with a well-defined phase. For a short time, many atoms participate in a synchronized vibration that can act as an internal metronome for the electrons.

Tai and colleagues used graphene on an iridium surface, written graphene/Ir(111), because that interface hosts image-potential states—electron states outside the surface that can be observed with time-resolved multiphoton photoemission. In the experiment, ultraviolet pump pulses with photon energy 4.71 eV excited coherent phonons and populated image-potential states. Infrared probe pulses at 1.57 eV then sampled how those states evolved as the time delay was varied.

The trick is that the expected phonon-driven sideband spacing is small. The paper notes that conventional energy resolution is typically around 30 meV, while the relevant phonon energy in this system is only about 5.4 meV. Rather than trying to see every sideband directly as a separated peak, the authors used quantum beat spectroscopy. When two close-lying coherent pathways interfere, their energy difference appears as an oscillation in time. Measuring the beat period can therefore reveal an energy splitting that is too fine for the static spectrum.

What the experiment saw

The central observation is a delay-dependent oscillation in the unresolved high-order image-potential-state region, labeled n ≥ 4 in the paper. The intensity is enhanced near 0.55, 1.4 and 2.2 picoseconds, and weaker near 1.05, 2.0 and 2.7 picoseconds. Fourier analysis assigns this oscillation a frequency of 1.3 THz, corresponding to 5.4 meV.

That number matters because it matches a phonon mode previously identified in the same graphene/Ir(111) system by helium atom scattering and high-resolution electron energy-loss spectroscopy. The authors identify it as a moiré-induced localized out-of-plane acoustic phonon of the graphene layer. In Floquet language, a periodic drive at angular frequency Ω produces quasienergy sidebands separated by ℏΩ. Here, the beat energy is the phonon energy.

The paper also reports an independent oscillation in the photoemission intensity of the Fermi-Dirac tail, spectrally distinct from the image-potential-state windows. Fitting this channel gives 5.40 ± 0.03 meV. The n = 3 and n ≥ 4 image-state beats show matching low-energy Fourier components and phase relationships to the coherent-phonon reference. This is important because a mere intensity wiggle would be ambiguous. The shared frequency and phase-locking argue that the electronic quantum beats are being driven by the same coherent lattice motion.

Why this differs from ordinary light-driven Floquet bands

In a conventional light-driven Floquet experiment, the optical field is the periodic drive. If the field is gone, the direct clock is gone. Coherent phonons offer a different architecture. A short optical pulse can launch a lattice oscillation, and the lattice itself can continue to modulate the electronic Hamiltonian for picoseconds or, in favourable systems, tens of picoseconds. The drive becomes partly material-internal.

This is not automatically better in every respect. Phonon energies are often much smaller than optical photon energies, so the sidebands are closer together. Coupling strengths, decoherence channels and material-specific modes all matter. A phonon-driven strategy also requires phase coherence; thermal phonons alone will not do the job. But if coherent vibrations can reliably dress electrons, they could reduce the dependence on extreme optical fields and open a route to more persistent driven phases.

For Floquet.ca’s energy-focused audience, the key point is not that this graphene/iridium surface is itself an energy device. The point is that long-lived periodic drives are a missing ingredient in many proposed quantum-energy applications. Quantum heat engines, quantum batteries, nonreciprocal thermal routers and beyond-Carnot research programs all depend on controlling when and how energy quanta move between subsystems. A coherent phonon is a natural candidate for a narrowband, material-compatible energy clock.

How this connects to practical quantum energy

There are three practical lessons from the experiment.

1. Time-domain probes can reveal hidden Floquet structure

Energy-domain spectroscopy is not always enough. If quasienergy spacings are below the instrument resolution, a driven state may be present but visually smeared out. Quantum beat spectroscopy converts the problem into a clock measurement: let the coherent pathways interfere, then read out the frequency. This could matter for low-energy drives such as phonons, magnons, plasmons and microwave modes.

2. Internal collective modes may be gentler drive sources

A material vibration can be launched with light but maintained by the material’s own coherent motion. That could reduce the need to keep applying a strong optical field throughout the entire Floquet window. In the long run, the most useful Floquet materials may be those in which a modest external pulse excites a robust internal oscillator that does the fine work of band dressing or energy routing.

3. Phase coherence is the resource

The paper repeatedly distinguishes coherent phonons from thermal phonons. This is exactly the language of quantum thermodynamics: energy by itself is not enough. The phase, timing and correlations of that energy determine whether it can perform structured work. In quantum batteries this distinction appears as ergotropy; in heat engines it appears as the difference between useful work extraction and dissipated heat. In Floquet materials, it appears as the difference between random lattice noise and a usable temporal symmetry.

The experiment reframes phonons from a nuisance heat bath into a possible Floquet resource—provided their motion is coherent, phase-locked and experimentally readable.

What to watch next

The immediate next steps are likely comparative. Researchers will want to know which materials support the longest coherent phonon clocks, which electronic states couple most strongly, and whether phonon-driven sidebands can be steered into topological, superconducting, thermoelectric or spintronic functionality. The new preprint already points toward multimode phonon scenarios, where more than one coherent vibration could generate richer quasienergy structures.

There is also a measurement challenge. The graphene/Ir(111) result benefits from image-potential states that are unusually clean to probe. Future quantum-energy devices will not necessarily provide such convenient spectroscopic handles. Translating this physics into useful platforms will require probes that can operate in more realistic materials and device geometries.

Still, the conceptual advance is clear. Floquet engineering is often portrayed as a battle against heating: drive matter strongly enough to change it, but not so strongly that the sample thermalizes into uselessness. Phonon-driven Floquet-Bloch states suggest another strategy: use the material’s own coherent motion as part of the drive architecture. If that strategy generalizes, Floquet engineers may gain a longer-lived, lower-energy knob for shaping quantum matter—and quantum energy flow—on demand.

Sources and further reading

• Yu-Chan Tai, Chih-Wei Luo, Noriaki Takagi, Hiroshi Ishida, Chun-Liang Lin and Ryuichi Arafune, “Phonon-driven Floquet-Bloch states probed by quantum beat spectroscopy”, arXiv:2606.30065 (2026).
• J. H. Shirley, “Solution of the Schrödinger Equation with a Hamiltonian Periodic in Time”, Physical Review 138, B979 (1965).
• H. Sambe, “Steady States and Quasienergies of a Quantum-Mechanical System in an Oscillating Field”, Physical Review A 7, 2203 (1973).
• Y. H. Wang et al., “Observation of Floquet-Bloch States on the Surface of a Topological Insulator”, Science 342, 453 (2013).
• J. W. McIver et al., “Light-induced anomalous Hall effect in graphene”, Nature Physics 16, 38 (2020).