Finite-Pulse Floquet Analysis

• Finite-pulse Floquet analysis is a framework that extends traditional Floquet theory by incorporating realistic, finite-duration control fields and pulse shapes.
• It captures experimental effects such as phase wrapping, quasienergy branch folding, and short-time occupations that influence observable dynamics.
• Multiple formulations—including one-cycle propagators and instantaneous Floquet states—enable precise pulse design and analysis of metrological limits.

Finite-pulse Floquet analysis denotes a set of Floquet-theoretic methods for periodically driven systems in which the control field is treated with its actual temporal structure—finite-duration pulses, slow envelopes, switching, phase steps, or finite sequence length—rather than as an ideal train of instantaneous kicks or an infinitely repeated monochromatic drive. In this setting, the relevant objects are not only quasienergies and effective Hamiltonians, but also one-cycle propagators, micromotion, branch structure, finite-time occupations, and the relation between stroboscopic observables and the underlying driven dynamics. In a dense diamond NV ensemble under periodic WAHUHA control, this perspective was used to show that increasing the effective stroboscopic dephasing time from $T_2^\ast \approx 0.9\,\mu\text{s}$ to $T_{2,\mathrm{eff}}^\ast \approx 31\,\mu\text{s}$ produced little improvement in dc magnetic-field sensitivity, because the observed long-lived signal arose from phase wrapping and quasi-energy branch folding rather than from a proportional increase in usable phase transduction (Nguyen et al., 9 Jun 2026).

1. Periodic driving beyond the ideal Floquet limit

Standard Floquet theory begins from a Hamiltonian satisfying $H(t+T)=H(t)$, with $\Omega=2\pi/T$, and writes solutions as $\psi_\alpha(t)=e^{-iE_\alpha t}u_\alpha(t)$ with $u_\alpha(t+T)=u_\alpha(t)$. The periodic part can be expanded in harmonics, $u_\alpha(t)=\sum_m e^{-im\Omega t}u_m^\alpha$, so that the drive generates Floquet replicas separated by integer multiples of $\Omega$, and quasienergies are defined modulo $\Omega$ (Giovannini et al., 2019). In this formulation, the periodic drive is the temporal analog of a lattice, and the quasienergy ladder plays the role of a Brillouin-zone structure in harmonic space.

Finite-pulse Floquet analysis arises because many experimentally relevant drives are not periodic for all times. Real protocols involve finite pulse width, switch-on and switch-off, slowly varying envelopes, and pulse sequences with non-continuous Hamiltonians. Several works therefore recast Floquet theory as an analysis tool for finite-duration dynamics rather than a theorem restricted to eternal periodicity. In pulse-laser-driven two-level systems, dynamics can be organized in terms of instantaneous Floquet states (IFSs), with adiabatic evolution along quasienergy branches interrupted by Landau–Zener-type transitions at avoided crossings; this description remains useful for shorter pulses down to about $2$ cycles in the model studied (Ikeda et al., 2022). In ultrafast photoionization of Ne, a Floquet state was reported to be established within about $T_{2,\mathrm{eff}}^\ast \approx 31\,\mu\text{s}$0 optical cycles, while pulses as short as $T_{2,\mathrm{eff}}^\ast \approx 31\,\mu\text{s}$1 cycles were still described by a finite-pulse Floquet-like model (Lucchini et al., 2022).

A central consequence is that finite pulses generally affect occupations, spectral weights, delay dependence, and the visibility of sidebands even when the underlying quasienergy structure remains informative. This suggests a methodological shift: the relevant question is often not whether a periodic effective Hamiltonian exists in the abstract, but whether the actual finite-time driven state is sufficiently close to a locally periodic dressed state for Floquet observables to be meaningful (Giovannini et al., 2019).

2. Mathematical formulations for finite-pulse Floquet dynamics

One common starting point is the one-cycle propagator

$T_{2,\mathrm{eff}}^\ast \approx 31\,\mu\text{s}$2

which defines a Floquet Hamiltonian $T_{2,\mathrm{eff}}^\ast \approx 31\,\mu\text{s}$3 through the logarithm of the exact single-cycle evolution (Nguyen et al., 9 Jun 2026). A closely related construction uses the one-cycle evolution operator $T_{2,\mathrm{eff}}^\ast \approx 31\,\mu\text{s}$4 and the effective Floquet Hamiltonian

$T_{2,\mathrm{eff}}^\ast \approx 31\,\mu\text{s}$5

whose eigenvalues are quasienergies defined modulo $T_{2,\mathrm{eff}}^\ast \approx 31\,\mu\text{s}$6 (Puente-Uriona et al., 2024). This formulation is particularly useful when finite geometries, basis truncations, or gauge consistency are essential.

A second formulation introduces two timescales. In pulse-driven two-level systems and finite-pulse band problems, the fast carrier oscillation is treated as periodic at fixed amplitude, while the envelope becomes a slow parameter. The state is expanded in an instantaneous Floquet basis, and the coefficients obey an effective equation of the form

$T_{2,\mathrm{eff}}^\ast \approx 31\,\mu\text{s}$7

so the diagonal quasienergies govern adiabatic phase accumulation and the $T_{2,\mathrm{eff}}^\ast \approx 31\,\mu\text{s}$8 term generates non-adiabatic transitions between Floquet branches (Gómez et al., 12 Feb 2025). The same structural decomposition appears in the IFS treatment of strong pulse lasers, where the effective Floquet-Sambe Hamiltonian contains $T_{2,\mathrm{eff}}^\ast \approx 31\,\mu\text{s}$9 on the diagonal and $H(t+T)=H(t)$0 as the envelope-driven coupling (Ikeda et al., 2022).

A third formulation generalizes Floquet theory to finite-duration and non-continuous Hamiltonians. Continuous Floquet theory writes the actual finite-time Hamiltonian as $H(t+T)=H(t)$1, where $H(t+T)=H(t)$2 is a rectangular window, and derives duration-dependent effective Hamiltonians. At first order, each Fourier component is weighted by

$H(t+T)=H(t)$3

so resonance conditions acquire finite width rather than being enforced only at exact $H(t+T)=H(t)$4 (Chávez et al., 2024). A related continuous-frequency-space construction replaces discrete harmonics by a Fourier transform, leading at second order to

$H(t+T)=H(t)$5

and thereby extends Floquet-style perturbation theory to non-periodic pulse design in solid-state NMR (Chávez et al., 2022).

3. Stroboscopic spectra, wrapped phases, and finite-pulse branch structure

The dense NV-ensemble study under WAHUHA provides a concrete finite-pulse Floquet analysis in which the measured signal is sampled stroboscopically after each control cycle. The one-cycle unitary is built from the full ordered product of pulse-segment unitaries, with finite-width microwave pulses, free-evolution segments, detuning $H(t+T)=H(t)$6, hyperfine shifts, residual interactions, and control imperfections all contributing to the net phase (Nguyen et al., 9 Jun 2026). After removing the global phase of the $H(t+T)=H(t)$7 one-cycle unitary, the relevant quantity is the Floquet eigenphase splitting

$H(t+T)=H(t)$8

The measured magnetization after $H(t+T)=H(t)$9 cycles is modeled as

$\Omega=2\pi/T$0

where $\Omega=2\pi/T$1 is the wrapped Floquet phase, and the stroboscopic spectrum has a peak at

$\Omega=2\pi/T$2

Because the signal is sampled only once per cycle, the one-cycle phase is subject to a Nyquist-like restriction. When the unwrapped phase exceeds the principal interval, it is folded back. The long-lived stroboscopic signal at large interpulse delay was therefore attributed to phase wrapping and quasi-energy branch folding of the one-cycle unitary, not simply to “better coherence” (Nguyen et al., 9 Jun 2026).

This branch structure has direct metrological consequences. In the weak-phase regime,

$\Omega=2\pi/T$3

and the signal derivative satisfies

$\Omega=2\pi/T$4

The dc response is therefore controlled by the detuning-to-phase transduction slope $\Omega=2\pi/T$5, together with contrast and lifetime. In the long-delay regime the wrapped phase becomes locally flattened in detuning, so $\Omega=2\pi/T$6 is reduced even though the envelope decays slowly (Nguyen et al., 9 Jun 2026).

Finite-pulse effects also generate short-delay spectral features beyond a simple linear model. These were captured by

$\Omega=2\pi/T$7

where $\Omega=2\pi/T$8 gives a detuning-axis offset and $\Omega=2\pi/T$9 represents a residual transverse Floquet component arising from finite pulse width, pulse distortion, or control imperfections. The associated zero-frequency crossing occurs at

$\psi_\alpha(t)=e^{-iE_\alpha t}u_\alpha(t)$0

In this framework, apparent $\psi_\alpha(t)=e^{-iE_\alpha t}u_\alpha(t)$1-period doubling is a consequence of branch folding near $\psi_\alpha(t)=e^{-iE_\alpha t}u_\alpha(t)$2, not evidence for a many-body discrete time crystal (Nguyen et al., 9 Jun 2026).

4. Finite-pulse corrections as a control and engineering problem

In pulse-driven NV platforms, finite-pulse corrections can be quantitatively dominant. For a single NV center driven by a Carr–Purcell sequence in a large-amplitude AC magnetic field, the finite-pulse Hamiltonian explicitly retains the pulse width $\psi_\alpha(t)=e^{-iE_\alpha t}u_\alpha(t)$3, with $\psi_\alpha(t)=e^{-iE_\alpha t}u_\alpha(t)$4, instead of replacing $\psi_\alpha(t)=e^{-iE_\alpha t}u_\alpha(t)$5 pulses by delta functions (Nishimura et al., 2022). In the ideal limit the synchronized sequential readout yields harmonic amplitudes

$\psi_\alpha(t)=e^{-iE_\alpha t}u_\alpha(t)$6

and harmonics up to the $\psi_\alpha(t)=e^{-iE_\alpha t}u_\alpha(t)$7-th order were observed. Numerical solutions of the full time-dependent Schrödinger equation showed that finite pulse duration and a small pulse error reproduce the irregular fluctuations in the data and explain weak even-order harmonics, while longer pulses progressively destroy the high-order response because the pulse bandwidth $\psi_\alpha(t)=e^{-iE_\alpha t}u_\alpha(t)$8 becomes too narrow (Nishimura et al., 2022).

Finite-pulse analysis also exposes symmetry breaking in more elaborate spin-control sequences. In PulsePol, previous treatments assumed very strong, near-ideal, instantaneous microwave pulses. A bimodal Floquet treatment of the finite-pulse regime showed that standard PulsePol loses selectivity because finite pulses distort the interaction-frame trajectory and break the quadrature and XY-time-reversal symmetries required for pure double-quantum or zero-quantum transfer (Jhamnani et al., 6 Apr 2026). The minimal remedy was to change the phase of the central $\psi_\alpha(t)=e^{-iE_\alpha t}u_\alpha(t)$9-pulse from $u_\alpha(t+T)=u_\alpha(t)$0 to $u_\alpha(t+T)=u_\alpha(t)$1, producing Q-PulsePol and restoring the symmetry relations $u_\alpha(t+T)=u_\alpha(t)$2 for the relevant transfer mode (Jhamnani et al., 6 Apr 2026).

A more general non-stroboscopic high-frequency expansion further enlarges the design space by allowing arbitrary periodic pulse shapes of finite duration. In that framework, the drive is written as $u_\alpha(t+T)=u_\alpha(t)$3 with $u_\alpha(t+T)=u_\alpha(t)$4, and the lowest-order kick operator is

$u_\alpha(t+T)=u_\alpha(t)$5

The effective Hamiltonian then depends on the full hierarchy of even moments $u_\alpha(t+T)=u_\alpha(t)$6, not only on a small set of Fourier coefficients, which permits pulse-shape-controlled engineering of effective XXZ, XY, or Ising interactions (Scott et al., 28 Apr 2025). A plausible implication is that “finite-pulse Floquet analysis” increasingly functions not only as a diagnostic language but also as a pulse-design language.

5. Finite-pulse Floquet analysis across experimental domains

In ultrafast and strong-field physics, finite-pulse Floquet methods often serve to interpret sidebands, dressed-state occupations, and transient interference. One approach extracts Floquet states and quasi-energies directly from a real-time wavefunction without solving the Floquet eigenproblem: the propagated state is Fourier analyzed in time, optionally after parity projection, to identify populated light-induced states and their quasi-energy ladders (Kapoor et al., 2011). This strategy was used to explain even harmonics in an inversion-symmetric system as interference between multiple populated Floquet states, to track gauge-dependent dressed-state populations, and to interpret channel-closing enhancements in above-threshold ionization (Kapoor et al., 2011).

In condensed-matter and band problems, finite pulses are commonly treated as approximations to an ideal Floquet regime whose validity must be assessed. For noninteracting lattice fermions driven by light pulses, the approach to the infinite-duration Floquet limit depends strongly on the Bessel factor $u_\alpha(t+T)=u_\alpha(t)$7, on the pulse width, and on the diagnostic used; the horizontal density of states can converge more cleanly than the diagonal one, while Gaussian pulses require additional averaging in frequency to expose the Floquet structure (Kalthoff et al., 2018). In monolayer transition metal dichalcogenides under finite-pulse polarized radiation, the $u_\alpha(t+T)=u_\alpha(t)$8 formalism separates fast carrier oscillations from the slow Gaussian envelope, yielding a time-dependent Floquet spectrum, transient valley polarization, and time-dependent circular dichroism (Gómez et al., 12 Feb 2025).

Photonic and wave-scattering problems motivate yet another variant. For a real, time-periodic photonic slab of finite spatial extent, the correct asymptotic object is a static Floquet scattering matrix $u_\alpha(t+T)=u_\alpha(t)$9 satisfying the pseudounitary relation

$u_\alpha(t)=\sum_m e^{-im\Omega t}u_m^\alpha$0

with $u_\alpha(t)=\sum_m e^{-im\Omega t}u_m^\alpha$1 encoding the sign of positive- and negative-frequency channels. This structure follows from conservation of wave action rather than energy, and it supports a Floquet Wigner–Smith matrix whose eigenstates are multifrequency spatiotemporal pulses shaped for optimal interaction with the modulated medium (Globosits et al., 2024). Finite extent, finite channel truncation, and explicit sideband conversion are therefore part of the Floquet analysis from the outset rather than small corrections.

6. Interpretive limits, misconceptions, and broader significance

A recurrent misconception is that a longer effective dephasing time under periodic driving necessarily implies better sensing or control. The NV-ensemble WAHUHA study gives a direct counterexample: $u_\alpha(t)=\sum_m e^{-im\Omega t}u_m^\alpha$2 increased by more than an order of magnitude, yet dc magnetic-field sensitivity improved only weakly, if at all, because the periodic drive suppressed the detuning-to-phase transduction slope through phase wrapping and quasi-energy folding (Nguyen et al., 9 Jun 2026). In a different register, apparent period doubling in the same experiment was traced to branch structure in a stroboscopically sampled Floquet spectrum, not to a many-body discrete time crystal (Nguyen et al., 9 Jun 2026).

A second limit concerns validity. Several finite-pulse formulations explicitly retain Floquet language even when exact periodicity is absent, but they do so under controlled assumptions: slowly varying envelopes relative to the carrier cycle, local periodicity, or a finite observation window during which Floquet coefficients become nearly time independent (Giovannini et al., 2019). Few-cycle pulse analyses show that this can remain quantitatively useful surprisingly far from the infinite-drive ideal, but they also identify breakdown regimes near abrupt pulses or when transitions are no longer localized near avoided crossings (Ikeda et al., 2022). In ultrashort photoionization, distinct sidebands cease to be well resolved below about $u_\alpha(t)=\sum_m e^{-im\Omega t}u_m^\alpha$3 optical cycles because the slow-envelope approximation breaks down and the XUV bandwidth becomes comparable to the IR photon energy (Lucchini et al., 2022).

A third limit is formal rather than physical. In finite-basis calculations, the quasienergy spectrum is gauge invariant only if the electromagnetic gauge transformation is implemented exactly. In finite systems, the exact gauge-covariant velocity-gauge Hamiltonian preserves quasienergy invariance, whereas the naive truncated velocity-gauge form can fail badly because the projected basis no longer realizes the canonical commutator exactly (Puente-Uriona et al., 2024). This suggests that finite-pulse Floquet analysis is inseparable from representation issues: quasienergies, micromotion, occupations, and effective Hamiltonians can all depend on how the finite-time drive is represented and sampled.

Taken together, these developments establish finite-pulse Floquet analysis as a practical framework for realistic periodically driven systems. Its distinctive contribution is not merely to correct ideal Floquet results perturbatively, but to specify what the experiment actually measures when pulse width, envelope, stroboscopic sampling, branch folding, finite system size, or finite sequence duration are integral parts of the dynamics.