Transverse-Field Floquet Ising Model

• The transverse-field Floquet Ising model is a periodically driven spin system combining Ising interactions and transverse fields to yield rich dynamical phases.
• Its experimental implementations, including Rydberg-dressed gases and superconducting qubits, enable engineering of both effective static and inherently Floquet phases.
• The model leverages methods like Jordan–Wigner fermionization to reveal phenomena such as topological edge modes, dynamical phase transitions, and diagnostic operator spreading.

A transverse-field Floquet Ising model is a periodically driven Ising spin system in which an Ising interaction and a field transverse to its ordering axis enter in explicitly time-dependent, stroboscopic, or effectively Floquet-engineered form. In the minimal binary realization, one period is generated by alternating Ising and transverse-field evolution, while in platform-specific implementations the drive often serves as Hamiltonian engineering so that the small-step limit reproduces a static transverse-field Ising Hamiltonian rather than an intrinsically anomalous Floquet phase (Shukla et al., 2020, Borish et al., 2019). The subject now spans exact free-fermion chains, fully connected ferromagnets, Rydberg-dressed atomic gases, superconducting-qubit simulators, quasiperiodic time crystals, nonunitary extensions, and defect-deformed or operator-Krylov reformulations (Arze et al., 2018, Russomanno et al., 2014, Su et al., 2023, Yeh et al., 2023).

1. Definition and scope

The defining structure is an Ising interaction along one spin axis together with a transverse field along a noncommuting axis, but with the Hamiltonian periodic in time. In a direct kicked chain one commonly writes a one-period unitary such as

$U=e^{-iH_{xx}\tau_1}e^{-iH_z\tau_0},$

with $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$ and $H_z=\sum_l \sigma_l^z$, so that the system is a genuinely driven transverse-field Ising chain rather than a static model viewed in a rotating frame (Shukla et al., 2020).

A central taxonomic distinction concerns whether the periodicity is itself the source of new Floquet phases or whether it is primarily a control protocol used to approximate a static transverse-field Ising model. In the Rydberg-dressed cesium experiment, the Ising interaction and the transverse field are applied in alternating time steps, and the small-step limit yields an effective static long-range transverse-field Ising Hamiltonian; genuinely Floquet phenomena such as Floquet criticality, symmetry-protected phases, and time-crystalline behavior are identified there as future directions rather than present observations (Borish et al., 2019). A closely related logic appears in superconducting-qubit proposals where a native $XY$ interaction is reshaped by fast modulation into an effective transverse Ising Hamiltonian in the Floquet basis (Kyriienko et al., 2017).

Not every stroboscopic transverse-field Ising construction is a periodically driven Floquet Ising model in the strict microscopic sense. A static long-range transverse-field Ising model in a large constant field can reproduce $XY$ dynamics at selected stroboscopic times by a rotating-wave approximation, but that setting is best regarded as a Floquet-like effective description rather than a genuinely periodically driven Floquet Ising Hamiltonian (Kiely et al., 2017). This distinction is methodologically important because spectral topology, heating, and quasienergy structure depend on actual periodic driving.

2. Canonical Floquet constructions and effective Hamiltonians

The main Floquet constructions differ in whether the drive alternates pre-existing Ising and field terms, modulates a native interaction into an effective Ising coupling, or engineers anisotropic transverse couplings whose tuned limit is Ising.

Setting	One-period construction	Outcome
Binary kicked chain	$U=e^{-iH_{xx}\tau_1}e^{-iH_z\tau_0}$ (Shukla et al., 2020)	Direct Floquet transverse-field Ising dynamics
Rydberg-dressed cesium gas	alternating $H_{ZZ}$ and $H_X$ with $H_{\mathrm{eff}}\propto \tau_R H_{ZZ}+\tau_X H_X$ (Borish et al., 2019)	Long-range ferromagnetic TFIM in stroboscopic form
Periodically modulated transmons	strong even-sublattice $x$-drive with $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$0 (Kyriienko et al., 2017)	Effective 1D TFIM in the Floquet basis
Tunable-coupler transmons	simultaneous pairing and hopping sidebands, $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$1 (Liang et al., 2024)	Effective $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$2 Ising chain with transverse $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$3 field

In the Rydberg setting, the dressed Ising part is

$H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$4

while the microwave transverse field is

$H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$5

One Floquet cycle alternates these two pieces, and in the regime $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$6 and $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$7 the effective Hamiltonian becomes

$H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$8

The corresponding dimensionless control parameter is

$H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$9

which organizes the mean-field bifurcation associated with the paramagnetic-to-ferromagnetic transition (Borish et al., 2019).

In superconducting transmons with native nearest-neighbor $H_z=\sum_l \sigma_l^z$0 exchange,

$H_z=\sum_l \sigma_l^z$1

a fast modulation on the even sublattice can eliminate the time-averaged $H_z=\sum_l \sigma_l^z$2 contribution. The explicit tuning condition is

$H_z=\sum_l \sigma_l^z$3

and the resulting infinite-frequency Floquet Hamiltonian is

$H_z=\sum_l \sigma_l^z$4

namely a transverse-field Ising model with Ising axis $H_z=\sum_l \sigma_l^z$5 and transverse field $H_z=\sum_l \sigma_l^z$6 (Kyriienko et al., 2017).

A later transmon implementation engineers not only number-conserving hopping,

$H_z=\sum_l \sigma_l^z$7

but also pairing,

$H_z=\sum_l \sigma_l^z$8

by simultaneous red- and blue-sideband modulation of tunable couplers. The effective chain Hamiltonian is

$H_z=\sum_l \sigma_l^z$9

and when $XY$0 with aligned phases, it reduces to

$XY$1

which is an effective Floquet transverse-field Ising chain (Liang et al., 2024).

3. Integrable chains, quasienergies, and Floquet phase structure

The integrable binary-driven chain is exactly solvable by Jordan–Wigner fermionization and Bogoliubov factorization. In momentum space, each $XY$2 sector is governed by a quasienergy angle $XY$3 satisfying

$XY$4

and the exact Heisenberg evolution of fermions is

$XY$5

Within this integrable Floquet transverse Ising system, the long-time-averaged longitudinal-magnetization OTOC distinguishes four known regions of the Floquet phase diagram: $XY$6-paramagnetic, $XY$7-paramagnetic, $XY$8-ferromagnetic, and $XY$9-ferromagnetic; the longitudinal OTOC is nonzero in the ferromagnetic and $XY$0-ferromagnetic regions and vanishes in the paramagnetic and $XY$1-paramagnetic regions (Shukla et al., 2020).

An exact Floquet Hamiltonian is available for the periodically quenched $XY$2 chain and therefore, at anisotropy $XY$3, for the transverse-field Ising limit. In that framework the exact real-space Floquet Hamiltonian contains hopping and pairing amplitudes $XY$4 whose decay is controlled by an unfolded quasienergy spectrum. The central result is an alternation between local and non-local Floquet regions produced by Floquet resonances. In the Ising limit the resonance lines reduce to

$XY$5

and crossing them changes the exact Floquet Hamiltonian from effectively local to one with $XY$6 tails (Arze et al., 2018).

Spatial quasiperiodicity enriches the same free-fermion framework further. In a two-step driven Ising chain with quasiperiodically modulated couplings,

$XY$7

with

$XY$8

the model supports $XY$9- and $U=e^{-iH_{xx}\tau_1}e^{-iH_z\tau_0}$0-Majorana edge modes, mobility-edge-like localization structure, and fully localized ferromagnetic phases near $U=e^{-iH_{xx}\tau_1}e^{-iH_z\tau_0}$1 and $U=e^{-iH_{xx}\tau_1}e^{-iH_z\tau_0}$2. In the fully localized ferromagnetic phase near $U=e^{-iH_{xx}\tau_1}e^{-iH_z\tau_0}$3, localized domain-wall excitations preserve eigenstate order and the $U=e^{-iH_{xx}\tau_1}e^{-iH_z\tau_0}$4-Majorana mode enforces exact $U=e^{-iH_{xx}\tau_1}e^{-iH_z\tau_0}$5-spectral pairing, yielding a $U=e^{-iH_{xx}\tau_1}e^{-iH_z\tau_0}$6 Floquet time crystal (Liang et al., 2020).

4. Experimental realizations

The clearest atomic realization is the Rydberg-dressed cesium gas. The effective spin-$U=e^{-iH_{xx}\tau_1}e^{-iH_z\tau_0}$7 basis is encoded in the clock states

$U=e^{-iH_{xx}\tau_1}e^{-iH_z\tau_0}$8

with only $U=e^{-iH_{xx}\tau_1}e^{-iH_z\tau_0}$9 dressed to a Rydberg level near $H_{ZZ}$0 by a $H_{ZZ}$1 laser. The many-body dressed Hamiltonian is

$H_{ZZ}$2

whose Ising part is

$H_{ZZ}$3

In the perturbative dressing regime,

$H_{ZZ}$4

and for $H_{ZZ}$5 the interactions are ferromagnetic, $H_{ZZ}$6. The experiment operates near a Förster resonance with defect

$H_{ZZ}$7

giving a soft-core radius $H_{ZZ}$8. Ramsey spectroscopy yields a mean-field interaction strength $H_{ZZ}$9, and the Floquet sequence uses, for example,

$H_X$0

The observed dynamical hallmark is the mean-field bifurcation of fixed points: for $H_X$1 only $H_X$2 is stable, while for $H_X$3 two ferromagnetic fixed points appear at

$H_X$4

with the effective critical condition shifted by contrast decay to $H_X$5 (Borish et al., 2019).

In superconducting hardware, a six-transmon chain with tunable couplers realizes the effective Hamiltonian

$H_X$6

with red and blue sidebands calibrated through synthetic-space Aharonov–Bohm interference. For the two-qubit demonstration,

$H_X$7

so the red and blue sidebands are

$H_X$8

With

$H_X$9

the chain reduces to

$H_{\mathrm{eff}}\propto \tau_R H_{ZZ}+\tau_X H_X$0

The experiment studies a finite-size dynamical phase transition through the average-time correlation

$H_{\mathrm{eff}}\propto \tau_R H_{ZZ}+\tau_X H_X$1

with $H_{\mathrm{eff}}\propto \tau_R H_{ZZ}+\tau_X H_X$2, and through the Loschmidt echo

$H_{\mathrm{eff}}\propto \tau_R H_{ZZ}+\tau_X H_X$3

for $H_{\mathrm{eff}}\propto \tau_R H_{ZZ}+\tau_X H_X$4. For $H_{\mathrm{eff}}\propto \tau_R H_{ZZ}+\tau_X H_X$5, the crossover sharpens near $H_{\mathrm{eff}}\propto \tau_R H_{ZZ}+\tau_X H_X$6 but remains finite-size rounded rather than singular (Liang et al., 2024).

5. Nonequilibrium many-body phenomena beyond static emulation

The fully connected periodically driven Ising ferromagnet provides a clean setting in which Floquet dynamics, thermalization, and classical chaos can be compared directly. The model

$H_{\mathrm{eff}}\propto \tau_R H_{ZZ}+\tau_X H_X$7

reduces for $H_{\mathrm{eff}}\propto \tau_R H_{ZZ}+\tau_X H_X$8 to a fully connected transverse-field Ising ferromagnet. When the corresponding $H_{\mathrm{eff}}\propto \tau_R H_{ZZ}+\tau_X H_X$9 classical dynamics is ergodic, the quasienergy statistics are Wigner–Dyson, Floquet states satisfy ETH, and local observables relax to the $x$0 value; when the classical dynamics is regular, quasienergy spacings are Poissonian and the long-time Floquet diagonal ensemble retains memory of the initial state (Russomanno et al., 2014).

Floquet perturbation theory can also produce constrained prethermal dynamics rather than heating. In a driven Ising chain with

$x$1

where $x$2 flips sign halfway through the period, the first-order Floquet matrix elements acquire a sinc filter,

$x$3

At $x$4 and $x$5, the surviving bulk dynamics is

$x$6

namely a PXP-type constrained model; under periodic boundary conditions, the largest surviving fragment is exactly the PXP sector at leading order. At $x$7 with $x$8, all first-order single-spin-flip channels are suppressed and $x$9, producing the regime termed Floquet freezing (Joshi et al., 24 Mar 2026).

Periodic driving can also accelerate collective tunneling. In one and two dimensions, the driven ferromagnetic Ising model

$H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$00

shows three regimes as the drive strength increases: weak drive with exponentially decaying tunneling rates and robust magnetic order, an intermediate regime with polynomial decay of tunneling and vanishing magnetic order, and a strong-drive regime in which the tunneling rate becomes nearly size-independent while time-averaged magnetic order reappears. The static 1D splitting scales as

$H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$01

whereas in the weak-drive Floquet regime the finite-size fit is

$H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$02

The strong-drive coexistence of rapid macroscopic tunneling and recovered order is explicitly identified as not captured by the average-Hamiltonian treatment alone (Grattan et al., 2023).

A different use of periodic driving is to change the operator content of the effective Hamiltonian. For

$H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$03

a rotating-frame transformation followed by the high-frequency approximation gives

$H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$04

In a 1D nearest-neighbor chain this reduces to

$H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$05

with

$H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$06

The resulting ordered phase is characterized by a nonlocal string order parameter rather than conventional magnetic order, and the transition occurs at $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$07 (Lee et al., 2016).

Transport control furnishes yet another Floquet-Ising use case. In a nearest-neighbor chain with

$H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$08

periodic modulation of a selected bond can suppress propagation across that location, giving an effective leading Floquet Hamiltonian consisting of two disconnected $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$09 chains. For a locally modulated transverse field,

$H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$10

the two couplings adjacent to the driven site are renormalized to

$H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$11

so choosing $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$12 with $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$13 makes $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$14 and acts as a Floquet quantum switch (Ahumada et al., 13 Jan 2025).

6. Diagnostics, extensions, and interpretive issues

The Floquet transverse-field Ising model is diagnostically rich. Exact and numerical OTOCs distinguish ballistic operator spreading from phase sensitivity and from chaos or integrability. In the kicked quantum Ising Floquet system, transverse-magnetization and longitudinal-magnetization OTOCs exhibit a characteristic region, a dynamic region with power-law growth rather than clean exponential growth, and a near-saturation region where integrable and nonintegrable behavior separate clearly. The same work emphasizes that the kicked transverse-field Ising model has four Floquet phases and that the longitudinal-magnetization OTOC, not the transverse one, acts as the useful order parameter for ferromagnetic versus paramagnetic Floquet regions (Shukla, 12 May 2025).

Allowing complex couplings yields a nonunitary Floquet TFIM,

$H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$15

with phases $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$16, $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$17, $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$18, and $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$19 distinguished by real edge modes at quasienergy $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$20 and/or $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$21. The steady-state entanglement scaling is area law when there are no real quasiparticle modes, logarithmic when there are finitely many, and volume law when there is an extensive set. The volume-law critical lines at $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$22 and $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$23 are tied to pseudo-Hermiticity (Su et al., 2023).

The Floquet TFIM also appears as a universal canonical model for operator dynamics. The operator Krylov space of any Hermitian operator under any Floquet unitary can be mapped exactly to the Krylov space generated by the edge Majorana of a 1D inhomogeneous Floquet transverse-field Ising model with open boundaries. In that mapping, the Krylov angles play the role of local transverse fields and Ising couplings, and the four topological phases of the homogeneous Floquet TFIM provide the universal language for long-lived $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$24- and $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$25-type operator modes (Yeh et al., 2023). A complementary generalization inserts a Kramers–Wannier duality defect into a deformed Floquet TFIM; the undeformed defect Floquet unitary hosts an isolated defect zero mode, while a weak self-dual deformation causes its memory to decay, with finite-size plateaus interpreted as residual finite-size effects rather than asymptotic stability (Yan et al., 2024).

Two interpretive cautions recur across the literature. First, a periodically driven Ising-plus-transverse-field protocol need not realize an intrinsically nontrivial Floquet phase: in both the Rydberg-dressed cesium system and the modulated-transmon construction, the drive is used primarily to synthesize an effective static TFIM, not to demonstrate anomalous quasienergy topology or time-crystalline order (Borish et al., 2019, Kyriienko et al., 2017). Second, exact free-fermion statements about quasienergy bands, edge modes, or OTOCs depend on integrability and specific drive protocols; once longitudinal fields, defect deformations, or non-Hermitian terms are added, the same terminology labels a wider class of near-integrable, nonintegrable, or nonunitary systems rather than a single solvable model (Su et al., 2023, Yan et al., 2024).

In aggregate, the transverse-field Floquet Ising model is less a single Hamiltonian than a structured family of periodically driven Ising problems. Its unifying content is the noncommuting interplay of Ising order, transverse spin rotation, and periodic time dependence; its technical diversity lies in whether that time dependence engineers an effective static TFIM, produces exact Floquet topological structure with $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$26 and $H_{xx}=\sum_l \sigma_l^x\sigma_{l+1}^x$27 sectors, induces transport blocking or constrained dynamics, stabilizes nonunitary steady phases, or furnishes a universal representation of Floquet operator dynamics itself (Shukla et al., 2020, Arze et al., 2018, Yeh et al., 2023).