Floquet Topological States

• Floquet topological states are non-equilibrium quantum phases in periodically driven systems characterized by stroboscopic dynamics and dynamical invariants.
• Researchers apply Floquet operator and transfer-matrix methods to compute reflection phases and winding numbers, distinguishing trivial from nontrivial topological gaps.
• These states have practical applications in photonic, cold-atom, and mesoscopic systems, fostering the design of robust, dynamically controlled quantum devices.

Floquet topological states are non-equilibrium quantum phases arising in periodically driven systems, characterized by topological invariants of the stroboscopic (Floquet) dynamics rather than those of static Hamiltonians. Unlike in equilibrium, these states can exhibit topological properties associated with quasienergy and quasimomentum gaps, support edge and interface modes with no static analogs, and allow for dynamical transfer or creation of nontrivial gap invariants in hybrid space-time dimensions. The classification, detection, and physical realization of these states rely on both explicit Floquet operator constructions and advanced transfer-matrix or scattering formulations, facilitating analysis in Hermitian and non-Hermitian systems. Applications span photonic, cold-atom, and mesoscopic platforms.

1. Floquet Theory, Hybrid Space-Time Formalism, and Transfer Matrices

Periodically driven systems with $H(t+T)=H(t)$ admit stroboscopic evolution described by the Floquet operator $U_F = \mathcal T \exp(-i\int_0^T H(t)dt)$, with the quasienergy spectrum defined by $U_F|\psi\rangle = e^{-i\epsilon T}|\psi\rangle$, where $\epsilon \in [-\pi/T,+\pi/T)$. In spatiotemporally modulated lattices, both spatial periodicity and temporal driving are present. The conventional approach constructs $H(k,\epsilon)$ in a 2D hybrid Brillouin zone and analyzes bulk topology via Berry curvature $F(k,\epsilon)$ and Chern numbers. This can be cumbersome—especially in non-Hermitian systems or those with strong modulation.

An alternative, leveraging transfer-matrix (TM) techniques, avoids full construction of $H(k,\epsilon)$. Instead, space-direction ($\mathcal{Y}_x(\epsilon)$) and time-direction ($\mathcal{Y}_t(k)$) transfer matrices directly encode scattering across the relevant band gaps. Explicitly, for a two-scatterer unit cell (U, S) with S-matrices $U, S$, the transfer matrix in space is

$$M(\epsilon,k) = \sqrt{M_U} \cdot e^{i(\epsilon/2)\sigma_z} \cdot M_S \cdot e^{i(\epsilon/2)\sigma_z} \cdot \sqrt{M_U},$$

and the time-direction analog is

$$\tilde{M}(k,\epsilon) = \sqrt{\tilde{M}_U} \cdot e^{i(k/2)\sigma_z} \cdot \tilde{M}_S \cdot e^{i(k/2)\sigma_z} \cdot \sqrt{\tilde{M}_U},$$

with matrix elements inherited from the scatterer S-matrices. This formulation allows direct computation of gap invariants from reflection phases at midgap values (Zhu et al., 2024).

2. Topological Invariants: Reflection Phases and Winding Numbers

In TM formalism, the topological characterization is encoded in reflection phases:

• Space-direction reflection amplitude $r(\epsilon)$ and phase $\phi_x(\epsilon) = \arg r(\epsilon)$.
• Time-direction reflection amplitude $\tilde{r}(k)$ with $\phi_t(k) = \arg\tilde{r}(k)$.

In full gaps ($|t|\ll 1$ for space, $|\tilde t_f|\ll 1$ for time), the sign or winding of $\phi_x$ and $\phi_t$ discriminates trivial and nontrivial topological phases. For example, at $\epsilon=\pi$:

• $\operatorname{sign}\phi_x(\pi)<0$ indicates a trivial quasienergy gap.
• $\operatorname{sign}\phi_x(\pi)>0$ signals a nontrivial (Floquet) topological gap.

Similarly, at $k=0$ or $k=\pi$, $\operatorname{sign}\phi_t$ encodes the topology of momentum gaps.

These signatures correspond to bulk winding numbers, such as

$$\nu_x = \frac{1}{2\pi i} \oint_{ \epsilon\in \text{gap} } d\ln r(\epsilon),$$

which reduce to $\nu_x\in\{0,1\}$ in two-band settings, reflecting $\mathbb{Z}_2$ topology (Zhu et al., 2024).

3. Floquet Topological Phases: Quasienergy, Quasimomentum, and Anomalous Gaps

Generalized Floquet topological states can possess:

• Quasienergy gaps, with topological invariants detected via $\phi_x(\epsilon)$.
• Quasimomentum gaps, with invariants in $\phi_t(k)$.
• Simultaneous gaps in both $\epsilon$ and $k$, depending on model parameters and unit-cell choices (e.g., asymmetric dimer structures).

Crucially, some Floquet topological gaps are "anomalous": their reflection-phase invariants are robust under shifting the unit-cell center. This anomalous Floquet quasimomentum gap is a dynamical feature—its topology emerges from the sequence U–S in the driving period, not from a static Hamiltonian (Zhu et al., 2024).

Concrete models demonstrate pure quasienergy gaps (e.g., non-Hermitian PT-symmetric dimer), pure momentum gaps, and coexistence regimes with simultaneous gaps. Phase diagrams in parameter space (coupling, gain/loss) sharply delineate transitions via exceptional points and gap-closing conditions (Zhu et al., 2024).

4. Beyond Hermitian Systems: Non-Hermitian Floquet Topology and Edge States

Non-Hermitian Floquet systems (i.e., with periodically modulated gain/loss) are captured seamlessly by the TM formalism: both Hermitian and non-Hermitian cases enter by directly substituting their S-matrix elements. Topological invariants in the non-Hermitian context are biorthogonal, with winding or Chern numbers defined using left and right eigenvectors. Edge states can be lossless, dissipative, or amplifying, but remain topologically protected against backscattering even when the bulk spectrum possesses complex exceptional points (Li et al., 2018).

Bulk-edge correspondence is preserved: the sign or winding of reflection phases (or biorthogonal Chern numbers) guarantees the existence of chiral edge modes.

5. Hybrid-edge Floquet Topological Insulators and Higher-order Floquet Phases

Floquet topological insulators admit robust edge states under nontrivial termination conditions. For example, in photonic honeycomb arrays with hybrid edges (alternating zigzag and armchair segments), Floquet topological edge states persist over large intervals in the Brillouin zone and propagate one-way with robustness to disorder and defects, including in the presence of Kerr nonlinearity and formation of nonlinear edge solitons (Ren et al., 2022).

More recently, higher-order Floquet topological phases have been engineered in periodically driven lattices with mirror symmetries, supporting lower-dimensional bound states—e.g., Floquet corner states, and—via stirring-drive protocols—dynamically generated bulk vortices hosting Floquet bound states (Rodriguez-Vega et al., 2018, Zhou, 2020). These phases are classified by mirror-graded winding invariants, with anomalous Floquet corner modes exhibiting period-doubling and robust bulk-corner correspondence.

The framework for classification utilizes hybrid Wannier representations and micromotion winding in the extended space-time torus, integrating both static and anomalous Floquet phases (Nakagawa et al., 2019).

6. Experimental Implementations and Outlook

Floquet topological phenomena are realized in a variety of platforms:

• Photonic circuits and waveguide arrays with time-periodic index modulation or Floquet gauge interfaces (Ren et al., 2022, Song et al., 2020).
• Cold-atom lattices with driven hopping amplitudes and optical shaking (Zheng et al., 2014).
• Mesoscopic circuits or fiber loops for discrete-time quantum walks (Bisianov et al., 2020).
• Acoustic and microwave systems with modulated loss/gain (Li et al., 2018).
• Nonlinear media supporting robust Floquet solitons and hybrid edge transport.

Edge and interface states, predicted by transfer-matrix gap invariants, demonstrate quantized transport, robust energy and spatial localization, and insensitivity to structural perturbations. Higher-order Floquet phases open avenues for dynamical control of corner and bulk bound states, with interventions such as stirring drive protocols.

The TM-based diagnostic is broadly applicable, directly revealing gap topology in hybrid space-time dimensions. This method circumvents full Hamiltonian reconstruction and is especially powerful for non-Hermitian and strongly modulated scenarios. It affirms the universality of Floquet engineering for topological phases and supports application across photonic, cold-atom, acoustic, and quantum simulator architectures (Zhu et al., 2024).

7. Summary Table: Key Invariants and Physical Observables

Type	Computation (TM/Scattering)	Observable Signature
Quasienergy gap	$\phi_x(\epsilon)$, sign at midgap	Edge (space-direction) chiral mode
Quasimomentum gap	$\phi_t(k)$, sign at $k=0$ or $k=\pi$	Edge (time-direction) chiral mode
Anomalous gap	Invariant under unit-cell center shift	Robust dynamical edge mode
Winding number	$\nu = (1/2\pi i)\int d\ln r(\epsilon)$	Number of protected edge states

Reflection-phase analysis yields direct, experimentally accessible predictions for topological phase boundaries and edge state existence, facilitating targeted design of non-equilibrium topological materials and devices.