Floquet mobility edges and transport in a periodically driven generalized Aubry-Andr'e model

arXiv:2604.21992 · cond-mat.dis-nn, cond-mat.str-el, quant-ph · Submitted 2026-04-23 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Condensed Matter. The day in condensed matter: materials, phases and the measurements under them.

Kai: We're looking at the paper: "Floquet mobility edges and transport in a periodically driven generalized Aubry-Andr'e model".

Mira: Here's the gist of the paper. Floquet mobility edges and transport in a periodically driven generalized Aubry-Andr´e model investigate the effect of a periodic electric field drive on the generalized Aubry-Andr´e model, also known as the Ganeshan-Pixley-Das Sarma (GPD) model, which is well known as a host of mobility edges The study uncovers the emergence of two distinct Floquet mobility edges, a delocalized–localized (DL) edge in the bounded regime, and a multifractal–localized (ML) edge in the unbounded regime Using analytical results derived from Avila’s global theory applied to the high frequency effective Hamiltonian, together with numerical diagnostics such as the fractal dimension and inverse participation ratio, we demonstrate that these mobility edges can be effectively controlled by the amplitude and frequency of the electric field drive We also identify drive-induced localization at specific values of the driving parameters, corresponding to dynamical localization points in the absence of quasiperiodic potential Furthermore, the dynamical study of the periodically driven GPD model demonstrates superdiffusive to almost ballistic transport in the bounded regime corresponding to the DL edges, whereas subdiffusive transport is observed in the unbounded regime associated with the ML edges.

Kai: Model Description and Static Properties The one-dimensional periodically driven Ganeshan-Pixley-Das Sarma (GPD) model describes non-interacting spinless fermions on a quasiperiodic lattice The time-dependent Hamiltonian for this system is given by Hˆ (t) = L X−1 j=1 J cˆ†j cˆj+1+H.c.

Mira: + X L j=1 Vj +Aj (t) cˆ†j cˆj, (1) The quasiperiodic on-site potential is given by Vj = 2λ cos(2παj + ϕ) 1 − β cos(2παj + ϕ), (2) The timeperiodic electric field drive is given by Aj (t) = −jK cos(ωt), (3) In the static limit (K = 0), for β = 0, the model described in Eq.

Kai: (1) reduces to the AAH model For β < 1, the potential is bounded, thereby making the Hamiltonian a bounded operator and consequently, all the eigenvalues of Hˆ are finite.

Paper summary: Kai: So, to recap, this paper is about taking the Ganeshan-Pixley-Das Sarma or GPD model, which is already known for having mobility edges, and studying how a periodic electric field drive affects them. The main claim they make is that this drive creates two distinct Floquet mobility edges: one delocalized to localized edge in the bounded regime, and another multifractal to localized edge in the unbounded regime.

Mira: They use analytical methods based on Avila's global theory applied to the high frequency effective Hamiltonian, and then they back up their predictions with numerical diagnostics like fractal dimension and inverse participation ratio.

Lev: What matters is that they demonstrate these mobility edges can be controlled by changing the amplitude and frequency of the electric field drive. That tunability is what really hooks me because it means we aren't just studying a fixed system anymore, but a controllable one.

Kai: They also found specific drive parameters that cause drive-induced localization points, which are special in the absence of any quasiperiodic potential at all. That's a significant finding because it points to inherent localization mechanisms driven by the drive itself.

Mira: The transport results show that in the bounded regime corresponding to those DL edges, you see superdiffusive to nearly ballistic spreading for both driving frequencies. But near those ML edges in the unbounded regime, things become subdiffusive with smaller growth exponents.

Lev: For a real system, that's like you have regions where information travels incredibly fast and other regions where it crawls along, which is a huge difference for any quantum computation or material science application.

Kai: It really shows that the way we drive these systems can fundamentally change the transport nature of the states based on the parameters we choose.

Mira: So this study connects static localization properties with driven dynamics in a new way by showing how periodic driving introduces this new spectral structure, which is what they call Floquet mobility edges.

Conclusion: Kai: So, to wrap up, we looked at the Floquet mobility edges and transport in this driven GPD model by Jayashis Das, Vatsana Tiwari, Manish Kumar, and Auditya Sharma. The main point is that the driving field introduces these two types of spectral features.

Mira: The implication is that these features aren't just noise; they are structural elements you can manipulate by tuning the drive parameters to control where states are localized or extended.

Lev: For me, it means we have a more precise roadmap for engineering driven systems because we know exactly which driving settings lead to which transport regimes.

Kai: It confirms that these localization properties from the static model don't just stay the same; they can be tuned by external periodic driving fields in a predictable way.

Mira: And it points toward experimental realizations where we can test this control over these mobility edges using different drive frequencies and amplitudes in our lab setups.

Department of Physics, Indian Institute of Science Education and Research, Bhopal · Department of Physics, Indian Institute of Technology Bombay

cond-mat.dis-nn, cond-mat.str-el, quant-ph

Submitted: 2026-04-23

Updated: 2026-11-10

Comments: 22 Pages, 17 Figures

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 86/100

The gist: Floquet mobility edges and transport in a periodically driven generalized Aubry-Andr´e model investigate the effect of a periodic electric field drive on the generalized Aubry-Andr´e model…

Key concepts

Floquet mobility edges
These are specific energy levels in a periodically driven system where the wave functions transition between different states of localization. They act as boundaries separating regions where particles spread out (delocalized) from regions where they are trapped (localized).
Generalized Aubry-Andr´e model (GPD)
This is a one-dimensional system describing non-interacting fermions on a lattice with both quasiperiodic on-site potential and a time-dependent electric field drive. It serves as the base model used to study how driving forces change the system's transport properties.
High-frequency effective Hamiltonian
When the driving frequency is very high, the complex time-dependent drive can be simplified into a simpler, static effective Hamiltonian. This simplification allows researchers to analytically determine where these mobility edges occur by finding conditions where the system's localization measure vanishes.
Fractal Dimension (Dq)
This is a mathematical tool used to characterize the complexity of quantum states. For extended states, it is 1; for localized states, it is 0; and for multifractal states, it falls between 0 and 1. It helps classify the nature of the wave functions near the mobility edges.

Terminology

Summary

Floquet mobility edges and transport in a periodically driven generalized Aubry-Andr´e model investigate the effect of a periodic electric field drive on the generalized Aubry-Andr´e model, also known as the Ganeshan-Pixley-Das Sarma (GPD) model, which is well known as a host of mobility edges The study uncovers the emergence of two distinct Floquet mobility edges, a delocalized–localized (DL) edge in the bounded regime, and a multifractal–localized (ML) edge in the unbounded regime Using analytical results derived from Avila’s global theory applied to the high frequency effective Hamiltonian, together with numerical diagnostics such as the fractal dimension and inverse participation ratio, we demonstrate that these mobility edges can be effectively controlled by the amplitude and frequency of the electric field drive We also identify drive-induced localization at specific values of the driving parameters, corresponding to dynamical localization points in the absence of quasiperiodic potential Furthermore, the dynamical study of the periodically driven GPD model demonstrates superdiffusive to almost ballistic transport in the bounded regime corresponding to the DL edges, whereas subdiffusive transport is observed in the unbounded regime associated with the ML edges. p. 0

Model Description and Static Properties

The one-dimensional periodically driven Ganeshan-Pixley-Das Sarma (GPD) model describes non-interacting spinless fermions on a quasiperiodic lattice The time-dependent Hamiltonian for this system is given by Hˆ (t) = L X−1 j=1 J cˆ†j cˆj+1+H.c. + X L j=1 Vj +Aj (t) cˆ†j cˆj, (1) The quasiperiodic on-site potential is given by Vj = 2λ cos(2παj + ϕ) 1 − β cos(2παj + ϕ), (2) The timeperiodic electric field drive is given by Aj (t) = −jK cos(ωt), (3) In the static limit (K = 0), for β = 0, the model described in Eq. p. 1 (1) reduces to the AAH model For β < 1, the potential is bounded, thereby making the Hamiltonian a bounded operator and consequently, all the eigenvalues of Hˆ are finite. p. 1

High-Frequency Effective Description and Analytical Edges

The linear drive in Eq. (3), corresponding to a time-dependent electric field, admits a natural gauge interpretation and can be treated analytically by transforming to a rotating frame Under this transformation, the Hamiltonian in the rotating frame is given exactly by HˆR(t) = Rˆ† (t)Hˆ (t)Rˆ(t) − iRˆ† (t) d t Rˆ(t), which evaluates to HˆR(t) = X j h JeiF (t) c†j c j+1+Je−iF (t) c†j+1c j i + X j Vj c†j cˆj. p. 2 (10) When the driving frequency ω is the dominant energy scale of the problem, i. p. 3e. ω ≫ J and ω ≫ λ (ħ is set to 1 overall), HˆF can be systematically approximated using a high-frequency (Floquet–Magnus) expansion To leading (zeroth) order, HˆF = Hˆ(0) F is given by the time average of the rotating-frame Hamiltonian Hˆ(0) F = 1/T Z T 0 HˆR(t) dt This yields the effective Floquet Hamiltonian in the leading order as Hˆ (0) F = Jeff X j c†j c j+1 + H. p. 3c. + X j Vj c†j cˆj, (13) In the high-frequency regime, the periodically driven system is effectively mapped onto a static GPD model with a continuously tunable hopping strength The analytical mobility edges correspond to the vanishing of the Lyapunov exponent, γ(E) = 0, yielding the conditions Ec = 2J J0(K) − 2λ / β, β < 1, and Ec = 2β J J0(K) − 2λ / β, β ≥ 1. p. 3

Numerical Verification and Localization Diagnostics

The fractal dimension Dq of a Floquet eigenstate is defined as Dq = - 1/q − 1 lim L→∞ lnPL j=1 ψn,j 2q lnL, (6) For q = 2 one obtains the standard fractal dimension previously defined in Eq. (6) In particular, for extended states Dq = 1, for localized states Dq = 0, and for multifractal states 0 < Dq < 1. p. 2 The fractal dimension serves as a scale-invariant diagnostic of localization properties In the high-frequency (ω = 10) regime, the Dq map clearly reveals DL edges [Fig. p. 5 2(a)] for β = −0.8, and an ML edge [Fig. p. 2 2(b)] for β = 2.0. The analytical mobility edges obtained from Eq. p. 1 (19), plotted as black dashed lines, are in excellent agreement with the exact numerics. p. 5

Dynamical Transport and Entanglement Probes

To characterize the dynamics and the nature of transport in the periodically driven system, we compute the root mean squared deviation (RMSD) and bipartite entanglement entropy For β < 1, away from the drive-induced localization point K2, the dynamics shows superdiffusive to nearly ballistic spreading for both driving frequencies. In contrast, near the ML edge for β ⩾ 1, the dynamics is governed by multifractal and localized eigenstates and exhibits subdiffusive behavior with smaller growth exponents. p. 6 The entanglement entropy analysis confirms this picture, showing rapid growth in the DL regime and subdiffusive growth in the ML regime. p. 1

Low-Frequency Corrections and Counterintuitive Findings

The second-order correction to the effective Floquet Hamiltonian introduces two distinct effects: first, it modifies the on-site potential, which is responsible for the multifractal band becoming more localized at lower frequencies Second, it generates next-nearest-neighbor hoppings, which tend to delocalize the eigenstates. This competition may shift the balance in favor of localization, potentially leading to quasidisorder-induced localization even among states that are otherwise delocalized A particularly counterintuitive phenomenon is enhanced localization within the multifractal band in the low-frequency regime. p. 9

Conclusion

In this work, we investigate the periodically driven Ganeshan-Pixley-Das Sarma (GPD) model subjected to a linear electric field drive and demonstrate the emergence of mobility edges in its Floquet spectrum These mobility edges closely resemble those of the static model, indicating that the underlying localization properties persist under periodic driving, while their structure can be tuned via the driving frequency and amplitude Our work also provides a route toward experimental realization of such models in the noninteracting limit.

Acknowledgments

We acknowledge the High Performance Computing (HPC) facility at IISER Bhopal, where the large-scale calculations for this project were performed We thank David Weld and Sayan choudhury for stimulating discussions and insightful comments M.K. acknowledges support from the Council of Scientific and Industrial Research (CSIR), India, through a PhD fellowship. p. 0

Appendix A: Transfer Matrix Decomposition and Lyapunov Exponent

The total Lyapunov exponent can be written as γ(E) = max[0, γA(E) + γB(E, 0)] This corresponds to Eq. (A9) Using Eq. p. 0 (A5) and (A8) we get γ(E) = max[0, Fβ(P)], where Fβ(P) = [ln P + p P squared − 4β squared / 2 1 + p 1 − β 2] for β 4β squared, and Fβ(P) = ln P + p P squared − 4β squared / (2 β) for β ≥ 1, P squared > 4β squared The mobility edges are given by Ec = 2J J0(K) − 2λ / β, β < 1, and Ec = 2β J J0(K) − 2λ / β, β ≥ 1. p. 3

Appendix C: Effect of Quasiperiodic Strength on Mobility Edge Structures

Increasing λ enhances the potential strength relative to the hopping energy set by Jeff, thereby expanding the localized region in the energy spectrum Increasing K also enhances localization through the suppression of the effective hopping At special points corresponding to J0(K) = 0, all states become localized.

Appendix D: Return Probability

Away from the drive-induced localization point K2, P(t) exhibits a rapid decay from unity, followed by oscillatory behavior at intermediate times before saturating at small values for both driving frequencies, indicating strong delocalization due to the dominance of extended states. At K = K2, P(t) remains close to unity for ω = 10, demonstrating strong confinement of the wave packet. p. 9

Appendix E: Derivation of Higher-Order Corrections in the Floquet–Magnus Expansion

The second order term introduces two distinct effects: first, it modifies the on-site potential, which is responsible for the multifractal band becoming more localized at lower frequencies Second, it generates next-nearest-neighbor hoppings, which tend to delocalize the eigenstates.

Appendix B: Mobility Edges Under Different Boundary Conditions

The bulk localization properties remain robust to the choice of boundary conditions In particular, the analytically predicted mobility edges consistently separate the localized, multifractal, and extended regions in both cases.

The return probability P(t) is defined as P(t) = ⟨ψ(0)ψ(t)⟩2, (D1) Away from the drive-induced localization point K2, P(t) exhibits a rapid decay from unity, followed by oscillatory behavior at intermediate times before saturating at small values for both driving frequencies, indicating strong delocalization due to the dominance of extended states. At K = K2, P(t) remains close to unity for ω = 10, demonstrating strong confinement of the wave packet

Away from the drive-induced localization point K2, P(t) exhibits a rapid decay from unity, followed by oscillatory behavior at intermediate times before saturating at small values for both driving frequencies, indicating strong delocalization due to the dominance of extended states At K = K2, P(t) remains close to unity for ω = 10, demonstrating strong confinement of the wave packet

Away from the drive-induced localization point K2, P(t) exhibits a rapid decay from unity, followed by oscillatory behavior at intermediate times before saturating at small values for both driving frequencies, indicating strong delocalization due to the dominance of extended states At K = K2, P(t) remains close to unity for ω = 10, demonstrating strong confinement of the wave packet.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved AI can perform Floquet engineering of quasiperiodic systems by analytically deriving an effective time independent Hamiltonian H F that depends continuously on driving parameters amplitude and frequency amplitude and frequency as stated in Section III.

  2. The system can predict mobility edges in the Floquet spectrum using Avila’s global theory, which provides analytical conditions Ec = 2J J0(K) - 2λ β for β < 1, directly linking the drive-induced localization to specific parameter values K where J0(K) is zero.

  3. The AI can dynamically tune localization by manipulating the driving amplitude K such that at the zeros of the Bessel function J0(K), the effective hopping vanishes, leading to drive-induced localization, as observed in Figure 2.

  4. The system can distinguish between delocalized, extended, and multifractal eigenstates numerically by calculating fractal dimension D2 and inverse participation ratio I2 for Floquet eigenstates across the quasienergy spectrum.

  5. The AI can predict transport regimes by analyzing the root mean squared deviation (RMSD) scaling exponent gamma; for DL edges in β < 1, it predicts superdiffusive to nearly ballistic spreading, while for ML edges in β >= 1, it predicts subdiffusive transport.

  6. The system can detect deviations from high-frequency effective description by examining low-frequency driving regimes where significant and counterintuitive deviations in both spectral properties and transport behavior are observed.

  7. The AI can predict the fate of quantum correlations by fitting the half-chain entanglement entropy S(t) to a power law S(t) t ν, where ν is determined from dynamics, showing how lower frequencies can enhance localization tendencies in the multifractal band.

Sources

Related papers