Floquet Spintronics: Using Light to Flip a Topological Surface’s Spin Response

Spintronics tries to move beyond ordinary electronics by using not only an electron’s charge, but also its spin. In principle, spin-based devices could switch faster, waste less heat and carry more information per operation. In practice, the difficult part is control: how do you turn spin currents on, off or around without building every device from bulky magnetic layers?

A new arXiv preprint posted on August 7, 2026 suggests a distinctly Floquet answer. In “Floquet spintronics: tuning the current-induced spin polarization of topological surface states with light,” Youngjae Kim, Aayushi Agrawal and Kwon Park show theoretically that high-frequency circularly polarized light can tune, and even reverse, the current-induced spin polarization on the surface of a topological insulator. Their model focuses on bismuth selenide, Bi2Se3, one of the canonical three-dimensional topological-insulator materials.

The central claim is not that light creates energy from nowhere. It is that a periodic optical drive can rewrite the spin texture of a conducting surface, changing how an electrical current converts into spin polarization.

For Floquet.ca, this belongs in the “practical energy applications” lane. Spintronic memory, logic and sensing are often motivated by lower dissipation than charge-only electronics. A light-tunable spin response would not be a power plant or a battery, but it could become part of the control stack for future low-energy information hardware.

Why topological surfaces are useful but stubborn

A three-dimensional topological insulator is insulating in the bulk but hosts conducting states at its surface. Those surface states are special because of spin-momentum locking: the direction an electron moves is tied to the orientation of its spin. Push a current through the surface and the moving electrons can produce a net spin polarization. This is the Edelstein effect, a key mechanism for all-electrical spin generation.

That sounds ideal for spintronics. It means a simple electric current can create spin polarization without needing to inject spin through a ferromagnetic contact. But topological protection is a double-edged property. The same robustness that keeps surface states stable against many local imperfections also makes them hard to tune. If a spintronic device is supposed to be programmable, “robust but fixed” is not enough.

Kim, Agrawal and Park ask whether illumination can solve the tuning problem. Their logic is connected to a famous Floquet idea: high-frequency circularly polarized light can turn graphene-like systems into Chern-insulator-like states by opening and reshaping gaps. Since topological-insulator surface states resemble a single spin-momentum-locked Dirac cone, the authors test whether a similar drive can reshape the surface spin texture itself.

The model: Bi2Se3 under a circular optical clock

The paper starts from an effective Hamiltonian for the topological surface states of Bi2Se3. In the model, momentum is measured in units set by the lattice constant, a = 4.14 angstroms. The phenomenological parameters used for the surface band are ζ = −0.0546 eV, λ = 0.0779 eV, ξ = 0.1653 eV and η = 0.0560 eV. The light enters through a Peierls substitution: the crystal momentum becomes time-dependent under a circular vector potential.

The normalized drive strength is written as A = eE0a/ℏΩ, where E0 is the electric-field strength of the light and Ω is the light frequency. In the main calculations, the authors take Ω = 8 eV, described in the paper as roughly 2,000 THz and 1,000 times the minimum quasienergy-band gap. That high-frequency limit lets the Floquet Hamiltonian separate into well-spaced sectors, making the driven surface analyzable as an effective band problem.

To compute the spin response, the authors use nonequilibrium Green’s functions and a Floquet generalization of the Kubo formula. They include impurity scattering through a self-consistent Born approximation. The quantity of interest is the electric spin susceptibility: the coefficient that says how much spin polarization appears when an electric field drives a current.

The headline result: the spin response can change sign

In the absence of illumination, the Edelstein effect gives a fixed relationship between current direction and induced spin polarization on the topological surface. Under the circularly polarized Floquet drive, that relationship becomes tunable. The paper’s Figure 2 shows that the transverse susceptibilities χsxy and χsyx reverse sign as A is increased through about 1.8. The longitudinal susceptibilities χsxx and χsyy remain essentially zero, while only a small out-of-plane component appears.

In plain language, the model predicts a light-controlled spin-current dial. An electrical current that would normally polarize spins one way could, under sufficiently strong circular illumination, polarize them the opposite way. That is why the authors call the idea “Floquet spintronics.” The periodic drive supplies a new control parameter that is neither a static gate voltage nor a permanent change in the material’s chemistry.

If the prediction is realized experimentally, a topological surface could become an optically programmable spin converter: same current, different spin response, selected by the drive.

The zero-field susceptibility scale reported in the paper is also useful for grounding the result. When expressed as induced magnetization per unit cell in response to an electric field, the authors give χ0 = 8.97 × 10−10 μB·m/V, roughly comparable to values computed for other materials. That does not make the proposal device-ready, but it puts the effect in a physically meaningful range rather than leaving it as an abstract sign change.

Why the sign flips: Floquet topology rewrites the spin texture

The sign reversal is not treated as a numerical accident. The authors trace it to a change in the momentum-space spin texture of the conduction band. In a topological surface state, the in-plane spin texture is what turns current into spin polarization. If that texture winds in the opposite sense, the Edelstein response changes sign.

The paper identifies two key drive strengths, A = 1.88 and A = 1.98, where the quasienergy gap closes through Dirac-like band touchings. Around these values, the Berry curvature changes dramatically: below A = 1.65 it is predominantly positive; near the critical points the gap closings mark topological transitions; above A = 2.05 the Berry curvature develops roughly balanced positive and negative regions. The authors argue that these Floquet-induced topological transitions reverse the spin texture, which in turn reverses the transverse spin susceptibility.

This is the deeper Floquet lesson. The laser is not merely heating the surface or shaking electrons around randomly. In the high-frequency regime, the drive modifies the effective Hamiltonian. That changes the quasienergy bands, the Berry curvature and the spin texture. The spintronic response then follows from the dressed band geometry.

An August cluster of light-tailored spin papers

The spintronics preprint appeared in the same week as several related Floquet-magnetism papers. On August 3, Tongshuai Zhu, Zixuan Li, Huaiqiang Wang, Su-Huai Wei and Jiawei Ruan posted a “Floquet spin-group framework” for light-tailored spin splitting in collinear magnets. Their framework combines crystal spin groups with the dynamical symmetries of light, showing how odd-parity, even-parity and mixed-parity spin splittings can be switched within the same material by tailoring the driving field.

On August 5, Hou-Jian Duan and colleagues posted a study of RKKY interaction as a probe of spin splitting and odd-parity character in Floquet collinear magnets. They propose that magnetic exchange signals, including sign changes in Heisenberg/Ising terms and a characteristic Dzyaloshinskii-Moriya pattern, could help detect light-induced odd-parity spin polarization. Together with the topological-surface paper, these works suggest that Floquet spin control is moving from isolated proposals toward a more systematic design language.

What this does, and does not, mean for energy

It is tempting to hear “light controls spin” and imagine an immediate low-power device. The responsible interpretation is narrower. The paper is a theory proposal. It uses a high-frequency drive and idealized surface-state model, and it estimates experimental parameter ranges rather than reporting a measured device. The authors note that present light fields can reach the order of 0.1 V/angstrom, while A of order unity would require stronger fields, potentially around 1 V/angstrom for the Bi2Se3 lattice scale.

Still, the energy relevance is real. Much of digital energy consumption comes from moving and dissipating charge. Spintronic approaches aim to encode, switch or read information with less waste, especially when spin can be controlled electrically. Floquet spintronics adds another possible layer: optical or terahertz control of the spin conversion coefficient itself. In future hybrid devices, that could let designers dynamically choose when a surface acts as a spin source, a spin sink or a reversed-polarity converter.

The connection to quantum thermodynamics is also subtle but important. A periodically driven material is an open energy-accounting problem: the drive supplies work, the electronic system absorbs and redistributes energy, and scattering plus phonons eventually dissipate heat. Any practical Floquet spintronic device will need the same discipline now being developed for Floquet master equations, energy-current definitions and heating control. Tunability is valuable only if the energy cost of the drive is justified by the function it enables.

What to watch next

The next milestone would be experimental evidence of the predicted sign reversal, ideally in a topological-insulator surface with pump-probe or transport readout of the Edelstein response. Researchers will also need to test how robust the effect is against disorder, finite temperature, bulk conduction and realistic laser pulses. The paper includes impurity scattering in the calculation, but devices bring additional complications.

Another watch item is the merger of Floquet spintronics with Floquet spin-group design. If light symmetry can select spin-splitting parity in magnets, and light strength can reverse spin response on topological surfaces, then the field may be approaching a genuine engineering toolkit: choose a material, choose a drive symmetry, choose a frequency window, and design a target spin response.

For smart non-specialists, the takeaway is simple. Floquet engineering is no longer only about making exotic bands for their own sake. It is becoming a way to program functional responses: heat flow, charge pumping, optical absorption and now current-induced spin polarization. The energy promise is not magic efficiency. It is control over when and where microscopic energy and information carriers are allowed to move.