Learning to shine: Neuroevolution enables optical control of phase transitions

TL;DR

This paper introduces a reinforcement learning-based method to actively control and stabilize non-thermal structural phases in solids using optically-driven electric fields, applicable in experimental settings.

Contribution

It presents a gradient-free, experimentally feasible approach employing Fourier Neural Networks to optimize electric fields for phase control in dissipative classical systems.

Findings

Successfully stabilized a symmetric phase in bismuth via simulations.

Demonstrated control using impulsive Raman scattering with continuous and pulsed light.

Method applicable solely based on experimental data, no gradient needed.

Abstract

We address the problem of active optical steering of structural phase transitions in solids. We demonstrate that existing reinforcement learning approaches can derive optimal time-dependent electric fields in optically-driven dissipative classical systems far beyond the harmonic regime, enabling the stabilization of non-thermal structural phases. Our approach relies on experimentally extractable metrics of the phase-space evolution and physically-interpretable Fourier Neural Network surrogates of the time-dependent electric field. Using first-principles simulations, we demonstrate the stabilization of a symmetric phase in bismuth through impulsive Raman scattering under continuous and pulsed light sources in the presence of dissipation. Importantly, the method is gradient-free, which enables optimization loops based solely on experimental data, such as the measures of half-periods of oscillations in transient spectroscopy. Our framework thus provides a practical route for controlling non-equilibrium structural dynamics with light, opening pathways to stabilize hidden and metastable phases in quantum materials.