Speaker: Rhiannon A. Zarotiadis, Simons Center for Computational Physical Chemistry; New York University

It is well established that nuclei explore electronic potential energy landscapes, and that this exploration governs key phenomena in chemistry and materials science. Theoretical advances, ranging from ab-initio molecular dynamics to semiclassical and nonadiabatic methods such as ring-polymer dynamics and surface hopping, have enabled increasingly accurate modelling. In contrast, light remains an often neglected component in chemical theory. Commonly treated as a probe, it has rarely been considered an active, quantum-mechanical participant. As a result, it has not undergone the same rigorous development in chemical theory as electronic and nuclear dynamics.

Recent advances in strong light–matter interactions, which have applications in numerous fields from catalysis in chemistry to photovoltaic cells in material science and photonic qubits in quantum information science, however, require accurate theoretical methods to capture light-matter dynamics. In the strong light-matter coupling regime, light and matter form hybrid states that cannot be separated, giving rise to complex, coupled light–matter landscapes. Exploring these hybrid landscapes presents a high-dimensional computational and theoretical challenge that demands new theoretical approaches.

Here, we develop a photon sampling scheme based on cavity Multi-Trajectory Ehrenfest (cMTE) dynamics, which is inspired by existing approaches for nuclei. The cMTE method uses Wigner sampling to retain the quantum character of light while propagating its dynamics (semi)classically for favourable scaling. Coupled with real-time electronic structure calculations, this framework allows us to explore complex light–matter landscapes in realistic molecules, introducing a new paradigm of ab-initio quantum optics. We apply this method to study quantum effects in molecules coupled to waveguides, which is relevant to photonic qubit formation. This framework also naturally extends to include nuclear quantum dynamics, therefore offering a unified path towards simulating coupled electron–nuclear–photonic systems.