PAH-water cluster anions: Investigating water-hydrocarbon interactions through ML dynamics

Water-hydrocarbon interactions are crucial to many chemical processes in astrochemical, atmospheric, and interfacial phenomena. Clusters composed of a charged polycyclic aromatic hydrocarbon (PAH) with one to four water molecules are model systems for closely studying the governing intermolecular interactions of water with the charged hydrocarbon surface. Vibrational spectroscopy of the OH stretch acts as a sensitive probe of the local chemical environment that can be directly compared to experiment. We employ machine learning interatomic potentials to simulate nanoseconds of dynamics that can reach electronic structure level accuracy at a more affordable computational cost. In this talk, I will demonstrate how, by applying a machine learned dipole model to the dynamics, we can capture electronic and dynamical effects that underly the experimental spectral peaks and line shapes. I will highlight how this methodology allows us to explore the roles of nuclear and electronic dynamics in spectroscopy, assigning different spectral features to water structure and to electronic dipole effects.