(455a) Accelerating Ab Initio Calculations of Chemical Bonding and Equilibria at Material Interfaces Using Machine Learning

The mechanism, rates, and chemical equilibria can be understood from free energy profiles of reactions at material interfaces, including heterogeneous catalysis, electrocatalytic reactions, and defect formation in functional materials. Predicting the energetics has involved a trade-off in accuracy between the enthalpic (i.e., interatomic potential) and the entropic (i.e., degrees of freedom) contributions to the free-energy. Traditionally, computational efforts in molecular reactions from ab initio calculations have been largely focused on the former. The latter, however, becomes increasingly important with increasing thermal effects and conformational entropy in many-body systems and high-temperature settings. Here, I introduce enhanced sampling approaches that combine machine-learning algorithms and electronic structure calculations, to bridge the gap between accuracy and system size for chemical reactions at material interfaces from first-principles. In particular, I present an enhanced sampling simulations tailored to ab initio calculations based on gaussian process models. I will demonstrate how the gaussian process method allows efficient sampling and direct computation of free energy landscapes with ab initio accuracy using examples from molecular reactions on transition metal catalysts and reactions under nanogap confinement, with applications in plastic waste conversion and chemical sensing.