Michael Chen

Michael Chen is an independent postdoctoral fellow in the Simons Center for Computational Physical Chemistry at New York University. He received his Bachelor’s degrees in chemistry and computer science from UC Berkeley before going on to conduct his PhD in chemistry at Stanford University where he was a fellow in the Center for Molecular Analysis and Design and received the Centennial Teaching Assistant Award. His research focuses broadly on the development and application of theoretical, simulation, and machine learning (ML) methods for elucidating the microscopic structure and dynamics of processes in the condensed phase including: charge transport in battery electrolytes, energy relaxation pathways of biological chromophores, and the self-assembly of colloidal crystals.

Abstract

Accelerating Condensed Phase Molecular Simulationswith Machine Learning: Modeling Charge Transport and Optical Spectroscopy Obtaining accurate atomistic simulations of disordered condensed phasesystems from first principles remains one of the forefront challenges of materials modeling. In my talk, I will highlight how we have leveraged machine learning (ML) to broaden the applicability of first principles simulations for investigating: (1) charge transport mechanisms of candidate proton-conducting battery electrolytes and (2) how the optical properties of chromophores are modulated by hydrogen bonding interactions with their environment. By employing ML-accelerated ab initio path integral molecular dynamics simulations, we demonstrated the importance of treating both the electronic and nuclear degrees of freedom quantum mechanically in order to accurately model the experimental properties of interest (e.g., conductivities and optical spectra) and unraveled the microscopic factors that affect these properties.