Thomas Warford, PhD Student, Department of Engineering, University of Cambridge
Supervisor: Prof Gábor Csány
Host: Prof Kristin Persson, Department of Materials Science & Engineering, UC Berkeley
Project: Electrostatics-aware Foundation Models for Scalable Prediction of Response Properties
Thomas is in the second year of his PhD, working with Professor Gábor Csányi at the Department of Engineering. He aims to develop the next generation of foundational machine-learned interatomic potentials (MLIPs) for atomistic simulation. MLIPs enable accurate simulation of materials at length and time scales many orders of magnitude beyond what is accessible to quantum mechanical methods, but until recently have been limited by their inability to capture long-range electrostatic interactions. Thomas is constructing foundational datasets for training electrostatics-aware MLIPs with self-consistent charge modelling, which promise to enable more physically accurate large-scale simulations of polar, charged, and interfacial systems. Thomas's Berkeley project is to apply these models to high-throughput prediction of electrical response properties, and to battery-relevant simulations, while contributing workflows for training data generation to the open-source atomate2 and pymatgen packages maintained by the Persson group.