Regensburg 2022 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
O: Fachverband Oberflächenphysik
O 46: New Methods and Developments 3: Theory
O 46.5: Talk
Wednesday, September 7, 2022, 16:00–16:15, H6
Machine learning potentials for complex aqueous systems made simple — •Christoph Schran1, Fabian L. Thiemann1,2, Patrick Rowe1, Erich A. Müller2, Ondrej Marsalek3, and Angelos Michaelides1 — 1University of Cambridge, UK — 2Imperial College London, UK — 3Charles University, Czech Republic
Simulation techniques based on accurate and efficient representations of potential energy surfaces are urgently needed for the understanding of complex systems such as solid-liquid interfaces. Here we present a machine learning framework that enables the efficient development and validation of models for complex aqueous systems. Instead of trying to deliver a globally optimal machine learning potential, we propose to develop models applicable to specific thermodynamic state points in a simple and user-friendly process. After an initial ab initio simulation, a machine learning potential is constructed with minimum human effort through a data-driven active learning protocol. Such models can afterward be applied in exhaustive simulations to provide reliable answers for the scientific question at hand or to systematically explore the thermal performance of ab initio methods. We showcase this methodology on a diverse set of aqueous systems. Highlighting the accuracy of our approach with respect to the underlying ab initio reference, the resulting models are evaluated in detail with an automated validation protocol that includes structural and dynamical properties and the precision of the force prediction of the models. Finally, we demonstrate the capabilities of our approach for providing insight into the solid-liquid interface for various systems.