Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
O: Fachverband Oberflächenphysik
O 28: Solid-Liquid Interfaces II: Structure and Spectroscopy
O 28.11: Vortrag
Dienstag, 19. März 2024, 13:00–13:15, H 1012
Efficient and Accurate Description of Solid-Liquid Interface Dynamics Using High-Dimensional Neural Network Potentials — •Knut Nikolas Lausch1,2, Marco Eckhoff3, Peter Blöchl4, and Jörg Behler1,2 — 1Theoretische Chemie II, Ruhr-Universität Bochum, Germany — 2Research Center Chemical Sciences and Sustainability, Research Alliance Ruhr, Germany — 3Theoretische Chemie, Eidgenössische Technische Hochschule Zürich, Switzerland — 4Institut für Theoretische Physik, Technische Universität Clausthal, Germany
Solid-liquid interfaces play a central role in many processes that are of utmost importance for a sustainable energy future. From energy storage solutions to heterogeneous catalysis, understanding interface dynamics and reactivity plays a key role in developing new materials. Molecular dynamics simulations of such interfaces rely on an accurate description of solid-liquid interactions, and density functional theory (DFT) can in principle provide reliable results. However, due to the high computational costs, time and length scales of ab initio simulations are severely limited. This can be overcome by employing machine learning potentials, which yield energies and forces several orders of magnitude faster while retaining chemical accuracy. Here, we present a high-dimensional neural network potential for lithium manganese oxide (LMO) in water as a model system for the electrocatalytic oxygen evolution reaction.
Keywords: Machine Learning Potentials; High-Dimensional Neural Network Potentials; Molecular Dynamics Simulations; Density Functional Theory; Lithium Manganese Oxide-Water Interface