Dresden 2026 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
O: Fachverband Oberflächenphysik
O 24: Solid-liquid interfaces: Reactions and electrochemistry – Poster
O 24.4: Poster
Montag, 9. März 2026, 18:00–20:00, P2
molecular dynamics simulations of belite-water interac tions using HDNNPs — •Usman Tafida1,2, Maite Böhm1,2, Henry Wang1,2, Bernadeta Prus1,2, and Jörg Behler1,2 — 1Theoretische Chemie II, Ruhr-Universität Bochum, Germany — 2Research Center Chemical Sciences and Sustainability, Research Alliance Ruhr, Germany
Advances in the development of machine learning potentials (MLPs), such as High-Dimensional Neural Network Potentials (HDNNPs), have enabled ab initio-level accuracy for molecular dynamics (MD) simulations of high dimensional systems at greatly reduced computational cost. In this work, HDNNPs are trained on DFT-D3 energies and forces data of beta-belite (Ca2SiO4) clusters in water, which serve as model systems for complex oxide-water interfaces. Here, we present the dataset construction scheme, training and validation of the potential. Moreover, MD simulations have been performed to study the structural and dynamical properties of interfacial water at these clusters.
Keywords: machine learning potentials; beta-belite; High-Dimensional Neural Network Potentials; oxide-water interfaces; molecular dynamics