Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

MM: Fachverband Metall- und Materialphysik

MM 12: Poster I

MM 12.12: Poster

Montag, 27. März 2023, 18:15–20:00, P2/OG1+2

Machine Learning Interatomic Potentials for amorphous mesoporous metallosilicates — •Julian Greif, Konstantin Gubaev, and Blazej Grabowski — Universität Stuttgart, D-70049, Stuttgart, Germany

Improving the efficiency of catalysis continues to be an important topic in modern chemistry. A promising approach currently under investigation is to utilize molecular catalysts in confined geometries. In the present project, we aim to model porous amorphous silica containing metal atoms on the pore surfaces that can act as co-catalysts. We conduct atomistic simulations using ab initio-trained machine learning potentials to obtain insights into the location of the metal atoms and the chemical configuration of hydrogen saturated surfaces in the silica structure. To that end, a potential capable of simulating the remelting of SiO₂ + Al was trained and checked on the bulk amorphous SiO₂-system against experimental data in terms of realistic densities, bond lengths and bond angles. Using this potential, new structures are then created by melting and quenching and a new potential is trained for simulations of mesoscale cells.