SurfaceScience21 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
O: Fachverband Oberflächenphysik
O 42: Poster Session III: Poster to Mini-Symposium: Machine learning applications in surface science I
O 42.9: Poster
Dienstag, 2. März 2021, 10:30–12:30, P
The data-driven search of new catalysts for an OCM reaction based on the properties of surface carbonates — •Aliaksei Mazheika1, Frank Rosowski1,2, and Ralph Kraehnert1 — 1BasCat, Technische Universitaet Berlin, Berlin, DE — 2BASF SE, Ludwigshafen, DE
The interest in oxidative coupling of methane (OCM) reaction is caused by the fact that this is a relatively simple way for conversion of methane to C2 products (ethane, ethylene). Despite quite many years spent for the search of an efficient catalyst, still a catalyst which would be commercially viable has not been found. Recently Wang et al. have experimentally observed the volcano-like dependence of OCM performance of oxide catalysts on decomposition of their carbonates [1]. In this study we develop a way for calculations of carbonates formation energies based on adsorption of CO2 on the surfaces of corresponding oxides. This allows us to reformulate experimentally observed volcano-like dependence in terms of theoretically calculated quantities. Based on this, we develop the strategy for high-throughput screening using artificial intelligence methodology - subgroup discovery [2] and SISSO [3]. With that we have done the screening of more than 800k materials, and obtained new materials promising for OCM reaction.
[1] H. Wang, PhD thesis, TU Berlin (2018).
[2] M. Boley et al., Data Min. Knowl. Disc. 31, 1391 (2017).
[3] R. Ouyang et al., Phys. Rev. M 2, 083802 (2018).