Berlin 2024 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
O: Fachverband Oberflächenphysik
O 85: Heterogeneous Catalysis I
O 85.2: Vortrag
Donnerstag, 21. März 2024, 10:45–11:00, TC 006
Machine-Learning Assisted Realistic Description of Catalytic Active Centers on the Ternary M1 (Mo,V)Ox(100) Surface — •Kyeonghyeon Nam, Yonghyuk Lee, Liudmyla Masliuk, Thomas Lunkenbein, Annette Trunschke, Christoph Scheurer, and Karsten Reuter — Fritz-Haber-Institut der MPG, Berlin
The surface structure and composition of complex heterogeneous catalysts can differ noticeably from assumed, idealistic cuts derived from the bulk material’s structure. As a crucial first step toward realistic models of the active surface, we explore the evolution of local atomic-scale structural motifs presented by the catalyst under conditioning and operating conditions. Our focus is on the industrial M1 catalysts used for the selective oxidation of light alkanes. The large primitive cell of the M1 catalyst poses a challenge for a detailed study of all surface terminations using predictive-quality first-principles calculations. To address this challenge, we deconstruct the primitive cell into ‘pillar structures’ of surface motifs with varying oxygen content. Machine-learned Gaussian approximation potentials, trained against this structural library [1], are used to identify operando stable (100) surfaces of ternary M1 catalysts via ab initio thermodynamics and comparison to experimental data from electron microscopy. Subsequent electronic structure calculations provide a detailed picture for various (hk0) surfaces, shedding light on the impact of surface stabilization on catalytic centers.
[1] L. Masliuk et al., J. Phys. Chem. C 121, 24093 (2017).
Keywords: Selective Oxidation; MLIPs; Density-functional theory; Catalysis