ALIs

kommt noch

Achtung: Die Navigationslinks auf dieser Seite funktionieren nicht, weil das nicht der Rahmen ist, der zu diesem Skript gehört. Wenn das Skript ordnungsgemäß dort installiert ist, wo es später laufen soll, funktionieren auch die Links in der dortigen Umgebung.

LINUX Cluster Project

First-Principles based Multi-Scale Modeling


Institution

  • Name: Lehrstuhl für Theoretische Chemie,TUM
  • Address: Lichtenbergstraße 4, 85748 Garching
  • Project Proposal Date: 2019-10-09 08:08:11

Abstract:

Study of the electronic non adiabatic effects in the dissociation of molecular oxygen on Silver (111). Reaching a high selectivity in surface catalytic processes with more than one thermodynamically feasible product is nowadays one of the major ambitions in catalysis. Compared to the overall activity our present atomic-scale understanding of the mechanisms driving such selectivity is very shallow. On the modeling side, this concerns notably quantitative simulations aiming at a first-principles based microkinetic description. Corresponding modeling unites an accurate description of the bond making and breaking in individual surface elementary processes (adsorption, desorption, diffusion, reaction) with a proper account of the statistical interplay of all processes in the catalytic cycle. Large-scale first-principles electronic structure theory calculations are employed for the prior, statistical mechanical approaches like kinetic Monte Carlo (kMC) are used for the latter. While corresponding modeling has already led to significant advances in the understanding of simple model reactions with one product (like the CO oxidation reaction), the approach is limited by the excessive number of elementary processes in more complex reaction networks offering different possible end products. An exhaustive determination of all kinetic parameters of all elementary processes from first-principles represents at present a prohibitive computational cost. We therefore explore an alternative approach based on a sensitivity analysis guided refinement. The statistical simulations are initiated with estimates for the kinetic parameters. Sensitivity analyses are then systematically employed to identify those rate-limiting steps in the network which require a more accurate first-principles treatment.