(329b) From Virtual Catalysis to Practical Applications: Understanding and Designing Catalysts in Silico

The importance of heterogeneous catalysis in modern applications that enhance the quality of life cannot be overstated: current estimates place the value of the catalyst market at more than $20 billion, and it has been estimated that every $1 spent on a catalyst can generate up to $1000 worth of product.  Computational methods are increasingly becoming crucial in gaining a fundamental understanding of the physicochemical phenomena that give rise to catalytic activity. Such an understanding makes it possible to design improved catalysts; however, the complexity encountered in catalytic systems poses several challenges in modelling them.

We present an ab-initio kinetic Monte Carlo (KMC) framework that uses concepts from graph theory to overcome these challenges and easily treat complex chemistries and surface phenomena. This approach allows one to define elementary reaction events characterized by complex neighbouring patterns of the sites involved. These sites can be of various types (such as steps or terraces) and occupied by monodentate as well as multidentate species. It also treats adsorbate-adsorbate lateral interactions in a generalised way by incorporating cluster expansions for the energetics.

We further showcase this approach by applying it to key catalytic reactions, in particular CO oxidation on Au nano-clusters, which is important for environmental and health reasons, and water-gas shift on single crystal Pt surfaces, which is relevant to H₂ production from biomass. We show that our approach can provide a wealth of information, which can be used to explain experimental observations and obtain insight in the design of materials with desired catalytic properties.