First-principles Thermodynamic Models in Heterogeneous Catalysis

In this chapter we describe and demonstrate computational approaches to modeling surface adsorption, a process fundamental to all heterogeneous catalysts that takes into account surface structure, adsorbate–adsorbate interactions, and reaction conditions. We begin by describing the development of supercell density functional theory (DFT) models of adsorption at a surface, taking as an example O adsorption at the stepped and kinked Pt(321) surface. We then discuss how these DFT simulations can be used as a basis to parameterize a cluster expansion (CE) model, an Ising-type Hamiltonian that accounts for structural heterogeneity and for adsorbate–adsorbate interactions on a lattice. When converged, the DFT and CE models provide a self-consistent description of the ground states of the surface–adsorbate system. We present a detailed thermodynamic analysis of the system and describe how this can be used to extract equilibrium surface properties from the converged database and provide access to coverage-dependent adsorption energies and surface phase diagrams. Further, the CE enables Monte Carlo simulations of more extended surfaces under fixed temperature and chemical potential conditions, and the average properties from these simulations provide access to average coverages, heat capacities, and phase behavior. Finally, we describe how these same tools can be applied further to relate surface properties with reaction conditions and to describe surface kinetic processes such as diffusion or adsorption.