Multiscale computational investigation of catalytic properties of zirconia supported noble metals

Abstract

This thesis work explores the catalytic properties of zirconia supported noble metals rhodium and platinum, using multiscale computational methods. Density functional theory (DFT) basedmethods were employed to gain atomic scale knowledge about industrially important catalytic phenomena over a wide size range of metal species. Models incorporating both the metal and the support, as well as first-principles atomistic thermodynamics and microkinetics were used in order to close the material and temperaturepressure gaps between theory and experiment. DFT results show that deposited Rh and Pt species on zirconia can enhance the reducibility of the support by accepting electrons from zirconia. A microkinetic model of the water–gas shift reaction was developed for Rh/ZrO2 and used to compare the roles the different domains of the catalyst, i.e. the metal, support, and their interface, play in the reaction mechanism. The interface of large nanoparticles, represented with a supported rod model, is shown to be a potential active site for the reaction due to its ability to bind and activate water more effectively than the pure Rh sites. A thorough screening of the water dissociation reaction on the perimeter sites of small supported clusters demonstrate that each interfacial site is unique which leads to non-scaling behaviour. A non-equilibrium nanothermodynamic framework was developed to assess the kinetics of initial stages of agglomeration of single-atoms and sub-nano clusters. Strong thermodynamic driving force is found for the process on ideal zirconia in the absence of CO. The results indicate that a CO atmosphere could stabilise Pt single-atoms and even induce cluster disintegration. Both metal single-atoms are strongly anchored by cationic and anionic defects on zirconia. These results give new atomic level insight into the behaviour of the Rh/ZrO2 and Pt/ZrO2 catalysts.