First Principles Calculations of Supported Catalysts: CO Oxidation on MgO Supported Gold Nanoparticles

Introduction Bulk gold is a chemically inert metal, and had previously been disregarded as a candidate for catalytic applications [1]. Haruta’s pioneering work [2] showed exceptional reactivity on gold nanoparticles of 2-5 nm in diameter. Among reactions, the low temperature CO oxidation [3] on Au has created a strong interest. Model catalyst samples, which consist of planar metal oxide surfaces (the ‘catalyst support’) onto which metal particles are deposited, have been used in experiments [4,5]. Research has revealed that reducible supports result in the best activity for CO oxidation on Au, while irreducible supports, such as SiO2 and Al2O3, are not as active [4]. TiO2 was found to be an extremely active support. For this reason, Au nanoparticles on TiO2 (110) have been a prototype catalyst [3,5]. Although it is widely accepted that gold nanoparticles ranging 2-5 nm in diameter exhibit the greatest reactivity [2,5], there is still much debate on the nature of active sites as well as on the reaction mechanisms. Theoretical studies on CO oxidation have explored metal-oxide supported gold clusters and some support the idea of an active metal particle/support interface [6]. Other calculations indicate that low coordination sites are the active centers [7]. A common feature of all DFT studies is that they have been limited to very small clusters due to the sheer computational cost. In this work, first principle calculations have been performed to examine the CO interaction with unsupported and supported Au clusters on a (001) MgO surface. The unsupported gold clusters ranged in diameter from a few angstroms to approximately 1.5 nm exposing various symmetries. Selective, very stable gold structures were supported on a MgO surface and their interaction with CO was investigated. Alkali doping of unsupported and MgO supported gold clusters was also investigated.