Chemistry Department Colloquium

The current state of the art of electronic structure methods allows us to tackle model systems on the order of 100s to 1000s of atoms with suitable speed and efficiency to perform statistical mechanical sampling on millions of configurations. This has opened the door to using theoretical tools such as ab initio molecular dynamics (AIMD) combined with enhanced sampling techniques) to discover novel emergent phenomena that result from chemical complexity. Nowhere is this more needed than in catalysis, where models need to include support materials, the catalysts, the reactants and products all at elevated temperature and pressures. In this context, both global and local anharmonicities on the potential energy surface can lead to unexpected phenomena that can be discovered through large scale simulation yet are often not accounted for in current theoretical studies. This will be illustrated with examples drawn from the chemistry of metal particles on reducible supports [1-3], Brønsted acid chemistry in confined spaces [4] and reactivity at solid-liquid interfaces [5,6]. In the context of metal particles supported on reducible metal oxides (such as TiO2, CeO2 and RuO2), we have found that there is a strong coupling between the redox state of the support and the redox properties of the nanoparticle [1,2] such that unique catalytic processes can occur including: Redox state dependent reaction energies [2,3]; Formation of transient single atoms, which are themselves catalytic [1,3]; For prototypical reactions such as CO oxidation, a complex mechanistic landscape where catalysis can occur by competing mechanisms involving both the nanoparticle and single atom sites at the same time. Similar methods have been used to investigate the confinement effect in zeolite [4] and understand the free energetics of acid catalysis in confined media. It is shown that due to the large anharmonic effects associated with molecules, such as ethanol, interacting with the walls of a s