The Chemical Physics of Cluster-Based Catalyst Systems

This proposal seeks renewal funding for research exploring the novel physical and chemical properties of atomic clusters. Clusters are small clumps of matter containing from tens to hundreds of atoms. Their properties are determined by an intricate interplay of factors related to their small size and can vary dramatically with the addition of even a few atoms. This sensitive size dependence creates the possibility of tailoring cluster properties for specific applications such as catalysis. The overall goal of our research is to use first-principles computational methods based on density functional theory to gain an improved understanding of the properties of clusters, particularly those that are important for applications. The results can be used to identify promising cluster systems for potential applications.

A main thrust of the project will be exploring the reactivity of mixed metal clusters. Experiments have shown that mixed metal clusters can have significantly higher reactivity than the corresponding single component clusters. Our specific interest involves dilute mixtures of Pd atoms with noble metals like Cu and Ag. Such clusters are analogs of the bulk “single atom alloy” (SAA) system consisting of a Cu surface with very small and well-separated islands of Pd atoms. Catalytic dehydrogenation reactions are found to occur readily at the Pd islands, with a spillover of H atoms onto the Cu surface. We will determine whether “SAA clusters” can be found that have high reactivity, but are more Pd atom efficient that pure Pd clusters. We will use techniques developed previously in our group to obtain accurate structural models for SAA clusters and systematically model their reactivity for dehydrogenation reactions. Among the issues to be addressed are the following: What is the relationship between reactivity and Pd concentration? Does the size of Pd islands affect reactivity – are single Pd substitutions sufficient, or are multiple Pd neighbors necessary? Do bulk-phase reaction descriptors also successfully predict cluster reactivity? Can scaling relations be found that will allow us to extrapolate properties of SAA clusters into the nanoscale range?

A second project thrust will be to further develop and refine a method established in our group to investigate the dielectric response of clusters. The method decomposes the cluster polarizability into contributions from individual atoms and expresses these in terms of local dipole and charge transfer components. The site polarizabilities have a straightforward physical interpretation in terms of changes in local charge densities in response to an applied electric field. Thus, effects like electrostatic screening can be read directly from the results. We will continue to develop the method in this project, with the goal of using it to address the question of metallicity in sub-nanometer systems and to extract atomic polarizabilities that could be used in molecular modeling applications.

The project will support two graduate students at Central Michigan University.