(509da) Nuclearity As a Descriptor for Site and Surface Selection

Binary transition metal intermetallic catalysts have a diverse range of structures, compositions, and possible surfaces. Active host elements embedded in an inactive host metal lattice can lead to surfaces with well-defined site nuclearity, which is the number of active interconnected surface elements. Previous detailed theoretical and computational studies showed that the nuclearity is directly related to the relative thermochemistry and kinetics for ethylene hydrogenation. We developed an open-source method to catalog and classify intermetallic surfaces for their nuclearity using graph network tools. This tool was used to systematically enumerate nuclearity for over 10⁵ low Miller index surfaces across a set of 30 possible active/inactive metal combinations from the AFLOW library, which is the largest database of its kind. These surfaces were further studied to establish correlations between nuclearity and other bulk and surface properties like composition, crystal structure, and shapes of the adsorption sites. The study enables prediction of the best catalyst surface candidates for any reaction, given its reactants and transition states are well understood. We also present an interactive tool to choose a suitable adsorption site from wide range of intermetallic surfaces based on the desired site geometry.