The synergy between Au nanoparticles and oxide, carbide, nitride, sulfide supports promotes catalytic turnovers at interfacial sites. This bifunctional gain is highly structure sensitive, being confined to active sites with specific coordination numbers. Designing improved bifunctional interfaces is limited by a lack of systematic understanding of how the interface structure and composition influences catalytic properties. Moreover, we cannot tailor structure sensitivity at interfaces such that a higher density of interfacial active sites participates in catalytic turnovers. Using a new class of site-specific scaling relations as a diagnostic tool, we holistically understand how the interfacial structure and composition influences adsorption energies of catalytic descriptors like CO*. We consider Au nanostructures ranging from single sites to epitaxial films supported on two- and three-dimensional carbides, oxides, sulfides, and nitrides. We classify site-specific scaling relations according to proximity to the interface, support termination, and support type. These scaling relations are constructed using data-driven techniques employing basic chemical properties of the interface as features. Our study reveals that perturbations in adsorption energies of CO* caused by the interface are strongest for low coordinated interfacial sites located at the interface. This interfacial perturbation decays at metal sites located beyond two lattice units from the interface. The scaling relation slopes represent changes in adsorption energies with variations in coordination numbers of interfacial Au sites, reflecting structure sensitivity. This structure sensitivity is highly tunable with interfacial modifications creating negatively charged Au sites being less structure sensitive. The decreased structure sensitivity increases the density of interfacial sites with similar catalytic properties, potentially boosting the overall rate per mass of precious metal. Validation tests across &gt; 126 different Au/support interfaces yield mean averaged errors of 0.11 eV, corroborating our scheme. We conclude by applying this scheme to design next generation Au interfaces.

Tailoring structure sensitivity of metal nanoparticles on 2D and 3D supports by controlling electrostatic interactions: Insights from data driven approaches

ACS Fall 2021

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.