(244g) Data Driven Approaches for Predicting Single Atom Alloy Nanoparticle Structures.

Single-atom alloys (SAA) are a subset of bimetallic alloy catalysts. By combining the catalytic activity of two metals, they not only form a superior catalyst compared to their monometallic counterparts but also complement the shortcomings of the individual metal catalysts. Most of the research on SAAs catalysts has focused on doping trace amounts of transition metals into the coinage metals (Cu, Ag, Au) for various reactions, including the synthesis of H₂O₂, the water-gas shift reaction, and coupling reactions. Here, we use density functional theory (DFT) calculations to explore AuPd and CuPd SAA catalysts. Specifically, we elucidate the preferential arrangement of Pd metal atoms in the host metal by examining 38–586 atom catalyst nanoparticles. DFT-calculated metal-atom exchange energies suggest that Pd prefers subsurface sites in Au nanoparticles and surface sites in Cu nanoparticles. In addition, exchange energies increase (less-stable exchange) with increasing Pd neighboring sites, suggesting that Pd would prefer to be dispersed when doped into Au and Cu. To expand the combinatorial space of SAA catalysts, we consider all possible combinations of bimetallic SAAs on slab models, spanning metals from groups 8 to 11. These slab models inform further alloy calculations in particle models. These data are being used to generate physics-based and machine learning models to describe exchange energies as a function of exchange site (e.g., corner atom, subsurface atom), local composition (i.e., number of Pd neighbors for M-Pd SAA catalysts), and overall composition. These models can be used to inform Monte Carlo simulations that will predict nanoparticle structure as a function of composition, particle size, and even reaction conditions by extending the above models to include the impact of adsorbates (e.g., CO*, O*). These structures will be used to gain insights into experimental results and as models for additional DFT calculations.