A Combined First-Principles\nand Data-Driven Computational\nFramework to Analyze the Surface Structure, Composition, and Stability\nof Binary Alloy Catalysts

Pt-based bimetallic alloys are considerably more active\nthan Pt\nin the oxygen reduction reaction (ORR). This increased activity has\nbeen attributed to the weakening of the adsorption of ORR intermediates\ndue to the presence of Pt “skins.” Density functional\ntheory (DFT) calculations have, in turn, pointed to the importance\nof surface segregation energies and Pt leaching on the formation and\nstability of the skins on close-packed surfaces of Pt alloys. The\ngeneralizability of these insights across different chemical environments,\nsurface compositions, and facets, however, remains a subject of active\nresearch and is the focus of this work. We present a generalized computational\nframework combining DFT calculations and data-driven methods to predict\nthe stability of different Pt<sub>3</sub><i>X</i> (<i>X</i> = Ni, Co, Fe, and Cu) alloy facets under vacuum conditions\nand in the presence of an electrochemical environment, wherein we\nanalyze the combined effect of segregation, intrasurface phase separation,\nleaching, and surface oxidation as a function of electrode potential.\nThe analysis reveals that a subtle interplay of these factors influences\nPt skin formation and stability, with Pt segregation being a strong\nfunction of the surface structure and continuous base metal dissolution\nbeing thermodynamically, although not always kinetically, favored\nat ORR-relevant voltages.