Metal-based electrocatalysts with data-driven designed particle size for hydrogen evolution

Abstract

Metal-based catalysts are currently at the forefront of hydrogen evolution catalysts, yet the influence of particle size on their catalytic activity is intricate. In this study, we compiled a database using actual test results and employed interpretable algorithms to analyze the relationship of catalyst factors with their catalytic properties. Our analysis revealed a bowl-shaped correlation between the measured overpotential and the particle size of metal-based catalysts, with the optimal size falling within the range of 1.5 to 3.0 nm. To corroborate the algorithmic analysis, we synthesized nanoalloys using an ultrathin carbon layer, which effectively restricted the movement of metal atoms at high temperatures. The series of materials prepared displayed a consistent pattern in their measured overpotential, aligning with our algorithm’s prediction. Notably, the 2.0 nm PtCoNi catalyst achieved superior catalytic performance with lower precious metal content than commercial Pt/C. Further data-driven analysis indicates that the size-dependent catalytic activity is shaped by the balance between metal-support interaction and the exposure ratio of the active site. This study underscores the efficacy of merging a data-driven approach with precise control of material synthesis to analyze complex catalytic systems, providing a potent tool for future exploration in the field.