Photocatalytic water splitting for hydrogen generation is a promising approach for converting solar energy to chemical energy. Platinum-doped vacancy-ordered perovskite Cs2SnBr6 exhibits high performance in photocatalytic hydrogen evolution, with peak activity at 25% doping. Nevertheless, the underlying microscopic mechanisms remain unclear. Herein, we systematically investigate Pt doping effects on Cs2SnBr6 photocatalysis using a multiscale approach combining first-principles calculations, nonadiabatic molecular dynamics, and machine learning. Results show that Cs2Sn0.75Pt0.25Br6 achieves the longest carrier recombination lifetime (134 ps) and fastest hot electron cooling rate (7 ps), revealing a synergy of rapid energy relaxation, prolonged charge separation, and efficient carrier transport that enhances the photocatalytic efficiency. Machine learning identifies the ∠Pt–Sn–Br bond angle as a key descriptor governing nonadiabatic coupling strength, influencing carrier lifetime, and yielding the lowest hydrogen evolution overpotential in both acidic and alkaline conditions. This work provides kinetic insights into the optimal photocatalytic performance and offers theoretical guidance for designing efficient perovskite-based photocatalysts.
Wei et al. (2026) studied this question.