| Abstract Scope |
Hydrogenation of ABO3 perovskites can enable proton-coupled electron transfer, but its thermodynamic and kinetic limitations remain unclear. Here, we combine density functional theory (DFT), ab initio molecular dynamics, and machine learning (ML) to investigate 118 pristine perovskites across cubic, tetragonal, orthorhombic, and hexagonal phases. DFT calculations show that B-site chemistry controls hydrogenation thermodynamics: redox-inert Zr and Ti produce high reaction energies, whereas Fe, Co, Mn, and Ce favor hydrogen incorporation. Orthorhombic phases often exhibit mismatched hopping and rotation barriers that trap protons, while cubic and hexagonal phases provide more balanced barriers and improved long-range transport. An ML model using bond dissociation energies and electronic descriptors achieved (R2 = 0.96) and an MAE of 0.22 eV, identifying (Eb(B-O)) and band gap as dominant features. Screening further highlighted hexagonal BaPdO3 and cubic BaSnO3 as promising candidates, establishing design rules that balance redox activity, phase stability, and proton-transport kinetics. |