Accelerated Discovery of High‐Performance PCFC Cathodes: Computational‐Experimental Optimization of Cobalt‐Substituted Ba_0.95La_0.05FeO_3‐δ

Abstract Developing high‐performance, cost‐effective protonic ceramic fuel cell (PCFC) cathodes involves navigating complex compositional landscapes to optimize multiple competing properties. This study presents a novel integrated methodology that combines computational screening with machine learning potentials, targeted experimental validation, and Bayesian optimization to accelerate the design of Co‐substituted Ba_0.95La_0.05FeO_3‐δ (BLF) cathodes. Utilizing Bayesian optimization, BLFC15 (15% Co) is identified as the optimal composition, achieving a 58% reduction in area‐specific resistance at 500 °C compared to pristine BLF. This integrated approach, which includes mechanistic studies through density functional theory (DFT) calculations and experimental characterization, illustrates the effectiveness of computational screening and Bayesian optimization in accelerating materials discovery and efficiently optimizing complex material properties for enhanced PCFC cathode performance.