Universal metrics for predicting the activity of a wide range of fuel-cell catalysts

Abstract

ABSTRACTThe increasing demand for universal metrics in material development, particularly in the context of material informatics (MI), emphasises the need for catalytic activity metrics in fuel-cell research, specifically for the oxygen reduction reaction (ORR). In this study, we comprehensively analyse 10 Pt/C catalysts using a constant-current density protocol and correlate the ORR activity with the constant kinetic current density, thereby demonstrating the effectiveness of this alternative metric. Our results reveal that the constant-current density protocol offers a reliable and consistent assessment of catalyst performance, providing a promising alternative to existing evaluation methods. By evaluating various protocols for extracting activity metrics from the datasets of fuel-cell catalysts, we establish an optimised protocol that accounts for experimental and analytical errors in obtaining universal activity metrics. Conventional methods relying on data obtained at a fixed potential are inadequate due to variable contributions of mass transport among the different materials. Instead, we propose a protocol using a constant kinetic current density obtained by the anodic scans of voltammetry (in the range of below 8 μA cm−2, normalized by absolute electrochemically active surface area, or 6 × 10−21 A per active site) to minimise influence of un-compensated resistance and non-ORR current, enabling robust analysis of catalyst activity trends. Depending on the purpose of the MI prediction, multiple metrics obtained at different conditions such as cathodic scans are preferable, and thus a range of metrics should be deposited in a database with their metadata. This unbiased comparison of catalysts emphasises the importance of revisiting assessment protocols. The resulting database of universal activity metrics is expected to be valuable for accelerating catalyst development through the application of materials informatics.

Keywords

• Fuel cells
• catalysts
• oxygen reduction reaction
• universal metrics
• materials informatics
• data science