Data‐driven structural descriptor for predicting platinum‐based alloys as oxygen reduction electrocatalysts

Owing to increasing global demand for carbon neutral and fossil-free energy systems, extensive research is being conducted on efficient inexpensive electrocatalysts catalyzing the kinetically sluggish oxygen reduction reaction (ORR) at cathode of fuel cells. Platinum (Pt)-based alloys are considered promising candidates replacing expensive Pt catalysts. However, current screening process Pt-based time-consuming labor-intensive, descriptor predicting activity catalysts generally inaccurate. This study proposed a strategy by combining high-throughput first-principles calculations machine learning explore used alloy with high utilization low consumption. Among 77 prescreened candidates, we identified 5 potential ORR overpotential. Furthermore, during second third rounds active learning, more based relationship between structural features their activity. In addition, highlighted role in found that difference electronegativity heteroatom, valence electrons number ratio heteroatoms around main factors affect ORR. More importantly, combination those can be as alloys. We believe findings this will provide new insight contribute exploring consumption experimentally.