How does the heteroatoms doping regulate the oxygen electrocatalysis performance of single atom catalysts?

Single-atom catalysts (SACs) with explicit metal-nitrogen-carbon moieties have attracted extensive attention over the past decade. Yet, understanding their catalytic origin toward the oxygen evolution reaction (OER) and oxygen reduction reaction (ORR) remains a great task due to the complicated coordination environment surrounding the active center. Here, we employed large scale density functional theory computations to examine the OER/ORR performance on 144 heteroatom-doped SACs (X-TMN4@Gra, TM=Fe, Co, Ni, Cu, Rh, Ir, X = B, N, O, P, S). Computational results unveiled the significant role of heteroatom doping in affecting the local electronic properties of TM centers and their catalytic activity, even leading to the transition of potential limiting steps. Machine learning (ML) enabled to delve deeper into the correlation between atomic characteristics and catalytic activity, suggesting that ΔGOOH* and ΔGO*-OH* can serve as vital descriptors for anticipating ORR and OER performance, respectively. These revelations not only elucidate the catalytic discrepancy of reported SACs but provide a rational way to optimize secondary coordination of SACs, thus paving new opportunities to discover the high-performance electrocatalysts.