M–N4–C single-atom catalysts (MN4) have gained attention for their efficient use at the atomic level and adjustable properties in electrocatalytic reactions like the ORR, OER, and HER. Yet, understanding MN4‘s activity origin and enhancing its performance remains challenging. Edge-doped substituents profoundly affect MN4‘s activity, explored in this study by investigating their interaction with MN4 metal centers in ORR/OER/HER catalysis (Sub@MN4, Sub = B, N, O, S, CH3, NO2, NH2, OCH3, SO4; M = Fe, Co, Ni, Cu). The results show overpotential variations (0 V to 1.82 V) based on Sub and metal centers. S and SO4 groups optimize FeN4 for peak ORR activity (overpotential at 0.48 V) and reduce OER overpotentials for NiN4 (0.48 V and 0.44 V). N significantly reduces FeN4‘s HER overpotential (0.09 V). Correlation analysis highlights the metal center’s key role, with ΔG*H and ΔG*OOH showing mutual predictability (R2 = 0.92). Eg proves a reliable predictor for Sub@CoN4 (ΔG*OOH/ΔG*H, R2 = 0.96 and 0.72). Machine learning with the KNN model aids catalyst performance prediction (R2 = 0.955 and 0.943 for ΔG*OOH/ΔG*H), emphasizing M–O/M–H and the d band center as crucial factors. This study elucidates edge-doped substituents’ pivotal role in MN4 activity modulation, offering insights for electrocatalyst design and optimization.
