Collective Effect in a Multicomponent Ensemble Combining Single Atoms and Nanoparticles for Efficient and Durable Oxygen Reduction

Metal single‐atom catalysts represent one of the most promising non‐noble metal catalysts for the oxygen reduction reaction (ORR). However, they still suffer from insufficient activity and, particularly, durability for practical applications. Leveraging density functional theory (DFT) and machine learning (ML), we unravel an unexpected collective effect between FeN4OH sites, CeN4OH motifs, Fe nanoparticles (NPs), and Fe−CeO2 NPs. The collective effect comprises differently‐weighted electronic and geometric interactions, whitch results in significantly enhanced ORR activity for FeN4OH active sites with a half‐wave potential (E1/2) of 0.948 V versus the reversible hydrogen electrode (VRHE) in alkaline, relative to a commercial Pt/C (E1/2, 0.851 VRHE). Meanwhile, this collective effect endows the shortened Fe−N bonds and the remarkable durability with negligible activity loss after 50,000 potential cycles. The ML was used to understand the intricate geometric and electronic interactions in collective effect and reveal the intrinsic descriptors to account for the enhanced ORR performance. The universality of collective effect was demonstrated effective for the Co, Ni, Cu, Cr, and Mn‐based multicomponent ensembles. These results confirm the importance of collective effect to simultaneously improve catalytic activity and durability. The collective effect between FeN4OH sites, CeN4OH motifs, Fe nanoparticles, and Fe−CeO2 nanoparticles in an active multicomponent ensemble has been unraveled with the capacity to simultaneously enhance ORR activity (a half‐wave potential of 0.948 V vs. RHE) and durability (negligible loss after 50k cycles) in alkaline. Density functional theory and machine learning reveal the necessary electronic interactions (ϵd, Wf, μB, Q) and geometric interactions (d and dFe−N) in a collective effect.