Abstract
Metal-nitrogen-carbon single-atom catalysts (SACs) have exhibited substantial potential for CO(2) electroreduction. Unfortunately, the SACs generally cannot generate chemicals other than CO, while deep reduction products are more appealing because of their higher market potential, and the origin of governing CO reduction (COR) remains elusive. Here, by using constant-potential/hybrid-solvent modeling and revisiting Cu catalysts, we show that the Langmuir-Hinshelwood mechanism is of importance for *CO hydrogenation, and the pristine SACs lack another site to place *H, thus preventing their COR. Then, we propose a regulation strategy to enable COR on the SACs: (I) the metal site has a moderate CO adsorption affinity; (II) the graphene skeleton is doped by a heteroatom to allow *H formation; and (III) the distance between the heteroatom and the metal atom is appropriate to facilitate *H migration. We discover a P-doped Fe-N-C SAC with promising COR reactivity and further extend this model to other SACs. This work provides mechanistic insight into the limiting factors of COR and highlights the rational design of the local structures of active centers in electrocatalysis.