Developing low-cost, high-performance metal-free oxygen reduction catalysts demands precise quantification of structure-performance relationships in nitrogen-doped carbons. Structural search algorithms combined with density functional theory (DFT) enable comprehensive analysis of catalytic performance versus structural features. Using nitrogen-doped graphene quantum dots (NGQDs) as models, we computationally resolve how nitrogen species types (pyridinic-N, graphitic-N) and substitution positions influence the stability and intrinsic activity. Boltzmann statistics quantify the contributions of all configurations to the current density during oxygen reduction. Furthermore, we identify dominant configurations grouping NGQDs into configuration lumps. Thermodynamically, nitrogen atoms preferentially occupy carbon atoms at defect-adjacent sites as pyridinic-N. Their spatial distribution controls local atomic charge redistribution and adsorption environments, thereby modulating intrinsic activity. These materials exhibit extreme configuration sensitivity: thermodynamically stable dominant configurations may contribute minimally to current density. Crucially, lumped pyridinic-N configurations dominate ORR performance. This work provides theoretical insights supporting carbon adjacent to pyridinic-N as the primary active site in N-doped carbon ORR catalysts. It establishes a universal framework for analyzing structure–activity relationships in nonmodel catalytic systems.
Yuan et al. (2025) studied this question.