Evaluating Potential Catalytic Active Sites on Nitrogen-Doped Graphene for the Oxygen Reduction Reaction: An Approach Based on Constant Electrode Potential Density Functional Theory Calculations

Discovering electrocatalysts that are active, stable, and inexpensive for the oxygen reduction reaction (ORR) is important for large-scale commercialization of proton-exchange membrane fuel cell technology. Recently, nitrogen-doped graphene has emerged as a promising material to catalyze this important reaction. In this study, we used density functional theory (DFT) combined with an electrochemical model, allowing one to fix the electrode potential during the entire reaction (Const-U), to study the ORR catalyzed by graphene doped with graphitic nitrogen (gN). First, we determined that the adsorption energies (ΔGad's) of the three ORR intermediates (i.e., *OOH, *O, and *OH) are U-dependent and can be fitted to the linear equation ΔGad(A) = βA + αAeU (A = *OOH, *O, or *OH). This result is in sharp contrast to the hypothesis used in previous DFT studies (performed under a constant charge of zero, Const-Q0), which assumed that the ΔGad's are independent of U. However, to our surprise, the well-known scaling relationship (ΔGad(*OOH) - ΔGad(*OH) ∼3.24 eV), which was developed using DFT at Const-Q0, still holds under Const-U conditions. Based on this scaling relationship and the fitted parameters, an equation to determine the ORR performance of a material based on a single parameter (β*OOH) was derived, and the optimum β*OOH value (-1.18 eV) to achieve the best performance is reported. Importantly, we find that β*OOH can be predicted by a simple additivity rule without performing any DFT calculations. This rule and the derived equation provide a method for achieving fast screening of the ORR centers with no computational cost. In addition, this approach was applied to search for highly active sites on graphene doped with either three or four gNs (containing more than 2000 reaction sites). Four carbon sites are predicted to be active for the ORR, and three of these sites were confirmed by our DFT calculations. Our work provides a method to quickly screen the ORR performance of reaction sites on gN-doped graphene and is useful for the rational design of this type of catalyst.