Sensitivity of Density Functional Theory Methodology for Oxygen Reduction Reaction Predictions on Fe–N4-Containing Graphitic Clusters

Density functional theory (DFT) was used to examine the O2 reduction reaction on Fe–N4-containing graphitic carbon clusters (Fe–N4–G) modeled after recent experimentally identified active sites, Mössbauer spin-state predictions and electrochemical reaction behavior in alkaline media. A detailed analysis of the O2, O, H2O, OOH, and OH adsorbate interactions on the Fe–N4–G cluster with solvation and/or dispersion corrections are considered. The total and partial density of states for the α- and β-spin orbitals are compared for the adsorbate of interest, Fe atom and surrounding graphitic cluster. Relative free-energy diagrams are constructed, which allow us to compare DFT predictions to experimental results for O2 reduction on systems containing embedded Fe–N4 clusters. For all reaction steps, different DFT functionals are explored and the respective geometries, energetics, and spin-states for each adsorbate interaction are reported for six commonly used functionals including B3LYP, M06-2X, M06-L, PBE, TPSSh, and ωB97X-D. Functionals with high fractions of exact exchange were found to favor higher spin-states, as well as stronger binding of the adsorbates, making these methodologies less feasible for Fe–N4-containing electrocatalysts when compared to experimental data. Pure functionals with and without empirical correlation exhibit different ground spin-states and geometries, however the free energy diagrams yield similar conclusions at relevant overpotentials. The activation energy for the O–OH bond scission step, as well as OH desorption from the Fe–N4–G cluster are discussed since the barrier could prohibit a pure 4e– ORR. Finally, we discuss the energetically unfavorable steps for select overpotentials, which provides the experimentalist with a tuning knob for electrocatalytic design.