Spin-State Energetics of FeII Complexes - The Continuing Voyage Through the Density Functional Minefield

Computing absolute spin-state energies continues to challenge quantum chemistry. Correlated wavefunction methods offer high accuracy but are expensive, whereas density functional theory (DFT) is faster, but its accuracy is variable. No universal functional has emerged, but predicting trends in spin-state energies is easier. Here, a simple DFT protocol is applied to the subtle variations in magnetic moments for Fe-II-(R,R')Pytacn complexes (Inorg. Chem. 2013, 52, 9229). Both BP86 and OPBE give satisfactory correlation coefficients (R-2 = 0.87 and 0.75, respectively), whereas the range-separated functional CAM-B3LYP performs worse (R-2 = 0.37). However, even for BP86 and OPBE, the p-CO2Et substituent is predicted to be too electron-donating. The relatively poor performance of CAM-B3LYP may be due to the more exacting test of DFT provided by mixed-ligand systems, where the competition between donors is an additional factor.