(Revised) SSB-D: A new functional for biological applications

Abstract

We recently reported a new functional (SSB-D) that combines the best of the non-empirical PBE density functional with the best of the empirical OPBE density functional. It was constructed based on the performance of OPBE and PBE for spin states and SN2 reaction barriers and how these are affected by different regions of the reduced gradient expansion. A previous study already showed how by smoothly switching from OPBE to PBE at a predefined point P of the reduced density gradient s, we could determine which part of the exchange functional determines its behavior for the different interactions. These prior results allowed us to design a new exchange functional, which was subsequently fine-tuned to give the SSB-D functional. It showed the good results of OPBE for reaction barriers and spin-state energies, and combined it with the good (H-bond, π-stacking) results of PBE for weak interactions. The SSB-D functional includes Grimme’s dispersion correction, a portion of the KT1 functional and the simplified PBE (sPBEc) correlation functional. The new functional was shown to give good performance for the spin ground-state of iron complexes, transitionmetal ligand distances, and small deviations compared to CCSD(T) for SN2 barriers (2.7 kcal·mol-1), geometries (0.005 Å), H-bond distances (0.012 Å), and weak interactions (S22 set, 0.5 kcal·mol-1).Moreover, it works excellent for two difficult cases for DFT as put forward by Grimme. More recently, it was shown to further improve upon OPBE for NMR shieldings, give good results for magnetizabilities, and work well for the relative stability of water-hexamer isomers