Which DFT functional performs well in the calculation of methylcobalamin? Comparison of the B3LYP and BP86 functionals and evaluation of the impact of empirical dispersion correction.

Controversy remains regarding the suitable density functionals for the calculation of vitamin B(12) systems that contain cobalt. To identify the optimum functionals, geometry optimization calculations were performed on a full-size model of methylcobalamin (MeCbl) using the B3LYP, B3LYP-D, BP86, and BP86-D methods in conjunction with the 6-31G* basis set. Single-point energy evaluations were also performed with the 6-311+G(2d,p) basis set. Consistent with previous studies, the BP86-optimized geometry showed fairly good agreement with the experimental geometry. Various factors that may influence the homolytic bond dissociation energy (BDE) of the Co-C bond of MeCbl were systematically evaluated with these methods. Our analysis demonstrated that dispersion was the largest correction term that influenced the magnitude of BDE. Previous studies have shown that B3LYP significantly underestimates BDE, whereas BP86 gives BDE values that are fairly close to the experimental values (36-37 kcal/mol). The same trend in the relative magnitudes of the BDEs was observed in the present calculations. However, BP86 underestimated the BDE for a full model of MeCbl. When the amount of Hartree-Fock exchange in the B3LYP functional was reduced to 15% and the dispersion correction was made (i.e., B3LYP*-D), the calculated BDE was in good accord with experimental values. B3P86-D also performed well. A detailed analysis was undertaken to determine which atoms in cobalamin have large dispersion interactions with a methyl fragment of MeCbl.