A multiconfigurational ab initio study of the zero-field splitting in the di- and trivalent hexaquo-chromium complexes.

A detailed analysis of the value of zero-field splitting for the di- and trivalent chromium hexaquo complexes is presented. The effect of the Jahn-Teller distortion was studied, for the case of the divalent complex, through the use of state-averaged CASSCF calculations, for the mapping of the potential energy surface along the e(g) normal modes. At the minima of the surface, multiconfigurational ab initio calculations (spectroscopy oriented configuration interaction, SORCI, and difference dedicated configuration interaction, DDCI) were used for the calculation of the D tensor and the analysis of the individual contributions to it. The final value calculated with the SORCI method (D = -2.45 cm(-1)) for the divalent complex is in excellent agreement with the experimental estimate (D = -2.3 cm(-1)). The importance of inclusion of the direct spin-spin coupling contribution to D is pointed out ( approximately 16%). At the same time, contributions of the higher than the lowest (3)T(1g) triplets were found to be non-negligible as well ( approximately 11%). The accuracy of second-order perturbation theory for the calculation of SOC was investigated and found to be satisfactory. For comparison, DFT calculations were performed with hybrid (B3LYP) and nonhybrid (BP86) functionals and were found to be inferior to the wave function based ab initio methods.