Journal of the American Chemical Society, Vol.122, No.21, 5100-5104, 2000
Electronic structure of gallium, copper, and nickel complexes of corrole. High-valent transition metal centers versus noninnocent ligands
Using regular nonlocal density functional theory (DFT) as well as combined DFT and configuration interaction calculations, we have carried out a first theoretical study of the electronic structure of metallocorroles. The valence orbital energy spectra and the calculated electronic absorption spectrum of (Cor)Ga (Cor(3-) = corrolato), a prototype non-transition-metal carrole, are qualitatively similar to those of a metalloporphyrin such as zinc porphyrin. The "four-orbital model" holds well for corroles. The a(2) and b(1) KOMOs of (Cor)Ga are crude analogues of the well-known a(1u) and a(2u) porphyrin HOMOs, respectively. Thus, as in the case of porphyrins, there are two nearly equienergetic pi-cation radical states for corroles. DFT also appears to provide a good description of the stabilization of high-valent transition-metal centers and of Ligand noninnocence, two intertwined and central themes in metallocorrole chemistry. The calculated ground state of (Cor)Cu is a diamagnetic d(8) Cu(III) state, with Cu(II) pi-cation radical states only slightly higher in energy, which faithfully mirrors the experimental scenario. In contrast, there are no known Cu(UT) porphyrin complexes. For (Cor)Ni, low-spin Ni(II) x-cation radical states are significantly lower in energy than a Ni(III) state, again consistent with experiment, reflecting the favorable energetics of d(8) square planar complexes. The various optimized geometries reveal significant, characteristic structural changes accompanying the formation of A(2)- and B-1-type corrole pi-cation radicals. We predict that the resonance Raman spectra of metallocorroles should reflect these structural features and, thereby, assist in the assignment of valence tautomeric states of transition-metal corrole complexes.