"True" iron(V) and iron(VI) porphyrins: a first theoretical exploration.

We present here a first theoretical characterization of iron(V) (S = (3)/(2)) and iron(VI) (S = 0) porphyrin intermediates. The Fe(V) calculations exhibit exceptionally narrow convergence radii and we believe that for this reason they have long eluded researchers working on high-valent iron intermediates. The Fe(V)-N(nitrido) bond distance in the DFT(PW91/TZP) optimized geometry of Fe(V)(P)(N) is 1.722 A, comparable to and slightly longer than the Fe(IV)-O bond distance of 1.684 A in Fe(IV)(P)(O) and the Fe(IV)-N(imido) bond distance of 1.698 A in Fe(IV)(P)(NH). In contrast, the Fe(VI)-N(nitrido) bond distances in [Fe(VI)(P)(N)](+) (S = 0) and Fe(VI)(P)(N)(F) (S = 0) are dramatically shorter, 1.508 and 1.533 A, respectively, consistent with the formal triple bond character of the Fe(VI)-N(nitrido) bond. The nitrido ligand appears to be uniquely capable of stabilizing a "true" Fe(V) center, in the sense defined in the paper. All three unpaired electrons in Fe(V)(P)(N) are completely localized on the Fe(V)-N(nitrido) axis, with the Fe and N gross atomic spin populations being 1.579 and 1.550, respectively. In contrast, an axial ligand set consisting of an oxide and a fluoride do not stabilize an Fe(V) ground state but favor an electronic structure best described as an Fe(IV)-oxo porphyrin pi-cation radical.