Abstract
We have performed density functional theory calculations on an oxo-iron porphyrin catalyst with chloride as an axial ligand and tested its reactivity toward propene. The reactions proceed via multistate reactivity on competing doublet and quartet spin surfaces. Close-lying epoxidation and hydroxylation mechanisms are identified, whereby in the gas phase the epoxidation reaction is dominant, while in environments with a large dielectric constant the hydroxylation pathways become competitive. By contrast to reactions with thiolate as an axial ligand all low-lying pathways have small ring-closure and rebound barriers, so it is expected that side products and rearrangements will be unlikely with Fe=O(porphyrin)Cl, whereas with Fe=O(porphyrin)SH some side products were predicted. The major differences in the electronic configurations of Fe=O(porphyrin)Cl and Fe=O(porphyrin)SH are due to strong mixing of thiolate orbitals with iron 3d orbitals, a mixing which is much less with chloride as an axial ligand. Predictions of the reactivity of ethylbenzene-h 12 versus ethylbenzene-d 12 are made.