Recent trapping and spectroscopic characterization of an O 2 adduct for the non-heme enzyme homoprotocatechuate 2,3-dioxygenase (HPCD) demonstrates it to be a Fe III -superoxo species. This proposal is in direct opposition to the consensus mechanism (., Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 7347-7352) in which the metal facilitates the transfer of electrons from the substrate to O 2 to form the reactive species in the mechanism without changing oxidation state. In this study we performed a detailed analysis of the electronic structure of the O 2 adduct for the mutant and native enzymes and the nature of oxygen activation in the reaction mechanism of HPCD using a combination of computational chemistry and theoretical M€ ossbauer spectroscopy. Our results are in agreement with the available experimental data and demonstrate that even for the native enzyme changes in the metal oxidation state are an important factor in oxygen activation.

This manuscript describes the formally iron(I) complexes L(Me)Fe(Py-R)(2) (L(Me) = bulky β-diketiminate; R = H, 4-tBu), in which the basal pyridine ligands preferentially accept significant unpaired spin density. Structural, spectroscopic, and computational studies on the complex with 4-tert-butylpyridine ((tBu)py) indicate that the S = 3/2 species is a resonance hybrid between descriptions as (a) high-spin iron(II) with antiferromagnetic coupling to a pyridine anion radical and (b) high-spin iron(I). When the pyridine lacks the protection of the tert-butyl group, it rapidly and reversibly undergoes radical coupling reactions that form new C-C bonds. In one reaction, the coordinated pyridine couples to triphenylmethyl radical, and in another, it dimerizes to give a pyridine-derived dianion that bridges two iron(II) ions. The rapid, reversible C-C bond formation in the dimer stores electrons from the formally reduced metal as a C-C bond in the ligands, as demonstrated by using the coupled diiron(II) complex to generate products that are known to come from iron(I) precursors.

Dinuclear metal systems based on sterically-hindered, three-coordinate transition metal complexes of the type ML3 where the ancillary ligands L comprise bulky organic substituents, hold great promise synthetically for the activation and scission of small, multiply-bonded molecules such as N2, NO and N2O. In this study we have employed density functional methods to identify the metal/ligand combinations which achieve optimum activation and/or cleavage of N2. Strong pi donor ligands such as NH2 and OH are found to produce the greatest level of activation based on N-N bond lengths in the intermediate dimer complex, L3Mo(mu-N2)MoL3, whereas systems containing the weak or non-pi donor ligands NH3, PH3, OH2 and SH2 are found to be thermodynamically unfavourable for N2 activation. In the case of the Mo-NH2 and W-NH2 systems, a fragment bonding analysis reveals that the orientation of the amide ligands around the metal is important in determining both the spin state and the extent of dinitrogen activation in the intermediate dimer. For both systems, an intermediate dimer structure where one of the NH2 ligands on each metal is rotated 90 degrees relative to the other ligands, is more activated than the structure in which the NH2 ligands are trigonally disposed around the metals. The level of activation is found to be very sensitive to the electronic configuration of the metal with d3 metal ions delivering the best activation along any one transition series. In particular, strong activation or cleavage of N2 was calculated for the third row d3 metals systems involving Ta(II), W(III) and Re(IV), with the level of activation decreasing as the nuclear charge on the metal increases. This trend in activation reflects the size of the valence 5d orbitals and consequently, the capacity of the metal to back donate into the dinitrogen pi* orbitals.

Obtaining an accurate theoretical model for the activation of dinitrogen by three-coordinate molybdenum amide complexes (e.g. Mo(NH2)3) is difficult due to the interaction of various high- and low-spin open-shell complexes along the reaction coordinate which must be treated with comparable levels of accuracy in order to obtain reasonable potential energy surfaces. Density functional theory with present-day functionals is a popular choice in this situation; however, the dinitrogen activation reaction energetics vary substantially with the choice of functional. An assessment of the reaction using specialized wave function based methods indicates that although current density functionals in general agree qualitatively on the mechanistic details of the reaction, a variety of high-level electron correlation methods (including CCSD(T), OD(T), CCSD(2), KS-CCSD(T), and spin-flip CCSD) provide a consistent but slightly different representation of the system.