The electronic structure of the family of sandwich complexes M(η6-C7H8)2 and M(η7-C7H7)(η5-C7H9) (M = Mo, Mo+, Zr) was investigated using density functional theory, and their geometries were accounted for. The mechanism for conversion of M(η6-C7H8)2 into M(η7-C7H7)(η5-C7H9) by hydrogen transfer was investigated, and the experimental trends were reproduced. The neutral Mo species was calculated to have an activation free energy 13 kJ mol-1 lower than the cationic Mo+. The rate-limiting step for the transfer in the neutral molybdenum species is breaking of the C−H bond to form a molybdenum hydride intermediate, which rapidly converts into the rearranged product. The transition state is late on the reaction coordinate, and C−H bond breaking occurs after ring slippage to form an (η6, η4) 16-electron species. The cationic molybdenum species has a similar reaction profile. The zirconium species is found to have two competing pathways both with higher activations energies: one with an (η7,η5) Zr−H intermediate, the other occurring by direct transfer of the hydrogen without forming Zr−H species. The effect of adding PH3 to the zirconium system has also been studied, which has the effect of lowering the barrier to the direct transfer of hydrogen between the ligands.
Density functional methods have been used to calculate the geometries, electronic structure and ionization energies (IE) of N-heterocyclic carbene complexes of palladium and platinum, [M(CN2R2C2H2)2](M = Pd, Pt; R = H, Me, Bu t). Agreement with X-ray structures (R = Bu t) was good. Calculated IE agreed well with the photoelectron (PE) spectra (R = Bu t); metal bands were calculated to be within 0.25 eV of the experimental values, whereas the higher lying ligand bands deviated by up to 0.9 eV. Spin-orbit methods were needed to achieve this level of agreement for the Pt complex, but the calculations were found to underestimate the spin-orbit splitting somewhat. The principal metal-ligand bonding is between the carbene lone pair HOMO and a (d(z2)+ s) hybrid on the metal. The metal p(z) orbital contributes very little to the bonding. The metal d(xz,yz) orbitals mix primarily with the filled pi3 orbitals on the ligands and secondarily with the empty pi5 orbitals. Consequently they are little stabilized in comparison to the metal d(xy,x2- y2) orbitals, which are non-bonding in the complex. The first PE band for both the Pd and Pt complexes is from ionization of a (s - d(z2)) hybrid orbital. The IE is greater for Pt than for Pd on account of the post-lanthanide relativistic stabilization of the Pt 6s orbital.
In contrast to [Cp(2)MoH(3)](+), which is a thermally stable trihydride complex, the ansa-bridged analogue [(eta-C(5)H(4))(2)CMe(2)MoH(H(2))](+) (1) is a thermally labile dihydrogen/hydride complex. Partial deuteration of the hydride ligands allows observation of J(H)(-)(D) = 11.9 Hz in 1-d(1) and 9.9 Hz in 1-d(2) (245 K), indicative of a dihydrogen/hydride structure. There is a slight preference for deuterium to concentrate in the dihydrogen ligand. A rapid dynamic process interchanges the hydride and dihydrogen moieties in complex 1. Low temperature (1)H NMR spectra of 1 give a single hydride resonance, which broadens at very low temperature due to rapid dipole-dipole relaxation (T(1) = 23 ms (750 MHz, 175 K) for the hydride resonance in 1). Low temperature (1)H NMR spectra of 1-d(2) allow the observation of decoalescence at 180 K into two resonances. The bound dihydrogen ligand exhibits hindered rotation with DeltaG(150) = 7.4 kcal/mol, but H atom exchange is still rapid at all accessible temperatures (down to 130 K). Density functional calculations confirm the dihydrogen/hydride structure as the ground state for the molecule and give estimates for the energy of two hydrogen exchange processes in good agreement with experiment. The presence of the C ansa bridge is shown to decrease the ability of the metallocene fragment to donate to the hydrogens, thus stabilizing the (eta(2)-H(2)) unit and modulating the barrier to H(2) rotation.
The electron spin g- and hyperfine tensors of the endohedral metallofullerene Sc@C82 are anisotropic. Using electron spin resonance (ESR) and density functional theory (DFT), we can relate their principal axes to the coordinate frame of the molecule, finding that the g-tensor is not axially symmetric. The Sc bond with the cage is partly covalent and partly ionic. Most of the electron spin density is distributed around the carbon cage, but 5% is associated with the scandium dyz orbital, and this drives the observed anisotropy.