Photoelectron spectroscopic investigation of isomeric methyl (isocyanide), η2-iminoacyl and η3-azaallyl complexes of molybdenum and tungsten

Abstract

Gas-phase ultraviolet photoelectron spectroscopy and extended-Hückel molecular-orbital calculations have been employed to investigate the valence electronic structures of twelve compounds drawn from three general isomeric classes; [M(Me)(CNR)(CO)₂(η⁵-C₅R′₅)][methyl (isocyanide)], [M{η²-C(NR)Me}(CO)₂(η⁵-C₅R′₅)](iminoacyl) and [M{η³-H₂CC(H)NR}(CO)(η⁵-C₅R′₅)](azaallyl)(M = Mo or W, R = Prⁱ or Bu^t, R′= H or Me). Although the methyl (isocyanide) compounds are unstable with respect to isomerisation to the corresponding azaallyls on the time-scale of the photoelectron spectroscopy experiment (ca. 4 h), sufficient data were acquired to support the assertion that RNC is a better π donor but poorer π acceptor than CO. The first two bands in the photoelectron spectra of all of the iminoacyl and azaallyl compounds are assigned to the two molecular orbitals containing the four electrons of predominant metal d character. The ionisation energies of these orbitals are ca. 0.5 eV lower in the iminoacyl than the azaallyl compounds. The energetic separation of these orbitals is traced to a destabilisation of the 2a′ M(CO)₂(η⁵-C₅R′₅) fragment orbital via a filled–filled interaction with the most stable π orbital of the iminoacyl and azaallyl ligands. The destabilisation is greater in the azaallyl compounds due to the additional involvement of an azaallyl fragment orbital which has been previously termed a ‘nitrogen lone pair’. Neither the photoelectron spectroscopic nor the theoretical data are consistent with this description.