A Density Functional Study of Metal-Ligand Bonding in [(PR(3))(2)M](+) and [PR(3)MCl] (M = Ag, Au; R = H, Me)...

Relativistic and nonrelativistic electronic structure calculations were carried out on [(PR(3))(2)M](+)and [PR(3)MCl] (M = Ag, Au; R = H, Me) complexes using the all-electron linear combination of Gaussian-type orbitals density functional (LCGTO-DF) method. The calculated relativistic metal-ligand bond lengths show good agreement with experimental values. The relativistic contraction of the M-P bonds in [(PR(3))(2)M](+) is about 8 and 22 pm for M = Ag and Au, respectively, resulting in Au-P bonds that are about 10 pm shorter than the Ag-P bonds in these species, in good agreement with recent crystallographic results for the cations [(PMes(3))(2)M](+) (M = Ag, Au). The relativistic contraction of the Au-Cl bond in [PR(3)AuCl] is significantly less than that of the Au-P bond, and this explains the experimentally observed differential Au-X and Au-P bond length contractions from [PR(3)AgX] to [PR(3)AuX]. The calculated relativistic bond lengths for corresponding PH(3) and PMe(3) complexes are very similar, confirming a previous conclusion that PH(3) is a good model for structural properties of larger tertiary phosphine ligands. However, the bond lengths for the PMe(3) complexes are all slightly longer than those for the corresponding PH(3) complexes, whereas the M-P dissociation energies are 20-40% higher for the PMe(3) complexes. These findings provide computational support for the concept of "longer but stronger bonds", which was recently proposed on the basis of experimental studies of transition metal complexes involving various substituted phosphine ligands.