Theoretical study of transition metal compounds with molybdenum- and tungsten-phosphorus triple bonds

Quantum mechanical calculations at the HF, MP2, DFT (B3LYP), and CCSD(T) levels of theory using quasirelativistic effective core potentials for the metal and valence basis sets of DZP quality are reported for the transition metal complexes [M(P)(NH2)(3)] (1, 2), [M(PS)(NH2)(3)] (3, 4), [M(P)(NH2)(3)(NH3>)] (5, 6), [M(P)(N3N)] (7, 8; N3N = [(HNCH2CH2)(3)N](3-)), and [(M(PS)(NH2)(3)(NH3)] (9, 10) with M = Mo, W. The B3LYP-optimized geometries of 1-10 are in good agreement with experiment. Bond dissociation energies for the LnMP-S bonds calculated at B3LYP an 8-10 kcal/mol higher than the CCSD(T) values. The LnMP-S and M-NH3 bonds of 9 and 10 are predicted to be stronger than the respective bonds of 3-6. P-31 NMR chemical shifts and the anisotropic components have been calculated using the IGLO and GIAO approaches. The results are in accord with experimental data. The bonding situation of the complexes has been analyzed with the help of the NBO partitioning scheme. The phosphido complexes LnMP have metal drop P triple bonds, while the phosphorus-sulfide complexes have LnM=P=S double bonds. This formally reduces the number of coordination sites at the metal, which explains the significantly shorter and stronger bond with an amine trans to the M=P=S moiety.