DFT calculations of the structures of various isomers (cis/trans) of [(CH3)(PH3)M-NH2-CH2-C(O)-O] and [(Cl)(PH3)M-NH2-CH2-C(O)-O] complexes with M = Ni, Pd and Pt

[(CH3)(PH3)M-NH2-CH2-C(O)-O] and [(Cl)(PH3)M-NH2-CH2-C(O)-O], M = Ni, Pd, Pt (P and N in cis and trans position) were computed at the density functional BPW91 level of theory using a 6-31++G(2d,p) basis set for H, C, N, O, P, Cl and Ni, and quasi-relativistic core potentials for Pd (ECP28MWB) and Pt (ECP60MWB) employing a (8s7p6d)/[6s5p3d] valence basis set. The structures were fully optimised without any symmetry constraints and were found to represent true minima (number of imaginary frequencies, NIMAG = 0). In the series of methyl complexes for the Ni and the Pd complexes, the trans isomers were found to be slightly more stable than the cis isomers by 1.9 (Ni) and 0.6 (Pd) kcal mol(-1). In the case of the Pt complex, both isomers were computed to be essentially identical in their total energy with the cis isomer being only 0.25 kcal mol(-1) more stable than the trans isomer. However, in the chloro series, the trans isomers were found to be more favourable than the cis isomers in all cases by 9.1 (Ni), 7.5 (Pd) and 6.5 (Pt) kcal mol(-1). This finding is in perfect accord with the experimental data for a series of palladium and platinum complexes of the type Cl(R3P)M(N,O-chelate) for which the trans-P-M-N structure has been determined by X-ray diffraction. These complexes crystallize obviously as the thermodynamically stable trans isomers.