Self-consistency versus "best-fit" approaches in understanding the structure of metal nitrosyl complexes

The metal nitrosyl complexes RuX(CO)(NO)L2+ (L = P(tBu)2Me, X = F-, Cl-, BF 4--, H2O, NCH, H-, no ligand, and CO) have been characterized by computing their structures, relative stabilities, and vibrational frequencies through Becke3LYP calculations on a RuX(CO)(NO)(PH3)2+ model complex. In the case of X = F-, Cl-, BF4-, and H 2O, a square pyramidal (SP) geometry with a bent nitrosyl ligand is preferred. In the case of X = NCH, H-, and CO two geometries exist as local minima: a trigonal bipyramid (TBP) with a linear nitrosyl ligand and a square pyramid with a bent nitrosyl ligand. The computed relative stabilities of such complexes cannot clearly identify the ruthenium coordination geometry. Nevertheless, the correlation between the experimental and theoretical ν NO stretching frequencies is conclusive in identifying Ru(H)(CO)(NO)L2 and Ru(CO)2(NO)L2+ as TBP structures and Ru(NCMe)- (CO)(NO)L2+ as SP. Three sets of additional calculations were also carried out on a selected system (X = H-). The computational level was increased to CCSD(T), the solvent effect was introduced with a PCM approach, and the real phosphine ligands were considered with a QM/MM ONIOM method.