| Abstract |
In this paper, we present a theoretical study based on DFT methods using functionals implemented in Gaussian03, to obtain geometry optimizations, harmonic frequencies, IR intensities and Raman scattering activities of the triad [Ni(tdas)(2)](z) [z = 0 (0); z = 1-(1); z = 2-(2)] and of [Au(tdas)(2)](-)(3) with the aim of elucidating the nature of the bonding in these complexes and checking whether C-C stretching vibrations are suitable spectroscopic markers to assign the innocent-non- innocent character of the ligand. Geometry optimization at the density functional theory level was performed, and geometrical parameters were obtained in good agreement with the experimental structural data for the Ni complexes. Following the geometrical changes upon reduction, the Ni-S, N-S and C-S distances undergo a small increase, while the C-C length does not change significantly. Thus, unlike what is generally found for non-benzenoid dithiolenes, no shift of the nu(C-C) vibration is expected on reduction as observed for 1 and 2. The vibrational spectra are very accurately reproduced by the calculations. The less satisfactory agreement obtained for the gold complex is probably due to the deviation of the structurally characterized sample from the D(2h) symmetry obtained in the calculations, and/or overestimation of the Au-S distances, typical of the DFT functionals used. Calculations indicated that the electroactive orbital, HOMO in 2 and 3, SOMO in 1 and LUMO in 0, is a highly delocalized pi-orbital with a predominant S-ligand character, and that the total spin density in 1 is delocalized over the ligands and the nickel atom, in agreement with EPR data. From the obtained results we can assign a non-innocent character to the tdas ligand and show that the C-C stretching vibrations are not suitable spectroscopic markers to reflect this feature. |
