The molecular and electronic structure of the planar nickel dithiolene (1c, R = H) and of related complexes derived from nickel dithiolene by replacement of Ni by Pd (palladium dithiolene, 2c, R = H) or by Pt (platinum dithiolene, 3c, R = H), or by replacement of S by NH (nickel diiminolene, 1a, R = H), O (nickel dioxylene, 1b, R = H) or Se (nickel diselenolene, 1d, R = H), were studied by density functional theory using the B3LYP functional and the valence triple‐zeta basis set 6‐311+G* for all atoms except Pd and Pt. For the latter atoms the quasirelativistic effective core potentials of the Stuttgart group were employed. The molecular structure of nickel dithiolene (1c, R = H) is satisfactorily reproduced by DFT calculations. The geometry of the corresponding platinum complexes 3a–3d is more sensitive to relativistic effects, resulting in the contraction of the X–Pt bonds. As shown with the metal dithiolenes, the two ligands are structurally related to mononegative ions of open shell structure. The C–C bond lengths of the complexes are close to those of aromatic and chain‐type polymethine structures (about 1.4 Å). The nickel dithiolene (1c, R = H) and related complexes have D2h symmetry and are 14 π‐electron systems with 10 π‐electrons at the ligands and 4 π‐electrons at the metal center. The natural population analysis has confirmed that metal M++ does accept electrons from the ligands but to a lesser extent than expected. The empty d‐orbitals of M++ are only partly occupied in the molecular ground state. The positive charge of the metal decreases in the order Ni > Pd > Pt. The 1H chemical shifts and the nucleus‐independent chemical shifts (NICSs) of the ring moieties calculated by GIAO‐DFT display a pronounced electron delocalization. In agreement with the calculated C–C bond lengths the 1H chemical shifts and the NICS values show a marked bond delocalization. The NICS values show a change of the aromatic delocalization in the order Ni > Pd < Pt and NH > O < S < Se. The wave numbers of the IR spectra of the complexes calculated by DFT are grouped in separate frequency regions. The very intense absorption of 1c (R = H) in the visible region of the spectrum is surprisingly well reproduced by ab initio single‐only configuration interaction calculations. While the color band of the palladium complex is predicted to be red‐shifted relative to the nickel complex, a blue shift is calculated on passing from the palladium to the platinum complex. The blue shift is, in part, due to the relativistic contraction of bond lengths in the Pt complexes.

A two-component relativistic density functional method based on the Douglas-Kroll-Hess transformation has been applied to the actinyls and hexafluorides of U and Np. All-electron scalar relativistic calculations as well as calculations including spin-orbit interaction have been compared to results obtained with a pseudopotential approach. In addition, several exchange-correlation potentials have been applied to examine their performance for the bond lengths and vibrational frequencies of the title compounds. The calculations confirm the well-known accuracy of the LDA approach for geometries and frequencies, which is corroborated for the hexafluorides where gas phase experimental data are available. Comparison with results of accurate wave function based methods provides further confirmation of this finding. Gradient-corrected functionals tend to overestimate bond lengths and underestimate frequencies also for actinide compounds. The results obtained with Stoll-Preuss (small core) effective core potentials agree very well with those of all-electron calculations, while calculations with Hay-Martin large core pseudopotentials are somewhat less accurate. For all molecules and properties considered, spin-orbit effects have been found negligible concomitant with the closed-shell electronic structure of the U(VI) compounds and the open-shell situation of the Np(VI) compounds with a single valence f electron.

No abstract

No abstract

No abstract