Lantern-type dinickel complexes: An exploration of possibilities for nickel-nickel bonding with bridging bidentate ligands.

Abstract

Many binuclear nickel complexes have NiNi distances suggesting NiNi covalent bonds, including lantern-type complexes with bridging bidentate ligands. This DFT study treats tetragonal, trigonal, and digonal lantern-type complexes with the formamidinate, guanidinate, and formate ligands, besides some others. Formal bond orders (ranging from zero to two) are assigned to all the NiNi bonds on the basis of MO occupancy considerations. A VB-based electron counting approach assigns plausible resonance structures to the dinickel cores. Model tetragonal complexes with the dimethylformamidinate and the dithioformate ligands have singlet ground states whose non-covalently bonded NiNi distances are close to those in their experimentally known counterparts. Trigonal dinickel complexes are unknown, but are predicted to have quartet ground states with NiNi bonds of order 0.5. The model digonal complexes are predicted to have triplet ground states, but the predicted NiNi bond lengths are longer than those found in their experimentally known counterparts. This could owe to inadequate treatment of electron correlation by DFT in these short NiNi bonds with their multiconfigurational character. All the NiNi bond distances here are categorized into ranges according to the NiNi bond orders of 0, 0.5, 1, 1.5, and 2, no NiNi bonds of order higher than two being identified. The NiNi bonds of given order in these lantern-type complexes are consistently shorter than the corresponding NiNi bonds in dinickel complexes having carbonyl ligands, attributable to the metalmetal bond lengthening effect of CO ligands.