Dinuclear metallacycles with single M-O(H)-M bridges [M = Fe(II), Co(II), Ni(II), Cu(II)]: effects of large bridging angles on structure and antiferromagnetic superexchange interactions.

The reactions of M(ClO4)2·xH2O and the ditopic ligands m-bis[bis(1-pyrazolyl)methyl]benzene (Lm) or m-bis[bis(3,5-dimethyl-1-pyrazolyl)methyl]benzene (Lm*) in the presence of triethylamine lead to the formation of monohydroxide-bridged, dinuclear metallacycles of the formula [M2(μ-OH)(μ-Lm)2](ClO4)3 (M = Fe(II), Co(II), Cu(II)) or [M2(μ-OH)(μ-Lm*)2](ClO4)3 (M = Co(II), Ni(II), Cu(II)). With the exception of the complexes where the ligand is Lm and the metal is copper(II), all of these complexes have distorted trigonal bipyramidal geometry around the metal centers and unusual linear (Lm*) or nearly linear (Lm) M-O-M angles. For the two solvates of [Cu2(μ-OH)(μ-Lm)2](ClO4)3, the Cu-O-Cu angles are significantly bent and the geometry about the metal is distorted square pyramidal. All of the copper(II) complexes have structural distortions expected for the pseudo-Jahn-Teller effect. The two cobalt(II) complexes show moderate antiferromagnetic coupling, -J = 48-56 cm(-1), whereas the copper(II) complexes show very strong antiferromagnetic coupling, -J = 555-808 cm(-1). The largest coupling is observed for [Cu2(μ-OH)(μ-Lm*)2](ClO4)3, the complex with a Cu-O-Cu angle of 180°, such that the exchange interaction is transmitted through the dz(2) and the oxygen s and px orbitals. The interaction decreases, but it is still significant, as the Cu-O-Cu angle decreases and the character of the metal orbital becomes increasingly d(x(2)-y(2)). These intermediate geometries and magnetic interactions lead to spin Hamiltonian parameters for the copper(II) complexes in the EPR spectra that have large E/D ratios and one g matrix component very close to 2. Density functional theory calculations were performed using the hybrid B3LYP functional in association with the TZVPP basis set, resulting in reasonable agreement with the experiments.