Oximato-Based Ligands in 3d/4f-Metal Cluster Chemistry: A Family of {Cu3Ln} Complexes with a “Propeller”-like Topology and Single-Molecule Magnetic Behavior

The organic chelating and bridging ligands 9,10-phenanthrenedione-9-oxime (phenoxH) and 9,10-phenanthrenedione-9,10-dioxime (phendoxH2) were synthesized and subsequently employed for the first time in heterometallic 3d/4f-metal cluster chemistry. The general reaction between CuCl2·2H2O, LnCl3·6H2O, phenoxH, and NEt3 in a 1:2:2:4 molar ratio, in a solvent mixture comprising MeCN and MeOH, afforded brown crystals of a new family of [Cu3LnCl3(phenox)6(MeOH)3] clusters (Ln = Gd (1), Tb (2), Dy (3)) that possess an unprecedented [Cu3Ln(μ-NO)6]3+ "propeller"-like core. Complexes 1–3 are the first {Cu3Ln} clusters in which the outer CuII and the central LnIII atoms are solely bridged by diatomic oximato bridges. The {Cu-N-O-Ln} bridging units are very distorted with torsion angles spanning the range 35.5–48.9° and 25.2–55.6° in 1 and 2, respectively. As a result, complexes 1–3 are antiferromagnetically coupled, in agreement with previously reported magnetostructural criteria for oximato-bridged Cu/Ln complexes. The magnetic susceptibility data for all complexes were nicely fit to an isotropic spin Hamiltonian (for 1) or a Hamiltonian that accounts for the spin of the CuII atoms, the spin component of the LnIII, the spin–orbit coupling (λ), an axial ligand-field component around the LnIII atoms (Δ), and the Zeeman effect (for the anisotropic 2 and 3). The resulting fit parameters were J = −1.34 cm–1 and g = 2.10 (1), J = −1.42 cm–1, gCu = 2.10, and Δ = −26.3 cm–1 (2), and J = −1.70 cm–1, gCu = 2.05, and Δ = −38.1 cm–1 (3). The reported fitting procedure, implemented in the PHI program, is here used for the first time. Even if this method is only valid in high-symmetry Ln environments, when it is properly used allows a very simple and efficient method to obtain the exchange parameters. In light of the negative anisotropy, compounds 2 and 3 were found to exhibit frequency-dependent tails of out-of-phase signals in the presence of a small external dc field, characteristic of the slow magnetization relaxation of a single-molecule magnet. By using the Kramers–Kronig equations, the effective energy barriers (Ueff) were derived and reported as Ueff = 10.1 and 5.4 cm–1 for 2 and 3, respectively.