Is [ReCl4(CN)2]2− a good Building Block for Single Molecule Magnets? A Theoretical Investigation.

Building blocks containing $5d$ spin centres are promising for constructing single molecule magnets due to their large spin-orbit interaction, but experimental and computational results obtained so far indicate that this might not be the case for Re$^\textrm{IV}$ centres in an octahedral environment. Density functional results obtained in this work for [ReCl$_4$(CN)$_2$]$^{2-}$ and trinuclear complexes formed by attaching Mn$^\textrm{II}$ centres to the cyano ligands indicate that zero field splitting in such complexes exhibits large rhombicity (which leads to fast relaxation of the magnetisation) even if there are only small distortions from an ideal geometry with a four-fold symmetry axis. This is already apparent if second-order spin-orbit perturbation theory is applied but even more pronounced if higher-order spin-orbit effects are included as well, as demonstrated by wavefunction based calculations. Computational results are cast into a ligand field model and these simulations show that especially a distortion which is not along the $C_4/C_2$ axeshas a large effect on the rhombicity. Quantum simulations on these complexes are difficult because the zero field splitting strongly depends on the energetic position of the low-lying doublets from the $t_{2g}^3$ configuration.