Ligand Rotation in [Ar(R)N]\u3csub\u3e3\u3c/sub\u3eM-N\u3csub\u3e2\u3c/sub\u3e-M′[N(R)Ar]\u3csub\u3e3\u3c/sub\u3e (M, M′ = Mo\u3csup\u3eIII\u3c/sup\u3e, Nb\u3csup\u3eIII\u3c/sup\u3e; R = \u3csup\u3ei\u3c/sup\u3ePr and \u3csup\u3et\u3c/sup\u3eBu) Dimers

Earlier calculations on the model N2-bridged dimer (µ-N2)-{Mo[NH2]3}2 revealed that ligand rotation away from a trigonal arrangement around the metal centres was energetically favourable resulting in a reversal of the singlet and triplet energies such that the singlet state was stabilized 13 kJ mol−1 below the D3d triplet structure. These calculations, however, ignored the steric bulk of the amide ligands N(R)Ar (R = iPr and tBu, Ar = 3,5-C6H3Me2) which may prevent or limit the extent of ligand rotation. In order to investigate the consequences of steric crowding, density functional calculations using QM/MM techniques have been performed on the MoIIIMoIII and MoIIINbIII intermediate dimer complexes (µ-N2)-{Mo[N(R)Ar]3}2 and [Ar(R)N]3Mo-(µ-N2)-Nb[N(R)Ar]3 formed when threecoordinate Mo[N(R)Ar]3 and Nb[N(R)Ar]3 react with dinitrogen. The calculations indicate that ligand rotation away from a trigonal arrangement is energetically favourable for all of the ligands investigated and that the distortion is largely electronic in origin. However, the steric constraints of the bulky amide groups do play a role in determining the final orientation of the ligands, in particular, whether the ligands are rotated at one or both metal centres of the dimer. Analogous to the model system, QM/MM calculations predict a singlet ground state for the (µ-N2)-{Mo[N(R)Ar]3}2 dimers, a result which is seemingly at odds with the experimental triplet ground state found for the related (µ-N2)-{Mo[N(tBu)Ph]3}2 system. However, QM/MM calculations on the (µ-N2)-{Mo[N(tBu)Ph]3}2 dimer reveal that the singlet–triplet gap is nearly 20 kJ mol−1 smaller and therefore this complex is expected to exhibit very different magnetic behaviour to the (µ-N2)-{Mo[N(R)Ar]3}2 system