A comparison of N 2 cleavage in Schrock's Mo[N 3 N] and Laplaza-Cummins' Mo[N(R)Ar] 3 systems

Abstract

The four-coordinate Mo-[N3N] complex, [N3N] = [{RNCH2CH2}3N], R = 3,5-(2,4,6-iPr 3C6H2)2C6H3 (HIPT), which is capable of converting N2 to ammonia catalytically, reacts with N2 in a similar manner to Mo[N(R)Ar]3 (R = tBu, Ar = 3,5-C6H3Me2) to form a dinitrogen-bridged dimer intermediate, but unlike its three-coordinate counterpart, N2 cleavage is not observed. To rationalise these differences, the reaction of N2 with the model Mo[NH 2]3[NH3] and full ligand Mo[N3N] systems was explored using density functional theory and compared with the results of an earlier study involving the model three-coordinate Mo[NH 2]3 system. Although the overall reaction is exothermic, the final N-N cleavage step is calculated to be endothermic by 75 kJ mol -1 for the model system when the Mo-amine cap bond length is fixed to mimic the constraints of the ligand straps, but exothermic by 14 kJ mol -1 for the full ligand system. In the latter case, the slightly exothermic cleavage step can be attributed to the destabilization of the N 2 bridged dimer relative to the nitride product owing to the steric effects of the bulky R groups. The activation barrier for N-N cleavage is estimated at 151 kJ mol-1 for the model system, more than twice the calculated value for Mo[NH2]3, and even greater, 213 kJ mol-1, for the full ligand [N3N]Mo system. A bonding analysis shows that although the binding of the amine cap helps to stabilize the intermediate dimer, at the same time it destabilizes the metal d-orbitals involved in backbonding to the π* orbitais on N2. As a result, backdonation is less efficient and N-N activation reduced compared to the three-coordinate system. Thus, the increased stability of the intermediate dimer on binding of the amine cap combined with the reduced level of N-N activation and higher kinetic barrier, explain why N-N cleavage has not been observed experimentally for the four-coordinate Mo[N3N] system.