Computational studies of spin-forbidden organometallic chemistry : ligand binding and C-H bond activation
Carreon-Macedo, Jose-Luis #
We present computational studies of organometallic reactions where reactants and products have different spin-states. In particular, we studied examples of ligand binding and C–H bonds activation. We aim to explain why some of these reactions are fast, why some are slow and why some do not occur. Along our work we use DFT methods as a standard for our calculations. This is done together with a technique for the localization of the minimum energy point of crossing between the different spin-state potential energy surfaces of reactants and products. The accuracy of the DFT calculations is tested against CCSD(T) calculations in several cases.
First, we study a simple spin-forbidden reaction: CpCo(CO) + L,
where L= CO, P(CH3)3 and C2H4. Our results indicate that these ligand additions are fast because the change of spin-state is a barrierless process. We then proceed to explain why two experimental examples of spin-forbidden reactions show opposite
reactivity. [N3N]WH + H2 → [N3N]W(H)3 , with N3N = [(Me3SiNCH2CH2)3N]3-, is slow because there is a spin-state change-induced energy barrier.
Tpi-Pr,MeCo(CO) + CO → Tpi-Pr,MeCo(CO)2, where Tpi-Pr,Me is a substituted trispyrazolylborate ligand, is fast because there is no energy barrier for the change of spin state. We then explain why CpCo(CO) does not activate C–H bonds. Its reaction with CH4 occurs without a significant barrier, but has an unfavourable free energy.
Exclusively using DFT calculations we explain trends in reactivity in the spin-forbidden C–H activation of methane using metallocenes of Mo and W. Our results show that heights in energy barriers for the change in spin state can be predicted with Hammond’s postulate. Our description of the potential energy surfaces for these reactions corresponds with observed kinetic isotope effects.
We obtain different corrections for the energetics of the reactions
Fe(CO)4 + L → Fe(CO)4•L, with L = Xe, CH4, H2, CO. With these results and for the case L = H2 we use a version of non-adiabatic transition state theory to calculate the spin-forbidden reaction rate coefficient. The calculated rate coefficient is in good agreement with the experimental one.
Our general conclusion is that the reactivity of spin-forbidden organometallic reactions can be explained reasonably well in qualitative and quantitative terms using DFT methods and provided that the accuracy of the calculations is carefully assessed.