Predictive models for metal-metal bond dissociation free energies between aluminum(III) and a series of transition metal carbonyls

It was previously shown that the complex, LAl(Me)Fp [where L is a beta-diketiminate and Fp = FeCp(CO)2] undergoes Al-Fe homolysis at ambient conditions, revealing a pair of metalloradical intermediates capable of cooperative small molecule activation. Towards expanding this platform to other Al-M derivatives, here we conducted an extensive computational screening of potential LAl(Me)M catalysts (where M is a series of common metal carbonyls) using DFT calculations combined with multivariate linear regression (MLR) analysis. Predictive models were established that anticipate the influences of different metals on Al-M bond dissociation free energies (BDFEs), utilizing electronic and steric parameters obtained from existing literature and from buried volume calculations (%Vbur), including a convenient model that can be applied without aid of further DFT calculations. Our investigations unveiled a partial correlation between Al-M BDFEs and M-H BDFEs, in that both depend on pKa and E degrees values of M. However, Al-M BDFEs demonstrate significant influences from both steric crowding and Al-M orbital overlap, neither of which contribute to M-H BDFE. By employing MLR to derive predictive models, extrapolation to other hypothetical complexes beyond the training set presented in this study is facilitated. Thus, our research will inform and enable the future design of sustainable catalytic processes involving heterobimetallic Al-M complexes with earth-abundant metals.