Comprehensive Thermodynamics of Nickel Hydride Bis(Diphosphine) Complexes: A Predictive Model through Computations

Prediction of thermodynamic quantities such as redox potentials and homolytic and heterolytic metal hydrogen bond energies is critical to the a priori design of molecular catalysts. In this paper we expound upon a density functional theory (DFT)-based isodesmic methodology for the accurate computation of the above quantities across a series of Ni(diphosphine)(2) complexes that are potential catalysts for production of H(2) from protons and electrons or oxidation of H(2) to electrons and protons. Isodesmic schemes give relative free energies between the complex of interest and a reference system. A natural choice is to use as a reference a compound that is similar to the chemical species under study and for which the properties of interest have been measured with accuracy. However, this is not always possible, as in the case of the Ni complexes considered here, where data are experimentally available for only some species. To overcome this difficulty, we employed a theoretical reference compound, Ni(PH(3))(4), which is amenable to highly accurate electron-correlated calculations, which allows one to explore thermodynamic properties even when no experimental input is accessible. The reliability of this reference is validated against the available thermodynamics data in acetonitrile solution. Overall the proposed protocol yields excellent accuracy for redox potentials (similar to 0.10 eV of accuracy), for acidities (similar to 1.5 pK(a) units of accuracy), for hydricities (similar to 2 kcal/mol of accuracy), and for homolytic bond dissociation free energies (similar to 1-2 kcal/mol of accuracy). The calculated thermodynamic properties are then analyzed for a broad set of Ni complexes. The power of the approach is demonstrated through the validation of previously reported linear correlations among properties. New correlations are revealed. It emerges that only two quantities, the Ni(II)/Ni(I) and Ni(I)/Ni(0) redox potentials (which are easily accessible experimentally), suffice to predict with high confidence the energetics of all relevant species involved in the catalytic cycles for H(2) oxidation and production. The approach could be extended to other transition metal complexes.