2021 Annual Meeting

(265f) When does the choice of DFT functional matter in computational catalysis? The case of methane-to-methanol

Authors

Vennelakanti, V. - Presenter, Massachusetts Institute of Technology
Nandy, A., Massachusetts Institute of Technology
The conversion of methane-to-methanol which involves the activation of C-H bond in methane is desirable, since methanol is a more efficient fuel. The C-H bond has a high bond dissociation energy, which makes the identification of appropriate catalysts challenging. Several metalloenzymes with mid-row transition metals in their active sites demonstrated efficient activation of C-H bonds which motivated the design of bioinspired catalysts with mid-row 3d transition metals. These metals with unpaired d electrons give rise to numerous spin and oxidation states which impact the reactivity and reaction selectivity. Fortunately, recent advances in computing power enable high-throughput screening to discover new catalysts which is especially important given the combinatorial scale of catalyst design efforts. However, there is uncertainty associated with the results of the first-principles modeling of catalysts with density functional theory (DFT) in terms of the choice of exchange correlation functional. Here, we carry out a computational study of 1,207 mid-row 3d transition metal complexes (TMCs) that catalyze all the four reaction steps on the methane-to-methanol conversion energy landscape. These TMCs consist of Cr, Mn, Fr, and Co metals in different oxidation and spin states. We demonstrate that the energetics of certain reaction steps exhibit functional dependence while those of others are unaffected by the choice of functional. We show that the catalysts of certain metals may not be ideal to catalyze methane-to-methanol conversion, as predicted by different functionals. We also elaborate the importance of spin and oxidation states of different metals in determining the energetics of the reaction landscape, which cannot be generalized across all the four metals studied. We then assess the effect of functional choice on the different reaction step energetics and the predicted ground states to understand which catalysts are best suited for spin-allowed methane-to-methanol conversion.