Combining Metal Sulfides and Organic Cocatalysts to Overcome the Challenges of Electrochemical CO2 Reduction

Abstract

Electrocatalytic reduction of CO2 is a potential method for converting CO2 to desirable products such as methanol and ethanol. Due to unfavorable scaling relations between the binding energies of catalytic intermediates on transition metals, such catalysts suffer from low current densities, high overpotentials, and low Faradaic efficiencies. Transition metal chalcogenides such as MoS2 have been shown to possess more favorable scaling relations for CO2 reduction – indeed, doped MoS2 in the presence of an electrolyte containing N-substituted imidazolium cations has been experimentally shown to have one of the highest reported current densities for CO2 reduction to CO. However, our density functional theory (DFT) calculations show that CO2 reduction pathways on this surface are hindered by high activation barriers for C-H bond formation. We hypothesize that a lower barrier pathway exists that involves the imidazolium cation functioning a cocatalyst by inducing polarity reversal (umpolung) on the carbon atom. As such, C-H bond formation can occur by proton transfer rather than by hydride transfer.We have used DFT to calculate barriers on such a cocatalyst, 1,3-dimethylimidazolium (DMIM), and found that CO2 can be activated by this molecule through electron-coupled addition to the carbon in the 2-position of the aromatic ring. The activated complex then undergoes several proton-coupled electron transfer steps and decomposes to eliminate CO. However, the activation barriers for this pathway are also too high for it to proceed at reasonable temperatures when the cocatalyst is not electronically coupled to the electrode. We further hypothesize that the DMIM cocatalyst chemisorbs on the MoS2 surface so that hybridization between the cocatalyst and surface orbitals lowers the electron transfer barriers enough for the proposed pathway to occur. Unfortunately, modeling these electron transfer reactions with DFT in the presence of the surface and electrochemical interface is challenging and we are currently extending the capabilities of an existing implicit solvation model (VASPsol) to be able to carry these out.