(10d) Understanding Solvation Effects on Hydrogenation Pathways for CO2 Reduction Using Multifaceted Approach – from Pyridine to Graphene.

Electrochemical CO₂ conversion into valuable hydrocarbons has garnered significant attention from numerous research groups, leading to the identification of a diverse spectrum of catalysts, ranging from organic molecules to metal oxides. Pourbaix diagrams are robust and useful tools for identifying relevant intermediates in CO₂ reduction; however, they only provide insights into the thermodynamics of different intermediate states and not the reaction barrier heights. In this study, we employ both the Potential of Mean Force (PMF) and mixed implicit/explicit solvation calculations to investigate different hydrogenation pathways of CO₂ and various pyridine-derived intermediates as catalysts. We demonstrate that using a mixed implicit/explicit modeling scheme with varying numbers of explicit solvent molecules, and then applying the Growing String Method (GSM) calculations to locate sequential transition states along a reaction pathway, enables us to calculate barrier heights with accuracy comparable to the more costly PMF calculations. Additionally, we use GSM calculations to assess how the catalyst molecule’s size affects the reaction mechanism and the energy barriers involved. We explore the conversion of CO₂ to formic acid using dihydropyridine, dihydroacridine, and nitrogen-doped nanoflake catalysts, and compare their reaction pathways, intermediate transition states, and the energetics of the final product states. Our results underscore the importance of incorporating explicit water molecules in computational models to accurately simulate the physical reaction pathway, as water molecules actively engage in the reaction, particularly by facilitating proton transfer. We identify nitrogen-doped nanoflakes as the most effective catalysts for reducing CO₂ to formic acid, with an activation barrier of approximately 10 kcal/mol. Finally, we propose using the Analytical Bond Energy from Bond Orders and Populations Model (BEBOP) as a surrogate to PMF or GSM calculations and demonstrate that the BEBOP model can predict barrier heights with accuracy on par with the more expensive GSM calculations for rapid investigation of hydrogenation processes.