| Photoelectrochemical CO2 reduction is a promising technology for the sustainable production of liquid fuels but presents some fundamental challenges. Overcoming these challenges requires the development of efficient catalysts, which could be accelerated by the discovery of the chemical mechanism by which existing successful catalysts operate. This dissertation uses quantum-mechanics-based simulations to investigate the CO2 reduction mechanism in a photoelectrochemical system that uses a p-GaP photoelectrode and a pyridine (Py)-based co-catalyst in an acidified aqueous solution. In particular, we develop accurate models of the electrode solution/interface and use density functional theory methods to compute relevant properties of species at this interface. Based on these properties, we gain mechanistic insights, assess the validity of previously proposed mechanisms, and hypothesize and test new mechanisms.;A long-standing mechanistic hypothesis is that the homogeneous reduction of pyridinium (PyH+) to the 1-pyridinyl radical (1-PyH •) is an essential step for formation of the active catalyst. Herein we show that 1-PyH• is unstable and spontaneously transfers its electron to the electrode surface. In addition, we find at most a very small thermodynamic driving force for this reduction step; PyH + is reduced instead more favorably to intermediates contributing to possible formation of adsorbed dihydropyridine (DHP), a co-catalyst proposed in an alternative mechanism. Moreover, we provide strong evidence based on our calculations and experimental observations that the mechanism cannot be fully homogeneous and must involve surface-bound intermediates.;Adsorbed DHP was proposed to form via surface hydride and aqueous proton transfer to adsorbed Py and to subsequently reduce CO2 by transferring hydride to it. In a combined experimental-theoretical characterization of the electrode/solution interface, we find that water dissociation is thermodynamically favorable at this interface, producing stable adsorbed protons that could reduce to surface hydrides under operating conditions. In another combined investigation, we find that adsorbed Py reactivity supports the proposed mechanism for adsorbed DHP formation. Additionally, our calculations show that the proposed hydride transfers, unlike adsorbed proton transfers, are thermodynamically favored. However, we find that adsorbed DHP formation is likely kinetically hindered. We therefore propose and investigate alternative co-catalytic intermediates that might form and react with CO2 more favorably. |
