Computational Analysis of Photocatalytic Nitrogen Fixation on Oxide Surfaces

The fixation of nitrogen is of great importance for producing chemicals and fertilizers, and is performed primarily through the Haber-Bosch process. However, the Haber-Bosch process is highly centralized and has a significant carbon footprint. Photocatalysis is one alternative method of generating fixed nitrogen for fertilizers in a distributed manner with lower carbon impact. In this approach photocatalytic reactors could be deployed to produce dilute fertilizer at the farm scale. Based on a preliminary technoeconomic analysis, it is estimated that performance targets for photocatalytic systems are much lower than traditional Haber-Bosch, requiring on the order of 0.1% solar to chemical conversion efficiency under the correct conditions. Rutile TiO2 was the first material observed to fix nitrogen photocatalytically, making it a good candidate for theoretical study. While experimental observations of this reaction on TiO2 have existed for several decades, the mechanism of the surface reaction was not previously known. Analysis of the nitrogen reduction surface reaction using density functional theory (DFT) on rutile (110) shows the reaction is thermodynamically improbable on pristine, oxygen vacant, and iron doped surfaces at room temperature. While oxidation appears thermodynamically feasible, it is hypothesized that kinetic barriers will make it challenging in practice. An alternative strategy of doping the surface with transition metals is also investigated, revealing that doping with Rh and Mo may yield an increase in reaction rates for electrocatalytic reactions. However, the thermodynamic barriers are still large under photocatalytic conditions. Investigation of potential carbon active-sites shows that there exist several carbon containing active-sites on rutile (110) that react strongly with molecular nitrogen. Analysis of the thermodynamics of the surface reaction on a carbon substitution site finds that it is feasible that this carbon active-site can catalyze the nitrogen reduction reaction on rutile (110). This active-site is metastable under reaction conditions and provides a plausible explanation of the experimental observations. Ambient pressure photoelectron spectroscopy (AP-XPS) experiments were performed by the Hatzell group on rutile (110) single crystals under illumination to directly probe the surface under reaction conditions. These experiments show reduced nitrogen is present on the surface only when adventitious surface carbon is also present, supporting the hypotheses that carbon based active site catalyze nitrogen reduction. These findings provide an atomic-scale explanation of photocatalytic nitrogen fixation, and lay the foundation for the discovery of new and more efficient catalyst materials.Ph.D