Hydrogen internal combustion engines necessitate the implementation of selective catalytic reduction of nitrogen oxides under O2-rich environments in exhaust. This is challenging as noble metals favor combustion, selectively burning H2 instead of reducing NO to N2. Yet, high NO conversion and selectivity to N2 can be achieved on Pt/Al2O3 catalysts at low temperatures. To rationalize this paradox, we construct a microkinetic model on Pt(111) using electronic-structure calculations and find that the model fails to capture experiments. We conjecture that high-performing reduction catalysts consist of small metal clusters. Density functional theory calculations on Pt dimers supported on γ-Al2O3(110) (Pt2/Al2O3) reveal lower activation energies for NO dissociation and N–N recombination than O2 dissociation and N–NO recombination, respectively. Electron density difference isosurfaces suggest that lower-lying and partially occupied π* orbitals of NO require significantly less electron back-donation from Pt2 for bond activation than those of O2, rationalizing the lower NO dissociation barrier compared to O2. The higher barrier of N–NO recombination is likely due to the NO binding in a stable Oδ−–Nδ+ state. State-based microkinetic model accurately describes experimental observations and reveals a spatially varying Pt oxidized state, where NO reduction occurs on metallic states. Interestingly, the oxidation–reduction selectivity reverses at the extremes of metal size.
Patel et al. (Mon,) studied this question.