Atomistic Mechanisms for Catalytic Transformations of NO to NH3, N2O,and N2 by Pd

The industrial pollutant NO is a potential threat to the environment and to human health. Thus, selective catalytic reduction of NO into harmless Nbegin{document}$_2$end{document}, NHbegin{document}$_3$end{document}, and/or Nbegin{document}$_2$end{document}O gas is of great interest. Among many catalysts, metal Pd has been demonstrated to be most efficient for selectivity of reducing NO to Nbegin{document}$_2$end{document}. However, the reduction mechanism of NO on Pd, especially the route of Nbegin{document}$-$end{document}N bond formation, remains unclear, impeding the development of new, improved catalysts. We report here the elementary reaction steps in the reaction pathway of reducing NO to NHbegin{document}$_3$end{document}, Nbegin{document}$_2$end{document}O, and Nbegin{document}$_2$end{document}, based on density functional theory (DFT)-based quantum mechanics calculations. We show that the formation of Nbegin{document}$_2$end{document}O proceeds through an Eley-Rideal (E-R) reaction pathway that couples one adsorbed NObegin{document}$^*$end{document} with one non-adsorbed NO from the solvent or gas phase. This reaction requires high NObegin{document}$^*$end{document} surface coverage, leading first to the formation of the trans-(NO)begin{document}$_2$end{document}begin{document}$^*$end{document} intermediate with a low Nbegin{document}$-$end{document}N coupling barrier (0.58 eV). Notably, trans-(NO)begin{document}$_2$end{document}begin{document}$^*$end{document} will continue to react with NO in the solvent to form Nbegin{document}$_2$end{document}O, that has not been reported. With the consumption of NO and the formation of Nbegin{document}$_2$end{document}Obegin{document}$^*$end{document} in the solvent, the Langmuir-Hinshelwood (L-H) mechanism will dominate at this time, and Nbegin{document}$_2$end{document}Obegin{document}$^*$end{document} will be reduced by hydrogenation at a low chemical barrier (0.42 eV) to form Nbegin{document}$_2$end{document}. In contrast, NHbegin{document}$_3$end{document} is completely formed by the L-H reaction, which has a higher chemical barrier (0.87 eV). Our predicted E-R reaction has not previously been reported, but it explains some existing experimental observations. In addition, we examine how catalyst activity might be improved by doping a single metal atom (M) at the NObegin{document}$^*$end{document} adsorption site to form M/Pd and show its influence on the barrier for forming the Nbegin{document}$-$end{document}N bond to provide control over the product distribution.