(144b) NO Oxidation and Selective Catalytic Reduction of NOx with NH3 On Cu/Zeolites

Further improvements in catalytic removal of nitrogen oxides (NO_x) continue to be driven by ever more stringent regulations. Copper deposited on ZSM-5, BEA and chabazite (CHA) zeolite based catalysts (Cu/SSZ-13 and Cu/SAPO-34) have adequate performance for selective catalytic reduction (SCR) of NO_x with NH₃. The chabazite structure, however, has a much higher hydrothermal stability and for this reason this structure was adopted for commercial use. We are trying to understand the nature and location of the Cu active site, the role of NH₃ and how it is activated on the zeolite, and the reaction mechanism. 

NO Oxidation

The SCR rate on Cu/ZSM-5 and SSZ-13 catalysts was found to be nearly two orders of magnitude greater than the NO oxidation rate (in the absence of ammonia), while catalysts which were shown to have undetectable rates for NO oxidation (in the absence of ammonia) maintained the expected rates per mole of Cu for SCR at the same temperature and partial pressure of the reactants. The NO oxidation reaction itself was shown to be strongly dependent on dimer sites in Cu/ZSM-5, which are formed efficiently only within a certain window of Cu/Al ratios. Isolated Cu sites and extraframework Cu on the other hand showed undetectable rates for NO oxidation. On SSZ-13, there is no indication that dimer sites are necessary for SCR activity and indeed the catalyst was active at varying Cu loadings, including low % Cu exchange. Although the rate of reaction is higher in the presence of NO₂, it appears that the SCR mechanism does not rely on NO oxidation as a mechanistic step.

Cu Active Site

Through a combined modeling and operando XAS study of Cu/SSZ-13, we found that isolated Cu atoms tended to be located within the 6-ring of SSZ-13, but that the addition of various adsorbates on that Cu atom changed the Cu oxidation state. Complementary XAS under the standard SCR condition (300 ppm NO, 300 ppm NH₃, 5% O₂, 5% H₂O and 5% CO₂) found that Cu remained in a mixed Cu(I)-Cu(II) oxidation state. Exposing the catalyst only to the oxidizing or reducing half-reaction of SCR, e.g. removing O₂, forced the catalyst into a reduced state with 75% of the total Cu in the Cu(I) oxidation state. Similarly, removing a reductant from the reaction mixture, forced the catalyst into the fully oxidized state. Additionally we saw evidence of the Cu mobility with the introduction of H₂O, a possible sign that while the model would suggest the 6-ring as the preferred active site, the Cu site could be more heterogeneous under reaction conditions. Both modeling of the active site in the presence of a more extensive set of adsorbates, as well as further experimental characterization of the nature of the Cu active site will be necessary.

NH₃ Storage and Activation

In studying the kinetics of the Cu/SSZ-13 catalyst, we found that the catalytic rate depends on NH₃ concentration until a particular gas phase concentration is reached. Thus, depending on the NH₃ concentration, the reaction will either be 0 order in NH₃ or >0 in NH₃. By combining various adsorption, reaction and desorption experiments into one sequential test, we were able to show that a large proportion of NH₃ adsorbed to the catalyst surface appears to be inactive and that this fraction does not change with NH₃ concentration.  These inactive NH₃ sites adsorb NH₃ gas before it can react, to the detriment of the initial SCR rate. Furthermore, we were able to measure the fraction of the adsorbed NH₃ that was active for the reaction. The amount of these ‘reactive’ NH₃ sites increased with increased NH₃ concentration until the NH₃ order goes to zero. The results suggest that there are multiple NH₃ storage sites, one of which is capable of activating NH₃ for SCR while the other binds NH₃ and makes it inactive. Understanding how to limit the number of inactive NH₃ sites is critical in developing a more capable SCR catalyst, especially in terms of cold start applications.