Kinetic and spectroscopic study of catalysts for water-gas shift and nitrogen oxide removal

Nitrogen oxides (NOx) are formed in high temperature combustion processes such as in power generation and motor vehicles. Increasingly stringent regulation of these harmful emissions continues to drive interest in developing, understanding and studying new catalytic formulations for exhaust aftertreatment. For mobile sources, predominantly heavy duty diesel engines, selective catalytic reduction (SCR) with NH3 has become the principal means of NO x abatement. An alternative technology developed, but now surpassed by SCR, is NOx Storage Reduction (NSR) catalysis. Both technologies have been studied in our laboratory and are the basis for this dissertation. We studied seven different lean NOx trap (LNT) monolith formulations for NSR ranging from 0.6 to 6.2 wt.% Pt and 4 to 20 wt.% Ba loadings on γ-Al 2O3. The noble metal component of a LNT oxidizes NO to NO 2 aiding in the storage of NO2 on the alkaline earth component. Before the storage component saturates, a reductant such as H2 is introduced into the vehicular exhaust and the stored NOx is released and reduced to N2. Once the storage component is free of NOx, reductant flow is ceased and storage is begun anew. Our research focused on understanding the effects that CO2 and H2O have on the storage capacity of the LNT over short as well as extended periods of time. We found that for high Ba loadings, CO 2 had a consistently detrimental effect on the fast NOx storage capacity (NSC), defined as the amount of NOx the catalyst can store before 1% of the inlet NOx is measured in the reactor outlet. Over long NOx storage periods, CO2 continued to inhibit storage compared to the same catalyst in CO2 free conditions. On low loadings of Ba, however, the inhibition of CO2 was significantly reduced. We found that the loading dependent characteristics of the Ba phase affected the way in which CO2 adsorbed on the storage component, which greatly affected the stability of the species on lower Ba loadings. The less stable adsorbed CO2 proved less competitive for NOx sorption sites, explaining the weak reduction of NSC by CO2 on low Ba loadings. Contrary to CO2, H2O exhibited both beneficial and inhibitory effects on the NSC. Over long periods of time, and at high Ba loadings, the addition of H2O in the feed increased the NSC, attributed to enhanced O2 spillover on the hydroxylated Ba surface allowing greater access to available NOx storage sites. When the Ba loading was reduced, the interaction sphere of Pt particles with the Ba storage component required for O2 to spillover to assist in NOx storage was reduced. Thus, despite the enhanced spillover capacity of oxygen on the hydroxylated storage component other NSC decreasing effects of H2O addition, such as Ba agglomeration, became more dominant and reduced the NSC. Recent developments in selective catalytic reduction have shown Cu and Fe/chabazite (CHA) based zeolites to be particularly well suited to sustaining high catalytic rates without degradation in the harsh environment of diesel engine exhaust. Little has been published about these catalysts as the academic community has just recently learned about the materials and their commercial implementation. Using operando X-ray absorption spectroscopy, combined with first-principles thermodynamics simulations and kinetic analysis, we have studied the nature of the Cu active site on Cu/SSZ-13, Cu/SAPO-34 and Cu/ZSM-5. Examining the catalysts under operando standard SCR conditions (300 ppm NO, 300 ppm NH3, 5% O2, 5% H2O and 5% CO2) showed the catalyst to be in a mixed Cu(I)-Cu(II) oxidation state. Neither the amount of Cu(I) nor Cu(II) individually correlated with the different rates measured on the various zeolite catalysts, and so we proposed that the SCR reaction progresses via a redox mechanism requiring both Cu(I) and Cu(II). First principles thermodynamic calculations found that the redox couple of Cu(I)H2O and Cu(II)(OH)2 were the most thermodynamically stable species of any of the OxHy variants modeled on Cu. The redox nature of the Cu active site was further investigated in a follow up study isolating the reducing portion of the SCR by removing O 2 from the reaction feed. Cutting off O2 drove the catalyst into a highly reduced state dominated by Cu(I) while removing a reductant drove the Cu into the fully oxidized state. Our research shows that not only is redox a vital part of the SCR reaction on Cu/zeolites, but that the oxidation state of the active site is highly sensitive to the gas environment. The water-gas shift (WGS) reaction is an industrially important step in H2 generation from steam reforming. I have had the opportunity to contribute to a number of studies in WGS by studying the catalysts in FTIR. We studied numerous catalytic formulations including Fe promoted Pd/Al 2O3 and Au/TiO2. We found that the Fe promoted the WGS rate of the catalyst by a factor of 160 compared to the Fe free Pd/Al 2O3. The reduced Fe promoter efficiently split H2O, typically the role performed by reducible supports, and the nearby noble metal particles provided spillover H2 to maintain the reduced Fe phase necessary to split H2O. Our study of Au/TiO2 involved the development of a modified operando transmission IR cell with ultra-low dead volume allowing for fast switching isotope experiments over the catalyst. The isotope switching experiments showed that only CO adsorbed on Au0 sites was an active surface intermediate at 120°C. Counting the amount of active surface Au atoms for the reaction ruled out the Au particle surface and perimeter atoms as the dominant active sites and confirmed our previous finding that the active site was composed mostly of low coordinated corner Au atoms