| In the process of producing hydrogen for PEM fuel cell applications from carbon-based fuels, an important stage is the catalytic water-gas shift (WGS) reaction, which is used to upgrade the hydrogen-rich fuel gas stream. The industrial Cu/ZnO catalyst lacks the required stability for this application which typically requires frequent shutdown-restart and may expose the catalyst to air and/or water. Based on numerous recent investigations, CeO2- and modified-CeO2-supported precious-metal catalysts, especially Pt and Au, have emerged as attractive alternatives possessing the required high activity for the low-temperature WGS, and no air or moisture sensitivity. The active sites comprise oxygen bound Pt or Au species, Pt-O-Ce or Au-O-Ce. Only a small number of these sites remain bound and active in realistic fuel gas streams; the number dictated by the oxygen potential of the gas and temperature. Another suitable support to bind Au-O sites is iron oxide. The apparent activation energy of the WGS reaction on gold is similar on both types of supports. From an economic viewpoint, it is interesting to design catalysts with just the right amount of a precious metal for a given set of operating conditions and avoid overdesigns that lead to destabilization, metal particle sintering, and therefore loss of the expensive precious metal. It is also interesting to examine whether non-reducible oxide supports can be used to stabilize Au-O and Pt-O species active for the WGS reaction. This information is also important from a mechanistic viewpoint, and if true, it would provide design flexibility and allow the use of abundant and cheap supports, such as silica and alumina. This hypothesis was tested in this thesis work.;Addition of alkali oxides on Pt-based WGS catalysts was effective in creating an active site comprising Pt-O and neighboring -OH groups that could be activated by CO at temperatures as low as 100 °C, similar to the Pt-O sites on ceria. It does not matter what the support surface is when alkali promoters are used with Pt; silica or alumina is as effective a carrier of the active sites as ceria. This is the major finding of this thesis. Such alkali-promotion was examined and identified on Au-based catalysts as well, but a detailed study of the latter was not conducted.;Characterization techniques such as XPS, aberration-corrected HAADF/STEM, XANES/EXAFS, CO chemisorption, and H2/CO-TPR were used to probe and improve our mechanistic understanding of the alkali-promotion of the Pt-based catalysts. Guided by these findings, DFT calculations were performed by Prof. Mavrikakis and his group at the University of Wisconsin-Madison, and a few plausible structures for the active site were proposed for K-promoted Pt-O catalysts. The cluster K-Ox-Pt-(OH)y retains Pt in oxidized state, with a ratio of K:O=1:1; binds CO weakly; and adsorbs/dissociates H 2O almost without energy demand (thermoneutral), similar to Cu (111), the best WGS catalyst.;In detailed studies with 1at%Pt-3at%Na on fumed silica, it was found that the active site stabilizes Pt as well as the alkali ions that comprise it. Thus, the alkali ions are not removed by repeated washings of the catalyst by de-ionized water at either ambient temperature or 70 °C. Stability in realistic gas streams was found; Pt retained its oxidized state as shown by in-situ XANES, and no activity loss was found in isothermal experiments up to 275 °C after 10-40 h-on-stream. Other alkali (Li, Cs) and alkaline-earth ions (Mg, Ca, Ba) added in small amounts were also effective, but Na provided the maximum promotion.;All the alkali-promoted Pt samples on the alumina or silica used here have similar apparent activation energy, 70 +/- 5 kJ/mol, for the low-temperature WGS reaction, as reported also for Pt on TiO2, CeO2, or ZrO2. Therefore, it is proposed that the key steps of the WGS reaction catalysis (CO adsorption and H2O activation) occur on Pt-Ox-(OH)y sites, irrespective of the type of support and the additive used.;Gold catalysts for the low-temperature WGS reaction were examined on various supports and their activity and stability was compared to that of the Au-CeO2 system. The reaction-relevant Au structures on Au-FeO x were followed by in-situ XANES and EXAFS spectroscopy. Similar to ceria, the number of Au-O sites that remain bound on iron oxide depends on the oxygen potential of the fuel gas and operating temperature. Highly reducing fuel gas mixtures destabilize the [Au-O-Ce, -Fe] sites and cause gold cluster/particle formation and deactivation. Activity recovery is possible by re-oxidation in air (350-400 °C). Catalyst stability is improved if gas mixtures with higher oxygen potential are used. Oxygen-assisted WGS operation was found to stabilize the Au-O-Fe sites, and is recommended for practical applications.;OH-rich or alkali-modified surfaces can create more binding sites and disperse the Au species better during the synthesis and thermal treatment steps. As a result, the WGS reaction activity is proportionately increased. The alkali-free and alkali-promoted Au-CeO2, Au-Fe2O 3, Au-La2O3 catalysts have the same apparent activation energy, which indicates a common active site for all Au-based catalysts for the WGS reaction, regardless of the support or additive used. By analogy to the Pt catalysts, we propose a cluster of the type Au-Ox(OH) y stabilized by Ce, Fe, La, or alkali ions. This needs to be tested by DFT calculations.;Although enhanced WGS reactivity over Au-based catalysts was manifested with OH-rich or alkali-modified surfaces, the active Au species are more sensitive to destabilization than Pt. The electronegativity of the oxide/additive plays an essential role on the stability of the Au active sites, and is thus recommended to use this to guide the choice of additives and the optimization of the synthesis conditions of gold catalysts. |
