Catalysts are essential for technological advances, because of their indispensable role in chemical and material manufacturing, energy conversion, and pollution control systems. Developing better catalysts is a highly desired goal that is impeded by the complexity of heterogeneous catalysts. This makes it extremely difficult to obtain information regarding active sites and reaction mechanisms, which is critical for improving catalyst design and performance. My research work has led to the understanding of how specific catalytic surface sites affect the performance of catalysts by constructing conceptually simpler planar model catalysts for kinetics and mechanism studies using model surface science tools and batch reaction testing. The work in this dissertation has demonstrated that planar model catalysts are versatile tools to probe reaction mechanisms on industrial catalysts. Supported gold nanoparticles have shown remarkable catalytic activity in a variety of reactions. However, many fundamental aspects of gold catalysts are still unclear, especially about the identity of active sites and oxidizing species. A Au(111) single crystal, the most stable and abundant facet on gold nanoparticles, is utilized to understand the reaction mechanisms of partial oxidation of 2-butanol and allyl alcohol. By controlling oxygen coverage on the surface, 100% selectivity to corresponding ketone and aldehyde, the desirable products, can be achieved. Two model catalysis systems, gold nanoclusters supported on a TiO₂(110) substrate and iron oxide dispersed on a Au(111) surface, were employed to understand the reaction pathways of CO oxidation and probe the role of the oxide/metal interface. The mechanistic and kinetic studies have shown that planar model catalysts are useful tools to probe reactions on industrial catalysts. The mechanistic understanding obtained from model catalyst studies can be used to create better catalysts. / text

Catalyst development is important to the contemporary world as suitable catalysts can allow chemical processes to proceed with reduced energy consumption and waste production. In order to design catalysts with improved performance, the fundamental studies that correlate catalytic properties with surface structures are essential as they can provide mechanistic insights into the reaction mechanism. Pd−Au bimetallic catalysts have shown exceptional performance for a number of chemical reactions, however, the interplay between the reactive species and surface properties are still unclear at the molecular level. In this dissertation, the catalytic chemistry of Pd−Au surfaces was investigated via model catalyst studies under ultrahigh vacuum conditions. A range of Pd−Au model surfaces were generated by annealing Pd/Au(111) surfaces and characterized/tested by surface science techniques. The findings in this dissertation may prove useful to enhance the fundamental understanding of structure-reactivity relation of Pd−Au catalysts in associated reactions.

This dissertation is focused on understanding heterogeneous metal catalysts supported on oxides using a model catalyst system of SiO2 thin film supported metal nano-clusters. The primary technique applied to this study is scanning tunneling microscopy (STM). The most important constituent of this model catalyst system is the SiO2 thin film, as it must be thin and homogeneous enough to apply electron or ion based surface science techniques as well as STM. Ultra-thin SiO2 films were successfully synthesized on a Mo(112) single crystal. The electronic and geometric structure of the SiO2 thin film was investigated by STM combined with LEED, Auger electron spectroscopy (AES), and X-ray photoelectron spectroscopy (XPS). The relationship between defects on the SiO2 thin film and the nucleation and growth of metal nano-clusters was also investigated. By monitoring morphology changes during thermal annealing, it was found that the metal-support interaction is strongly dependent on the type of metal as well as on the defect density of the SiO2 thin film. Especially, it was found that oxygen vacancies and Si impurities play an important role in the formation of Pd-silicide. By substituting Ti atoms into the SiO2 thin film network, an atomically mixed TiO2-SiO2 thin film was synthesized. Furthermore, these Ti atoms play a role as heterogeneous defects, resulting in the creation of nucleation sites for Au nano-clusters. A marked increase in Au cluster density due to Ti defects was observed in STM. A TiO2-SiO2 thin film consisting of atomic Ti as well as TiOx islands was also synthesized by using higher amounts of Ti (17 %). More importantly, this oxide surface was found to have sinter resistant properties for Au nano-clusters, which are desirable in order to make highly active Au nano-clusters more stable under reaction conditions.

This dissertation is focused on understanding heterogeneous metal catalysts supported on oxides using a model catalyst system of SiO2 thin film supported metal nano-clusters. The primary technique applied to this study is scanning tunneling microscopy (STM). The most important constituent of this model catalyst system is the SiO2 thin film, as it must be thin and homogeneous enough to apply electron or ion based surface science techniques as well as STM. Ultra-thin SiO2 films were successfully synthesized on a Mo(112) single crystal. The electronic and geometric structure of the SiO2 thin film was investigated by STM combined with LEED, Auger electron spectroscopy (AES), and X-ray photoelectron spectroscopy (XPS). The relationship between defects on the SiO2 thin film and the nucleation and growth of metal nano-clusters was also investigated. By monitoring morphology changes during thermal annealing, it was found that the metal-support interaction is strongly dependent on the type of metal as well as on the defect density of the SiO2 thin film. Especially, it was found that oxygen vacancies and Si impurities play an important role in the formation of Pd-silicide. By substituting Ti atoms into the SiO2 thin film network, an atomically mixed TiO2-SiO2 thin film was synthesized. Furthermore, these Ti atoms play a role as heterogeneous defects, resulting in the creation of nucleation sites for Au nano-clusters. A marked increase in Au cluster density due to Ti defects was observed in STM. A TiO2-SiO2 thin film consisting of atomic Ti as well as TiOx islands was also synthesized by using higher amounts of Ti (17 %). More importantly, this oxide surface was found to have sinter resistant properties for Au nano-clusters, which are desirable in order to make highly active Au nano-clusters more stable under reaction conditions.

Die heterogene Katalyse spielt in der industriellen chemischen Synthese sowie in umwelttechnischen Prozessen eine herausragende Rolle. Viele Katalysatoren zeichnen sich durch eine hohe strukturelle Komplexität aus, welche ein detailliertes Verständnis von entscheidenden Parametern sowie zugrunde liegenden Reaktionsmechanismen meist verhindert. Daher ist die Untersuchung von geeigneten Modellsystemen unerlässlich. Im Rahmen dieser Arbeit wurde ein dünner, kristalliner Aluminiumoxid-Film auf NiAl(110) mittels Rastertunnelmikroskopie (STM) und -spektroskopie (STS) bei 5K untersucht. Dieser Film konnte bereits in zahlreichen Studien als Modell für Alumi-niumoxid-Trägermaterialien etabliert werden, obwohl seine atomare Struktur nicht bekannt war. Hier wurden nun atomar aufgelöste STM-Bilder des Films aufgenommen, die sich später verschiedenen Lagen des Films zugeordnen ließen. Ferner konnten Antiphasendomänengrenzen (APDB), d.h. regelmäßig auftretende Liniendefekte des Oxidfilms, mit STM und STS charakterisiert werden. Es gelang somit, deren elektronische und geometrische Struktur zu korrelieren. Im zweiten Teil der Arbeit wurde das Adsorptionsverhalten des Aluminiumoxid-Films gegenüber einzelnen Au-Atomen untersucht. Nach der Präparartion waren Au-Monomere, Dimere und kleine Cluster auf der Oberfläche vorhanden, die mit STM und STS untersucht wurden. Die Ergebnisse zeigen, dass das NiAl(110)-Substrat an der Bindung von Au-Atomen auf dem Oxid-Film beteiligt ist, und dass dünne Filme offenbar ein anderes Adsorptionsverhalten aufweisen können als die entsprechenden Bulk-Oxide. Im dritten Teil dieser Arbeit wurden die Eigenschaften von Metall-Clustern (Ag, Pd) auf dem Aluminiumoxid-Film in Abhängigkeit von ihrer Größe untersucht. Leitfähigkeits-Spektren zeigen eine charakteristische Signatur, die am besten mit einer Coulomb-Blockade erklärt werden kann. Somit reflektieren die Spektren eher Eigenschaften des Tunnelkontakts als intrinsische Cluster-Eigenschaften. / Heterogeneous catalysis plays an important role in industrial synthesis and in environmental chemistry. Due the difficulties related with the investigation of working catalysts, the study of well-defined model systems is very important to gain a fundamental understanding of the principles and reaction mechanisms. Within the scope of this work, a well-ordered, thin alumina film on NiAl(110) has been investigated by scanning tunneling microscopy (STM) and spectroscopy (STS) at 5K. This film was established as model for bulk alumina supports in previous studies, though its exact structure remained unknown. Here, atomically resolved STM images of the film have been obtained, which could later be assigned to distinct atomic layers. Furthermore, anitiphase domain boundaries (APDB), regularly appearing line defects in the oxide film, were characterized by STM and STS. These studies provide a detailed understanding of how their geometric and electronic structure are related. The second part of this thesis addressed the adsorption behavior of the alumina film toward single Au atoms. The sample preparation yielded Au monomers, dimers and small clusters on the surface, which were investigated with STM and STS. Accordingly, the NiAl(110) substrate participates in the binding of Au atoms, demonstrating that adsorption properties of thin oxide films can deviate significantly from bulk oxides, whereby the metal adatom seems to play an important role. The third part of this work presents size-dependent STM/STS studies on metal clusters (Ag, Pd) deposited onto alumina/NiAl(110). Conductance spectra reveal a distinct signature, which can be explained by a Coulomb blockade effect. Another interpretation based on quantized electronic levels, is also discussed, but cannot account for all experimental findings. Thus, the spectroscopic data reflect most likely no intrinsic properties of the metal clusters but are due to the specific behavior of a double-barrier tunneling junction.

Vinyl acetate (VA) is an important monomer used in the production of paints, surface coatings and adhesives. Synthesis of VA is usually carried out over supported Pd alloy catalysts with a selectivity as high as 96% and described as C2H4 + CH3COOH + ½ O2 -> C2H3OOCCH3 + H2O Although the VA synthesis reaction has been industrially carried out for many years, the nature of the active sites and the reaction mechanism is still unclear. The goal of this study was to acquire a fundamental understanding of the VA reaction mechanism by carrying out detailed kinetic and spectroscopic investigations on single crystals and supported Pd catalysts, and to detail the role of alloying in optimizing the selectivity of this important industrial reaction. A combination of surface science techniques and kinetic measurements has been used to address the mechanism. Supported catalysts, 1 wt% Pd/SiO2 and 5 wt% Pd/SiO2, and 1 wt% Pd-0.5 wt% Au/SiO2, were prepared by an incipient wet-impregnation method and characterized using XRD and TEM. On Pd-only catalysts the reaction rates were found to be: Pd(100) < 5 wt% Pd/SiO2 (dpd = 4.2 nm) < 1 wt% Pd/SiO2 (dpd = 2.5 nm). Particle size-dependence of the reaction rates is evident for the Pd-only catalysts, which suggests a degree of structure sensitivity of the reaction. There is an increased availability of uncoordinated, edge atoms on small particles. With a Pd single crystal, fewer less-coordinated surface sites are present compared to a comparable area on a small Pd particle on a supported Pd catalyst. The formation of Pd carbide (PdCx) during the synthesis of VA was investigated over Pd/SiO2 catalysts with two different Pd particle sizes, as well as over a Pd-Au/SiO2 mixed-metal catalyst. XRD data indicate that smaller Pd particles show greater resistance to the formation of PdCx. The alloying of Au with Pd is apparently very effective in preventing PdCx formation in Pd-based catalysts for VA synthesis. Addition of Au to Pd/SiO2 catalysts significantly enhances the VA formation rate and selectivity. Infrared reflection absorption spectroscopy (IRAS) of CO on Pd/Au(100) and Pd/Au(111) confirms the presence of Pd as isolated monomers on a Au-rich surface. A pair of Pd monomers is the most favorable active site for the formation of VA. The spacing between the two active isolated Pd atoms is critical and is demonstrated by the relative rates of VA formation on Pd/Au model catalysts, i.e. Pd/Au(111) < Pd/Au(100). The role of Au is to isolate the surface Pd atoms and thus suppress the formation of by products, CO and CO2. A pair of Pd monomers required for VA synthesis is further confirmed by the results from model studies of Sn-Pd.

The aim of this thesis is preparation of inverse model catalyst CeOx/Pt(111) and its investigation using combination of surface physics methods and electrochemistry. New electrochemical cell was designed and built for electrochemical experiments. CeOx/Pt(111) samples were prepared and studied in UHV using STM and XPS methods. After that, samples were transferred to the electrolyte environment and studied by means of cyclic voltammetry and AFM. For high surface coverage of CeOx, new reaction was observed. We have identified this reaction as a combination of the reduction of cerium(IV) oxide by interaction with hydrogen adsorbed on the Pt(111) surface and oxidation of cerium(III) oxide by dissociative adsorption of water molecules. Powered by TCPDF (www.tcpdf.org)

"Catalytic growth of Carbon Nanotubes (CNT) provides important advantages of controlling their diameters and possibly chirality. Our work involved growing CNT on model and supported catalyst by catalytic decomposition of carbon source such as benzene, methane and propylene. On supported catalyst, iron nitrate was deposited on alumina and reduced to form metallic iron clusters. These were reacted at 700 - 950 C under varying benzene concentrations. Multi Walled CNT (MWNT) grew below 800 C and Single Wall CNT (SWNT) are observed at 850 C and above as confirmed by TEM and Raman. Model catalysts were studied by producing CNT from ferrocene which acted as the carbon and catalyst source on Silica/Si (100). Large yield of MWNT was observed at 900 C. MWNT grew perpendicular to the model support as seen by SEM. In another model catalyst study, iron salt clusters were deposited on silica/Si (100) by spin coating, controlling their diameters by solution concentration and speed of spinning. Agglomeration of clusters at high temperatures produces only MWNT on silica/Si (100). Cluster agglomeration can be reduced with strong support metal interaction such as with alumina. We deposited alumina on silica/Si (100) by atomic layer deposition, with conformal coatings on surface and low relative roughness. Alumina film was stable under reaction temperatures of 900 C. Cluster deposition on alumina by spin coating was difficult because of different surface acidity compared to silica. Clusters on alumina did not agglomerate at high reaction temperatures. We report effect of parameters such as the temperature of reaction, conditions of pretreatment such as reduction and oxidation of catalyst precursor, type of precursor, type of carbon source, and type of support material on growth of CNT. The role of spin coating in controlling the diameter of salt clusters deposited is discussed. We also report deposition of alumina on top of silica/Si (100) by atomic layer deposition and the effect of deposition and calcination temperatures on the alumina film integrity."

In dieser Arbeit wurde ein Modell-Katalysator-Ansatz angewandt um Mischoxide mit Molybdän und Titan zu untersuchen. Die Schichten wurden auf TiO2(110) Einkristallen durch Verdampfen der Metalle in einer O2 Atmosphäre und in UHV und Nachbehandlung der Filme vorbereitet. Verschiedene Präparationen wurden in der Studie untersucht und diese werden in sechs Kategorien dargestellt. Wenn Mo und Ti in O2 gemeinsam aufgedampft wurden, wurde das meiste Mo an der Oberfläche abgeschieden mit einer nur geringen Mo-Konzentration in tieferen Schichten. Eine Mischung von Mo in TiO2 war sehr begrenzt, und die stimmt mit dem Phasendiagramm MoO2 und TiO2 überein. Mo6+ und Mo4+ sind die dominierenden Oxidationsstufen in den meisten der Schichten, wobei Mo6+ stets näher an der Oberfläche war als Mo4+. Schichten, in denen Mo vollständig in TiO2 gelöst ist, konnten durch Abscheidung von Metallen in UHV und Post-Oxidation der Filme erstellt werden. Im Inneren des TiO2 Gitters hat Mo die Oxidationsstufe 4+. Aufdampfen von Mo in O2 bei Raumtemperatur und anschließendes Tempern in UHV führte zur Bildung zweier Arten von Merkmalen in den STM-Bildern. Diese waren im UHV stabil bis mindestens 1000 K. Die Schichten mit hoher Mo-Konzentrationen erschienen uneinheitlich in den STM-Bildern aber sie zeigte das TiO2(110)-(1x1) LEED-Muster. Der Anstieg in der Mo-Konzentration führte zur Blockierung der Überbrückung Sauerstoffleerstellen (BOVs), was durch STM und Wasser-TPD nachgewiesen wurde. Die Reaktivitäten der Schichten wurden mit Methanol- und Ethanol-TPD getestet. Eine unerwartete Formaldehyd/Methanbildung bei hohen Temperaturen (~650 K) wurde bei der Methanol-TPD von reinem TiO2(110) beobachtet und mit BOVs in Verbindung gebracht. Der Anstieg der Mo-Konzentration unterdrückte diesen Effekt sowie die Ethylenbildung (~600 K) beim Durchführen von Ethanol-TPD. Sowohl in Ethanol als auch Methanol-TPD wurden neue Reaktionswege zu Ethylen und Methan-Bildung bei ~500 K beobachtet. / In this work, a model-catalyst approach has been taken to study oxide mixtures containing molybdenum and titanium. The layers were prepared on TiO2(110) single crystals by evaporating the metals in an O2 atmosphere or in UHV and post treating the deposited material. Different preparation procedures were employed in the study and these are presented in six categories. When Mo and Ti were co-deposited in O2, most of the molybdenum stayed at the surface with only a small Mo concentration in deeper layers. Mixing of Mo into TiO2 was very limited, consistent with the phase diagram of MoO2 and TiO2. Mo6+ and Mo4+ were the dominant oxidation states in most of the layers and Mo6+ was always nearer to the surface than Mo4+. Layers where Mo was completely mixed into TiO2 could be prepared by depositing metals in UHV and post-oxidizing the deposited material. Inside the TiO2 lattice, Mo had an oxidation state of 4+. Depositing Mo in O2 at room temperature and post annealing in UHV led to the formation of two types of features in the STM images. These features were stable in UHV up to at least 1000 K. The layers with high Mo concentrations appeared patchy in the STM images but they still exhibited the TiO2(110)-(1x1) LEED pattern. The increase in Mo concentration led to the blocking of the bridging oxygen vacancies (BOVs) as evidenced by STM and water TPD. The reactivities of the layers were tested by methanol and ethanol TPD. An unprecedented high temperature (~650 K) formaldehyde/methane formation channel was observed in the methanol TPD of clean TiO2(110) and was associated with BOVs. The increase in the Mo concentration led to the vanishing of this channel as well as the ethylene formation channel (~600 K) in the case of ethanol TPD. In both ethanol and methanol TPD, new reaction channels towards ethylene and methane formation at ~500 K appeared.

Les propriétés catalytiques des nanoparticules métalliques peuvent être améliorées par effet d’alliages. La synthèse en solution par voie colloïdale permet de préparer des nanoalliages homogènes en taille, en forme et en composition chimique, de structure ordonnée, désordonnée ou cœur-coquille. La nucléation et la croissance des nanoalliages de Pt-Pd sont étudiées ici par microscopie électronique en transmission, en condition standard, puis in situ dans une cellule liquide formée par des feuilles d’oxyde de graphène. La cinétique de croissance des nanoalliages de Pt-Pd correspond à l’incorporation directe des monomères en solution, compatible avec un processus limité par la réaction de surface, sans phénomène de coalescence, contrairement à la croissance du Pt pur. La structure théorique à l’équilibre des nanoalliages de Pt-Pd est déterminée par des simulations Monte Carlo. La structure la plus probable correspond à une surface riche en Pd et à une sous couche atomique riche en Pt, stable à des températures élevées. L’effet de l’adsorption de gaz oxydants ou réducteurs sur la forme des nanoparticules, est étudié in situ par microscopie environnementale sous pression de quelques mbar, dans un porte objet environnemental. On observe des changements de formes sous oxygène, dus au développement de facettes d’indices plus élevés. La réactivité des nanocubes de Pd@Pt est étudiée pour l’oxydation du CO en fonction du recouvrement de Pt à la surface. La réactivité maximale pour un faible recouvrement est interprétée par une baisse de l’énergie d’adsorption du CO liée au désaccord paramétrique entre le Pt et le Pd et à la modification de la structure électronique du Pt lié au Pd. / The catalytic properties of metal nanoparticles can be improved by the alloying effect. Nanoalloys homogeneous in size, shape and chemical composition can be prepared with the colloidal synthesis method, with an ordered, random or core-shell chemical structure. Nucleation and growth of colloidal Pt-Pd nanoalloys were studied by transmission electron microscopy (TEM), in standard conditions and in situ with the aid of a graphene oxide liquid cell. The growth kinetics of homogeneous Pt-Pd nanoalloys corresponds to the direct incorporation of the monomers in solution. It was compatible with a process limited by the surface reaction, without coalescence (Lifshitz-Slyozov-Wagner mechanism). On the contrary, coalescence occurs during the growth of pure Pt nanoparticles. The theoretical structure of Pt-Pd nanoalloys is determined by Monte Carlo simulations. The most stable structure corresponds to a Pd surface and Pt subsurface layer, which is stable up to high temperatures. The effect of adsorption of oxidizing or reducing gasses on the shape of pure Pd nanocubes and core-shell Pd@Pt nanocubes is studied in situ by TEM with an environmental cell. The observed changes in a few mbar of oxygen are due to the development of higher index facets. The CO oxidation reaction is used to compare the reactivity of homogeneous Pt-Pd nanoalloys and core-shell Pd@Pt nanocubes with increasing coverage of Pt at the surface. A maximal reactivity is attained for a low coverage. The effect is interpreted by a decrease in adsorption energy of CO, due to electronic effects originating from the lattice mismatch between Pt and Pd and the mixed Pt-Pd bonds.