The preparation of a supported metal catalyst is a complex process where some steps are often not completely understood. The work of this thesis is focused on understanding and improving the processes involved in the preparation of a catalyst, and in particular aims to find a route for the preparation of homogeneous supported alloy catalysts and to establish how particle size can be controlled during the preparation. One of the difficulties preparing homogeneous alloys with traditional methods is the formation of single metal particles. To avoid this, a method based on the selective reduction of a second metal on the surface of a single metal supported catalyst (parent) was chosen. This method has the advantage of producing bimetallic particles where the two metals are in close proximity and can be subsequently alloyed by thermal treatment. Due to the difficulty in characterising supported alloy nano-particles, a combination of different techniques (in situ XRD, XPS, temperature programmed reduction, pulse chemisorption, temperature programmed hydrogen desorption etc.) were used to investigate the prepared catalysts and to study the degree of alloying as a function of the reduction temperature. The catalysts were also tested in a fixed bed gas phase reactor for the hydrogenation of crotonaldehyde. To achieve a background of knowledge on the single metal parents used for the selective deposition method, Pd-Al2O3, Ni-Al2O3, and Co-Al2O3 were prepared and characterised. Some important aspects developed in this preliminary work were the solubility of hydrogen in Pd (and its use in characterizing Pd particle size and degree of alloying) and the interaction between Co and Ni with the alumna support. After the single metal study the deposition of Ni over Pd and of Pd over Ni were attempted. The latter system led to the successfully preparation of core-shell particles where the Pd covers the Ni. The Pd-Ni-Al 2O3 bimetallic catalyst was then reduced at different temperatures in order to alloy the particles. The results revealed that alloys could be formed at relatively low temperature (300 °C). The last part of the work diverged from the main path of my project. It was a work of pure characterisation using in situ techniques. An interesting series of XRD and Raman in situ experiments on vanadium phosphorus oxide (VPO) catalysts revealed the formation of a metastable phase (co-VOPO4). The study of the transformation of this phase induced by various reactants under reaction conditions provided useful information in understanding the structure of the active catalyst and cast new light on the on role of the reactant and product gas mixtures in forging the structure of the catalyst.

<super>1</super>N.F.Dummer, R. Jenkins, X. Li, S.M. Bawaked, P. McMorn, A. Burrows, C.J. Kiely, R.P.K. Wells, D.J. Willock, G. Hutchings, J. Catai, 243 (2006) 165. 2 S. Lavoie, M.-A. Laliberte, I. Temprano, P H. McBreen, J. Am. Chem. Soc, 128 (2006) 7588.

The planar extended shape of the phthalocyanine macrocycle results in a strong tendency of its derivatives to form densely packed co-facial aggregates. The strategy to avoid co-facial self-association that forms the basis of this thesis involves the use of substituents that can introduce severe steric crowding adjacent to the phthalocyanine core. Previous work showed that the introduction of 2,6-di-/jo-propylphenoxy groups on the peripheral positions of the phthalocyanine seems to be perfect for this purpose. Of particular interest is zinc octa(2,6-di-/$o-propylphenoxy) phthalocyanine (PclZn), which forms a remarkable cubic crystal structure, containing interconnected solvent-filled voids 2 nm across. The aim of the research programme was to investigate the crystal forming properties of related phthalocyanine derivatives containing different metal cations and bulky phenoxyl substituents. A range of metal cations were introduced into 2,3,9,10,16,17,23,24-octa(2',6'-di-/s0-propylphenoxy) phthalocyanine (Pel) to establish which, if any, were compatible with the formation of the nanoporous cubic crystal observed for the zinc derivative. It was found that any metal capable of binding to a ligand at its axial site formed the cubic crystal including metals of primary catalytic relevance such as cobalt, iron, manganese and ruthenium. Single crystal X-ray diffraction studies demonstrated the exchange of axial ligands to confirm the interconnectivity of the nanovoids, which is essential for the potential exploitation of these molecules in heterogeneous catalysis. Of particular interest is the introduction of bidentate ligands, which act as structural wall ties that bind metals between cubic subunits. Since loss of crystallinity occurs after removal of the guest solvent from the cubic clathrates, the introduction of substituents at the 4-position of the phenoxy groups was also investigated in order to induce stronger dipole-dipole (e.g., R = Br, CI, CN, OMe) or hydrogen bond interactions (e.g., R = OH), which might stabilise the crystal structure. Unfortunately, these derivatives formed non-cubic crystals, although in each case solvent was included within the structure to form novel clathrates.

Electron Paramagnetic Resonance (EPR) and Electron Nuclear Double Resonance (ENDOR) spectroscopies have been used to study the fundamental nature of chromium-based selective oligomerisation catalysts. A series of 'pre-catalyst' complexes were fully characterised CW-EPR revealed each complex to possess an axial g matrix (g > ge > g ) and superhyperfine coupling to two equivalent 31P nuclei, consistent with a low-spin cf species of approximate Cjv symmetry, where the metal contribution to the SOMO is primarily dxy. The isotropic component to the 31P coupling was of a larger magnitude in those Cr(I) complexes bearing PNP ligands than those bearing PCP, indicating that the phosphorus 3s character in the SOMO was higher for the former. CW-ENDOR demonstrated that subtle structural differences in the complexes, namely in the phenyl ring conformations, occurred as a function of ligand type. Pulsed experiments proved that the technique is valid and viable for further work on the activated system. Upon activation of the pre-catalyst with an alkylaluminium, four distinct paramagnetic centres were identified. A Cr(I) bis-arene complex was firstly detected it was found to form either via intramolecular co-ordination of the ligand phenyl groups, or preferentially via solvent-based arene co-ordination, if such groups were available. Two further species (I and III) were subsequently observed at low temperatures the spin Hamiltonian parameters extracted for both showed that a significant modification to the structure of the pre-catalyst had occurred. Half-field transitions indicated the possibility of a dimeric nature to Species III. ENDOR measurements detected an exceptionally large proton coupling in the activated system, possibly due to the co ordination of alkyl fragments to the metal centre. A final, fourth paramagnetic centre (Species IV), was detected and classed as an intermediate species, due to the greater similarity between its g and A matrices with those of the parent complex, than the other activated species. Finally, a preliminary investigation into analogous pre-catalyst complexes bearing N-heterocyclic carbene ligands was performed, due to their similar employment in oligomerisation catalysis CW-EPR spectra revealed information on both their electronic and structural natures.

The current steam reforming process for the production of CH₃OH is complicated and difficult, and therefore the direct partial oxidation of CH₄ to CH₃OH would be economically desirable. In previous work a design approach for a selective partial oxidation catalyst has been investigated, which comprises the combination of components with a desired reactivity, producing a successful selective partial oxidation catalyst. In this approach, it is considered a successful partial oxidation catalyst must activate methane, activate oxygen and not destroy the desired product, methanol. All these properties could not be found in a single catalyst, so it was proposed that two synergistic components could be combined, one responsible for methane activation and the other for oxygen activation/insertion. Previous work has studied the CH₄/D₂ exchange reaction as an indication of the ability of a metal oxide surface to activate CH₄. Two metal oxides demonstrated appreciable activity for the activation of CH₄, these being Ga₂C₃ and ZnO. These oxides were then doped with different metals in order to try and increase the activity of the catalyst. The doping of Ga₂O₃ with Zn or Mg did not improve the methane oxidation properties of Ga₂C₃, and the doping of ZnO with Ga significantly lowered the light off temperature, the temperature at which CH₄ was first detected, and increased its oxidative capacity. The addition of precious metals significantly affected the catalysts ability to activate CH₄. The addition of Au to the Ga and Zn catalysts dramatically reduced the light off temperature, and increased its rate of oxidation at lower temperatures, with the optimum loading 2% for both catalysts. For GaO(OH) and ZnO, the addition of 1%Au and l%Pt by coprecipitation produced a synergistic effect, producing lower light offs and higher CH₄ conversion than the singly doped catalysts with Au and Pt separately. When the methane activation catalysts were combined with MoO₃ in a physical mixture, a number of the mixtures produced higher methanol per pass percentage yields than its constituent parts. It is concluded that the increased methane activation properties beneficially interact with the oxygen activation and insertion properties of MoO₃. However, none of the yields reported were significantly higher. A dual bed system, with the lower layer comprising the methane activation catalysts, and the upper layer consisting of MoO₃ was tested. The results for this system were promising, with the low temperature activation of CH₄, combined with the oxygen insertion ability of MoO₃, producing high selectivities of CH₃OH at much lower temperatures. The best results were obtained when the ratio of the two layers was 50:50 with respect to 2%Au ZnO and MoO₃. In previous work a design approach for a selective partial oxidation catalyst has been investigated, by combining components with a desired reactivity to produce a successful selective partial oxidation catalyst, which must activate methane and oxygen, and not destroy methanol. All these properties could not be found in a single catalyst, so it was proposed that two synergistic components could be combined, one responsible for methane activation and the other for oxygen activation/insertion. The doping of ZnO with Ga significantly lowered the light off temperature, and increased its oxidative capacity, an effect which was not seen with the doping of Ga₂O₃ with Zn or Mg. The addition of Au to the Ga and Zn catalysts dramatically reduced the light off temperature, and increased its rate of oxidation at lower temperatures, both with optimum loading of 2%. The addition of l%Au and l%Pt produced a synergistic effect, producing lower light offs and higher CH₄ conversion than the singly doped catalysts with Au and Pt separately. When the methane activation catalysts were combined with MoO₃ in a physical mixture, a number of the mixtures produced higher methanol per pass percentage yields than its constituent parts. It is concluded that the increased methane activation properties beneficially interact with the oxygen activation and insertion properties of MoO₃. The dual bed system, with the lower layer comprising the methane activation catalysts, and the upper layer consisting of MoO₃ produced promising results, with the low temperature activation of CH₄, combined with the oxygen insertion ability of MoO₃, producing high selectivities of CH₃OH at much lower temperatures. The best results were obtained when the ratio of the two layers was 50:50 with respect to 2%Au ZnO and MoO₃. (Abstract shortened by UMI.).

Using rapid supercritical CO2 antisolvent precipitation, a set of nanoscaled oxide catalyst precursors have been prepared, which were further calcined to give metal oxides, such as CuMnOx(hopcalite), CeC>2, Ti02 and ZnO etc. CuMnOx(hopcalite) was used as catalyst without further modification and other single metal oxides were used as supports for gold catalysts. These materials have been characterized using a full range of techniques: XRD, Raman, FT-IR, BET surface area, SEM, DTA/TGA and XPS. Element distribution in the as-precipitated materials was evaluated by TEM-EDX. The catalytic activity of our products was evaluated by low temperature (0 25 C) CO oxidation. For CuMnOx (hopcalites), an amorphous homogenous precursor mainly composed of acetates has been prepared. Following calcinations, separated copper nanocrystals supported on manganese oxide (Cu/MnOx) has been obtained. The preliminary catalytic data show the intrinsic activity for CO oxidation of the catalyst derived from this precursor is considerably higher than the conventional CuMn204 catalysts prepared by coprecipitation, and also currently available commercial catalysts. The results clearly show that a catalyst with enhanced activity can be prepared without the presence of intimately mixed copper and manganese oxide components. In addition, using mixed solvents, such as water-ethanol and water-DMF, crystalline heterogeneous precursors have been produced. With the addition of more water to the precursor solution, there appears to be a reaction between the metal acetates CO2, and H2O. Therefore, carbonates of the metals are precipitated instead of the acetate composition. Following calcination, less crystalline or even amorphous phase-separated nanostructure final catalysts retain the high surface area, which leads higher catalytic activities than that of the current commercial hopcalite catalysts. Furthermore, using 30% H2O2 as an oxidizer and ethanol as solvent, catalysts have been prepared as well. All the as-prepared catalysts exhibit higher catalytic activities on CO oxidation when compared to those from solvents in the absence of H2O2. Novel nano-polycrystalline Ce02 was produced. When it was used as a support for gold and gold palladium nanoparticles, the catalytic data show that the activity and catalyst lifetime for CO oxidation of a gold catalyst supported on this material is much greater than that for gold supported on regular Ce02 derived from the direct calcination of cerium acetylacetonate. In addition, the Au-Pd catalysts supported on Ce02 prepared using supercritical antisolvent precipitation are amongst the most active catalysts yet reported for the selective oxidation of alcohols and the direct oxidation of hydrogen to hydrogen peroxide. Similarly, T1O2 has been produced by supercritical process. When using it as a support for gold nanoparticles, the activity and stability for CO oxidation of a gold catalyst supported on this material is much greater than that for gold supported on regular Ti02 derived from the direct calcination of titanium oxide acetylacetonate. Finally, ZnO was prepared using supercritical process and then was used as supports for gold particles as well. The catalytic data show that it can give very high activity for CO oxidation.

The use of nitrate precursors for the production of mixed oxide catalysts such as iron molybdates and copper manganese oxides has been well documented in the literature. Nitrates have been used because of their high solubility and the ease of removal of the nitrate anion during calcination. In the case of copper manganese oxides, the use of chlorides and sulphates leads to the retention of the anion on the surface of the catalyst, using nitrates leaves no residue on the catalyst. The use of nitrates, however, is unfavourable due to the environmental issues associated with them. The main aim of this project was to provide a nitrate free route for the production of iron molybdate and copper manganese oxide catalysts. Using supercritical antisolvent precipitation (SAS) provided an alternative preparative route for the production of these catalysts. The catalysts produced exhibited activity when tested in their respective reactions. Some catalysts produced were comparable to those currently prepared using the conventional preparation methods i.e. the use of nitrate precursors. The research performed focussed on the production of theses mixed metal oxides using supercritical carbon dioxide. Optimisation of the supercritical process by varying operational parameters was also investigated to determine their effect on the activity of the catalyst obtained. By utilising operating parameters and understanding the supercritical antisolvent process, the results obtained showed that it was possible to produce mixed metal oxides, and to tune their catalytic activity.

This thesis embraces two main sections that examine alternative architectures for the iminium ion catalysis of the Diels-Alder reaction - comprising acyclic and cyclic systems. Chapter 1 provides an overview of iminium ion and enamine catalysed processes. Focus is on the structural requirements of catalysts for effective and efficient transformations. Chapter 2 highlights work carried out previously within the group and provides a more in depth review of the Diels-Alder reaction, the protocol adopted for comparing catalyst activity and methods by which results were analysed. Chapter 3 describes the design and synthesis of a series of five membered cyclic catalysts based on the pyrazolidine subunit. These systems incorporated aspects of acyclic catalysts developed previously within the group to decrease reaction times and catalyst loadings. Reaction conditions were optimised with regards to solvent and acid co-catalyst. SAR studies of the cyclic and acyclic architectures lead to the expansion of the acyclic series within Chapter 4. This resulted in the most active acyclic catalyst to date with a conversion of 98% after of six hours within the iminium ion catalysed Diels-Alder reaction. Chapter 5 shows a combination of SAR studies and computer modelling that led to the design and synthesis of a series of six-membered cyclic analogues containing an a-heteroatom. These systems catalysed the Diels-Alder reaction more efficiently producing a conversion of 89% after three hows and 99% after six and would provide an effective architecture for the development of a chiral catalyst A trend was discovered relating the electronic environment of a p-carbonyl incorporated into the catalyst structure and its associated activity with supporting evidence from an associated computational project Finally, Chapter 6 describes preliminary studies into the three proposed component steps of the iminium ion catalysed Diels-Alder reaction. 1H NMR and UV spectroscopy was (and continues to be) used to examine these steps to gam valuable information iiyrilUig the mechanism of the catalytic cycle.

The scope of this thesis covered three main areas of this emerging area of chemistry. A reaction pathway is proposed for the formation of iminium ion intermediates from the reaction of secondary amines with carbonyl compounds. This suggests the deprotonation of the amine as the rate-determining step. The effect of modification of amine structure on this pathway was studied, and rationalised using Atoms-in-Molecules analysis. Molecular properties were used to describe a range of secondary amines with an aim to find a relationship between composition and reactivity. A number of catalytic candidates were suggested for synthesis. Two reaction types that directly and indirectly utilise iminium ions as catalysts or intermediates were covered. This included studies of model amines, as well as those in current use as asymmetric catalysts. The Diels-Alder reaction was studied in depth for a number of iminium ions. The alpha-acyloxylation of enamines was also studied, and reaction profiles in gas-phase and solvent identified, and possible mechanisms of asymmetric induction explored.

This thesis describes the synthesis of a series of mixed N,0,S donor ligands, their co-ordination chemistry using palladium, rhodium and iridium metal precursors and their catalytic behaviour. First synthesised were a series of potential tripodal oxygen donor ligands mimicking the synthesis of meTAM but incorporating a fourth additional group that was hoped to preferentially co-ordinate to a metal centre. A variety of synthetic pathways were unsuccessfully employed to synthesise mixed donor ligands using aliphatic bidentate backbones. This work was followed by the synthesis, structure and co-ordination chemistry of novel tripodal ligands containing bis(l-methylimidazol-2-yl)methyl methane in combination with an ether or sulfide functionality. These N,N,0 and N,N,S ligands were expected to demonstrate properties intermediate of those of the Trofimenko and Klaui ligands. As expected these ligands found application in both co-ordination chemistry and organometallic chemistry, as a series of metal complexes of the N,N,0 and N,N,S hemilabile ligands were isolated and fully characterised crystal structures of Pd, Rh and Ir complexes are illustrated above. The Pd(II) complexes were found to be active catalysts in the Heck coupling of alkenes with aryl halides.