The work presented in this thesis highlights the influence of water soluble tertiary amines in the synthesis of transition metal substituted heteropolyoxometalates (TMSPs), with particular emphasis on polyanions based on the Keggin structure and lacunary derivatives thereof. In particular, an understanding is gained into the role that N,N’-bis(2-hydroxyethyl)piperazine (bhep) and morpholine (morph) play in the polyanion self-assembly process, with the successful isolation of molecular species, chains, layers and three dimensional frameworks. Each of the ligands were used in two different sets of synthetic investigation to probe their influence on the complete self-assembly of transition metal substituted tungstophosphates using Na2WO4 as the tungstate source, and by using the [XW10O36]8- X = Si, Ge precursors. Indeed, at no point in this work has the same polyanion been isolated using both the bhep and morpholine ligands, with this observation being indicative of each ligand having structure directing properties. Synthesis of TMSPs based on the phosphotungstate clusters in the presence of (bhep) has resulted in the isolation of clusters with nuclearities {W6}, {W7}, {W9}, {W10}, and {W11} which were only possible through fine control of the reaction pH which is aided by the ligands high buffer capacity. Additionally, in these compounds the ligand is observed to adopt a variety of coordination tendencies ranging from monodentate coordination through the ligands hydroxyl groups to chelation between nitrogen and oxygen donors. Specifically, the chelation of the ligand to CoII and NiII ions has allowed the rare inclusion of transition metal complexes into lacunary polyanions through a complimentary self-assembly process. Furthermore, the (bhep) ligand has been used to directly functionalise the [γ-SiW10O36]8- polyanion through the controlled formation of W-OC linkages via a dehydration process. Interactions of the Morpholine ligand with heteropolytungsate clusters and transition metals in aqueous solution results in an array of markedly different structural motifs, which are highly depedant on the synthetic approach. The high basicity of the morpholine ligand results in its function as a buffer and cation within a pH domain that is underdeveloped in TMSP chemistry. Consequently, novel molecular species and functional materials have been isolated through the exploitation of this knowledge. The most significant of all the findings is the isolation of five mesoporous frameworks, which are composed solely of manganese or cobalt substituted Keggin polyanions which act as trigonal and tetrahedral nodes in the purely inorganic frameworks. Furthermore, the manganese containing frameworks have been shown to have cation exchange properties and are also redox active with single-crystal to single-crystal transformations being observed.

The host-guest chemistry of the isopolyoxometalate {W36} crown-like cluster is investigated. Previous reports showed that the cluster can specifically bind primary amines via a central binding pocket formed by six terminal oxo ligands on the all - inorganic, cation binding host. This assembly strategy was transformed to utilise the incorporation of long chain alkyl diammonium guest cations to physically define the supramolecular structure of the clusters with respect to each other and demonstrate the structure direction as a function of alkyl chain length. The systematic variation of chain length gave access to five unprecedented supramolecular assemblies. Diprotonated 1,8-diaminooctane molecules link the {W36} clusters into infinite 1-D zig - zag chains whereas huge trimeric assemblies were observed for 1,9-diaminononane and 1,10-diaminodecane linkers. Dumb - bell shaped dimeric units as a result of direct centre – to - centre linkages between the {W36} clusters using diprotonated 1,12-diaminododecane were prepared and triply protonated bis(hexamethylene)triamine was employed to obtain linear 1-D chains of directly connected {W36} cluster units. In an alternative approach, the linking of polyoxometalate – based secondary building blocks was investigated using transition metal linkers in organic solvents. It has been previously shown that Ag(I) is an excellent candidate due to its versatility and ability to readily adopt a wide variety of coordination environments. A series of novel architectures have been prepared including a supramolecular dimer based on decavanadate {V10}, an interconnected 2-D network incorporating the decatungstate species {W10}, extensively interconnected systems based on the unique {W19} cluster and unusual, highly symmetrical frameworks incorporating non – classical Dawson type units {SbW18} and {BiW18}. The newly discovered silver - tungstate systems, which feature varying degrees of attractive argentophilic silver - silver interactions, have proven to be effective precursors for the formation of interesting nano / sub - micro materials. For the ultimate study, three titanium embedded polyoxotungstates with unprecedented structural features are presented: a monotitanium containing tungstoantimonate featuring an unusual {Ti=O}2+ moiety, a hexatitanium containing tungstoarsenate featuring a highly charged {Ti4(H2O)10}16+ core and a decatitanium containing tungstoarsenate exhibiting the second largest structure of its type ever reported. Due to their highly unique nature, the systems have been fully investigated using DFT calculations yielding important results. The first cluster has been optimised with five sodium cations in the belt position, which in addition to reducing the high charge, influence a stabilisation of the antimony lone pairs. Furthermore, electrostatic potential calculations highlight the high electronegativity of the terminal oxygen on the single titanium centre, enhancing real potentiality as a reactive site for catalysis. The second cluster could also be isolated as an n-tetrabutylammonium salt and has been thoroughly investigated by ESI-MS studies providing direct information on the protonation state of the overall system.

This thesis discusses the synthesis of transition metal chalcogenide nanostructures (where the chalcogen is either sulfur, selenium or tellurium) through the use of standard chemical vapour transport (CVT) and chemical vapour deposition (CVD) techniques. The resultant structures are characterised with a variety of methods and comparisons of their properties are made with their bulk counterparts. A discussion into how some of these structures form during the reaction is also given. Highly symmetrical, isotropic, nickel disulfide (NiS2) nanocubes have been synthesised via a Physical Vapour Transport (PVT) method in which sulfur vapour generated in situ is reacted with nickel-coated silica substrates. Systematic studies demonstrate the effect of the reactant ratio, substrate, metal layer thickness and reaction temperature on the synthesis and growth process. The evolution of structure and composition has been followed by diffraction and scanning electron microscopy (SEM). The size of the NiS2 cubes can be varied from below 200 nm to 1 -2 1m across. Magnetic properties of the disulfide nanomaterials have been determined using superconducting quantum interference device (SQUID) magnetometry. Initial experiments also demonstrate that related CVT techniques can be exploited to produce alternative compositions in the Ni-S system with varying morphologies that can be controlled via chemical and physical reaction parameters. Surface Assisted Chemical Vapour Transport (SACVT) methods have been employed to grow flower-like nanostructures of titanium disulfide (TiS2) and titanium trisulfide (TiS3) on titanium coated silica substrates. Systematic studies demonstrate the role of the reactant ratio and reaction temperature on the synthesis and growth process. The evolution of structure and composition has been followed by powder X-ray diffraction (PXD) and electron microscopy techniques such as and transmission electron microscopy (TEM). Magnetic properties of the disulfide nanomaterials have been determined using SQUID and Raman spectroscopy has been used to confirm the identity of the sulfides. Investigations into nanostructured materials of the group IV transition metals zirconium and hafnium resulted in the successful synthesis of nanostructures of zirconium trisulfide/selenide (ZrS3/Se3) and hafnium trisulfide/selenide (HfS3/Se3). The unusual effects on structure that can occur when reactant time and synthesis temperature are varied and when a balance between these two factors is successfully found, nanostructures other than tubes and wires can be formed. Each of these systems were characterised with a variety of techniques including, TEM, PXD and SQUID.

Due to intrinsic limitations of conventional silicon based devices the trend of miniaturisation cannot continue indefinitely, thus molecular devices are being used to develop smaller, faster and higher storage density memory devices. We present a thermally activated, switchable hetero-polyoxometalate (HPOM) cluster immobilised on a highly polarisable gold surface which has potential as such a device. This cluster consists of a nanometre sized Mo(IV) oxide “shell” which encapsulates two electronically active pyramidal sulfite (SIVO32-) groups, and has the ability to reversibly interconvert between two electronic states. In the passive state, at cryogenic temperatures (77 K), the two SO32- groups are non-bonding with respect to the sulfur centres, however upon thermal activation, i.e. when the temperature is increased to room (298 K), two electrons are ejected from the active sulfite anions and delocalised over the metal oxide cluster cage. This has the effect of switching it from a fully oxidised to a two-electron reduced state, along with the concomitant formation of an S-S bonding interaction between the two sulfur centres inside the cluster shell. This process does not occur in the crystalline state and to proceed requires the stabilising effects provided by an image charge, generated as a consequence of being adsorbed onto a metal surface. The prototypical enantio-selective heterogeneously catalysed reaction involves the hydrogenation of the α-ketoester, methyl pyruvate on Pt. Using TPD, XPS and UPS we have investigated this compound’s behaviour on a model Cu(111) single crystal surface. Monolayers of methyl pyruvate at 180 K consist predominately (ca. 66%) of a chemisorbed methyl pyruvate moiety, with its keto-carbonyl bonded to the surface in a η2configuration, this moiety desorbs intact at 364 K. The rest of the monolayer contains weakly adsorbed methyl pyruvate, which desorbs at 234 K, and interacts with the surface through the lone pair electrons of the oxygen atoms of the C=O groups, adopting a η1 configuration. The observation of a strongly chemisorbed moiety in the present study is attributed to the activation of the keto-carbonyl by the electron withdrawing ester group, and is consistent with the homogeneous inorganic chemistry of ketones. It is widely assumed that the α-ketoester needs to be π-bonded to the surface for the enantio-selective hydrogenation to proceed, consequently, given both the formation of a η2 bonded methyl pyruvate moiety on Cu(l11) and the known activity of Cu as a selective hydrogenation catalyst, it is suggested that it is maybe worthwhile considering the possibility of testing the effectiveness of chirally modified supported Cu as an enantio-selective catalyst. The thermal and electron induced chemistry of (S)- and (R)-methyl lactate (MLac) on Cu(111) was investigated; both enantiomers exhibited similar behaviour. MLac adopts one of two adsorption modes on the terraces of a Cu(111) crystal, which desorb molecularly at 209 K and 220 K. Concerning the molecules adsorbed at defect sites, as the temperature is increased over the range 250 – 300 K, a fraction desorb intact, while the majority lose a hydrogen atom to form the more strongly bound alkoxy species on the surface. Of these, some recombine with the hydrogen and proceed to desorb as MLac at 360 K, while a larger proportion are dehydrogenated further and methyl pyruvate and hydrogen are ejected from the surface at 380 K. When a monolayer of MLac is irradiated with a low energy electron beam, the molecules at the terrace sites are electronically excited and desorb as intact molecules, while those at the defect sites undergo electron induced hydroxyl O-H bond cleavage. Subsequent to electron bombardment there is consequently a decrease in molecularly adsorbed MLac and an increase in the number of strongly bound alkoxy species on the surface, entities which are not susceptible to ESD. We believe the ESD excitation mechanism is dissociative electron attachment. Low energy electrons of <1 eV are prevalent in the secondary electron background and can excite the hydroxyl O-H stretch, facilitating its cleavage at a threshold of 1.4 + 0.7 eV. The cross sections for the electron induced processes are high, 3.0 + 0.4 x 10-16 cm2 for 50 eV electrons, thus MLac is extremely susceptible to electron stimulated desorption. The enantio-specific adsorption of both the (S)- and (R)- enantiomers of methyl lactate on the chiral Cu(643)R surface has been investigated. The results from the (111) surface enabled us to assign the features in the TPD profiles. The peaks arising from molecular desorption at terrace and step sites occurred at the same temperature for both enantiomers, however, those attributed to desorption from the kink sites differed by 13 K, representing an enantio-specific difference in desorption energies of 0.94 kcal mol-1. This value is significantly larger than those observed in previous experimental work, although it is consistent with theoretical studies. Furthermore, we also observed enantio-specific surface reactions. It was found that there was a greater tendency for the (R)- enantiomer to undergo both the alkoxide recombination reaction and further dehydrogenation to methyl pyruvate, while the (S)-enantiomer had a greater proclivity to undergo total decomposition. We have discovered, to the best of our knowledge, the first example of enantio-specific surface chemistry initiated by a beam of non-chiral low energy electrons. When (S)- and (R)-methyl lactate molecularly adsorbed at the chiral kink sites of a Cu(643)R substrate is irradiated with 50 eV electrons, it has been found that (R)-methyl lactate is more receptive to both electron induced desorption of the parent molecule and electron induced cleavage of the hydroxyl O-H bond. This behaviour has been attributed to the (S)-enantiomer forming a more intimate bond with the kink site than the (R)-enantiomer, as evidenced by its higher desorption temperature. Consequently the substrate is more effective at providing relaxation channels to the electronically excited adsorbate, which reduces the probability of ESD occurring. Starting with a racemic mixture, we have demonstrated a 20% enantiomeric enrichment in the molecular adsorbates at the chiral kink sites, after only 30% depletion of the initial population. As a control, the initial rates of desorption from terrace and step sites were found to be unaffected by enantiomeric identity, which was to be expected because these sites are achiral, and as such both enantiomers interact to a similar degree with each. When the monolayer is considered as a whole, it was found that electron irradiation drives desorption more completely with an (R)-MLac covered surface than with (S). It has been suggested that this property of the system could be exploited in the laboratory as a method for separating racemic mixtures, and that in an astrochemical context, it could provide insight into the origins of biohomochirality.

Some purple bacteria species, such as Rhodopseudomonas palustris 2.1.6, produce light harvesting antenna (LH2) with unusual absorption spectra when they are grown under low-light intensities. This ability is often related to the presence of multiple genes encoding the LH2 apoproteins. This thesis describes isolation of pure stable LH2s from Rhodopseudomonas palustris 2.1.6 grown at different light intensities, determination of the polypeptide composition of high- (HL) and low-light (LL) LH2 complexes and characterisation their spectroscopic properties using various optical spectroscopies. The question of whether rings with a heterogeneous apoprotein composition exist has been addressed by single-molecule spectroscopy. For the first time, direct evidence that individual LL LH2 complexes have a heterogeneous αβ-apoprotein composition has been found. Such mixed rings feature Bchl a molecules with both B820-like and B850-like site-energies. This finding was supported by a femtosecond study on the energy transfer reactions and exciton relaxations within both HL and LL LH2 complexes. This thesis also describes attempts to crystallise the HL and LL LH2s. Even though three-dimensional crystals of both HL and LL LH2 complexes only diffracted to low resolution, it was possible to use molecular replacement to obtain structures that suggest both these types of LH2s are nonamers.

The main focus of this research was to examine the molecular complexes formed by the proton sponge material DMAN [1,8-bis(dimethylamino)naphthalene]. It is well known that DMAN will readily protonate – forming the DMANH+ – leaving the second molecule in a DMAN complex negatively charged. In addition to the preparation of materials by crystallisation techniques, full characterisation was carried out using X-ray diffraction of the single crystal samples obtained. Neutron diffraction was also carried out on some of the samples for which the hydrogen atom information was most important to obtain, though this was restricted by the amount of beamtime available on the central facilities used. Several families of DMAN complexes were studied, including DMAN : benzoic acid (chapter 4), DMAN with a wide series of halo-substituted benzoic acids (Chapter 4), DMAN with hydroxyl-, nitro- and methyl- substituted benzoic acids (Chapter 5), and its complexes with a selection of other organic acids (Chapter 6). Several general patterns are found in this work, including the following: • The DMAN is usually protonated (DMANH+), with the hydrogen atom held in a short N−H···N intramolecular hydrogen bond, as observed previously; • The carboxylic acids tend to form dimers, but with a carboxyl and a deprotonated carboxylate group involved, these DIMER− units are formed through a single, very short, strong, charge-assisted O−H···O hydrogen bond; • a wide range of C−H···O and C−H···π weaker hydrogen bonds are present, together with consistent stacking of the DMAN rings, including π···π interactions, but with additional stacking at a range of interplanar distances; • the overall packing in several complexes includes chains or sheets of DIMER− units with DMANH+ included in pockets in the structure. There are some exceptions to these patterns, for example: • varying stoichiometry, resulting in the first example of the occurrence of both protonated (DMANH+) and unprotonated DMAN in a single complex • the inclusion of solvent water in some hydrated complex, which interrupts some of the patterns of intermolecular interactions In addition to the studies of DMAN, a ternary hydrogen bonded complex was studied using X-ray and neutron diffraction and the new molecular porous material 4-phenoxyphenol, containing a range of solvents, was fully characterised using these diffraction techniques as well as the thermal methods TGA and DSC.

Citrus fruits are of special interest as they accumulate large amounts of flavonoids and are consumed in substantial quantities worldwide. Despite extensive research on citrus flavonoids, many compounds remain unidentified in tropical citrus species. High performance liquid chromatography with mass spectrometry and electrospray ionisation (HPLC-MS–ESI) has proved to be a powerful tool for flavonoid characterisation. This study describes how HPLC-MS-ESI and HPLC with a photodiode array detector (PDA) was used to identify and quantify flavonoids in the tropical species Citrus microcarpa, Citrus hystrix, Citrus medica var. 1 and 2, Citrus suhuiensis and Citrus medica var. sarcodactylis. Among the major compounds detected were apigenin-6-8-di-C-glucopyranoside, apigenin-8-C-glucoside-2′-rhamnoside, phloretin-3′,5′-di-C-glucopyranoside, diosmetin-7-O-neohesperidoside, hesperetin-7-O-rutinoside, diosmetin-7-O-rutinoside and hesperetin-7-O-neohesperidoside. Most of the C-glycosylated flavones and dihydrochalcone have not been found previously in tropical citrus. C. microcarpa contained a high amount of phloretin-3′,5′-di-C-glucopyranoside that was shown to possess a high Trolox Equivalent Antioxidant Ratio (TEAR) value due to its 2,4,6-trihydroxyacetophenone structure. In general, most of tropical citrus flavanones were neohesperidoside conjugates, which are responsible for the bitter taste of the fruits and juices. Tropical citrus essential oils were extracted by steam distillation (SD) and simultaneous distillation extraction (SDE) and analysed using gas chromatography-mass spectrometry (GC-MS). More than 40 compounds were identified and the major component in almost all citrus fruits and leaves was R–(+)-limonene. Citronellal was detected and was highest in C. hystrix leaf (87.8%). The radical-scavenging activity of each oil was assessed using the 1,1-diphenyl-2-picrylhydrazyl (DPPH) method. Essential oils obtained from all citrus fruits showed very high radical-scavenging activity against the DPPH radical, well in excess that in leaves. The antioxidant activities of standard flavour compounds were also determined. The antioxidant activities of essential oils of selected tropical citrus fruits were attributed mainly to the presence of gamma-terpinene and terpinolene. These compounds showed high free radical-scavenging activity against DPPH radicals and were more active than vitamin E analogue. The colonic bacterial contribution to the catabolism of the flavanoids in a range of beverages was tested by comparing the urine profile of intact adults and adults with ileostomies and also using in vitro fermentation studies. Analysis of catabolic products i.e. free phenolic acids in human urine and faecal slurries following ingestion and supplementation of green tea and green tea catechins was conducted using GC-MS. Pyrogallol, pyrocatechol, 4-hydroxyphenylacetic acid, (-)5-(3',4'-dihydroxyphenyl)-gamma-valerolactone, (-)-5-(3',4',5'-trihydroxyphenyl)-gamma-valerolactone, 3-(3-hydroxyphenyl)propionic acid and 5-(3,4-dihydroxyphenyl)-gamma-valeric acid were among catabolites produced during a 4-48 h fermentation of green tea catechins. In total, 58.4% of (-)-epigallocatechin gallate, 25.2% of (-)-epigallocatechin and 50.1% of (-)-epicatechin were catabolised to phenolic acids in the human faecal slurries in the presence of glucose. Hippuric acid and 4-hydroxyphenylacetic acid were found to be abundant after drinking green tea, but did not appear to be major phenolic acids produced from the catabolism of green tea catechins. Calculation based on the levels of selected phenolic acids excreted in urine of healthy volunteers after drinking green tea indicated excretion was equivalent to 23.8% of flavan-3-ols intake. The mean 0-24 h excretion of 163 ± 48 μmole phenolic acid for healthy volunteers who drank green tea was significantly higher (p<0.05) than the excretion of phenolic acids by volunteers who drank water (20 ± 3 μmole). Analysis of phenolic acids in brewed green tea was also carried out in this study. Gallic acid and p-coumaric acid were by far the most abundant phenolic acids present both in free and conjugate forms, but the amounts were small compared to the quantities of phenolic acids excreted in urine 0-24 h after drinking green tea. This study also describes the pharmacokinetics study of urinary excretion of phenolic acids in human volunteers (with and without a colon) after drinking instant coffee. All volunteers displayed a substantial increase in excretion of 3-(3-hydroxyphenyl)-3-hydroxypropionic acid, dihydroferulic acid, ferulic acid and 3-hydroxyhippuric acid 8-24 h after drinking coffee. The mean amount of urinary phenolic acids excreted after drinking coffee was 134 ± 43 μmole, which was equivalent to 29.4% of chlorogenic acid intake and significantly higher (p<0.05) than phenolic acids excreted when water was consumed instead coffee. The influence of the food matrixes (yoghurt and milk) on the catabolism of polyphenols is described in this study. 3-Methoxy 4-hydroxyphenylhydracrylic acid, 3-(3-hydroxyphenyl)-3-hydroxypropionic acid, hippuric acid, 3-hydroxyhippuric acid, 3-dihydroxyphenylacetic acid and dihydroferulic acid were found in human urine after drinking orange juice enriched with hesperetin-7-O-rutinoside. The total 0-24 h urinary excretion of flavanone-derived phenolic acids increased by a statistically significant nine-fold (p<0.05) following ingestion of orange juice. After consumption of orange juice with yoghurt, only small amounts of phenolic acids were excreted. A study on the effect of milk on the catabolism of cocoa flavan-3-ols showed low concentrations of phenolic acids were found in human urine after drinking hot cocoa and phenolic acid excretion was suppressed in healthy volunteers after drinking cocoa with milk. On the other hand, studies on cocoa polyphenol catabolism using human faecal slurries revealed breakdown to phenolic acids. This may indicate that in vivo cocoa polyphenols may form complexes with proteins, which reduces the extent to which they are degraded to phenolic acids when they reach the colon.

The mechanism of strand exchange by Tn3 resolvase was studied using FRET based methods. For this purpose Cy3 dye and a Cy5 dye were attached to modified thymines within Tn3 res site I, the DNA substrate of Tn3 resolvase. The dyes formed a FRET pair and followed the movements of the substrates, resulting in changes of the distance between both dyes and therefore changes in the FRET efficiency of the FRET pair which were monitored using fluorescence spectroscopy. A library of short, double-stranded substrates containing one copy of Tn3 res site I and Cy3 and/or Cy5 dyes attached to at different positions within site I was generated. Fluorescent substrates which do not interfere with the formation of synapses and with the recombination process were selected using gel-based assays. Fluorescent dyes attached at positions about five bases from the centre of site I were found to allow uninhibited recombination. Short double-stranded substrates with dyes attached at the selected positions within site I were studied in ensemble FRET experiments. Recombination of substrates containing a FRET pair with an excess of non-fluorescent substrates resulted in a strong decrease of the FRET efficiency due to the formation of recombinant products containing one dye each. This observation suggested that recombination and strand exchange could indeed be studied using FRET-based methods. The ensemble FRET experiment and FRET based experiments showed that the short substrates are usually recombined in both parallel and anti-parallel orientation. The lack of control over the orientation of the substrates in the synapses motivated the development of U-shaped substrates which consisted of two double-stranded arms, each containing one copy of site I, and a single stranded linker which connected both arms. In gel based assays, the U-shaped substrates were found to prefer the intramolecular recombination of both sites in one defined orientation. A U-shaped substrate containing a Cy5 and a Cy3 dye was studied in a single-molecule FRET experiment. In the presence of an activated Tn3 resolvase mutant, several specific FRET states and transitions between FRET states had been observed. The FRET states and transitions differed in the presence and absence of MgCl2 allowing the identification of a FRET state transition potentially corresponding to the conformational changes during ligation and several transitions corresponding to intermediate steps during strand exchange.