Whole document restricted, see Access Instructions file below for details of how to access the print copy. / The literature on polymer gels and hydrogels since the beginning of last century has been reviewed, and the properties of polymers and gels based on 2-acrylamido-2-methylpropanesulphonic acid (AMPS) summarized. Copolymer gels were synthesised, using gamma radiation initiation, from aqueous solutions of the comonomers acrylamide and AMPS. The progress of copolymerisation and crosslinking was followed by measuring the viscosity of reaction mixtures that had been irradiated for different times and hence had received different doses. The variation of viscosity with dose has been interpreted in terms of main chain growth and the onset of crosslinking and gelation. It was observed that the viscosity of irradiated reaction mixtures decreased substantially on standing at ambient temperature. That behaviour has been attributed to breaking of non-permanent bonds formed during irradiation and subsequently broken over the ageing period. The environmental responsiveness of the gels formed by gamma radiation initiation was assessed by observation of the volume phase transition triggered by increasing the proportion of acetone in aqueous acetone media in which the gels were immersed. Free-radical copolymerisation of aqueous solutions of comonomer AMPS and Nisopropylacrylamide (NIPA) was initiated with ammonium persulphate at 70°C. Reaction times that varied with comonomer mixture composition gave a series of copolymers corresponding to low conversion. The copolymer composition data set was used in conjunction with the Kelen-Tϋdos and Finemann-Ross methods, assuming the terminal model to be valid, to give the monomer reactivity ratios r(AMPS)=0.24 and r(NIPA)=1.12. The reactivity ratios show that copolymers made from comonomer mixtures with more than 25 mol% AMPS consist of relatively short monomer sequences. The dependence of glass transition temperature on copolymer composition was determined using differential scanning calorimetry and the experimental data compared with Tg predicted from the Fox equation. High-field (400 MHz) 1H-NMR spectroscopy was used to probe the early stages of polymerisation and gelation at 70°C in a D2O solution of AMPS and NIPA, crosslinker N,N1-methylenebisacrylamide (MBAA) and poly(ethylene glycol) (PEG). The change in signal intensity of vinyl protons of the monomers and crosslinkers, and the methylene proton signals from PEG were recorded as a function of reaction time. The relative initial rates of reactant consumption showed that the crosslinking monomer was consumed more rapidly than either AMPS or NIPA, so that crosslinking of the copolymer chains took place predominantly in the early stages of copolymer chain growth. The PEG signal was found to be invariant with time, confirming that PEG did not participate in the polymerization reaction but acted as the linear interpenetrating component in semi-interpenetrating network gels. The kinetics of gel formation and other rheological properties of copolymer and SIPN pregel solutions were analysed using parallel plate rheometry. The variation of shear moduli and other rheological parameters during reaction under dynamic shear was interpreted in terms of chain growth and crosslinking reactions. In relation to the sol-gel transition, viscosity, tan(δ), G' and G" were used to assess the kinetics of gelation. The densities of copolymer and SIPN dried gels were measured for a range of copolymer chain compositions and found to be similar in magnitude (within about 10%) for gels with the same copolymer chain composition. The surface morphology of both types of gel, in dried form, was examined using scanning electron microscopy. The swelling ratio of SIPN gels in water was found, in general, to be proportional to the AMPS content of the copolymer chains. The influence of initiator and crosslinker concentrations in the pregel solutions on the rheological properties of gels was investigated in detail. G', G", tan(δ) and critical strain were correlated to initiator concentration, but the crosslink density of the gel was essentially independent of initiator concentration. As expected, the effect of crosslinker concentration, hence crosslink density in the gels, on rheological properties, was significant. With increasing crosslinker concentration, the critical strain (at gel rupture) and tan(δ) decreased markedly. The surface roughness of dried gels also decreased with increasing crosslinker concentration. Effects of comonomer mixture and copolymer chain composition for copolymer and SIPN systems on the initiation temperature of pregel solutions and rheological properties of gels were compared. A feature of the data is that the copolymer gels exhibited much larger values of G' and G" than the corresponding SIPN gels, whereas the SIPN gels had larger equilibrium storage modulus (G'e), indicating that stronger covalent or physical crosslinks may exist in the SIPN gels. Copolymer and SIPN gels examined at constant frequency and shear stress exhibited a phase transition associated with the water in the gels. Both types of gels were found to have an ice-dominant structure at -6°C with G' smaller than that of ice at -5°C, and there was an ice phase to gel phase transition induced by shear strain at -6°C, which was not found in gels at -4°C. Temperature-time effects on gel formation kinetics and post-gel properties were determined and appear to be correlated to rates of gel formation as well as the final gel microstructure. As expected, the induction time for gel formation is greater at lower reaction temperature, and PEG enhances this effect and reduces the viscosity compared to the corresponding PEG-free system. The post gel microstructure has been examined in terms of G', angular frequency, reaction temperature and time. The most significant effect is that increasing reaction temperature for a fixed time builds a more complete gel network. Creep and stress relaxation of gels were investigated. Both copolymer and SIPN gels show a high degree of elasticity, and at specific frequencies the gels behave as ideal elastic solids. SIPN and copolymer gels were characterized using spectroscopic, thermal and morphological techniques (DSC, TG, FTIR, XPS, XRD and SEM). Interactions between PEG and copolymer network were found to markedly influence the properties of SIPNs. Thermal analysis revealed that the PEG component of the SIPNs showed melting behaviour that was similar to that of pure PEG, but modified by interactions between PEG and the copolymer network. Composition and temperature effects on thermal and morphological properties were investigated in detail. Protonation of a substantial proportion of the amide groups in the dried SIPN gets was detected by X-ray photoelectron spectroscopy, and mainly attributed to proton transfer from the sulphonic acid groups of AMPS units to the amide groups of NIPA units. Sodium diclofenac was used as a model drug in design of a drug encapsulation/release system based on use of PEG/poly(AMPS-co-NIPA) SIPN hydrogels. The thermal stability of the drug and the SIPN gel used for its encapsulation was determined as a prerequisite to the development of a direct drug loading technique whereby the drug was encapsulated during accelerated formation of the SIPN matrix in a suspension polymerisation system at 100°C. Direct loading of the drug was found to be superior, in terms of the distribution of the drug in the gel matrix, to the conventional diffusion loading technique. Drug release in pH7.4 phosphate buffer solutions and in pHl.0 simulated gastrointestinal liquid was determined as a function of time.

The work has been divided and abstracted under three headings: (1) The Absorption of Sulphur Dioxide by Aqueous Solution. The rate of absorption of low concentrations of SO2 (~ 500 µg.m-3) was examined, varying a number of parameters. The results were expressed as deposition velocities (Vg = flux/gas concentration). Vg did not vary with relative humidity. Deposition into acid solutions (pH <3) was controlled by diffusion processes in the liquid film while under more alkaline conditions (pH >4) the process was governed by the rate at which SO2 could diffuse to the interface through the gas boundary layer. Oxidizing agents (e.g. H2O2) were found to increase the rate of diffusion in the liquid boundary layer. Salts generally exerted a more complex effect on liquid phase diffusion. Vg for SO2 deposition into aqueous solution under calm conditions and at pH values expected in the atmosphere was calculated to be about 0.7 cm.s-1 from these experiments. Efflux of SO2 from solution was examined at low pH values under boundary conditions similar to the deposition work. At high S(IV) concentrations in solution (~10-3N) the half lives for SO2 desorption from solution were 600, 1020, 3300 s for pH values 2.0, 2.5, 3.0 respectively. The consequences of the experimental results are discussed with respect to the atmosphere at large and to scrubbing of flue gases. Poor absorption rates for SO2 at low pH values suggested that sea water with its high pH would be an excellent scrubbings agent and its buffering ability was investigated. Experiments showed 10-3M S(IV) could be absorbed with the pH dropping only as far as pH 5.5. (2) The Oxidation of Sulphur Dioxide in Aqueous Solution The rate of oxidation in 10-5M of SO2 solutions was examined. The results showed the reaction to be very sensitive to traces of metallic ions, which made it impossible to determine the uncatalysed rate. The half life for oxidation of S(IV) in (10-5M) in triply distilled water was 36, 8, 5 hours at pH 4.6, 5.2, and 6 respectively. The rate increased upon addition of traces (10-6M) of Fe(III) and Mn(II) ions (pH 4 – 5) while Cu(II) seemed to act as an inhibitor. In solutions of ammonium sulphate Fe(III) was found to be an excellent catalyst. The reaction appeared to have a variable order with respect to S(IV) being 1st order at pH <4 and 2nd order at pH >5. Results in potassium sulphate were similar, but the reactions somewhat slower. At pH >5 the reaction was sensitive to the alkali used for neutralization, the oxidation rate being much greater if ammonium hydroxide was used rather than sodium hydroxide. Solutions of ammonium chloride and sodium bicarbonate showed a second order dependence on S(IV) concentration at high pH values. The Fe(III) catalysed oxidation was faster in 'Chloride' solutions than in 'sulphate' solutions at pH <4. A radical mechanism previously proposed by a number of workers was found to give considerable insight into the complex experimental results. (3) The Dissolution of Iron from Ferric Oxide and Pulverized Fuel Ash The aqueous oxidation of SO2 requires the presence of a catalytic metal ion. The most abundant catalyst in the atmosphere would appear to be iron. However, it would have to be present in solution to be an effective homogeneous catalyst. The rates of dissolution of iron from Fe2O3 were investigated and found to be extremely slow even in the presence of H2SO4 or 'H2SO3'. The dissolution of iron from pulverised fuel ash (a possible atmospheric source of iron in polluted areas) is quite rapid even under only Moderately acid conditions (pH 3.5) where solutions 1.9 x 10-4 M in Fe(III) can be reached in a matter of 20 minutes. The high rate of dissolution is thought to arise from the fusing of iron oxides with other alkaline oxides in the furnace.

This thesis is concerned with the synthesis, and aspects of the chemistry, of carbon-donor complexes of osmium as organo-metallic models for CO reduction intermediates. In Chapter 1 some aspects of ligand reactivity of the carbon-donor ligands CO, CS, CSe, CTe and carbenes in transition metal complexes are reviewed. The reduction reactions of these ligands are emphasized. The preparation and structure of the osmium η2-formaldehyde complex Os(η2-CH2O)(CO)2(PPh3)2 is described in Chapter 2. This complex has proved to be a useful synthetic precursor for stable osmium formyl, hydroxymethyl, methoxymethyl and halomethyl (-CH2X, X = Cl, Br, I) complexes and some facets of the reactivity of these ligands have been investigated. A general synthetic route to neutral osmium formyl complexes Os(CHO)X(CO)2(PPh3)2(X = halide or alkyl) has been developed. The facile preparation of a stable example of an intermediate formed during decarbonylation of a simple aldehyde by a transition metal, the osmium monohapto-acetyl-hydrido complex Os(η1-C[O]CH3)H(CO)2(PPh3)2, has been demonstrated. A preliminary study of the reactions of the osmium iodomethyl complex Os(CH2I)I(CO)2(PPh3)2 is reported in Chapter 3. The typical reaction of this species is nucleophilic substitution; in many respects reactivity is analogous to an electrophilic methylidene complex. Reaction with a variety of nucleophiles [e.g. OR-, H-, EH- (E = S, Se, Te), NH2R and PR3] has been investigated and the ligand reactivity of some of these derivatives studied. A synthetic route to η2-CSeS and η2-CSe2 complexes of osmium without the use of molecular CSeS or CSe2 has been developed and the isolation of the geometrical isomers of the η2-CSeS complex Os(η2-CSeS)(CO)(CNR)(PPh3)2 (arising from η2-C,S or η2-C,Se coordination) has been achieved. The synthesis of a stable osmium hydrido-selenocarbonyl complex, OsHCl(CO)(CSe) (PPh3)2, has allowed the direct observation of hydride transfer from metal to CSe ligand. These latter results are discussed in Chapter 4.

The kinetics and mechanism of the anodic process on lead in aqueous 35% H2SO4 (4.65 mol L-1) were studied. The lead anode is of interest as it is used in the form of the sponge lead negative plate in the lead-acid battery. Discharge of both planar and porous lead electrodes result in the electrocrystallisation of PbSO4. In the past, the processes occurring at the lead electrode have been shown to include: dissolution, precipitation, and passivation. The techniques used were cyclic voltammetry, potential-step, and potential-ramp/potential-hold. The resultant transients were recorded both as analogue and digital waveforms. Investigations were carried out at a constant acid strength over a range of temperatures (-18° to 30°C), and in the presence of additives. Significant variation was found in cyclic voltammetry with regard to the cathodic potential prehistory immediately prior to measurements. After extreme cathodic polarisation (below -1500 mV us SHE), voltammograms were recorded with relatively high anodic peak currents and charges. The extra charge was associated with the growth of a multilayer, and dissolution from the bare electrode. This multilayer is not reduced on subsequent moderate cathodic excursions (above -1000 mV). Hence for cyclic voltammograms in the absence of extreme cathodic polarisation one observes relatively small currents and anodic charges. This is due to transmission of Pb2+ ions through the multilayer under resistance control. Passivation occurs in both cases by precipitation of microcrystalline PbSO4 on the multilayer. In contrast to cyclic voltammetry, potential-step transients were insensitive to cathodic polarisation and displayed current peaks due to growth of the surface film. One also observes a dissolution current from the bare electrode before it is covered by the growing film. The current first increases, passes through a maximum, and then decays with time. The increasing current was associated with growth of the multilayer, and the decaying current due to passivation by PbSO4. A new model was successfully applied to the potential-step and potential sweep results. The model consists of several charge-transfer reactions: growth of a multilayer, dissolution from the bare electrode, and transmission of Pb2+ through the multilayer. The kinetics of the multilayer are pivotal to the other charge-transfer reactions. The inclusion of film transmission is the novel aspect of this model. Quantitative analysis of the model resulted in a set of optimised parameters that follow plausible variation with anodic potential. Both the potential step and potential sweep transients are adequately described by the model. The effect of chloride added to the electrolyte was studied. The anodic and cathodic peaks in cyclic voltammetry are enhanced by this anion (100% more anodic charge at higher chloride concentrations) and the cathodic to anodic charge ratio is markedly increased. The general form of the potential-step transients are not modified by chloride. Evidence of a monolayer of PbCl2 was found (q = 500 µC cm-2) in both cyclic voltammetry and potential step experiments. The monolayer of PbCl2 must underlie the subsequent growth of the multilayer as the monolayer is first formed in potential-step experiments and first to be reduced in cyclic voltammetry at a small underpotential. The effect of methyl orange on the lead anode was explored and found to enhance the charge capacity. Battery tests confirmed these observations, but methyl orange was destroyed by the oxidising PbO2 at the positive plate. Hence a derivative, designed to be insoluble in H2SO4, was synthesised. However, experiments on both planer and porous electrodes showed that the derivative, lauryl orange, C12H25NHC6H4NNH+C6H4SO3- (pKa = 4.06) was found to act as an inhibitor for the anodic reactions. In particular there was no contribution from multilayer growth in potential step experiments, and only dissolution (markedly diminished) and passivation were evident. Lauryl orange exhibited the required capabilities to bind to lead metal. The structure of lauryl orange was confirmed by n.m.r. and mass spectrometry.

The synthesis of novel C9-halogenated anthracyclinones by Lewis acid promoted cyclisations of ortho-methallyl anthraquinonyl chiral dioxanes has been investigated. Tin tetrachloride-N,N-dimethylformamide promoted cyclisation of the dioxane (30) proceeded with excellent stereoselectivity to give an 82% yield of the diastereomerically pure chlorotetracycle (78). Stronger Lewis acids gave poorer selectivity and boron trifluoride etherate reacted slowly with (30) but with high selectivity to give the diastereomerically pure fluorotetracycle (91) in 58% yield. The presence of a β-methoxy group in the substrate leads to decreased stereoselectivity due to competition between chelation and non-chelation controlled pathways. Stereochemistry was assigned to the products using a combination of X-ray crystallography, conformational analysis, nuclear Overhauser enhancements and chemical degradation. Attempts were made to extend the highly selective acetal cyclisation methodology to the intramolecular reaction of the allylsilane (121), cyclisation of which was expected to provide enantioselective access to the alkene (146). The synthesis of (121) was achieved despite its instability, but attempted cyclisations with Lewis acids or fluoride ions were unsuccessful. The racemic alkene (146) was prepared by an ene cyclisation and subjected to asymmetric dihydroxylation (AD). The products from the AD reactions are of considerable biological interest and have been characterized. Their stereochemistry has been assigned by 1H nmr comparisons with the parent diphenols, which have resulted in revisions of assignments made by earlier workers.

This thesis contains two parts. The first part describes the synthesis of orthogonally protected L-2,3-cis-3,4-trans-DHP 1.44 in eleven steps starting from L-arabinose via double displacement of bis-mesylate 2,8. The second part describes the synthesis of the spiroacetal fragment of the alkaloid broussonetine H. (±)-2-(3-Bromopropyl)-1,7-dioxaspiro[5.5]undecane (3.38) was prepared in nine steps from 1,4- butanediol (1.49), propargyl bromide (3.33) and δ-valerolactone (1.51). Enantioenriched material, namely (2S)-2-(3-bromopropyl)-1,7-dioxaspiro[5.5]undecane (4.2) was synthesized in thirteen steps from (2S)-pentane-1,2,5-triol (4.14), trimethylsilylacetylene (4.22) and δ-valerolactone (1.51). The enantiomeric excess of the derived alcohol, (4S)-3-(1,7-dioxaspiro[5.5]undec-2-yl)propan-l-ol (4.28) was measured by its conversion into a Mosher ester derivative 4.28. This thesis also describes the progress towards the synthesis of the dihydroxypyrrolidine fragment 5.28 of broussonetine H from L-sorbose (1.48).

The chemistry of transition metal silyl complexes still remains relatively undeveloped despite the recent advances that have been made in this area. This thesis describes the synthesis and reactivity of new ruthenium and osmium silyl complexes. A general survey on the bonding, preparation and reactivity of transition metal silyl complexes is given in chapter one as an introduction to the chemistry discussed in the following chapters. Several reviews relating to transition metal-silicon chemistry have been published, 1-5 and therefore only the major features and more recent developments are highlighted in this overview. Emphasis has been placed on those aspects not previously reviewed. Coordinatively unsaturated transition metal silyl complexes are uncommon and the chemistry of these complexes has been virtually unexplored. The synthesis of the novel coordinatively unsaturated ruthenium and osmium silyl complexes, M(SiR3)Cl(CO)(PPh3)2 (M=Ru, Os; R=Cl, alkyl group, alkoxy group), is described in chapter two. The reaction of M(Ph)Cl(CO)(PPh3)2 (M=Ru, Os) with a silane, in some instances, provided a facile, high yield route to complexes of the type M(SiR3)Cl(CO)(PPh3)2 (M=Ru; SiR3=SiMe3, SiEt3, SiCl3, SiMe2Cl; M=Os; SiR3=SiCl3, SiMe2Cl). The chlorosily1 complexes were also prepared by treatment of MHCl(CO)(PPh3)3 (M=Ru, Os) with HSiCl3 and HSiMe2Cl. X-ray crystal structures were obtained of Ru(SiEt3)Cl(CO)(PPh3)2 and Os(SiCl3)Cl(CO)(PPh3)2. Complexes containing chlorosily1 ligands proved synthetically useful, undergoing a variety of nucleophilic substitution reactions at silicon with retention of the transition metal-silicon bond. New complexes synthesized by this method included Ru[Si(OEt)3]Cl(CO)(PPh3)2 and Os(SiMe3)Cl(CO)(PPh3)2. X-ray crystal structures of these two compounds were also obtained. These exchange reactions illustrate a route to new transition metal silyl compounds that has rarely been utilised. Also described in chapter two is the unprecedented formation of the novel compound Os(SiMe2C6H4PPh2)(C6H4PPh2)(CO)(PPh3), resulting from phenylation at silicon. The structure of this complex was determined by X-ray crystallography. In chapter three, addition of the Lewis bases CO and CN(p-tolyl) to coordinatively unsaturated silyl complexes, M(SiR3)Cl(CO)(PPh3)2 (M=Ru, Os), is described. The coordinatively saturated osmium silyl complexes Os(SiR3)Cl(CO)L(PPh3)2 (SiR3=SiCl3, SiMe2Cl, SiMe2OEt, SiMe3; L=CO, CNR) were generated in this way. Similarly, addition of 4,4'-dimethyl-2,2'-bipyridine to Ru(SiEt3)Cl(CO)(PPh3)2 afforded Ru(siEt3)Cl(dimethylbipy)(CO)(PPh3). The six-coordinate complex Ru(SiEt3)(η2-S2CNMe2)(CO)(PPh3)2 was synthesized by displacement of the labile chloride ligand from Ru(SiEt3)Cl(CO)(PPh3)2 by the dimethyldithiocarbamate anion. Characterization of Ru(SiEt3)(η2-S2CNMe2)(CO)(PPh3)2 included an X-ray crystallographic analysis. The exchange of silyl groups at the metal was also observed in several reactions. For example, heating Ru(SiMe3)Cl(CO)(PPh3)2 in the presence of excess HSiEt3 yielded Ru(SiEt3)Cl(CO)(PPh3)2. These reactions illustrate aspects of the chemistry that can occur at the metal centre of coordinatively unsaturated transition metal silyl complexes. Another route to coordinatively saturated osmium silyl complexes was via oxidative addition of a silane to Os(CO)2(PPh3)3, which yielded Os(SiR3)H(CO)2(PPh3)2 (SiR3=SiMe3, SiEt3, SiPh3, SiPh2H). The synthesis of these complexes is also discussed in chapter three. An X-ray crystal structure determination of Os(SiEt3)H(CO)2(PPh3)2 confirmed the presence of mutually trans carbonyl ligands. The first thiocarbonyl-containing transition metal silyl complexes, M(SiMe2Cl)Cl(CS)(PPh3)2 (M=Ru, Os), were prepared by treating M(Ph)Cl(CS)(PPh3)2 or MHCl(CS)(PPh3)3 with HSiMe2Cl. The Si-Cl bond in these compounds reacted readily with nucleophiles, yielding M(SiMe2OR)Cl(CS)(PPh3)2 (OR=OEt, OMe, OH). In contrast to the addition of CO to Os(SiR3)Cl(CO)(PPh3)2, addition of CO to Ru(SiMe2R)Cl(CS)(PPh3)2 (R=Cl, OEt, OMe, OH) and to Os(SiMe2OEt)Cl(CS)(PPh3)2 afforded dihapto-thioacyl complexes, M[η2-C(S)SiMe2R]Cl(CO)(PPh3)2, via a migratory insertion reaction involving the silyl group and the thiocarbonyl ligand. This reaction represents the first formal insertion of CS into a transition metal-silicon bond. The structure of Ru[η2-C(S)SiMe2OEt]Cl(CO)(PPh3)2 was obtained by X-ray crystallography and confirmed that bonding of the thioacyl ligand occurred in a dihapto fashion. These reactions are described in chapter four. When Os(Ph)Cl(CO)(PPh3)2 was reacted with HSiMe3, the formally osmium(IV) silyl complex Os(SiMe3)H3(CO)(PPh3)2 was produced. Few compounds of this type are known. The synthesis, characterization and reactivity of Os(SiMe3)H3(CO)(PPh3)2 are discussed in chapter five. The crystal structure of Os(SiMe3)H3(CO)(PPh3)2 is also depicted. In solution, Os(SiMe3)H3(CO)(PPh3)2 appeared to be in equilibrium with the highly reactive, coordinatively unsaturated species OsH2(CO)(PPh3)2. The reaction of Os(SiMe3)H3(CO)(PPh3)2 with HSiR3 (R=Et, Ph), HSn(p-tolyl)3 and HC2Ph was carried out, using Os(SiMe3)H3(CO)(PPh3)2 as an in situ source of OsH2(CO)(PPh3)2. These reactions afforded Os(SiR3)H3(CO)(PPh3)2 (R=Et, Ph), Os[Sn(p-tolyl)3]2H2(CO)(PPh3)2 and OsH(C2Ph)(CO)(PPh3)3 respectively, most likely via a series of oxidative addition-reductive elimination reactions involving OsH2(CO)(PPh3)2. Transition metal hydroxysilyl complexes are extremely rare. Only three systems have been reported containing a hydroxysilyl group bonded to a transition metal. 6,7,8 Chapter six deals with the formation of compounds of this type. Hydroxysilyl-containing complexes of ruthenium and osmium were obtained via the hydrolysis of M(SiMe2Cl)Cl(CO)(PPh3)2 and M(SiCl3)Cl(CO)(PPh3)2. The complexes M[Si(OH)3]Cl(CO)(PPh3)2 are the first trihydroxysilyl-containing transition metal complexes and therefore represent a new class of transition metal silyl compounds. Characterization of Os[Si(OH)3]Cl(CO)(PPh3)2 included an X-ray crystal structure which showed that, remarkably, no inter- or intra-molecular hydrogen bonding of the type O(H)···O or O(H) ··Cl was associated with the Si(OH)3 group. In contrast, intermolecular hydrogen bonding was found by X-ray crystallography for the dicarbonyl derivative, Os[Si(OH)3]Cl(CO)2(PPh3)2. Subsequent reactions involving Os[Si(OH)3]Cl(CO)(PPh3)2 led to the synthesis of the diosmium tetrahydroxydisiloxane complex, [OsCl(CO)(PPh3)2Si(OH)2-]2O. For example, [OsCl(CO)(PPh3)2Si(OH)2-]2O was isolated after Os(SiCl3)Cl(CO)(PPh3)2 was added to a dichloromethane solution of Os[Si(OH)3]Cl(CO)(PPh3)2. Formation of this tetrahydroxydisiloxane compound is significant and models the first condensation reaction generating a Si-O-Si linkage in the hydrolysis of organochlorosilanes to polysiloxanes. The Si-O-Si linkage was clearly visible in the X-ray crystal structure of [OsCl(CO)(PPh3)2Si(OH)2-]2O. Chapter seven provides a short conclusion, highlighting the important features of the work discussed in this thesis and identifies areas for future investigation.

Reductively triggered internal cyclisation reactions have been investigated as a prodrug system for the hypoxia selective release of aromatic nitrogen mustards. The observed pseudo-first-order rate coefficients of cyclisation of several model 2-aminoaryl-acetamides and propanamides have been measured. Cyclisation was observed to be strongly influenced by stereochemistry, whereas electron withdrawal from the amine-bearing ring resulted in a comparatively modest slowing of the rate of cyclisation. Protonation of the leaving group appeared to increase the rate of cyclisation, while changes in 4-substitution on the leaving amine had little effect on this rate. The cyclisation of 2-(2-aminophenyl) alkanamides was found to be subject to general catalysis by acidic buffer components, and rate determining formation of the tetrahedral intermediate has been proposed. Ring closure reactions of several 2-hydroxylaminophenylalkanamides have been studied by γ-radiolysis. HPLC methods have been developed for the separation of reduction and cyclisation products. Reduction stoichiometry implicates the hydroxylamine as the predominant reduction product of radiolysis of the 2-nitrophenylalkanamide precursors, which varied in the nature of substitution of the nitrobenzyl ring, 4-substitution of the leaving aniline, and overall geometry. Cyclisation via the hydroxylamino was observed to be significantly faster than that of its amino counterparts, and was similarly influenced by changes in geometry. The hydroxylamine undergoes a base catalysed, oxygen dependent reaction under aerobic conditions. This reaction did not appear to be influenced by the geometry of the compound. Substitution of the hydroxylamine-bearing ring with a carboxamide group (CONHR σp = 0.36) lowered the pH at which hydroxylamino-amide cyclisation was slowest, compared with its unsubstituted counterpart. The reaction was found to be aided by electron-withdrawal from the leaving amine. Rate determining breakdown of the tetrahedral intermediate has been proposed. Preliminary investigations have been made on 2-nitrophenyl alkyl esters and a 2,6-dinitrophenylamide prodrug system. Rapid, reductively triggered release of coupled phenols and amines has been observed from the nitro-esters and -amides, respectively. In contrast to amino-amide and hydroxylamino-amide cyclisation, gem-dimethyl substitution did not facilitate reductive release from the nitro-ester. The inability to measure the rates of reductive release in radiolysis solutions suggests that these reactions occur significantly faster than hydroxylamino-amide ring closure. Molecular mechanics calculations have been undertaken to investigate relationships between the geometry of 2-aminoarylalkanamides, and rates of cyclisation. The distance between, and angle of approach of the nucleophilic and electrophilic centres in the calculated minimum-energy conformer did not display a correlation with cyclisation rates. / Whole document restricted, but available by request, use the feedback form to request access.

This thesis is concerned primarily with a study of the oxidation of aromatic steroids employing either chromium trioxide-aqueous acetic acid or chromium trioxide-aqueous sulphuric acid-acetone mixtures. The nature of the chromic acid oxidation of various functional groups is discussed first and this is followed by an outline of the two methods used for the synthesis of the aromatic steroids which were oxidised in this study. The second part of the discussion deals with the oxidation of these aromatic steroids. It was found that chromic acid oxidation of ring-A aromatic steroids containing a strong electron-donating C3-substituent, such as methoxyl, gave the corresponding 9-hydroxy-11-oxo derivative (ketol). However, a ketol was not formed if a C3-methoxyl substituted ring-A aromatic steroid also contained a substituent at C1. When a C3-methoxyl substituent was present, the 6-oxo-ring-A aromatic steroid was a minor oxidation product but such compounds were the major products from the chromic acid oxidation of ring-A aromatic steroids containing a weak C3-electron-donating group, such as acetoxyl. The oxidation of a ring-A aromatic steroid containing a C2-methoxyl substituent gave an almost quantitative yield of the corresponding 6-oxo compound. Suzuki103 has claimed that the major oxidation product of 17β-acetoxy-3-methoxyestra-1,3,5(10)-triene (34b) is 17 β-acetoxy-9α- hydroxy-3-methoxyestra-1,3,5(10)-trien-11-one (123). Physical and chemical evidence are presented to show that this product is in fact the 9β-hydroxy epimer and a reaction pathway for its formation is proposed. An examination of the oxidation products of a ring-B aromatic steroid and a ring-C aromatic steroid shows that no ketols were formed.

The Claisen rearrangement of 1,4-bis(prop-2'-enyloxy)anthraquinone under reducing conditions is investigated. The elaboration of the product of double Claisen rearrangement towards an anthracyclinone skeleton is reported. The insertion of methyl ketone and aldehyde functions into the C-2 and C-3 positions of the anthraquinone skeleton and the subsequent elaboration towards an anthracyclinone skeleton is investigated. Methods for the conversion of hydroxymethylanthraquinones to side chains suitable for anthracyclinone syntheses are reported. A total synthesis of 7,9-didesoxydaunomycinone from 1,4-dihydroxy-2-methylanthraquinone is described.