A review is presented of the Claisen, Cope and Carroll rearrangements, with consideration of the mechanisms operative in and the synthetic applications of these reactions. The Claisen orthoester rearrangement of several systems derived from gamma,delta-unsaturated alpha-hydroxyphosphonates is described. The presence of the phosphoryl substituent appears to deactivate the Claisen system. A corresponding acetoacetic ester derivative failed to undergo a Carroll rearrangement. The [2,3] sigmatropic rearrangement of allylic sulphenates derived from gamma,delta-unsaturated alpha-hydroxyphosphonates is described. The rearrangement of Claisen systems generated by the addition of allylic alcohols to diethyl (3-methyl-buta-1,2-dienyl) phosphonate is reported. Where the participating alcohol is allyl alcohol the two possible ketonic products of Claisen rearrangement are formed at room temperature. The ready rearrangement observed in this case is explained in terms of the reaction proceeding via the first-formed allylic anion intermediates rather than via the neutral adducts. An example of a Cope rearrangement involving the participation of an allenic phosphonate ester is described., The product of this rearrangement, a mixture of isomeric dienes, was converted into the corresponding Diels Alder adduct upon reaction with N-phenyl maleimide. The synthesis of a vinylallenic phosphonate with the potential to undergo an intramolecular ene reaction was achieved. Although the product of the ene reaction, a l,3,5-hexatriene, was not isolated, species resulting from the cyclisation and aromatisation of this product were characterised. A modification to this system produced a vinylallenic phosphonate amenable to function as the diene fragment in a Diels Alder cycloaddition reaction. The formation of an adduct with N-phenyl maleimide is described. A short review of the use of PPh3/CCl4 as a reagent for dehydration and condensation is presented as an introduction to Part 2. The reaction between simple phosphites and elemental iodine is investigated and the use of the intermediate species formed as reagents for dehydration and condensation is described.

No description available.

The Precambrian basement rocks of the Lowder-Mudiah area are divided into three main belts separated by two northeast-striking ductile shear (thrust) zones. The central belt consists of highly metamorphosed granitic gneisses intruded by granites. The eastern belt comprises greenschist-grade northeast-trending metavolcanics; these rocks are deformed in open antiforms and synforms. The western belt is composed of granite (740 22 Ma), diorite and gabbro traversed by two extensive dyke swarms, forming 75-90% of the total rock volume. The major swarm (709 21 Ma) trends NE-SW and the minor (587 18 Ma) E-W to NW-SE. Three deformational episodes are recognized. Northeast trending recumbent isoclinal folds were followed by open folds with vertical axial planes and northwest trending crossfolds and kink bands. The thrust-lineation relationships suggest that there was an early phase of thrusting which produced down-dip lineations, and that these thrust surfaces were later reactivated to become dextral trans current ductile shear zones which gave rise to the present predominant shallow-plunging lineations. Tertiary block faulting occurred along reactivated Precambrian thrust planes. Peraluminous, high silica, "A-type" granites formed from LILE-and LREE-enriched crustal melts in an extensional tectonic regime. These granitoids are similar to the felsic Mudiah metavolcanics, the felsic Mukeras dyke swarm and the "A-type" young granitoids (686 - 517 Ma) of Central Hijaz, Saudi Arabia. The Mudiah metavolcanics and the NE-SW Mukeras dyke swarm are bimodal formed from LILE- and LREE-enriched source(s), and they exhibit Andean-type convergent margin characteristics. However, a large degree of compositional overlap may be expected between lavas in highly magmatic, rapidly distending rifts, and Andean-type convergent margins. Moreover, the Mudiah metavolcanics and the Mukeras dyke swarms are chemically similar to the dyke swarms in the Northeastern Desert of Egypt and the "sequence" A volcanics of Central Arabian Shield. It is concluded that the late Precambrian accretion of arc systems resulted in a thickening of the crust, which led to deep crustal melting, which was probably triggered by upper mantle diapirism, caused by adiabatic decompression, updoming, stretching and crustal thinning and the injection of mantle-derived basaltic magma, that gave rise to the basaltic Mudiah metavolcanics and the basaltic Mukeras dyke swarms. In contrast, crustal melts' are represented by the Lowder-Mudiah "A-type" granitoids, the felsic Mudiah metavolcanics, and the felsic Mukeras dyke swarms. The host rocks of the dykes (granite (740 22 Ma), quartzdiorite, diorite and gabbro) have a calc-alkaline subduction zone signature, and they were probably emplaced in the roots of contemporary island arcs, similar to the old calc-alkaline granodiorite association (720 Ma) of Central Hijaz of Saudi Arabia. Carbonatite bodies, mainly those which occur as dykes, which are either apatite-bearing or RE-mineral-bearing (group III and group IV) are confirmed as carbonatites. However, petrochemistry suggests that other carbonate sheets are sedimentary marbles enveloped by granitoid gneisses.

The work contained in this thesis examines the aziridination of a number of ?-diketones via their enol forms using oxidative addition of N-aminoquinazolin-4(3H)-ones, to give N-(quinazolinone-N'-y1)-N-acyl-a-aminoketones which result from C-C cleavage of the intermediate aziridine ring. Products which have the N-N bond di-acylated are primed for reductive cleavage. One example has been studied in which reduction was successfully accomplished using aluminium amalgam to afford an N-acylated amino acid ester in good yield. The N-acetoxyaminoquinazolinone has been shown to function as an NH2+ equivalent. Aziridination of 3-methylpentan-2,4-dione using oxidation of N-aminoquinazolinone in the presence of trifluoroacetic acid (TFA) afforded the aziridine C-N cleavage product. The effects of retarded N-N bond rotation in many of the products from these aziridinations were observed in their n.m.r. spectra: the barriers involved were quantified in some cases but were too large to be measured in others. The good leaving group ability of the quinazolinone ring was revealed in attempted alkylation reactions under basic conditions a- to the exocyclic nitrogen with the N-(quinazolinone-N'-y1)-N-acyl-a-aminoketones above. Synthesis of N-(2-isopropylquinazolin-4(3H)-one-3-y1)-6-ethoxycarbony1-2-pyridone was accomplished with the expectation that elimination of the quinazolinone ring could be stereoelectronically inhibited. However, attempted methylation of this compound also resulted in elimination of the quinazolinone ring. In their reactions with stabilised anions, the N-acetoxyaminoquinazolinones function as synthetic equivalents of the NH2+ synthon. Oxidation of the N-aminoquinazolinones with lead tetra-acetate (LTA) in the presence of pyridine and its derivatives gave N-(quinazolin-4(3H)-one-3-y1)pyridinium ylides. These pyridinium ylides, on heating, functioned as aziridinating agents e.g. for styrene and diethyl fumarate. From the n.m.r. spectra of one of these pyridinium ylides at various temperatures, the barrier to rotation around both N-N bonds could be calculated. An investigation of the possible mechanisms of aziridination of alkenes using these pyridinium ylides was undertaken.

The mechanisms of the enzymes of the shikimate pathway, and the mechanisms of phosphoryl transfer by kinases in general are reviewed in chapter 1. In chapter 2, the preparation of B,y-[18O]ATP, alpha,beta-[18O]ADP, alpha,beta-[18O]ATP, and alpha1,beta-[18O]Ap3A are reported. [18O]ATP is used to investigate the mechanism of phosphoryl transfer by shikimate kinase. In the presence of shikimate kinase and shikimate, randomisation of the -18O label with the O2 occurs as expected. No significant randomisation was observed in the presence of a novel fluoro shikimate analogue or in the absence of any cosubstrate. An associative Sn2(P) mechanism of phosphoryl transfer is thereby assumed. In chapter 3, the design, synthesis, and characterisation of analogues of shikimate are reported. An epoxy-shikimate analogue is observed to be a weak irreversible inhibitor of shikimate kinase. Substrate protection studies suggest that the modification is most likely at the ATP binding site. Cis-diol and trans-diol shikimate analogues, lacking the C-5 hydroxyl group, are observed to be substrates of the enzyme. The presence of the C-5 hydroxyl group is found to be important for specificity with a maximum contribution of 3.9 kcal mol-1 to binding energy. A fluoro shikimate analogue, also lacking the C-5 hydroxyl, is found to be a competitive inhibitor, with a Ki of 21 mM. The design, synthesis and characterisation of shikimate analogues with altered hydroxyl stereochemistry at C-3 and C-4, prepared from (5S,6R)-5,6-dihydroxy-l-(trifluoromethyl)-cyclohexa-1,3-diene are reported in chapter 4. Each of the analogues demonstrates similar properties, showing weak non-competitive inhibition with shikimate kinase. The importance of the carboxylate group for the recognition of shikimate is discussed. In chapter 5, shikimate kinase was observed to be inactivated by the reagents iodoacetamide, N-ethyl maleimide, iodoacetate, 2,3 butanedione and diethylpyrocarbonate at pH 7.0. N-ethyl maleimide was the most potent inhibitor, being 60- to 6000-times more potent than the other reagents. The presence of ATP was observed to give total protection against modification by iodoacetamide, N-ethyl pyrocarbonate, and diethylpyrocarbonate, suggesting modification at the active site of the enzyme. Shikimate was found also to partially protect against inactivation by each of these three reagents. A cysteine residue within the probable ATP binding site is tentatively assigned as the site of modification, and its possible roles discussed.

This thesis is a continuation of previous work done in the Jenkins group to produce taxanes from glucose. The reactions investigated by R. Bonnert and J. Howarth are improved upon to enable the large scale synthesis of the diol (1). After the successful protection of (1) with two tert-butyldiphenylsilyl protecting groups the benzylidene ring was opened by bromination with N-bromosuccinimide. This compound then underwent fragmentation with zinc metal to open the acetal ring and yield the highly functionalised key intermediate chiral cyclohexane (2). This aldehyde was then reduced with sodium borohydride and protected with a triethylsilyl group. The alkene side chain underwent ozonolysis cleanly followed by the addition of lithium-2,2,3,3-tetramethyl-bromocyclopropane. The cyclopropane addition was improved upon by the addition of cerium trichloride. Compounds (3) were isolated from this reaction and a series of studies were done to test the viability of these compounds to undergo further reactions, in particular to rearrange to give dienes. Compounds with bromine in were found to rearrange and those with chlorine in did not. It is believed that most of this work was done with compounds containing chlorine, although at the time this was not realised. Detailed spectroscopic studies verified the stereostructures as those desired for taxanes at the key chiral centres. Alternative sugar routes were also considered and this work gave rise to a new Robinson reaction on a carbohydrate, to give compound (4). Unfortunately this compound was found to have the wrong stereostructure for the synthesis of taxanes from it.

The threat to the Earth's ozone layer posed by chlorofluorocarbons (CFCs) arises because of the photolytic generation of chlorine atoms in the stratosphere. New methods of rendering CFCs less harmful by dechlorinating them, are therefore of environmental importance. Preliminary investigations have suggested that silylene chemistry may be useful in this respect. Silylenes have also been shown to be formed as gas-phase intermediates during the Direct Synthesis of methylchlorosilanes. This work reports on a study of the possibility of using these silylenes for the dechlorination of CFCs using the technology of the Direct Synthesis, and the effect that CFCs would have upon methylchlorosilane product distribution. The gas-phase reactions of silylenes with chlorofluorocarbons were studied. Silylenes (:SiMe2, :SiMeCl and :SiCl2), were generated thermally from appropriate disilane precursors in the presence of halomethanes (CF2Cl2 and CFCl3). The reactions proceeded via radical mechanisms, initiated by the decomposition of an adduct formed between the CFC and silylene, resulting in chlorine abstraction to yield chlorosilyl and halomethyl radicals. Abstraction proceeded most efficiently with :SiMe2. The mechanisms were propagated by the resultant halomethyl radicals which acted as chain carriers. Although silylene insertion reactions were observed, radical reactions were dominant. The reactions of silylenes with methyl chloride were also studied but were found to be less extensive, with die greater strength of the C-Cl bond allowing insertion reactions to become competitive. All reactions were found to proceed much more efficiently in the presence of a relatively weak Si-H bond. To understand further the role of radical reactions in the dehalogenation of CFCs, silyl radicals were generated directly using mercury photosensitisation. Similar mechanisms were indicated, although reaction was more extensive as a result of the increased number of gas-phase radical species. A laboratory scale reactor was used to simulate the addition of CFCs to the Direct Synthesis, and to study their effect on product distribution. Surface reactions dominated, with the adsorption of CFCs leading to greater chlorination of the methylchlorosilane products.

Aziridination of some vinyl silanes and vinyl stannanes using 2-substituted 3-acetoxyaminoquinazolinones are reported. The desilylation and destannylation of two of the resultant aziridines with caesium fluoride in the presence of potassium cyanide proceeds with loss of the quinazolinone unit via an azirine intermediate; subsequent addition of the cyanide ion furnishes a nitrile-substituted aziridine. This aziridine-azirine-aziridine transformation was also accomplished using an enantiopure 2-substituted-3-acetoxyaminoquinazolinone to provide an optically active aziridine of unknown enantiopurity. Attempts to bring about the above aziridine-azirine-aziridine conversion by deprotonation of 1-(quinazolinone-3-yl)-2-ester- and 1-(quinazolinone-3-yl)-2-nitrile-substituted aziridines are reported. The synthesis of 3-amino-2-trifluoromethylquinazolinone and the anomalous stability of the derived 3-acetoxyaminoquinazolinone are reported. This stability of 3-acetoxyamino-2-trifluoromethylquinazolinone leads to unusually high yields in its aziridination reactions with alkenes compared to those obtained using 2-alkyl-3-acetoxyaminoquinazolinones. In aziridination of cyclohexenol, this 2-trifluoromethyl-3-acetoxyaminoquinazolinone gives the syn aziridination product completely stereoselectively. To probe the origin of this unexpected stability of 2-trifluoromethyl-3-acetoxyaminoquinazolinone, the stability of a number of other 3-acetoxyaminoquinazolinones bearing electron-withdrawing substituents in the 2-position was examined. From these investigations and from the deleterious effects of acetic acid upon the stability of 3-acetoxyaminoquinazolinones an explanation for the greater stability and greater yields obtained in aziridinations using 2-trifluoromethyl-3-acetoxyaminoquinazolinones is offered.

The N,N-diacylation of 3-aminoquinazolinones results in the formation of imides. X-ray crystal structures of some of these compounds confirm the presence of N-N chiral axes. Both enantiopure diastereoisomers of one imide have been separated and each behaves as a chiral acylating agent. The faster eluted diastereoisomer reacts with methylbenzylamine with partial kinetic resolution to give a 3.6 : 1 ratio of diastereoisomers of the amide. The ability of 3-diacetylaminoquinazolinones to selectively acetylate one amine in the presence of another amine is also illustrated. Attempted aziridination of cyclic B-ketoesters, diketones and enol silyl ethers with 3-acetoxyamino-2-isopropyl-quinazolinone proceed in good yield with C-N bond cleavage of the intermediate aziridiue. The products from treatment of these (quinazolinon-3-yl)amino cyclic ketone derivatives with LTA are dependent on the solvent used. In methanol, ring-cleavage occurs for 5- and 6-membered rings to give imine-esters. In dichloromethane, 5-membered ring ketones undergo a ring-expansion reaction whereas when 6-membered ring analogues are used, no expansion occurs and the only product isolated is the benzoxazinone. A mechanism which accounts for this dependence on solvent is presented; radicals do not appear to be involved. The reaction of 3-aminoquinazolinones with ketones results in disparate ratios of C=N isomers of imines. The barriers to N-N bond rotation in some of these compounds have been calculated. The potential use of these imines for the preparation of enantiopure amines is examined. Addition of triethylamine (or other tertiary amines) to a solution of 3- acetoxyaminoquinazolinone, results in the formation of triethylammonium imdde. Solutions of this compound react with alkenes at -30° C to give aziridiaes; the reactivity of the aziridinating intermediate is consistent with its formulation as an azaimide (N-nitrene). Aziridination by the addition of 3-acetoxyaminoquinazolinones to alkenes in the presence of HMDS results in increased yields of aziridine products compared to those obtained in the absence of HMDS. The role of HMDS in these reactions is not clearly understood, although it seems likely that it is behaving in a similar way to trifluoroacetic acid which also brings about an enhancement of yields in these aziridinations.

Organomercurial esters 2a-c obtained from cyclopropyl alcohols la-c in three or five (depending on whether the system is cis or trans) steps, react with organocuprates via an intramolecular addition across the carbonyl bond to generate the corresponding lactolate. Quenching of the latter intermediate with water leads to the lactol 3a, c in the cycloheptane and cyclopentane series, whereas the open hydroxyketone 4b is formed in the cyclohexane series. Quenching of the lactolate with BF3 Et2O gives different products, as demonstrated for the cyclohexane series. The Pd(II)-catalysed carbonylation of chloromercurio alcohols 5a-c allows the construction of either cis- or trans-fused lactone rings 6a-c. The stereochemistry of the lactone annulation is controlled by the Hg (II) reagent to initially open the cyclopropane ring, in conjunction with oithogonal protection.