Carbon monoxide adsorption, surface speciation and particle size distributions have been studied in platinum, palladium, and rhodium hydrosol systems using Fourier transform infrared spectroscopy, X-ray photoelectron spectroscopy and transmission electron microscopy. In transmission electron micrographs of unprotected platinum and palladium hydrosols, particle necklacing believed to arise from sintering during preparation, is apparent. The average particle sizes of all hydrosols studied were in the range of 2 to 12 nm. X-ray photoelectron spectra of the metal hydrosols revealed evidence for (Pt-O)ads, Pt(II) and Pt(IV) oxides on platinum hydrosol particles whereas Pd(II) and Pd(IV) oxides were detected on the surfaces of palladium hydrosol particles. These surface oxides are found to be important in influencing hydrosol surface processes such as CO adsorption as a function of pH, inhibition of CO adsorption by alcohols and surface corrosion products resulting from the addition of iodide and cyanide. Fourier transform infrared spectra of CO-treated metal hydrosols revealed bands due to CO linearly adsorbed on the metal particles at ca. 2070 cm-1 (Pt), 2067 cm-1 (Pd) and 2045 cm-1 (Rh) whereas bands due to bridge-adsorbed (B2) CO were detected at ca.1950 cm-1 (Pd) and 1890 cm-1 (Rh). The use of CO as a spectroscopic probe molecule enabled the study of changes in the surface properties of the metal hydrosols which were induced by changes in the dispersion medium. For example, v(CO)ads was observed to decrease in infrared spectra of CO-treated platinum and rhodium hydrosols as pH was increased by KOH or other dissolved salts yielding alkaline solutions. This suggested a reduction in CO coverage resulting from hydroxyl adsorption and consequent increased oxide growth on the particles. In contrast, CO adsorption on platinum and rhodium hydrosols was enhanced in acidic media possibly as a result of the neutralisation of surface hydroxyls. The spectroscopic behaviour of adsorbed CO on platinum and rhodium hydrosols was only comparable to that of CO adsorbed at an electrode surface in acidic media when protecting agent was present which prevented aggregation of the hydrosol in such media. Inhibition of CO adsorption on platinum hydrosols was induced by the addition of aliphatic alcohols, poly(vinyl alcohol) and poisoning anions such as CN- and SH-. Correlations of v(CO)ads with CO coverage suggested that island formation of adsorbed CO was occurring for CO adsorption on unprotected palladium hydrosols and protected platinum and rhodium hydrosols as a function of pH. In allied investigations, an infrared spectroelectrochemical study of corrosion of a nickel electrode in aqueous cyanide media has revealed that [Ni(CN)4]2- is detected at potentials more cathodic than 200 mV vs. SCE. Cyanide was oxidised to cyanate (OCN-) and then successively to carbon dioxide at potentials more anodic than 200 mV vs. SCE. The appearance of features at 2094 cm-1 (HCN) and 2256 cm-1 (HNCO) were attributed to pH changes associated with the oxidation of cyanide to cyanate. The appearance of a band at ca. 2218 cm-1 in infrared spectra of the thin layer at very high potentials (> 1000 mV vs. SCE) was believed to arise from an unstable nickel(II) isocyanate complex.

Chapter One of this Thesis reports an investigation of the synthesis of diterpenoids (as potential ligands for cyclomanganation reactions) from podocarpic acid, and one from dehydroabietic acid. A number of monocyclic aromatic and chalcone derivatives were synthesised to act as model compounds in the subsequent complexation reactions. The second chapter describes successful complexations of the above ligands containing aldehyde, ketone, ester, or oxime groups as directing functionalities to form the resulting tetracarbonylmanganese sigma adducts in moderate to excellent yield. Attempted complexation of amide-containing ligands proved generally to be unsuccessful. Complexation of ligands containing either two potential sites for manganation or two potential ligating groups were investigated and the structures of the isolated complexes established unequivocally by N.M.R. spectroscopy or by X-ray crystallography. Chapter Three describes the successful coupling reactions of the tetracarbonylmanganese complexes with ethene and with substituted electron-poor olefins to give pentaannulated derivatives of podocarpic acid in very good yields. The stereochemistry of two cyclised alcohols was assigned unambiguously by single-crystal X-ray diffraction experiments. Various methods of activating the tetracarbonyl complexes were investigated, including chemical oxidation, and palladium(II)-mediated or thermally-promoted reactions. Two of the pentaannulated adducts were converted successfully into benzannulated analogues in good yield. Analogous coupling reactions of the activated complexes with ethyne and substituted alkynes are discussed in Chapter Four. Chapter Five reports an investigation of the reactions of the tetracarbonylmanganese complexes with halogen electrophiles such as Br2, ICl, or ICl3. When these reactions were performed in non-protic solvents the corresponding aryl halides were isolated in moderate to excellent yield. However, in protic solvents carbonyl insertion occurred to afford γ-lactones in very good yield. The stereochemistry of one of these lactones was assigned unambiguously by a single-crystal X-ray diffraction experiment. Oxidative cleavage of the C-Mn bond with a number of reagents proved generally to be unsuccessful, the isolation of oxygenated diterpenoids being observed only from reaction with Pb(OAc)4. Chapter Six describes various attempts at functionalising ring C of podocarpic acid derivatives via alternative methods to cyclomanganation. These include an investigation into the formation of η4-dienol ether and η4-diene tricarbonyliron complexes derived from the Birch reduction products of podocarpic acid derivatives; attempted regioselective chlorination and mercuration studies of ring C arenes; functionalisation of chlorinated podocarpic acid derivatives via their tricarbonylchromium complexes; and attempted oxidation of ring C arenes with meta-chloroperbenzoic acid.

Whole document restricted, see Access Instructions file below for details of how to access the print copy. / The adsorption and decomposition of Mo(CO)6 in zeolites X, Y, KL, Na-mordenite and Na-ZSM5 had been followed gravimetrically and spectroscopically. Mo(CO)6 was readily adsorbed into the pore structures of faujasites and zeolite KL but not into Na-mordenite and Na-ZSM5. Infrared studies showed that the adsorption produced a common adsorbed Mo(CO)6 species whose Oh symmetry was lowered. Subsequent activation in-vacuo at 100°C caused loss of approximately two carbonyl ligands per molybdenum which could be reversed by addition of carbon monoxide. Heating at or above 200°C caused complete and permanent decarbonylation. Zerovalent molybdenum clusters were produced within the pore structures of alkali metal-exchanged zeolites but for HY and LaY, oxidised molybdenum ions were also produced. Further uptake of Mo(CO)6 occurred on subsequent exposure of faujasite samples containing molybdenum to the vapour at room temperature. By repeated adsorption and decomposition of Mo(CO)6, zeolite sample containing up to 157 molybdenum per unit cell had been prepared. The carbon monoxide hydrogenation activity and selectivity of molybdenum-zeolite catalysts varied with zeolite structure and exchanged cation. Molybdenum-alkali metal-exchange zeolite catalysts were more selective to alkenes and liquefied petroleum gas range hydrocarbons (C2-C4) than molybdenum-HY and molybdenum-LaY catalysts. Using a combination of gas chromatography and infrared spectroscopy, a mechanism was postulated for the formation of hydrocarbons.

Part 1. Some aspects of NMR spectroscopy: A review of coupling mechanisms is given and notable examples of long range coupling in the literature are discussed. Methyl protons can undergo long range coupling with protons which are 4s bonds away, and the mechanism of this process as related to the known examples is discussed. The effects of electronegative substituents on long range coupling is also outlined. A Discussion of the present work follows, wherein the description of the preparation of many bornane derivatives, including several deuterated derivatives, is presented. The line widths (Wh/2) of the methyl groups were measured. It was found that the width of the methyl groups decreased in the order C8>C9>C10, and the C8 and C9 methyl signals were always appreciably broader than the C10. Long range coupling between the protons of the C8 and C9 methyl groups is postulated as being responsible for those two being broader than C10. Long range coupling between the C8 methyl protons and the C5 and C6 endo protons, and between the C9 methyl and the C3 endo proton is tentatively ascribed to the differences in the widths of the C8 and C9 methyl groups. Although the line widths of the tertiary methyls were found to vary with solvent, the relative widths of the three methyl signals of the bornane derivatives was found to be the same. Thus measurement of line widths is a method of correlating the methyl signals in spectra run in different solvents. The effects on the chemical shifts of the tertiary methyl groups of bornanes when various substituents are placed in the basic skeleton, is also described. The presence of an ethylenedioxy group at C2 or C3 results in a desheilding of C8 and a shielding of C9, while introduction of hydroxyl groups has no uniform effect because of different orientations of the hydroxyl group from compound to compound. An introduction to the mechanism of benzene induced solvent shifts is given. Benzene appears to form a stereospecific complex with ketones, and other molecules, which can undergo a dipole-induced dipole-interaction. The effects of this interaction on the n.m.r. spectra of various ketones, quinones and aromatic compounds is described. The present work describes the measurement of solvent shifts at various concentrations of ketones and ketals in benzene. Previous similar work is discussed in the light of the results obtained, and it has been shown that some assumptions previously used in correlating the data with the interactions, have been unjustified. The solvent shifts versus the concentration have been plotted to give curves which have been interpreted as indicating specific solvent-solute interactions, i.e. 1:1, 1:2, etc., and not random associations. This has been shown by comparing calculated curves with the experimental results, with which agreement is good. The third part of this section of the work deals with the spectra of some hydroxymethylene ketones, and commences with an introduction to the general behaviour of 1,3-diketones in solution. The rates of exchange of enolizable protons is discussed, and their use in determining the relative proportions of enol forms is outlined. The present work describes the behaviour of 3-formylbornan-2-one in carbontetrachloride-D2O solution, and an explanation for the preference for one enol form is given. Also discussed is the coupling between ‘formyl’ and hydroxyl protons in hydroxymethylene ketones, and it is shown that the coupling depends strongly on the H-O-C-H angle, which can be altered by slight changes in molecular geometry, thus leading to changes in coupling constants. No change in coupling constants caused by temperature differences was observed, and it is speculated that exchange processes do not affect the value of the coupling constant, but only whether coupling occurs. Part 2: Acid catalysed rearrangements in cyclic systems A comprehensive review of the dienone-phenol rearrangement is given and it is shown that any cross-conjugated cyclohexadienone will rearrange to an aromatic product, the substitution of which depends strongly on (i) the substitution pattern of the original dienone, (ii) peculiarities in the molecular structure, and (iii) the rearrangement conditions. Rearrangements of some dienols and some cross-conjugated methylenecyclohexadienes or ‘semibenzenes’ are described, and in general these follow the pattern of the dienone-phenol rearrangement. Rearrangement of some steroidal α-substituted ketones is also mentioned briefly. Two cross-conjugated methylenecyclohexadienes were synthesised by reaction of methylmagnesium iodide on the corresponding dienones – this reaction had previously been reported to give only the products of rearrangement of the intermediate semibenzenes. The semibenzenes were rearranged using perchloric acid in acetic anhydride-ethyl acetate to give the 1,4-disubstituted aromatic product. The attacking species was shown to be a proton, and the reaction, unlike the dienone-phenol rearrangement, showed a clear preference for methylene rather than methyl migration. Several methyl substituted steroidal 2α-hydroxy-4-en-3-ones were prepared by the rearrangement of the corresponding 4,5-epoxy-3-ones with sulphuric acid in aqueous acetone, herein shown to be a general method. The hydroxyenones were rearranged using three sets of conditions: Toluene-p-sulphonic acid in methanol, toluene-p-sulphonic acid in benzene, and perchloric acid in acetic anhydride-ethyl acetate. Only one compound, 2α-hydroxycholest-4-en-3-one gave an aromatic product; the other substrates underwent rearrangement through the various enols, to give attack usually at C6 of the carbon skeleton. A mechanism is given and discussed for each rearrangement. No general pattern seemed to emerge. Part 3: The Clemmensen reduction of 1,3-diketones A review of the Clemmensen reduction is given with a specific reference to reduction of difunctional compounds, where rearrangements may occur. The present work shows that in the Clemmensen reduction of 1,3-diketones, the intermediate cyclopropanediol which is first formed, undergoes acidolysis to give an α-hydroxy ketone and a β-hydroxy ketone, each of which may then undergo further Clemmensen reduction to give the observed rearranged and unrearranged monoketones. The intermediacy of a β-hydroxy ketone was proved by observing the rate of formation of the products, and by measuring the rate of reduction of an independently synthesised α-hydroxy ketone. Reduction of two acyclic ketones was shown to give a ring expanded product only in the case of 2-acetylcyclopentanone. Part 4: Intramolecular hydrogen abstraction reactions from aromatic alkyl groups A short resume of intramolecular hydrogen reactions using lead tetraacetate or the hypoiodite process is given, with some detail being applied to the conditions, and the stereochemistry needed for the reaction, and to fragmentation processes which may occur. Brief mention is also made of the new reaction; hydrogen abstraction using decomposition of a hypobromite with silver ions. The lead tetraacetate reaction was applied without success to abstraction of hydrogen from an aromatic methyl group, utilizing an α-benzylic hydroxyl group. Fragmentation reactions were generally obtained. Application of the bromine-silver ion method gave good yields of cyclic ethers in two cases, indicating that an abstraction reaction had taken place. The conditions for such a reaction seem to be critical: A tertiary hydroxyl group is required, and the abstraction seems only to take place from a primary centre.

The anodic growth of a Cu2O film on polycrystalline copper surfaces was studied by a potentiostatic method. Plots of anodic current versus time allowed the growth law to be determined and the Cu2O film thickness was found to vary linearly with the cube root of time. The same growth law applied when Cu2O films were formed by aqueous oxygen attack at the copper surfaces. A mechanism has been proposed explaining this law, based on assumptions of Cabrera and Mott for gaseous oxidation of copper. The cyclic voltammetry technique was used to study the growth, dissolution and reduction of Cu2O and the higher copper oxides over the pH range 5.20 to 13.83. Cu2O was the first oxide formed, and depending on the pH, either CuO or Cu(OH)2 was formed over the underlying Cu2O at higher potentials. In NaOH solutions, of concentration 0.10M and greater, presence of the oxide, Cu2O3 was noted at potentials corresponding to oxygen evolution. During cathodic reduction of the films formed anodically, the Cu(II) product was initially reduced to Cu2O, followed by Cu2O reduction to Cu metal. Solid state reactions for the growth, dissolution and reduction processes are included. No explanation for peak formation during anodic film growth by the cyclic voltammetry technique can be found but plateau formation is predicted. The inclusion of chloride in the anodically formed films increased the conductivity of the oxide films greatly. A 0.100M Na2B407 solution containing 9.6 x 10-4M benzotriazole was found to inhibit the formation of films at the copper surface probably by the formation of a Cu(BTA) polymer. The presence of sodium hexametaphosphate had no effect on oxide growth kinetics at pH 9.18. Potentials of specimens of copper piping corroding in Auckland's supply water were measured against a saturated calomel electrode. The values depended largely on the conditions of water flow and stagnation. Illumination and surface preparation of the samples had little effect. The corrosion potentials reached a maximum after approximately 30 days and the high potentials observed were attributed to the resistance of the films of corrosion products covering the metal surface. A mechanism for the corrosion process has been proposed with the electrophoretic deposition of a layer of silica gel over the underlying Cu2O film causing local cell action and the formation of anodic and cathodic areas over the copper surface. The gel allows diffusion of Cu2+ into the solution from the anodic regions and OH-from the cathodic regions, forming a loosely adherent flocc of Cu(OH)2 the metal surface. This flocc is readily scoured from the surface leading to Auckland's green water problem.

The heterolytic chlorination of 1-arylnaphthalene (aryl = 1(p-nitrophenyl), (p-chlorophenyl), 1-phenyl, and 1(p-methylphenyl) ) yields the appropriate 1-aryl-r-1,c-2,t-3,t-4-tetrachloro-1,2,3,4-tetrahydronaphthalene as the major product of addition. The 1H n.m.r. spectra of these compounds show that there is no significant change in the conformation adopted in solution accompanying the change in the para-substituent. The major product of the alkaline dehydrochlorination of each of these tetrachlorides using sodium methoxide in methanol-acetone (80:20 v/v) or benzene-methanol (75:25 v/v) is the 1-aryl-2,3-dichloronaphthalene. The first stage of the reaction is rate-controlling and both stages appear to be bimolecular. The observed rate-coefficients have been dissected to derive rates for the two competing elimination processes involving either initial loss of the 4-chlorine and 3-hydrogen or of the 1-chlorine and 2-hydrogen. The value of the Hammett reaction constant for the first of these processes was found to be 0.60 in methanol-acetone and 0.85 in benzene-methanol. The corresponding values for the second process were found to be 1.48 and 1.58. Both these results indicate that the effect of substituents in stabilizing the incipient negative charge developing at the β –carbon in the transition state is an important feature of the elimination processes involved. A slow, unimolecular solvolysis at the 1-chlorine has been observed with 1-Ph-NTC and 1(p-Me-Ph)NTC in methanol-acetone. It is proposed that the major product of this solvolysis results from the replacement of the 1-chlorine by a methoxy-group with retention of configuration. In contrast, the major product of solvolysis assisted by silver(I) ions is that possessing the inverted configuration. Alkaline dehydrochlorination of the methoxy-substituted compounds derived from 1(p-Me-Ph)NTC leads to fully aromatic products. This requires the elimination of the components of a molecule of methanol under comparatively mild alkaline conditions. Dehydrochlorination from the 1-arylnaphthalene tetrachlorides promoted by aluminium chloride in carbon disulphide at reflux temperature results in a mixture of the 1-aryl-2,3- and 2,4-dichloro- products. The proportion of 2,4-dichloro- product varies from 49% (1(p-NO2-Ph)NTC) to 65% (1-Ph-NTC).

PART I A STUDY OF MECHANISTIC VARIATIONS IN THE BROMINATION AND CHLORINATION OF SOME ALKYL PHENOLS, PHENYL ETHERS AND PNENYL ESTERS. The brominations of some 2,6-dialkylphenyl acetates in nitro-methane and in acetic acid have been shown to give (a) products of substitution at both the C-3 and C-4 nuclear sites (b) products of substitution with rearrangement leading to the formation of various brominated alkylphenols and (c) uniquely in acetic acid, a product of alkyl side-chain substitution. The bromination of the corresponding benzoate derivatives realised only the products of 3- substitution. Various products of (a) substitution and (b) substitution with rearrangement were determined in the bromination of 3,4-dimethylphenyl acetate, phenyl acetate, and 2,3,5,6-tetramethylphenyl acetate. The bromination of 2,4,6-dimethylphenyl methyl ether gave only the expected product of nuclear substitution. Evidence for the likely involvement of addition-elimination sequences in the formation of products of substitution with rearrangement, has been adduced. The possible role of various other processes leading to the formation of deacylated products is discussed. The chlorination of 2,6-dimethylphenyl acetate, 2,6-dimethylphenyl benzoate and of 2,3,5,6-tetramethylphenyl acetate in both acetic acid and nitromethane gave only the expected product of nuclear substitution. The reaction of chlorine with 3,4-dimethylphenol, its acetate and methyl ether in acetic acid led to the formation of certain enone and dienone adduct derivatives which could be shown to rearrange under the conditions of the reaction to give (a) in the case of the phenol-products of substitution and (b) for the acetate and methyl ether derivatives, products of substitution with rearrangement. The structures of the intermediates were determined from their spectral properties. Various mechanisms likely to be involved in processes of chloro-deacylation and of chloro-demethylation are discussed; a consideration of certain structural features of the substrates studied led to the conclusion that chloro-deacylation and chloro-demethylation were essentially processes initiated by ortho- attack of the chlorine molecule. PART II REACTION PATHS IN THE BROMINATION OF CHOLEST-4-EN-3-ONE, CHOLEST-5-EN-3-ONE AND 3-ACETOXYCHOLESTA-3,5-DIENE The bromination of cholest-4-en-3-one can occur by a number of reaction paths. Conditions have been established under which reaction can be established as occurring in part by electrophilic attack on (a) 3-hydroxycholesta-4,6-diene (b) 3-hydroxycholesta-2,4-diene or its anion and (c) cholest-4-en-3-one; among the products can be included 2α-, 4-, 6β-, and 6α-bromocholest-4-en-3-one; 2α,6β-, 2α,6α-, and 4,6β-dibromocholest-4-en-3-one; and 2α,4,6β-tribromocholest-4-en-3-one. Unstable adducts involving the 4,5-double bond are implicated in a number of cases. The bromination of cholest-5-en-3-one is equally complicated, and can involve addition to the 5,6-double bond. Knowledge of the range of conditions and reagents through which 3-acetoxycholesta-3,5-diene is brominated to give 6β-bromocholest-4-en-3-one has been extended. Intermediate adducts involving the 5,6-double bond have been shown to be components of the reaction mixture from the acetoxydiene and bromine in acetic acid. Acid catalysed isomerisation and rearrangements of the kinetically controlled products, and dibromination, can be controlled by selective choice of the conditions of bromination and work-up. Complications associated with the reaction paths have been partly elucidated. The rates and mechanism of reversible isomerisation of 6β- to 6α-bromocholest-4-en-3-one, and of 2α,β- to 2α,6α-dibromocholest-4-en-3-one, have been studied and a proposed mechanism is outlined.

A study of the influence of several "soft donor" ligands on the cupric ion has been completed. Nitrogen and sulphur donors were chosen since these are considered to be present in "biological copper" situations, some of the conditions of which it was hoped to duplicate. Complicated reactions were found with several different types of ligands. Those containing imine functions, when complexed to copper(II), were invariably found to cleave due to the presence of external nucleophiles and it was considered that for such donors to be important in "biological" situations they would need to be shielded by bulky ligand systems. The cleavage reactions were found to have their origin in electronic mechanisms and earlier suggestions based on steric interactions were criticised. A study of sulphur donors showed that, whereas thioethers had a poor affinity for copper(II), thiols bonded very strongly. Reduction to copper(I) invariably took place, but with ligands containing TT-delocalised aromatic thiols as well as amine donors mixed-valent complexes of copper were formed. The properties of most of these complexes were unusual and one, especially, (N-(2-thiophenol) pyridine-2-aldiminato)copper(II) chloride, showed reactions and properties that suggested it may be a suitable model for cupro-enzymes. Complexes with thiol donors were termed "anomalous" in view of their complicated physical properties, whereas compounds with other donors showed "normal" behaviour. Absorption spectral studies were undertaken in an attempt to understand the electronic states of the "soft donor" complexes; at the same time an assessment of a recent scheme for assigning stereochemistries from visible spectra as applied to "soft donor" systems was possible. An X-ray structural determination of the complex μ-chloro-chlorobis(3,4-bis(3-amino-1-thiapropyl)toluene)dicopper(II) diperchlorate was undertaken, partly because of its unusual synthetic and physical characteristics. Two copper centres, one five- and the other six-co-ordinate, were found to be linked by a bridging chloride ion. The geometries about each copper ion were considerably distorted from square-pyramidal and octahedral respectively; the ligand systems also showed unusual electronic effects. Comparison with other recent structural investigations suggested that five-co-ordination is favoured by the copper ion in "soft donor" environments, and this was thought to be significant in terms of oxygen-carrying ability, where an increase in co-ordination number must occur. Several attempts to oxygenate prepared complexes were made, but no conclusive results were obtained, although several somewhat unstable crystalline products were produced.

This thesis is concerned with the synthesis and properties of fluorocarbene complexes of ruthenium and osmium. Chapter 1 is a review of transition metal CF3-complexes and illustrates a number of unusual features of these compounds which makes them suitable for modification into CF2-complexes. To place the fluorocarbene complexes into context, Chapter 2 contains a brief review of transition metal carbene complexes with the emphasis on other halocarbene species. As described in Chapter 3, CF3-complexes of ruthenium can be prepared by the addition of Hg(CF3)2 to Ru(CO)3(PPh3)2. The x-ray crystal structure of Ru(HgCF3)(CF3)(CO)2(PPh3)2 shows longer C-F bonds in the ruthenium bound CF3-group than in the mercury bound CF3-group. Reactions of fluorocarbon complexes with Lewis acids have been explored and RuCl(CF3)(CO)2(PPh3)2 reacts with BCl3 to form the known RuCl2(CO)(CCl2)(PPh3)2. Similarly OsCl(CF2H)(CO)2(PPh3)2 reacts with one equivalent of BCl3 to form initially a cationic carbene complex in which the carbene carbon atom substitutes first into a benzene ring of one PPh3-ligand and, through further reaction with BCl3, into a ring of the second PPh3-ligand. The x-ray crystal structures of OsCl(CO)2(PPh2C6H4CHCl)PPh3 and OsCl(CO)2(PPh2C6H4CH)(PPh2C6H4) are reported. RuCl(CF3)(CO)2(PPh3)2 in many of its reactions behaves as a CF2-complex precursor and Chapter 4 describes the reaction between the related RuCl(CF3)(CH3CN)(CO)(PPh3)2 and Me3SiCl to form the first isolable CF2-complex, RuCl2(CO)(CF2)(PPh3)2. RuCl2(CO)(CF2)(PPh3)2 contains an electrophilic carbene ligand and reacts rapidly with hard, oxygen and nitrogen containing nucleophiles. A comparison of the reactivity of RuCl2(CO)(CF2)(PPh3)2 and RuCl2(CO)(CCl2)(PPh3)2 is presented and reasons for the differing reactivities proposed. Chapter 5 shows that d8 CF2-complexes can be prepared directly from reaction between the more active CF3-transfer reagent, Cd(CF3)2-glyme, and various zerovalent complexes. The x-ray crystal structures of M(CO)2(CF2)(PPh3)2 M = Ru, Os and OsCl(NO)(CF2)(PPh3)2 are described and the bonding characteristics of the CF2-ligand discussed. The carbene ligand of M(CO)2(CF2)(PPh3)2 M = Ru, Os has some nucleophilic character and reacts readily with electrophiles, for example reaction with HC1 yields MCl(CF2H)(CO)2(PPh3)2 M = Ru, Os. A comparison between RuCl2(CO)(CF2)(PPh3)2 and Ru(CO)2(CF2)(PPh3)2 reveals that the reactivity of the carbene ligand is controlled by the electron density at the metal centre, i.e. is oxidation state dependent.

The synthesis and ozonolysis of the 6- and 7-substituted 12-hydroxypodocarpatrienes (26), (86), (90), and (98), as a possible route for the synthesis of drimanes from 12-hydroxypodocarpa-8,11,13-trien-19-oic acid (podocarpic acid) (9) has been investigated. Modification of the aromatic ring of (9) has given 11- and l4-hydroxypodocarpatrienes which were ozonised to give the congeners (31), (34), and (36) of the naturally occurring drimanes confertifolin (6), isodrimenin (7), and winterin (8). The oxidation of a 12,13-dihydroxypodocarpatriene (37) and its dimethyl ether (38) have also been investigated. Routes for the conversion of the enone (45) and oxime (230), derived from the ozonolysis products of methyl podocarpate (10), to drimanes have also been investigated.