A new method for the generation of radical anions has been developed. Homolytic cleavage of an azo compound bearing a negative charge $\beta$ to the N=N linkage creates a radical anion intermediate. The radical anions so generated have been shown to undergo quantitatively the same chemical reactions as their neutral counterparts, namely disproportionation and recombination. The cage effects for anionic azo compounds are essentially the same as for their neutral counterparts. The negative charge causes a slight decrease in the overall stability of the azo compounds.

2-Chloro-1,3-didehydronaphthalene was generated as a reactive intermediate in both the open (diradical) and closed (bicyclo{3.1.0}-hexatriene) forms by treating either 1-bromo-3,4-benzo-6,6-dichlorobicyclo{3.1.0}hexane or 3,4-benzo-1,6,6-trichlorobicyclo{3.1.0}hexane with potassium tert-butoxide in tetrahydrofuran at 0(DEGREES)C. The open form reacts with solvent to produce 2-chloronaphthalene, 2-chloro-1-(2'-tetrahydrofurfuranyl)naphthalene, and 2-chloro-3-(2'-tetrahydro-furfuranyl)naphthalene. The closed form reacts with nucleophiles to produce 2,3-dichloronaphthalene, 2-bromo-3-chloronaphthalene, 1,2-dichloronaphthalene, 1-bromo-2-chloronaphthalene, 1,3-di(tert-butoxy)naphthalene, 1-tert-butoxy-2-chloronaphthalene, and 2-tert-butoxy-3-chloronaphthalene. Naphtho{a}cyclopropene was generated in low yield by treating 1-bromo-2-(p-boluenesulfonatomethyl)naphthalene with n-butyl lithium in tetrahydrofuran at -15(DEGREES)C. This compound was characterized by both low temperature nuclear magnetic resonance spectroscopy and by its reaction with tert-butyl alcohol in the presence of silver tetrafluoroborate to produce a mixture of 1 and 2-(tert-butoxymethyl)-naphthalene. The reactions of 2,2-dichloro-1-methylenecyclopropane with potassium tert-butoxide were studied. The products of this reaction were cis and trans-1-tert-butoxybut-1-ene-3-yne. 2-Chloro-1-methylene-cyclopropene is proposed as a reactive intermediate in these reactions.

A novel approach to natural product synthesis is presented herein which encompasses three original functionalization processes. These are; (a) the direct photochemical conversion of alkenes to spiro-activated cyclopropanes, (b) the homoconjugate addition of organo-magnesium reagents to said cyclopropanes, and (c) the ozonolysis of isopropylidene malonate enolates. An application of this new methodology to the partial synthesis of racemic brefeldin A is then described.

The triplet sensitized photolysis of methylazo-(1,1-dimethyl-2-propene) (3) and azo-bis-(1,1-dimethyl-2-propene) (4) have been examined due to their unique ability to decompose under those conditions. The cis isomer was found by two independent experiments not to lie on the triplet sensitized decomposition pathway. Instead, the azo triplet decomposes directly to radicals plus N(,2). This result encouraged an attempt to observe a SCE (spin correlation effect). By means of a crossover experiment, it was established that the cage effect was substantially lower for triplet sensitized decomposition than for direct irradiation, which is consistent with a SCE. However, the data are also "consistent" with a number of other hypotheses as well. These alternate proposals are discussed in some detail; from this discussion emerges a hybrid postulate that incorporates an SCE and a long-lived diazenyl radical in the triplet sensitized decomposition of these azo compounds. Two stable cis azoalkanes were isolated (azoadamantane and azonorbornane). The activation parameters for the disappearance of the cis isomer of both compounds were determined. These activation parameters for cis-azoadamantane disappearance were compared to those obtained by others. The postulate that decomposition of azoalkanes is sensitive to the nature of the incipient radical is bourne out by our results. Finally, an interesting discrepancy was noted in which the 1-adamantyl radical is estimated as unfavorable due to the reluctance of azoadamantane to decompose thermally. On the other hand, the 1-adamantyl radical is judged as similar to the t-butyl radical by other means.

The bis(triphenylphosphine)nickel dichloride-catalyzed reaction of 1-vinylcyclohexanol with phenylmagnesium bromide produced 1-penyl-1-vinylcyclohexane as the major product. An analogous product was obtained with 1-hexynylmagnesium bromide and this allylic alcohol, while the employment of trimethylsilylethynylmagnesium bromide led to substitution at the opposite terminus of the allyl system, yielding trimethylsilylethynylethylidenecyclohexane. Vinylmagnesium bromide initiated a disproportionation reaction of 1-vinylcyclohexanol which was investigated through deuterium labeling experiments. An intramolecular variation of these reactions was carried out with A yielding tricycles B. Indolymagnesium halides could be involved in these reactions as exemplified by the conversion of indolylmagnesium iodide to 3-allylindole in the presence of allyl alcohol, and to C and D in the presence of dimethylallyl alcohol. A reaction between skatole-derived Grignard reagent and allyl alcohol led to 2-allyl-3-methylindole. The nickel-catalyzed reaction of phenylmagnesium bromide with cyclohex-2-en-1-ol and 1-isopropenylcyclohexanol yielded 1-phenyl-cyclohex-2-ene and 1-phenyl-1-isopropenylcyclohexane, respectively. Both methylmagnesium, and phenylmagnesium bromides reacted with 1,3-dimethylcyclohex-2-en-1-ol to produce largely bi-1,3-dimethylcyclohexenyl. In a nickel catalyzed reaction of methylmagnesium and phenylmagnesium bromides with E, 1-isopropenylcyclohexanol and F were produced, respectively. Nickel-catalyzed substitution of the alkoxy groups of 1-methoxycyclohexene, 1-methoxy-4-tert-butylcyclohexene, 4-methoxyheptene, 1-methoxyheptene, dihydropyran, and 2-methoxynaphthalene by the alkyl or aryl groups of Grignard reagents could also be effected. Enamines, enolates, and highly hindered enol ethers proved ineffective in the latter process. An analysis of the stereochemistry associated with these substitutions revealed the original enol ether stereochemistry to be retained in most cases. A catalytic cycle explaining the enol ether substitution process is proposed.

The activation of the carbon-oxygen bonds in ethers and the carbon-hydrogen bonds in alkanes is accomplished by reaction at cryogenic temperatures with metal atoms. Calcium, strontium, and barium react with dimethyl ether to produce yellow-orange pyrophoric solids. Hydrolysis of these solids yields hydrocarbon products and the metal hydroxide. The cocondensation reactions of transition metals with dimethyl ether are also investigated. Chromium, manganese, iron, cobalt, and nickel react to give hydrocarbons upon hydrolysis but copper and zinc fail to react. The photolytic insertions of metal atoms into the carbon-hydrogen bonds of methane are studied using infrared matrix isolation spectoscopy. The initial product is a methyl-metal hydride. The reaction takes place for the metals: manganese, iron, cobalt, copper, silver, gold, and zinc. No reaction was observed for calcium, titanium, chromium, or nickel.

Part I. The photochemistry of 2,3-diazabicyclo{2.2.2}oct-2-ene (DBO) and 11 of its derivatives has been investigated; the syntheses of two new derivatives of this class are described. Introduction of either radical stabilizing substituents at the bridgehead position or strained rings anti to the azo moiety have been found to increase the photoreactivity of the normally photoinert DBO skeleton by up to 70 fold as measured by the quantum yield of nitrogen formation. Strained rings syn to the azo function have little effect. The increase in photoreactivity with structural changes is accompanied by a corresponding decrease in fluorescence lifetimes and quantum yields. The quantum yield of nitrogen formation has been found to correlate with the activation parameters for thermal decomposition of the azoalkanes; that is, the compounds that are labile photochemically are relatively labile thermally and vice versa. This is taken as evidence that a barrier to photochemical nitrogen loss from S(,1) exists which can be lowered with suitable structural changes. Strained rings syn to the azo moiety and bridgehead methyl substitution exert little effect on photoreactivity but do affect the fluorescence quantum yield and lifetime relative to DBO; this is accounted for by slight changes in the rate of intersystem in these relatively photoinert DBO derivatives. An investigation of products from photolysis of two DBO analogues, one with monovinyl and the other with divinyl bridgehead substitution, revealed the complete lack of formation of the corresponding bicyclo{2.2.0}hexanes, the product expected from simple collapse of the 1,4 biradical. To account for this lack of ring closure product, it is proposed that these compounds could decompose via a one bond homolysis to form a diazenyl radical as a discrete intermediate. This species then collapses to the most stable product. The effect of temperature on the photoreactivity and photophysics of several cyclic and bicyclic azoalkanes is described. For azoalkanes whose quantum yields for nitrogen formation are less than unity at room temperature, it has been found that raising the temperature at which the photolysis is performed can be an effective means of increasing photoreactivity. A simple model in which the rate constant for excited state nitrogen loss is treated with the Arrhenius equation yields activation parameters for nitrogen loss. For the compounds investigated here, these values fall between 0-10 kcal mol('-1). At least for the 2,3-diazabicyclo{2.2.2}oct-2-ene derivatives investigated, those which are more photoreactive exhibit lower activation energies than those which are relatively photoinert. Part II. The synthesis and thermal rearrangement of 1-vinylbicyclo{2.2.0}hexane is reported. A synthesis of this compound via a route involving a tosylhydrazone decomposition was unsuccessful but a novel tosylhydrazone fragmentation reaction was discovered in its stead. In addition, a photochemical cleavage of the strained 1,4 bond in 1-acetylbicyclo{2.2.0}hexane was discovered.

The total synthesis of isocomene has been accomplished in seventeen steps. Robinson annelation of 2-methylcyclopentanone with 1,4-dimethoxy-2-butanone gave 1,2,3,6,7,7a-hexahydro-4-methoxy-7a-methyl-5H-indene-5-one. Lithium aluminum hydride reduction of the latter followed by a Simmons-Smith cyclopropanation reaction with ethylidene diiodide and diethyl zinc resulted in the formation of an (alpha)-oxycyclopropylcarbinol. This was elaborated by conventional methods into the crucial intermediate, 2(alpha),4,5,7(beta)-tetramethyl-3(alpha)-methoxy-4-hydroxytricyclo{5.3.0.0('1,3)}dec-5-ene, whose cyclopropylcarbinyl rearrangement product was a cyclobutanone. Ring enlargement of the latter formed 3-oxoisocomene, from which isocomene was obtained readily.

The bis(triphenylphosphine) nickel dichloride-catalyzed reaction of thiophenols and aryl sulfides with phenyl-magnesium and methylmagnesium bromides produced biphenyls and toluenes, respectively. The coupling of p-t-butylthioanisole with ethylmagnesium bromide and cyclohexylmagnesium bromide in the presence of (1,3-diphenylphosphinopropane)nickel dichloride produced p-t-butylethylbenzene and p-t-butylcyclohexylbenzene, respectively, while the coupling with isopropylmagnesium bromide led to the formatin of n-propyl as well as isopropyl aryl derivatives. The same sulfide reacted with isopropylmagnesium and cyclohexylmagnesium bromides in the presence of anhydrous nickel dichloride and triphenylphosphine to produce t-butyl-benzene. With ethylmagnesium bromide t-butylbenzene and p-t-ethylbutylbenzene were produced. p-t-Butylthiophenol afforded t-butylbenzene as the sole product, when treated with isopropylmagnesium bromide, nickel dichloride and triphenylphosphine, and t-butylbenzene as the major product, when treated with isopropylmagnesium bromide and (1,3-diphenylphosphinopropane)nickel dichloride. Ketenethioacetal C(,6)H(,5)CH(,2)CH=C(SEt)(,2) underwent reaction with Grignard reagents in the presence of phosphine-nickel complexes to produce vinyl sulfides in which the alkyl group is oriented trans to the initial benzyl group. When exposed to an excess of methylmagnesium bromide and (1,3-diphenylphosphinopropane)nickel dichloride, the thioacetal produced 1-phenyl-3-methyl-2-butene. Reaction of enol sulfides of ketones with Grignard reagents in the presence of the aforementioned nickel catalyst produced di- and tri-substituted olefins with retention of configuration at the sulfur-bearing olefinic carbon site. The reaction between 1,1-diethylthioethylene and phenylmagnesium bromide in the presence of nickel dichloride and triphenylphosphine led to (alpha)-ethylthio styrene, 1,1-diphenylethene and unreacted ketenethioacetal.

The structure of gerberinol, isolated from the Indian plant Gerbera lanuginosa, has been elucidated as 3,3'-methylenebis-(5-methyl-4-hydroxy-2H-1-benzopyran-2-one). Furanoeremophilane has been synthesized from the known ketone, 3,4,4a,5,6,7,8,8a(beta)-octahydro-4a(beta),5(beta)-dimethylnapthalene-2(1H)-one, using oxycyclopropane chemistry in the key sequence for the preparation of the 3-methylfuran moiety.