Amidocyclohexadienes in synthesis of biologically active compounds

Abstract

A summary of tin hydride mediated reactions in generating radicals in organic synthesis is presented, together with some of the many alternative methods available for conducting radical reactions. Particular attention has been given to the development of tin-free organic radical precursors. This is followed by three chapters describing research on the use of pro-aromatic 1-carbamoyl-l-methylcyclohexa-2,5-dienes as free-radical precursors. A variety of 1 -carbamoyl-l-methylcyclohexa-2,5-dienes, TV-benzyl protected-amide, free-radical precursors have been prepared and utilised in subsequent experiments. EPR studies confirmed the formation of the delocalised cyclohexadienyl radical at low temperatures. At higher temperatures cyclohexadienyl radicals underwent homolysis with release of the associated aminoacyl (carbamoyl) radicals. Measurement of radical concentrations in DTBP solution from the aquired EPR spectra allowed us to calculate the rates of dissociation for several of these aminoacyl radicals. We concluded that the dissociations of cyclohexadienyl radicals, and the carbamoyl radical cyclisations, were both fast enough for chain propagation to be sustained. The formation of pyrrolidin-2-ones was tested by examining the decomposition of but3-enyl amides induced thermally with dibenzoyl peroxide. 3-Alkyl-TVbenzylpyrrolidin-2-ones were obtained in moderate yields, along with significant amounts of the corresponding formamides. Reactions with other initiators, including photochemical processes with di-/-butyl peroxide, showed no significant advantage. A cyclohexadienyl amide containing bis-methyl substitution of the butenyl chain was prepared; in the hope that ring closure of the carbamoyl radical would be more efficient because of a Thorpe-Ingold effect. In practice, the yield of the corresponding pyrrolidinone was a disappointing 20 %. Evidence from characterisation of byproducts suggested that radical-radical reactions were important. The use of cyclohexadienyl amides for the preparation of (β-lactams was also investigated. The cinnamyl-substituted amide 144 (Chapter 3) was chosen because the phenyl substituent confers resonance stabilisation on the cyclised radical and hence should expedite the 4-exo-cyclisation step. However, as an additional consequence of this resonance stabilisation the hydrogen-atom abstraction step was disfavoured. The thermal and photochemical reactions gave the corresponding azetidin-2-one, in low yields. Addition of a good H-atom donor enabled the ring closed benzyl type radical to be trapped. Reactions initiated with di-/-butyl peroxyoxalate, including 1.2 mol equiv. of methyl thioglycolate (RSH) produced 1,3-dibenzylazetidinone (148, Chapter 3) in increased yields. When a catalytic amount of RSH was employed, the (3- lactam yield increased, but radical-radical reactions again became important. Use of lauroyl peroxide as initiator, in concert with 1 equiv. of RSH, led to a greatly improved (3-lactam yield. It is likely that polarity reversal catalysis played a part in enhancing the yield of (3-lactam. The azetidinylbenzyl radical is resonance stabilised and nucleophilic. Hence a polar effect favoured hydrogen-abstraction from the electronegative RSH. The electrophilic thiyl radical (RS*) generated in this way, in turn, abstracted more readily from the cyclohexadienyl site thus regenerating RSH and continuing the chain. In the lauroyl peroxide initiated reactions, products derived from the initiator included undecane and docosane which significantly increased the viscosity of the reaction medium thus slowing radical-radical reactions which are normally diffusion controlled. Overall, the induced carbamoyl-cyclohexadiene decompositions represent novel, comparatively 'clean', tin-free radical routes that are applicable for the preparation of a range of lactams from secondary amine starting materials. Inclusion of methyl thioglycolate in the reaction medium leads to increased lactam yields. Cyclisation onto an oxime ether C=N double bond was faster and more efficient, and the product heterocycles contain N-functionality at C(3), exactly as required in many antibiotics. Amidocyclohexadiene precursors containing oxime ether functionality were therefore designed and assembled. The radical-induced decompositions were studied by EPR spectroscopy which showed the expected cyclohexadienyl radicals at low temperatures. At higher temperatures, alkoxyaminyl radicals were also observed. Cyclisations of the corresponding carbamoyl radicals were so fast that they were not observable by EPR spectra. When reactions were carried out preparatively, amino benzopyrrolidinone derivatives were isolated in good yields. Addition of methyl thioglycolate in dilauroyl peroxide mediated radical reactions, again improved product yields. An elegant use of O-trityl oxime ethers, in which the oxime functionality was ultimately regenerated had attracted our attention. We sought to adapt this tactic to our system and, accordingly, we prepared an O-trityl oxime ether substituted cyclohexadienecarboxamide precursor. The standard initiation followed by (3-scission of the cyclohexadienyl radical, produced the corresponding carbamoyl radical which ring closed to yield the alkoxyaminyl radical. However, the reaction took a novel route because trityl alkoxyaminyl radicals containing O-trityl moiety, underwent a second [3- scission to extrude the highly stabilised trityl radical with production of a nitrosocompound. The use of our strategy was planned for a tin-free, radical-mediated synthesis of benzepril (ACE inhibitor). 7-exo-ring closure onto the oxime ether radical acceptor would be sufficiently rapid to facilitate the otherwise difficult ring closure. Carbamoyl radical released from a specifically designed 1-methylcyclohexadiene carboxamide containing O-trityl oxime ether moiety would undergo 7-exo-cyclisation to afford the desired 7-membered lactam oxime ether (tautomer of the corresponding nitrosocompound). Due to time limitations, work in this area could not be completed but this remains a promising avenue for future work.