Investigations into metal-oxo reagents in organic syntheses and towards the syntheses of (-)-dysibetaine using aza-[2,3]-Wittig rearrangement-cyclisation

Part 1 of this thesis describes the continued development of a new method for the synthesis of olefins from sulfur reagents and carbonyl compounds. The proposed catalytic olefination investigated the reaction of sulfur ylides or sulfines with carbonyl compounds to produce the corresponding alkene, catalysed by a transition metal oxo complex. A variety of literature trioxo rhenium and molybdenum di-oxo and oxo-imido complexes were reacted with dimethylsulfoxonium methylide and generally led to degradation. The reaction of diphenylsulfonium benzylide with the rhenium and molybdenum oxo complexes gave no reaction or led to degradation. Diphenylsulfine was also reacted with the metal-oxo complexes and produced benzophenone and degradation of the metal oxo complex. Part 2 of this thesis describes the investigation into the novel complex MoO(NtBu)(2,6-R2C6H3O)2py [where R = iPr (72a) or Me (72b)], which was developed in the Anderson group, as an epoxidation catalyst. The initial epoxidation conditions were optimised for trans-stilbene. The optimised conditions were used to epoxidise a variety of alkenes. The catalyst 72 is effective at epoxidation of electron rich alkenes, will selectively epoxidise allylic alcohols and does not epoxidise electron deficient alkenes. The rate of epoxidation of cis-stilbene using catalyst 72a and 72b were compared and showed that 72b had a faster rate of epoxidation than 72a. The rate of reaction for catalyst 72b was also compared with MoO2(2,6-Me2C6H3O)2py2 98 for the epoxidation of cis-stilbene and found to be similar, indicating that the catalytically active species may be common to both. Part 3 of this thesis describes an investigation into the synthesis of (-)-dysibetaine using the aza-[2,3]-Wittig rearrangement-cyclisation protocol. The initial retrosynthesis led to a -lactone enolate 57 in the aza-[2,3]-Wittig rearrangement cyclisation. However the desired cyclised product 58 was not obtained and instead ring opening of the -lactone to the acrylic acid 80 was observed. A revised route used a phenyldimethylsilyl amino acid 106 as a masked hydroxyl group to mimic a serine amino acid that would not undergo -elimination. The aza-[2,3]-Wittig precursor 109 was prepared and subjected to the standard aza-[2,3]-Wittig rearrangement-cyclisation protocol and the desired cyclised product 110 was obtained in 43% yield. This product contained the correct dysibetaine skeleton and would require functional group transformations to complete the synthesis