Radical macrocyclisations in cembranoid synthesis
In recent years the cembranoid family of natural products has emerged as an interesting class of biologically active compounds. The origins and properties of these compounds are first outlined, and a summary of methods previously used to synthesise the 14-membered cembranoid ring system is then given. In connection with a synthetic programme towards cembranoids, based on a radical macrocyclisation as the key step, a number of suitable radical precursors were prepared and their radical macrocyclisations were attempted. Cyclisation via electropinacolisation of the α,ω-dialdehyde (69) was found to be an unsuitable method. Similarly, attempts at a reductive cyclisation of the acetylenic aldehyde (70) also proved unsuccessful. A modified precursor, the halo-enone (118) possessing an activated olefin function, was then chosen. This precursor also failed to cyclise on radical initiation. Ultimately it was found necessary to design the radical precursor such that the β-position of the activated olefin was unsubstituted. In order to favour the required macrocyclisation. Thus the 14-membered cembranoid type ring was eventually prepared by radical cyclisation of the allylic iodo-dienone (127) to give the cyclic enone (132). This preparation then constituted a formal synthesis of the natural cembranolide (37). Despite other possible modes of cyclisation this particular allylic radical cyclisation was found to give selectively only the 14-membered product as a mixture of 10E- and 10Z-isomers. Attempts to extend this methodology, to include an epoxide function in the substrate, proved unsuccessful. Finally a convenient synthesis of the natural cembranoid mukulol (41) is described, using a radical macrocyclisation of the iodo-dienone (149) to give the cyclic enone (148) as the key step. In the appendix an investigation of low valent titanium mediated couplings as a means of preparing dihydrothiophenes and dihydrofurans is described. The diketo ether (184) was found not to cyclise under any conditions. Conversely the diketothioether (176) was found to cyclise to give a cyclic diol (178). We were unable to subsequently convert the diol to the corresponding dihydrothiophene.