To achieve the oxidation of dione 2 to the endione 23 we had to develop a suitable method, because the originally chosen dehydrogenation with DDQ failed under acidic as well as neutral conditions.We succeeded when oxidizing under rather drastic basic conditions.We assume the electron rich conjugated bisenolate as the reactive intermediate and its reluctant formation is due to the low kinetic as well as thermodynamic acidity of the proton geminal to the acetyl group. We arrived at this conclusion by examining the X-Ray diffraction of crystalline 2. Its conformation, which corresponds well with the main conformation in solution according to NMR-spectroscopic data, shows the hindered proton geminal to the acetyl group perpendicularly positioned to the pi-electrons of the carbonyl group. We succeeded using a mixture of potassium tertiary butoxide and potassium hydride in THF for deprotonation and anhydrous ferric chloride in DMF for oxidation.
The enedione 23, a highly light sensitive compound [8], was reduced immediately with sodium boronhydride and cerium chloride. Oxymercuration of 24 yielded the unsaturated compound 26. The organomercury compound could serve as intermediate in our planned nodusmicin synthesis. With Jones reagent the unsaturated alcohol 26 was oxidized to the tricyclic enone 3.