Synthesis of long-chain fatty acids and some derivatives

An attempt has been made to compare anodic synthesis, malonation, and chain-extension with enamines as procedures for converting readily-available acids to higher homologues. Anodic synthesis was successfully employed to prepare esters of several olefinic acids (20:1 11c, 22:1 13c, 20:2 11c 14c, 22:2 13c 16c, 20:3 8c 11c 14c), acetylenic acids (18:1 16a, 18:1 17a), and saturated acids of 'odd' chain length (13:0, 15:0, 17:0, 19:0, 21:0). The desired esters were recovered from the reaction products by column chromatography and further purified when necessary by crystallisation. Yields were usually in the range 40-50%. Some difficulty was encountered with acids having △5 and △6 unsaturation and this made the chain-extension of γ-linolenic acid by this procedure unsatisfactory. A possible reason for this is discussed. An electrolytic procedure for converting long-chain acids (RCO H) to their nor-alcohols (ROH) has been examined: oleic, linoleic, and ricinoleic acid were converted to the corresponding C₁₇ alcohols. Oleic, linoleic, ɑ-linolenic and γ -linolenic acids were successfully converted to their C₂₀ homologues in 70-80% yield by an improved malonation procedure which was recently described. In one case the malonation was repeated to give a C₂₂ acid. The enamine chain-extension procedure was briefly examined using cyclopentanone and cyclohexanone to furnish heneicosanoic and eicbsanoic acids respectively The ¹³C NMR spectra of the starting materials and products have been recorded. It has proved possible to allocate the observed resonances (with greater or lesser certainty) to all the constituent carbon atoms and some interesting generalisations are presented. A minor study was concerned with the preparation of long-chain hydroperoxides and peroxides by reaction of the mesylates of some primary alcohols (stearyl, oleyl, linoleyl) with hydrogen peroxide and t-butyl peroxides. Some secondary peroxides were formed by a similar reaction with allylic bromides