Some recent advances in the synthesis of carotenoids.

A route to unsymmetrical polyene dialdehydes is described. The utility of these intermediates is illustrated by a synthesis of okenone. The stereochemistry assigned to methyl natural bixin has been confirmed by preparation of the cis-4 isomer. Support has been obtained for the cis-6 formulation of natural cis-crocetin. Oxidation of astacene with manganese dioxide yields violaxanthin. The synthesis of stereoisomers of alloxanthin and crocoxanthin has confirmed the structures of these acetylenic carotenoids, and identified the principal cis isomers formed on stereomutation. The synthesis has been achieved of racemic forms of the allenic ketone from grasshoppers, and of an allenic degradation product from fucoxanthin. INTRODUCTION In this paper I propose to review some recent developments in the use of synthesis for studying the structures of carotenoids. I shall cover partial, as well as total synthesis, and, at times, the not unrelated technique of partial degradation. I shall also have to outline briefly the scope of some of the newer classes of carotenoids so as to indicate the relevance of the synthetic studies themselves. POLYENE DIALDEHYDES The value of symmetrical polyene dialdehydes, in particular of the C10and C20-dials [4] and [5] and their central acetylene analogues, as intermediates for the synthesis of carotenoids, is well known. The carbon skeleton of many symmetrical carotenoids can be built up quite simply by reacting one of these dialdehydes with an excess of the appropriate Wittig reagent. Fortunately, as was predicted on electronic grounds, the nucleophilic attack of the first aldehyde group by the Wittig reagent is appreciably faster than that of the second. It is therefore possible to carry out a selective reaction at one end of the dialdehyde with a Wittig reagent, and then react the remaining aldehyde group with a different reagent. In this way many unsymmetrical carotenoids have also been prepared'. The most convenient Wittig reagents for the introduction of the two end groups of some unsymmetrical carotenoids may, however, be of different size. We have therefore explored the possibility of preparing and using the unsymmetrical polyene dialdehydes. The Wittig reagent [3] was first made by Makin2 from 1-ethoxy-2-methylbuta-l,3-diene, but can be obtained