258 CAROTENOID CARBOXYLIC ACIDS XIII 



H3C0-0C-CH=CH-C= 



CH3 



As early as 1909, van Hasselt^^ observed the formation of w-xylene during 

 the dry distillation of bixin. This finding was later confirmed by Herzig and 

 Faltis^^ and indicates the presence of the grouping : 



= CH-C=CHCH=CH-C= 

 CH3 CH3 



KuHN and co-workers subjected bixin to oxidative degradation with per- 

 manganate and, later, with chromic acid, and deduced the presence of 4 side 

 chain methyl groups in both cases^^. When the structure of the ozonisation 

 products I and II obtained by Rinkes^^ 



CHaOOC- CH=CH- C=CH- CHO OHC- CH=C- CHO 



I CH3 II CHj 



had been determined, Kuhn and Winterstein^' suggested the above structural 

 formula for bixin, which is also based on the recognition of the symmetrical 

 structure of carotene, lycopene and squalene (Karrer). The correctness of this 

 formula was proved by the investigations of Karrer, in the course of which 

 the position of the two terminal methyl groups was established by the oxidative 

 degradation of partially hydrogenated bixin^^. By the degradation of perhydro- 

 norbixin, Karrer and co-workers^^ obtained 3:7:12: i6-tetramethyloctadecan- 

 i:i8-dialIII, 



OHC • CH2CHCH2CH»CH2CHCH2CH,CH2CH2CHCH2CH2CHi,CHCHjjCH0 



II II 



CII3 CH3 CH3 CH3 



III 



which was oxidised to the dicarboxylic acid and converted into perhydro- 

 crocetin^^ (cf. p. 280). 



The elucidation of the constitution of bixin was completed by the synthesis 

 of perhydronorbixin by Karrer and co-workers^^ and by the conversion of 

 perhydrocrocetin into perhydronorbixin. The structure of both these pigments 

 was thus confirmed^". The formation of m-toluic acid by the thermal decompo- 

 sition of bixin^^, is also in accord with the above formula. 



The synthesis of perhydronorbixin was carried out as follows: 



CHo CHo 



I I 



H0CH,CHCH„CH,CH»CHCH20H 



References p. 2go-2g4. 



