ISOLATION OF RACEMIC 2,4-PENTANEDIOL AND 2,5-HEXANEDIOL FROM COMMERCIAL MIXTURES OF RACEMIC AND MESO ISOMERS BY WAY OF CYCLIC SULFITES

Enantiomerically pure diols with C-2 symmetry such as 2,3-butanediol, 1,2,4-pentanediol, 2, and 2,5-hexanediol, 3, are useful chiral auxilia ries, but they are expensive because chemists lack good synthetic rout es that eliminate both the meso isomer and one enantiomer. Enzymic res olutions efficiently separate the enantiomer, but do not remove the me so isomer. To simplify enzymic resolutions of 2 and 3, we developed si mple methods to isolate the racemic isomer from commercial mixtures of racemic and mesoisomers. For 2, the meso isomer selectively reacted w ith SOCl2 to give a cyclic sulfite that was removed by column chromato graphy to leave (+/-)-2, 92% de, 1.4 g, 55% yield. For 3, both meso an d racemic isomers reacted with SOCl2 to give cyclic sulfites, but the sulfite derived from the meso isomer rearranged to trans-2,5-dimethylt etrahydrofuran under acidic conditions. Hydrolysis of the remaining su lfite gave (+/-)-3, 84% de, 1.1 g. 37% yield. Resolution of (+/-)-2 an d (+/-)-3 using lipase from Pseudomonas cepacia yielded (2R,4R)-2-diac etate, 78% ee, >97% de, 40% of theory and (2R,5R)-3-diacetate 94% ee, >97% de, 47% of theory. Previously reported acetylations of 2 and 3 by lipase from Candida antarctica (CAL) or by lipase from Pseudomonas sp . (Amanolipase AK) are more enantioselective and thus, the best route to enantiomerically and diastereomerically pure 2 and 3 is removal of the meso isomer by way of cyclic sulfites followed by resolution with CAL or Amano lipase AK.