Dehydrochlorination of 3,4,6-tri-O-acetyl-2-chloro-2-deoxy-~-~-glucopyranosyl chloride with diethyla~nine yielded 3,4,6-tri-0-acetyl-2-chloro-D-glucal. The latter was further chlorinated to 3,4,6-tri-0-acetyl-2-deoxy-2,2-dichloro-~-~-urubino-hexopyranosy chloride, which on deacetylation provided the first example of a stable, free glycopyranosyl chloride.A 3,4,6-tri-0-acetyl-2-chloro-2-deoxymannopyranosyl chloride, also isolated in this work, was characterized by properties which correspond closely to the recently reassigned anomer.Canadian Journal of Chemistry, 46, 3001 (1968) Although moilohalogen substitution on asymmetric carbon atoms of cyclic sugars has been achieved in ilumerous instances (1) and with each of the four halogens, cyclic sugars with gemdihalogen substitution on ring carbons are comparatively rare. The earliest examples appear to be acetylated 2-deoxy-2,2-dichloro-D-erythropentopyranose and its glycosyl chloride, synthesized by Vargha and Kuszmann (2). More recently Brown and Jones (3) have reported a 3-deoxy-3,3-di-iodo derivative of D-ribo-hexo-1 furanore. Our present work concerns the iilvestigation of synthetic routes to 2,2-dichloro substitution of hexopyranose structures. Electronegative substitution at this position is of particular interest to us in kinetic studies of the 1 acid hydrolysis of glycopyranosides (4).
IA long-established route to 2-chlorodeoxy substitution of hexopyranoses has been via chlorination of the 1,2-glycals to 2-chlorodeoxy-D-glycosyl chlorides. In the earliest investigation of this reaction Fischer, Bergmann, and Schotte (5) passed chlorine into a cold solution of tri-0-acetyl-D-glucal (1) in carbon tetrachloride and obtained a product described as 'triacetylglucal dichloride'. Recent investigations of this substance using modern analytical techniques show that it is a mixture of gluco (2) and n7anno (3) epimers. Lemieux and Fraser-Reid (6), using nuclear magnetic resonance spectroscopy 1 (n.m.r.), reported the a-D-gluco dichloride as the main product in greater than 80% yield, but did not detect the manno dichloride. Subsequently I Lefar and Weill (7) subjected a crystalline sample I of the dichloro-triacetyl-glycal to thin-layer chromatography (t.1.c.) and showed that it was a mixture of two major components which they identified by n.m.r. as the a-D-gluco and a-Dmanno dichlorides in a ratio of roughly 4 : 1. The results of the present work are qualitatively similar to those of Lefar and Weill, as far as the ratio of the two compoilents is concerned, but the minor component exhibits significantly different physical properties and n.m.r. spectrum (vide infra). Very recently, since the completion of our work, Igarashi, Honma, and Imagawa have reported (8) mainly by a four-centered transition state or a concerted mechanism, which is in contrast to the two-stage oxocarbonium mechanism proposed by . In 1963 Vargha and Kuszmann (2) reported a series of reactions leading from 3,4-di-0-acetyl-D-arabinal to an acetylated 2,2-dichloro-2-deoxy-D-pentose. Chlorinatio...