CHEMISTRY OF RIBOSE AND DEOXYRIBOSE 17 



unable to obtain crystalline ribose, its p-bromophenylhydrazone or crystal- 

 line derivatives (i.e., phenylhydrazide or cadmium salt) of ribonic acid 

 from syrups obtained by treating d- or L-arabinose with alkali. Alberda van 

 Ekenstein and Blanksma heated L-arabinose in N sodium hydroxide solu- 

 tion and then oxidized the product to a mixture of L-arabonic and L-ribonic 

 acids which were separated by fractional crystallization of their phenyl- 

 hydrazides. Except that they used calcium hydroxide (0.04 N) as alkali, 

 Austin et al. employed the same procedure. Explanations of the different 

 results reported by these two groups of workers have been offered in terms 

 of difTerent temperatures for epimerization, nature of the alkali, and con- 

 centration effects, but these are not entirely convincing. 



Ribose has been obtained from arabinose by Gehrke and Aichner^" 

 by application of a method originally developed by Bergmann and Schotte" 

 during their researches with glucal. D-Arabinose (III) was converted via 

 its tetraacetate (VII) and 2,3,4-tri-O-acetyl-D-arabinopyranosyl bromide 

 (VIII) into 3 , 4-di-O-acetyl-D-arabinal (IX). (This compound can also be 

 named 3,4-di-O-acetyl-D-ribal.) After deacetylation, the product (X, 

 D-arabinal = D-ribal) in chloroform solution was hydroxylated at 0° with 

 perbenzoic acid to give a mixture pi D-arabinose (III) and D-ribose (I), 

 with the latter predominating. The ribose was purified via its crystalline 

 benzylphenylhydrazone. The sequence of reactions was also carried out 

 using L-arabinose as the initial material and in this case the syrupy product 

 was converted by oxidation into crystalline L-ribonolactone. The method 

 was extended and improved by Austin and HumoUer^® (cf. Neker and 

 Lewis" for related work), purification of the product being achieved by 

 direct crystallization and through formation of the p-bromophenylhydra- 

 zone. It was shown that treatment of L-arabinal with perbenzoic acid in 

 ethyl acetate at 0° yields 5 parts of L-ribose to 1 part of L-arabinose and 

 the overall yield of crystalline ribose, based on the arabinose used, was 

 nearly 10% of theoretical. Workers in other laboratories^^- "^ have also 

 employed this method to prepare ribose and frequently obtained the arabi- 

 nose needed as starting material from calcium D-gluconate by descent of 

 the sugar series. Hudson and Richtmyer*" have used this latter method to 

 convert the calcium salt of D-altronic acid (XI) (for the preparation of 



7" M. Gehrke and F. X. Aiehner, Ber. 60, 918 (1927). 

 75 M. Bergmann and H. Schotte, Ber. 54, 450 (1921). 

 'SW. C. Austin and F. C. Humoller, J. Am. Chem. Soc. 54, 4749 (1932); 56, 1152 



(1934). 

 " H. T. Neker and W. L. Lewis, J. Am. Chem. Soc. 53, 4411 (1931 ). 

 '8 R. Kuhn, K. Reinemund, F. Weygand, and R. Strobele, Ber. 68, 1765 (1935). 

 " P. Karrer, B. Becker, F. Beiiz, P. Frei, H. Salomen, and K. Schopp, Helv. Chim. 



■Acta 18, 1435 (1935). 

 «» C. S. Hudson and N. K. Richtmyer, U. S. Pat. 2,162,721 (July 20, 1939). 



