iiiiii - ACS Publications - American Chemical Society

Department of Chemistry, University College, University of Alabama in Birmingham,. University Station, Birmingham, Alabama 35294. Received August 6,19...
7 downloads 0 Views 738KB Size
3562

J.Org. Chem., Vol. 42, No. 22,1977

Cantrell, Kiely, Abruscato, and Riordan

8-Dicarbonyl Sugars. 5. A Novel Synthesis of a Branched-Chain Cyclitol' Charles E. Cantrell,2 Donald E. Kiely,* Gerald J. Abruscato, and James M. Riordan Department of Chemistry, University College, University of Alabama in Birmingham, University Station, Birmingham, Alabama 35294 Received August 6,1976

The cyclization of tri-0-acetyl-1,7-dideoxy-1,7-bis(diazo)-~y~o-2,6-heptodiulose (5), in acetic acid solution, to ~~-3-C-acetoxymethyl-2,4,5,6-tetra-0-acetyl-2,3,4,6/5-pentahydroxycyclohexanone (8) is described. This conversion, considered to take place by way of a carbene, represents a new synthesis of the cyclitol ring system. The reactive diketone (5) was prepared by a diazomethane chain extension sequence originating with D-xylose (1).Reduction of the keto carbonyl of 8 followed by the acetylation of the resulting products gave DL-2-C-acetoxymethyl1,3,4,5,6-penta-0-acetyl-epi-inositol(12). A minor product from the decomposition of 5 was identified as penta-0acetyl-xylo-2,6-heptodiulose (9). T h e biosynthetic pathways t h a t lead t o the carbocyclic ring system as it is found in the cyclitols3 are thought to involve enzyme-catalyzed aldol condensations of appropriate 6-dicarbonyl monosaccharides. We have recently discovered t h a t some synthetically prepared derivatives of this class of carbohydrates can be chemically induced t o undergo these same ring closure^.'^^ This paper describes a new route t o the acetylated hydroxyniethylcyclitol DL-2-C-acetoxymethyl1,3,4,5,6-penta-O-acetyl-epiinositol (12) by way of an unusual cyclization of a bisdiazo ketone ( 5 ) derived from xylo-2,6heptodiulose.

Results and Discussion T h e bisdiazo ketone 5 was prepared by a standard diazomethane chain extension sequence beginning with D-xylose (1) (Scheme I). In order t o obtain satisfactory yields of xylaric acid (2) and tri-0-acetylxylaric anhydride (3), it was necessary t o modify the Wolfrom and Usdin procedure for the synthesis of these compounds.5 When the oxidation of D-xylose with nitric acid was completed, excess oxidizing agent was destroyed with 2-propanol and crystalline xylaric acid was obtained in 44% yield. Deletion of t h e 2-propanol addition step in t h e workup procedure resulted in the isolation of a syrup that did not crystallize and eventually decomposed. When the anhydride 3 was prepared by refluxing zinc xylarate in acetyl ~ h l o r i d e the , ~ average yield of the product from the reaction was only 30%. Furthermore, t h e anhydride obtained in this way gradually underwent a n irreversible phase change t o an oil. However, stable 3 was synthesized in reasonable yield (70-8096)by treating 2 with acetic anhydride t h a t contained a catalytic amount of sulfuric acid. Syrupy tri-0-acetylxylaryl dichloride (4),6 prepared by refluxing t h e half sodium salt of 2 triacetate with thionyl chloride, was treated with an ether solution of diazomethane and tri-0-acetyl-l,7-dideoxy-1,7-bis(diazo)-xylo2,g-heptodiulose ( 5 ) crystallized as yellow needles directly from the reaction mixture (68%). Thin-layer chromatography of the mother liquors showed a three-component mixture. Silica gel column chromatography of this residue gave first dimethyl tri-0-acetylxylarate (6). Next off the column was a mixture (ca. 40% of the material from the column) of 7 and an unidentified compound. T h e characterization of 7 is described in the following paper (ref 18).T h e third component, 5, was also isolated (5%). T h e bisdiazo ketone 5, when dissolved in acetic acid t h a t contained cupric acetate, rapidly decomposed (5 min) a t 70 "C t o give 8 (32%),which crystallized directly from the reaction mixture (Scheme 11). The conversion was also accomplished without the presence of copper ion, but required a higher temperature (80-90 "C), and a significantly longer reaction time (10 h).

Scheme I H

OH

I

\ c=o

?=O

I

I

H-C-OH HO-C-H

H3

H-C-OH + HO-C-H

I

I

I H-C-OHI

+A&

I

H-C-OH

I

I c=o I

CH,OH

3

OH 2

1

c1

CHN,

c=o I C -OAC

c=o I

I

-

2

H-

l

A&-C-H3 'H -C

I

1 I

3H-C-OA~ AcO-C-H4

I -0Ac I c=o

I

H-

I

I I c=o I

C-OAC

c1

CHN,

4

5

OCH,

I

'H+A&-C-H3

c=o I c-OAC

I

1

4H-C -0Ac

I c=o

CHN,

I I H- C-OAc + A&-C-H I c=o

H-C-

I

OAC

I

I

I

I

C -CHzC1 O