Anhydro compounds inversions

In succeeding years a considerable number of papers have been written on Walden inversion and the formation of anhydro compounds in the sugar series. Since this material has been summarized fairly recently by Peat and by Isbell, the following discussion will be very limited in this respect. 

In each case it was assumed, without attempt at experimental verification, that no Walden inversion had occurred and that the product had the configuration of the parent sugar. Since an anhydro compound is the probable intermediate, this assumption is open to some question. This same author treated potassium methyl mercaptide with one (denoted by E. Fischer as form I) of the two isomeric methyl 2-bromo-/3-D-glucosides and obtained a nicely crystalline thiomethyl compound which was hydrolyzed to the free sugar. 

In contrast to other 2,5-anhydroaldoses (which exhibit mutarota-tion, possibly due to the formation of hemiacetals28), 2,5-anhydro-D-glucose does not show any mutarotation.27 The importance of this compound as a potentially useful precursor to C-nucleosides warrants a reinvestigation of the deamination reaction, and the definitive proof of the structure of the compound. The readily accessible 2,5-anhydro-D-mannose (11) does not possess the cis-disposed side-chains at C-2 and C-5 that would be required of a synthetic precursor to the naturally occurring C-nucleosides, with the exception of a-pyrazomycin (8). The possibility of an inversion of the orientation of the aldehyde group in 11 by equilibration under basic conditions could be considered. 

The synthesis of other anhydro sugar derivatives is usually accomplished by an internal displacement of a leaving group with inversion, as has already been emphasized for those anhydrides in which the anhydride bridge engages the anomeric position. Three such compounds that have been used in polymerizations deserve specific mention 5,6-anhydro-l,2-0-isopropylidene-a-D-glucofuranose,12 13 its 3-methyl ether,11 and 3,5-anhydro-l,2-0-isopropylidene-a-D-xylo-furanose,14 all of which were synthesized by p-toluenesulfonylation of the primary hydroxyl group of the parent, isopropylidene derivative, followed by treatment with base. 

The stereochemistry at C-5 of each of the hexitols (6) and (7) was determined in the following way. Periodate oxidation of these hexitols afiForded dialdehydes (8) and (9), which, on subsequent reduction with sodium borohydride, afforded the enantiomeric trihydroxy ethers (10) and (11). The configuration at C-3 of the ether (10) was then correlated with that of C-4 of the known l,4-anhydro-5-deoxy-D-arabino-hexitol (12), as follows. Periodate oxidation of (12), followed by reduction of the resulting dialdehyde with sodium borohydride, yielded a trihydroxy ether that was identical with (10). If it is assumed that no inversion of configuration at C-3 or C-4 of 3,4-di-O-acetyl-D-xylal occurs during the oxo reaction, then compounds (6) and (7) are l,3-anhydro-4-deoxy-L-JC /lo-hexitol and l,5-anhydro-4-deoxy-D-arobino-hexitol, respectively. 

These compounds are rather rare. One of the first examples was presented by Helferich who prepared 3,5-anhydro-l,2-0-isopropylidene-Q -D-xylofuranose (82) from the appropriate 3,5-dimesyl derivative by treatment with ethanolic potassium hydroxide [1]. Another convenient approach to such derivatives utilized the reaction of l,2-0-isopropylidene-5-0-triflate-a-D-glucofuranos-S,6-lactone (79) with base (K2CO3) in methanol, which produced the corresponding L-ilactone ring with formation of the alcoholate at 3-OH, which substituted the triflate with inversion of the configuration at the C-5 atom (O Fig. 17). 

Addition of the anion derived from uracil to l,5 2,3-dianhydro-4,6-0-benzyli-dene-D-allohexitol, followed by inversion of configuration of the resulting 3-OH group by displace,ment of its mesylate derivative with sodium hydroxide, afforded 12. Compound 12 was a useful intermediate for making l,5-anhydro-2-deoxy-D-mannitol-containing pyrimidine nucleoside analogues. ... 

Condensation of the glycal ester (15) with the epoxide (16), catalysed by boron trifluoride, then gave an unsaturated disaccharide derivative from which the pentadeoxy-compound (17) was made. This is a derivative of a disaccharide that occurs in anthracyclin antibiotics. The closely related dimer (18) was also prepared using the same glycal derivative and methyl 2,3-anhydro-6-bromo-6-deoxy-0 -D-allopyranoside. Inversion at C-5 of the product was effected by hydrogenation of the derived 6-deoxy-5-enose compound. This work also led... 

---