small amounts ( < 5%). The exolendo ratio did not depend on the steric requirements of the substituents on C4 (OSiPh,, OSiBn,, OSitBuMe,). The em-directing influence arose exclusively from the effect of the 1,6-anhydro configuration. This observation has been confirmed with related monocyclic pseudoglucal derivatives. The oxirane ring was preferentially oriented trans to the alkoxy substituent on C 1 . ["] The endo oxirane 2 was formed stereoselectively from the same starting material if a formally analogous hypobromite was used for the electrophilic attack on the double bond in 3 b instead of a peracid. The addition of acetylhypobromite (AcOBr) to 3 b in CC1, gave the hitherto unknown compound 5 a with a trans-diaxial arrangement of the newly introduced substituents. Compound 5 a could be isolated from the reaction mixture by chromatography in 62% yield.[*] The bromohydrin 5 b was obtained quantitatively by mild, selective hydrolysis of the acetyl group with K,CO, in aqueous methanol (90%). After complete reaction (about 3 h, monitored with TLC), 5 b was converted (without workup) in a slow ( 5 d) and clean reaction into the endo oxirane 2.IY1 The latter was obtained in 85 YO yield, based on 5a.
In the fully analyzed ' H N M R spectrum of compound 5aL81 all vicinal ' J couplings are less than 2 Hz, with the exception of 'J(5,6,,,), 5.8 Hz. The relatively high values for 4J and ' J (0.7-1.7 Hz) are quite remarkable and indicate a largely coplanar arrangement of the bonds. These results corroborate an exclusive and rigid ' C, conformation for the acetate, in which all ring protons are in equatorial positions and the substituents are in an antiperiplanar arrangement.
An equilibrium with the possible Bo, , conformation (equatorial substituents) in 5 a does not exist, at least not at room temperature. The considerable shift of the resonance frequency of proton H 3 to lower field (AS ca. 2) in comparison with that in I and 2 is caused by the deshielding effect of the carbonyl group of the axial acetyl substituent on C3. This observation confirms the correct position of the acetyl group and our original assumption that the addition of hypobromite to 3 b takes place regio-and stereoselectively. Thus, 5 a is predestined for ring closure to the endo oxirane.
The most notable aspect of our new synthetic route for the Cerny epoxides is the evident simplicity of the methods employed. Laborious protecting group chemistry is not needed for preparing the synthetic building blocks 1,2,3, and 5 from tri-O-acetyl-~-glucal. These compounds, which were previously unknown or could only be prepared by complicated routes, now provide an easy and generally accessible entry to hexoses selectively substituted at C 2 to C4.