Examination of the stereochemistry of the nucleophilic addition of a dialkyl phosphite anion to a configurationally and conformationally defined nitrone on the one hand and of the 1,3-dipolar cycloaddition of the same nitrone on the other hand should reveal the direction of approach of both the anion and the 1,3-dipolarophile. The spironitrone 7 (Scheme 2) in which the nitrone function and the C(4)-0 bond (corresponding to the C( 1)-0 bond in 1) are about perpendicular to each other seemed to be appropriate. 3 4 5
For its preparation, the spirolactone 3 [3] (Scheme 2) was first reduced with diisobutylaluminium hydride (DIBAH) to give a crystalline hemiacetal4 (88%). Treatment of 4 with excess hydroxylamine in MeOH at reflux gave rapidly and quantitatively the oximes 5 (see below). Replacing the solvent by PrOH gave, after 20 h at reflux, a mixture consisting of 5, the dioximes 6, and some nitrone 7. During chromatography on silica gel, 6 was transformed into 7, which was obtained as a colourless oil in yields of up to 50% from 4.
The hemiacetal 4 showed mutorotation ([~]g = +46.5" -+39.9", CHCI,, 20 h). A 11 :9 ratio of two main anomers was evident from 'H-NMR spectroscopy. The configuration of the crystalline material is not known. The monooximes 5 are a 3:2 (E)/(Z)-mixture of mainly one (a-D?) hcmiacetal. In the 'H-NMR of 6, two signals for H-C(I) showed the presence of at least two diastereoisomers ( ( E ) / ( Z ) 3:2). The IR spectrum of 7 showed a characteristic nitrone band at 1580 cm-' and no OH or C=O bands. The MS showed M + at m /i 31 3 and further peaks at m / z 298 and 282, indicating loss of a CH, group. and of a CH, group and monooxygen, respectively.
The 'Hand T -N M R spectra of 7 showed the presence of a single diastereoisomer. No other diastereoisomer was detectable among the reaction products. The configuration ') We thank Dr. J . H . Bieri and Mr. R. Prewo for the X-ray analysis of 8 and 15. A more detailed report of these structures will be published later.