摘要：

J. CHEM. SOC. PERKIN TRANS. I 1985 S,2 and S,2' Alkylation of some Gibberellin Allylic Lactones by Lithium MethyIcu prates Michael H. Beale AFRC Research Group, School of Chemistry, The University, Bristol BS8 I TS A detailed study is described of the reaction of methyl gibberellate (2), methylisogibberellate (4) and their corresponding tetrahydropyranyl ethers (3) and (5) with three cuprates, (LiCuMe,),, Li,Cu,Me,, and (LiCu,Me,),. The structures of the products, produced by displacement of the allylic lactone system by SN2or SN2'synand anti mechanisms, were deduced by extensive n.m.r. studies. In a recent paper Taylor and MacMillan' described the preparation of 2P-alkylgibberellins via the anti SN2' alkylation of the allylic lactone system in gibberellic acid (1) by lithium dialkylcuprates.This reaction was intriguing to us in connection with our project aimed at the synthesis of 10-methyl compounds.2 We realised that it might be possible to influence this reaction such that the displacement occurred to yield C-10 methylated products. This paper describes a detailed study of the reaction of methyl gibberellate (2) and some of its derivatives with three different lithium methylcuprates. These investigations resulted in conditions where the reaction could be made to proceed by &2 or &2' (syn and anti) mechanisms to give intermediates useful for the synthesis of 2P-, 2a-, or lop-methylgibberellins. (1) R-H ,R1=H (4)R = H (2) R= Me,R'= H (5) R THP (3 1 R- Me,R'-. THP Results and Discussion A careful re-examination of the product from the treatment of methyl gibberellate (2) with lithium dimethylcuprate in ether- tetrahydrofuran by t.1.c.and capillary g.1.c.-mass spectrometry revealed the presence of a second isomeric product (5%) along with the 2P-methyl compound (7)(95%) described by Taylor and MacMillan.' The new product was isolated, after treatment with diazomethane, by preparative layer chromatography and assigned the 10-Me structure (11) (see Scheme) on the following n.m.r. evidence. The 'H n.m.r. in deuteriochloroform showed two singlet methyl signals at 6 1.24 and 1.30, two methyl ester singlets at 6 3.63 and 3.68, an allylic carbinol proton (3-H) at 6 4.51 and a two-proton olefinic signal at 6 5.61 (1-H and 2-H) which was resolved into two doublets (J 10 Hz) at 6 5.34 and 5.52 when the spectrum was run in deuteriochloroform plus deuteriobenzene. The 13C n.m.r.(see Table 3) supported the assigned structure with C-1 and C-2 appearing as doublets in the off-resonance decoupled spectrum at 128.1 and 133.6 p.p.m. and the introduced C-10 methyl group as a quartet at 29.1 p.p.m. Confirmation of the lop-stereochemistry in (11) was achieved spectroscopically and chemically (see Scheme) as follows. 'H N.0.e. difference spectra were obtained for irradiation at each of the two methyl singlets at 6 1.24 and 1.30. This revealed an equal enhancement (11%) of the doublet at 6 2.78 (5-H) and no enhancement of the doublet at 6 2.30 (6-H) when either methyl group was irradiated.Thus both methyl groups are on the same side of the molecule as 5P-H. The higher field methyl singlet (6 1.24) was assigned to the C-lop methyl group on account of an observed n.0.e. of 5.3% to the olefinic signal at 6 5.61 and a similar enhancement of the broad signal at 6 1.82 (assigned to 9-H). Very little n.0.e. was observed from the 4P-methyl group (6 1.30) to 3a-H at 6 4.51. Treatment of (11) with manganese dioxide in acetone gave an enone (12) which was acetylated with acetic anhydride and toluene-p-sulphonic acid to give (13) (see Scheme). Reduction of this enone acetate with L-Selectride gave the 3a-allylic alcohol (14) and the saturated alcohol (15) in 47 and 30% yields respectively. For the allylic alcohol (14) n.0.e.difference spectroscopy as above revealed enhancements of the doublet at 6 2.79 (5-H) and of 3P-H at 6 4.29 when the methyl group at 6 1.25 (4P-Me) was irradiated, whereas irradiation at the other methyl singlet at 6 1.21 (10-Me) gave n.O.e.3 to 5P-H, to 1-H at 6 5.79 and to the b

展开