Optically active (1S)-l-methylbicyclo[2.2.l]heptan-2-one arise. The ee of both products (42-45 %) was found to agree (25) was prepared by starting from (lR)-bicycl0[2.2.l]heptan-within experimental error. These data point to the bridged 2-one (21). Photolysis of the tosylhydrazone 8 of 25 genera-2-methyl-2-norpinyl cation ( However, charge-destabilizing groups at C-3L41 or C-2rsl of 1 and similarly charge-stabilizing groups at C-1L61 or C-6m can cause (partial) endo-2-norbornyl + 2-norpinyl rearrangement (1 + 3), particularly when nitrogen (X = N2f) serves as the leaving group (Scheme I).Nitrous acid deamination of 1 -methylnorbornan-endo-2-amine (7) in water affords 17.2% of 1-methylnorbornan-endo-2-01 (14) and 19.6% of 2-methylnorpinan-2-01 (16) along with 62% of endo-2-methylnorbornan-exo-2-01 (18) whereas the analogous exo amine 9 yields 18 exclusively[*]. These data indicate that the endo diazonium ion 10 dissociates with participation of C-7 to give the bridged norpinyl cation 13 from which 14 is obviously derived. On the other hand, 16 can arise from 13 and/or from the open 2-methyl-2-norpinyl cation 17 (Scheme 2). If the origin of 16 were known, we would be able to assign the relative stability of 13 and 17. Why should we address this seemingly marginal problem? As a rule, tertiary carbocations are, at best, weakly bridged (15 is a thoroughly studied example ceptional since enhanced ring strain compensates for charge stabilization by the methyl group. To our knowledge, no analogous case has been completely analyzed so far (see below). Hopefully, exploration of the 13/17 pair will help us to understand the influence of structural effects on the stability of carbocations[l01.The symmetry of 17 (CJ, as opposed to the chirality of 13 (Cl), is instrumental in distinguishing these species. A previous approach using deuterium labels was frustrated by overlapping signals of 6-D and 7-D in 1 6 L ' l . This failure prompted us to reexamine the problem by means of a nonracemic substrate. To this end, it is sufficient to employ the tosylhydrazone 8 rather than the amine 7. Photolysis of 8 in 0.2 N NaOH affords 8.4% of 14 and 9.9% of 16 (about half of the yields obtained from 7), i.e. the diazo compound generated from 8 is protonated to give the diazonium ions 10 and 12 in a = 1 : 1 ratio[*]. Therefore, nonracemic 1 -methylbicyclo[2.2.l]heptan-2-one (25) was prepared.According to a published procedure, bicyclo[2.2. Ilhept-5-enendo-2-yl acetate (19) was treated with lipase from Candida cylindruceaI"1. After 30% conversion, the unreacted acetate 19 was separated from the alcohol (+)-20 which was hydrogenated and oxidized to ( In the present system, the (rapid) 1,2-alkyl shift (WagnerMeenvein rearrangement) does not affect the enantiomeric purity whereas the (slow) 6,2-H shift of the intervening carbocation(s) leads to racemization. For a compromise between yield and ee, the reaction was stopped at an intermediate stage (35% of 23, 40% of 22, and 25% of 24). Oxidation of the mixture with PCC provided (19-25 (88% ee) from...
