Synthesis of some brominated cage molecules as possible precursors to pentaprismane

Eaton and co-workers ( 7 ) have recently described a synthesis of pentaprismane (15) via homopentaprismanone, obtained by photocyclization and hydrolysis of a diene ketal identical with 12 except that it lacks the bromine. Consequently, the synthesis of this last compound would furnish a very simple atternative route to pentaprismane. Our attempts to thus obtain pentaprismane foundered on the steps 11 or 14 12 but afforded a variety of bromoketals shown in Scheme I , some of which may prove to be useful intermediates in the synthesis of other cage or half-cage compounds. The preparation of the bromocyclopentadienone ketal 2 by bromination of 1, ldimethoxycyclopentane to 1, followed by dehydrobromination, was described by Eaton and Hudson in a preliminary communication without experimental details (2). I f left in solution, the diene 2 dimerizes, and the product 3 (from partial hydrolysis during workup) is obtained. However, the diene 2 can be intercepted by strong dieneophiles such as maleic anhydride, benzoquinone, or 2,3dichloro-p -benzoquinone to give the adducts 4-8, respectively. (Adduct 5 was reported by Eaton and Hudson (2)). The endo-endo stereochemistry of compounds 5 and 8 was proved by their photocyclization to the cage products 8 and 9. The anhydride 4 probably has similar stereochemistry; it was hydrolyzed by dilute potassium hydroxide to the dicarboxylic acid 7. Reduction of the diketone 8 with sodium borohydride gave the endo-endo diol 11, the stereochemistry being established by formation of the cyclic sulfite 13 (cf. ref 3). The diol reacted with p-toluenesulfonic acid in refluxing xylene to give the keto ether 10 (cf. ref 4) and with methanesulfonyl chloride in pyridine to give the dimesylate 14. Attempts to convert the dimesylate 14 to the diene 12 by treatment with iodide under conditions favorable in other reactions (3, 5) gave intractable tars.