Recently, there has been considerable interest in the synthesis of cyclooctatrienediones['I and their valence bond isomerslZ1. In the thermolysis of bicyclo[4.2.0]octa-3,7-diene-2,5-dione ( 1 ) which gives tropone, we have suggested the intermediacy of cis,cis,cis-2,5,7-cyclooctatriene-1,4-dione (6) or its cis,cis,trans Quite recently, ChaLoner et a[. have reported the synthesis of the bis(ethyleneketa1) of (6)13]. We here report the synthesis of (6).
131Reduction of (I)[Zb1 with zinc in acetic acid at 90°C gave bicyclo[4.2.0]octa-7-ene-2,5-dione (2) (yield 65 %) whose synthesis by an independent and different method has recently beende~cribed~~l. Flash thermolysis of (2) The conditions for successful bromination were found as a result of the following observation: The 'H-NMR spectrum of (3) in CF3COOH-CF3SO3H (2: 1) exhibits signals at 6=1.50 (dd, J = l l , 9Hz, lH), 4.17 (dd, 11, 9Hz, lH), 5.32 (t, 9Hz, 1 H), 7.5 (m, 3H), and 8.04ppm (dd, 12, 7Hz, 1 H), thus indicating that the species in this solution is 1,6-dihydroxyhomotropylium ion ( 4 ) . Therefore, the appreciable downfield shift of olefin protons of (3) (Japan) [**I To whom all correspondence should be addressed. those in CDCl3 seems to reflect fast equilibrium between (3) and ( 4 ) . Bromonium ions would be expected to react at C-7 of ( 4 ) .Actually, treatment of (3) with NBS in CFKOOH-CH2CIZ(1 : 2)at room temperaturegave the unstable 2-bromo-5,7-cyclooctadiene-l ,4-dione (5). Dehydrobromination of (S), without isolation, with triethylamine afforded, after chromatography on silica-gel (CH,CI,-ethyl acetate, 95 :5), 2,5,7-cyclooctatriene-1,4-dione (6) The 'H-NMR signals of the diene protons of (6) and of (3) are almost the same, thus indicating that contribution of the polar form (6') can be neglected, i.e. (6) is not aromatic.(6) can be hydrogenated with PtO, to 1,4-cyclooctaned i~n e [~* 6J.Compound (6) is stable in aprotic solvents; however, it is unstable in protonic, particularly in acidic, solvents, giving complex mixtures (thin layer chromatography). Dissolution in CF,COOH at room temperature led to complete decomposition of (6).Although the mass spectrum of (6) is very similar to that of ( I ) , thermal decomposition of (6) at 500°C is only slight. This result rules out the intermediacy of (6) in the thermolytic conversion of ( I ) into tropone. The cis,cis,trans-isomer of (6). which, according to Wood~urd-Hqffmann[~~, is expected as the initial product of thermal ring-opening of ( I ) still, however, remains as a possible intermediate. Dilute solutions of the orange-red l-dimethylamino-2,2-bis(9-fluorenylidenemethyl)ethylene ( I )[I1 are slowly decolor-
