Transformations in the benzosuberone series

This investigation concerned transformations in the benzosuberone series, directed to the formation of non-benzenoid aromatic compounds related to benzotropone (VIII) having, however, the oxygen atom attached to the six membered ring (IX). Several routes were attempted to obtain the substituted benzotropilidenes required as intermediates. (p-methoxyphenyl) valeric acid could not be cyclised to m-methoxybenzosuberone. Different methods gave either the starting material or polymers. Ultimately this ketone was synthesised by a different and tedious route of little practical value. Unsuccessful attempts were made to introduce a functional group (nitro) at the m-position in 1:2 benzocycloheptene 3:7-dione; in contrast it was found possible to introduce (in low yield) a keto group at the 7-position in m-nitrobenzosuberone side products were also isolated and identified. Dehydrogenation of m-hydroxybenzocycloheptene was also investigated and found impossible with conventional agents, as the starting material was recovered except in the case of triphenylmethylperchlorate which gave a red pigment of unconfirmed structure. This lead to investigation of the stability and the effects on triphenylmethylperchlorate on various solvents. The best route to m-hydroxybenzotropilidene was found to proceed from the p-acetoxybenzosuberone, whith the photobrominated and dehydrobrominated to the olefin (XXIII R=OAc). The latter on borohydride reduction followed by dehydration gave m-hydroxybenzotropilidene; which on treatment with triphyenlmethylperchlorate gave the required tropylium salt. However, on basification of the salt the keton (IX) could not be purified, although its presence was inferred by its colour, and by reconversion to the parent tropilident. The corresponding acetoxy and hydroxyl benzotropone was synthesised by standard procedure. Reduction of 1:2 (3:6 dihydroxybenzo) cycloheptene 3:7 dione gave unexpected and interesting results at variance with recently published Polish work on the subject, but this diketone could not be converted to the corresponding dihydroxybenzotropilidene