The scope and limitations of a modified Bogert-Cook cyclization was investigated. I t was found that the sequence is applicable only for the unsubstituted and the meta-substituted series.La genkralite et les limitations d'une reaction de cyclisation de Bogert-Cook modifiee ont t t e examinees. 11 a ete trouve que la sequence est applicable uniquernent aux series de composes non substitues ou substitues en m&ta.[ Recently we reported on a sequence (1) which could be regarded as a modification of the Bogert-Cook (2) reaction (1 -t 2 -+ 3). For obvious reasons this reaction would be more useful if it could be extended to compounds in which the aromatic nuclei are carrying potential functional groups. The present paper gives a full account on the unsubstituted example and reports on all compounds carrying one methoxy group in the aromatic ring. It has been found that the sequence is applicable only to the metasubstituted and the unsubstituted series. Starting with the ortho-and para-substituted Nef adducts 7a and b led to no useful products.
Results and DiscussionThe addition of the appropriately substituted sodium phenylacetylide to 2,6-dimethylcyclohexanone in liquid ammonia led to the acetylene alcohols l a and l b and 7a and 7b. Apparently only one stereoisomer was formed in these Nef reactions corroborated by the fact that 7a and b were obtained in crystalline form. Although o-methoxyphenylacetylene was previously prepared from benzofurane using a sealed tube (3, 4), we developed a more practical preparation starting with o-methoxyacetophenone. The so far neglected traditional method gives a 35% yield after purifying the product as the silver salt.The Rupe rearrangment of acetylenic alcohols l a and b gave unsaturated ketones 2a and b, respectively, in 60% yield. In both reactions a less polar unidentified polymer was also formed 'To whom correspondence should be addressed. whose formation could not be avoided under various conditions. Cyclization of unsaturated ketone 2a to the tricyclic product 3a was carried out in polyphosphoric acid in 90% yield. The appearance of a saturated carbonyl group (i.r. 1710 cm-') and a three-proton singlet at 8.93 T corresponding to the angular methyl group are evidence for the tricyclic structure in 3a.From 2b the cyclization in polyphosphoric acid led to a crystalline mixture of monoketone 36 and diketone 46. The latter compound must have been formed as a result of atmospheric oxidation. Due to the experimental features of the reaction it was impossible to exclude oxygen completely. Since the two compounds could not be separated by re-crystallization or chromatography the entire sample was subjected to SeO, oxidation leading to the pure diketone 4b in 75% yield.For diketone 4b there is an alternative structural possibility 6 ; a similar structure is possible for the corresponding monoketone 3b. It is well known from the vast information concerning aromatic substitutions that the formation of a 1,2,3-vicinal trisubstituted isomer is disfavored in comparison to the 1,2,4-trisubs...