In the mass spectrum, the highest mass ion was found at 615 and the base peak at 532 m/e. Although the material fragmented into a great many ions, the highest concentration of them clustered into two groups: one ranged from 530 to 536 and the other from 262 to 270 m/e. Since the starting material has a molecular weight of 302, the possibility that the ion at 615 m/e is the parent ion of a dimer of the starting material may be ruled out. The ion could possibly arise, however, from the fragmentation of a trimer formed from an intermediate generated during the reaction. The ions in the groups at rnle 530-536 and 262-270 could then be due to dimers and monomers, respectively, fragmenting from the trimers or present in the mixture obtained from the reaction.
Of the many structures possible, three monomeric ones that can be written for the ions and that appear to be consistent with the various spectral data and with the starting material are shown in IX, X, and XI.
Structures such as X and XI can arise from a rearrangement similar to that given for tbe dienol-benzene rearrangement (1 1).
It seems, therefore, that the initial reaction is a Friedel-Crafts reaction on the allylic alcohol group of I with the formation of an ally1 carbonium ion. This carbonium ion can be stabilized not only by the excess aluminum chloride but also by the nitrobenzene (12, 13). The stabilized carbonium-ion complex, conjugated with one or more double bonds, results in the red color observed in the analytical reaction. Red oils were reported previously in reactions of aluminum chloride, or bromide, with aromatic hydrocarbons (14). When the complex is destroyed by the addition of water, polymeric material is formed by attack of the carbonium ions on the various double bonds initially present in the molecule or generated during the reaction.