CYCLIALKYLATIONS OF SOME PHENYLALKANOLS 3571 ratio was detected until the amount of olefin exceeded 10%. Above 10% small differences were noted and the values indicated that the cis alcohol (equatorial OH) was being dehydrated a t a faster rate than the trans alcohol.Control Experiment.--The isolated product obtained from the reaction of methylmagnesium bromide (0.1 M ) with 4-t-butylcyclohexanone (Table I) was dissolved in 10 ml of ether and added to dimethylcadmium or dimethylzinc reagents under the conditions previously described for the reaction of in situ reagents with ketone. After hydrolysis arid work-up, percentages obtained by glpc were unchanged within experimental error.Registry methylzinc, 1, methylmagnesium iodide, methylmagnesium bromide, methyllithium, Ac~owledgment,-we thank the Central university Research Fund (University Of New Hampshire) for partial support.Tertiary 3-and 4-phenylalkanols underwent cyclialkylation readily a t room temperature with sulfuric acid to yield the corresponding indans and tetralins. Primary and (ordinary) secondary 3-phenylalkanols did not undergo cyclialkylation even a t high temperature with phosphoric acid; l,3-diphenyl-l-propanol cyclized readily. 3-P14ethyl-3-phenyl-l-butanol was dehydrated (mainly) by phosphoric acid a t 230°, but some cyclialkylation ocwirred. The requirement of a stable secondary or tertiary carbonium-ion intermediate is implied by these results; in the case of the last example, a l13-phenyl shift provides the required intermediate.
Some time ago Bogert and ~o w o r k e r s~~~investigated the cyclialkylations of some phenylalkanols and related alkenes. They identified their cyclization products by means of qualitative elemental analyses, comparison of physical constants, and characterization, of oxidation products. Since such identification methods are qualitative a t best and probably liable to failure in detecting minor components, we decided to reinvestigate some of the reported results, applying modern instrumental methods of separation and identification. Some of the experiments are repetitions of cyclialkylations reported previously, 2 , 3 but others are new ones designed to provide more insight into the mechanisms of these reactions.
Results and DiscussionThe conditions and results of our cyclization experiments are summarinied in the accompanying Table I. 3-Phenyl-1-propanol (1) and 4-phenyl-1-butanol ( 2 ) were reported by Bogert and Davidson2 to yield polymer and pure tetralin, respectively, upon treatment with phosphoric acid at high temperature. We obtained a product from the former that was shown to contain a trace of indan and a little n-propylbenzene, but the major constituents were the three isomeric phenylpropenes resulting from normal dehydration. Confirmatory evidence for the structure of the latter isomers was obtained by catalytic reduction which converted them to n-propylbenxene, The product from 2 was found to consist of tetralin (80%) and three lower boiling unidentified products (20%).4-Phenyl-2-butanol (3) was report...