Electrophilic aromatic reactivities via pyrolysis of 1-arylethyl esters. Part 13. Carbon–carbon hyperconjugation and the origin of the Baker–Nathan effect

Abstract

From rates of gas-phase pyrolysis of the appropriate 1-arylethyl acetates, σ+ values have been determined for the following substituents : p-cyclohexyl (–0.380); p-t-butyl (–0.365); m-t-butyl (–0.19). These values are the highest recorded for these substituents, whereas for analogous reactions carried out in solution, decreasing values of σ+ are required as the solvent is made more polar. The gas-phase value of σ+(p-But)/σ+(p-Me) is 1.26. the highest recorded for a reaction possessing an electron-deficient transition state which may be stabilized by resonance. In solution reactions the value of σ+(p-But)/σ+(p-Me) decreases with increasing solvent polarity. The data show conclusively that the Baker–Nathan order of electron release by alkyl groups is a solvation phenomenon, and also that carbon–carbon hyperconjugation is greater than carbon–hydrogen hyperconjugation; this latter is confirmed by the greater value of σp+–σI for t-butyl (0.291) than for methyl (0.244). It is therefore a paradox that Baker and Nathan proposed the important and well established concept of hyperconjugation, but for the wrong reason.In solution, electrophilic aromatic substitutions carried out in trifluoroacetic acid, e.g. hydrogen exchange, give results which most closely parallel those obtained in the gas phase; for hydrogen exchange the required σ+ values are: p-cyclohexyl, –0.338; p-t-butyl, –0.320; m-t-butyl, –0.175. The solvolysis of αα-dimethylbenzyl chlorides is seen to give exceptionally low σ+ values for bulky substituents (even more so for meta-positions than for para-) and this reaction (especially when carried out in aqueous acetone) is therefore exceptionally sensitive to steric hindrance to solvation, which is more important when the suhstituent is in the meta-position than when it is para. In this respect therefore, the solvolysis is not a good model for determination of electrophilic substituent constants, and recent conclusions based on this reaction and concerning electron release from the 2-exo- and -endo-norbornyl substituents may be invalid.