1452 Rates and Relative Rates of Chlorination of Benzene, Toluene, and t-Butylbenzene in Carboxylic Acid Solvents. The Influence of Solvent on the Baker-Nathan Effect' zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQ Albert Himoe2 and Leon M. Stock Contribution from the George Herbert Jones Laboratory, Department zyx of Chemistry, The Unioersity of Chicago, Chicago, zyxwvu Illinois Received September 19, 1968 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCB 60637. Abstract: Rates, relative rates, and isomer distributions are reported for the noncatalytic chlorination of benzene, toluene, and t-butylbenzene in carboxylic acid solvents. The reaction solvent influences both the absolute and relative rates for chlorination. The rate ratio, zyxwvuts kp-~elkp-t.~~, is 2.8 in pivalic acid, 1.8 in propionic acid, 2.0 in acetic acid, 1.1 in formic acid, and 0.69 in trifluoroacetic acid; km-Me/km-t.Bu is 0.84 in acetic acid, 0.45 in formic acid, and 0.18 in trifluoroacetic acid. Added nucleophilic cosolvents havea modest influence on kp-~e/kp.l.Bu. The values 0.84 and 1.05 were determined for chlorination in trifluoroacetic acid-22 M water and zyx 51Sz trifluoro- acetic acid-acetic acid, respectively. The influence of reagent selectivity on the rate ratio was examined by study of the ferric chloride catalyzed chlorination of the hydrocarbons in nitromethane. The relative rate, kp.hfe/kp.t.Bu, is 1.3 for this reaction in accord with the expected dependence on reagent selectivity. Ground-state solvation effects appear significant. The partial molal heats of solution of toluene and t-butylbenzene in acetic acid and in trifluoroacetic acid (determined by Arnett and Douty) differ significantly. These results are discussed with ref- erence to the interpretations advanced for the Baker-Nathan effect. he experimental observation that the order of reac- T tivity of alkyl substituted compounds is altered from the "normal inductive order" of t-Bu zyxwvuts > i-Pr > Et > Me is known as the Baker-Nathan e f f e ~ t . ~ In- terpretations emphasizing the importance of hyper- conjugation* or of a polar effect in the inductive order coupled with steric hindrance to solvation5 are well known. A third, less familiar idea is that the Baker- Nathan effect originates in steric hindrance to bond contraction.6 The evidence for the effect and the rela- tive merits of these viewpoints have been thoroughly discu~sed.~-~ The apparent conclusion reached in these critical reviews is that the rate sequence for elec- trophilic substitution of the alkylbenzenes (p-Me > p-t-Bu >> m-t-Bu > nz-iMe > H) would offer secure evidence for the somewhat greater importance of C-H hyperconjugation relative to C-C hyperconjugation if other factors, in particular solvent-transition-state in- teractions, could be e x c l ~ d e d . ~ , ~ The suggestion that the reaction medium plays a major role in the origin of the Baker-Nathan effect5 (1) Acknowledgment is made to the donors of the Petroleum Re- search Fund, administered by the American Chemical Society, for sup- port of this work. (2) Fellow of the Petroleum Research Fund, 1961-1963. (3) J. W. Baker, "Hyperconjugation," Oxford University Press, Oxford, 1952. (4) R. S. Mulliken, C. A. Rieke, and W. G. Brown, J. Am. Chem. zyxwvutsr SOC., 63, 41 (1941). (5) W. M. Schubert and W. A. Sweeney, J. Org. Chem., 21, 119 (1956); W. M. Schubert, J. Robins, and J. L. Haun, J. Am. Chem. SOC., 79, 910 (1957); W. M. Schubert, J. Craven, R. G. Minton, and R. B. Murphy, Tetrahedron, 5,194 (1959); W. M. SchubertandR. G. Minton, J. Am. Chem. SOC., 82, 6188 (1960). (6) A. Burawoy and E. Spinner, J. Chem. SOC., 3752 (1954). (7) Conference on Hyperconjugation, Tetrahedron, 5, 105 (1959). (8) M. J. S. Dewar, "Hyperconjugation," The Ronald Press Co., (9) An excellent review is presented in the introductory section of the W. M. Schubert and D. F. Gurka, J. Am. Chem. New York, N. Y., 1962. accompanying article: SOC., 91, 1443 (1969). prompted our investigations lo of solvent effects on the nuclear, noncatalytic chlorination of the alkylbenzenes. Chlorination was chosen for study because, among other reasons,IO the arenonium ions formed in the course of this reaction fulfill the requirement for isovalent, non- sacrificial hyperconjugation. We measured the rates and relative rates for chlorination and used eq 1 where pfMe and the paru partial rate factors, are defined by the rate constants for substitution in the para position, kp-bIe and kp+Bu, relative to one position in benzene, to assess the variations in the substituent effects of the alkyl groups in different solvents. An approach of this kind is essential because the simple rate ratio, kp.Me/kp-L-Bur depends on the selectivity of the reagent, that is, on the reaction constant, p. Presumably, p measures the charge deficiency in the aryl fragment of the transition state. This charge deficiency is, accord- ing to current thought. determined by the electrophilic properties of the reagent and by the degree of nucleo- philic solvation in the transition state. Reagent elec- trophilicity is, in turn, solvent dependent. The fact that a linear free energy relationship correlates electro- philic substitution datd implies that the contributions of the alkyl groups, i.e., the ~TT values, should not de- pend on reagent electrophilicity. Thus, substituent effects originating in important variations in the nucleo- philic solvation of differently substituted transition states and manifest in the relative free energy of solva- tion of ground and transition state should be revealed by solvent dependent values of log pf""/log pff-Bu. We first examined the chlorination of benzene, toluene, and t-butylbenzene in six aqueous acetic acid (10) (a) L. M. Stock and A. Himoe, ibid., 83, 1937 (1961); (b) L. zy M. (11) R. S. Mulliken, Tetrahedrok 5, 253 (1959); 6, 68 (1959). Stock and A. Himoe, ibid., 83, 4605 (1961). Journal ojthe American Chemical Society / 91.6 March 12, 1969