THEMECHANISM OF THE SPECIAL SALTEFFECT
Feb. 20, 1961
was omitted. Results of gas chromatographic analysis are given in Table I, run 35. A mixture of 88.3% 2-methyl-1-phenylpropene and 11.7y0 2-methyl-3-phenylpropene (total 0.013 mole) was refluxed for 35 hours with 2 moles of acetic acid containing 0.02 mole of sodium acetate. After the usual work-up, the prcduct was analyzed by gas chromatography (Table I , run 33). Gas Chromatographic studies were carried out on two instruments: Aerograph model A-90-C and Aerograph model A-110-C (Wilkins Instrument and Research, Inc.). Chromosorb (Johns-Manville, Inc.) was the stationary phase on all columns. Most of the work was done on two Carbowax 20M (Union Carbide Co.columns: 10' x a polyethylene glycol having a molecular weight of about 20,000) and 10' X 114"Ucon Polar (Wilkins Instrument and Research, 1nc.-a polypropylene glycol of unstated molecular weight). Approximate operating conditions for the Carbowax 20M column were 200 ml. He/min. a t a column temperature of 120" and for the Ucon column 60 ml. He/min. at 160'. Good separations of the olefins from each other and from the alcohols were obtained, except for the failure to separate V and VIIb. Attempts to separate sec-butylbenzene and isobutylbenzene with a wide variety of column materials failed. Analysis for the small amounts of VIIa required special
[CONTRIBUTION FROM
THE
DEPARTMENT
OF
855
techniques. Except with large samples, VIIa appeared a s a slope change on the tailing end of IV. Therefore a large (1.5-2.0 ml.) sample was introduced and the olefins and alcohols collected separately. Re-introduction of the olefins a t a higher injector temperature usually gave essentially complete resolution of IV and VIIa. In one run enough VIIa was collected for an infrared spectrum. The sample was shown, by comparison with authentic samples, to be mainly VIIa, with some IV as the only detectable contaminant. Cnder the usual conditions of the gas chroniatographic analyses, neophyl chloride partially decomposes in the injector block to hydrogen chloride and rearranged olefins Consequently, the products from neophyl chloride solvolyses were run through the gas chromatograph a t lower injector temperatures t o remove unreacted neophyl chloride (usuall) about 30% of the total product). The olefin fraction wa5 then resubmitted with the injector a t its usual temperature (about 50' above the column). LVhen reaction products containing formate esters were introduced without prior conversion of the esters to alcohols, partial decomposition to olefin and alcohol occurred (Table I, runs ll,lS,19). That this occurred in the gas chrom,ctograph was shown by submission of pure ester, which still gave olefin and alcohol peaks. The alcohols were stable under the conditions adopted for the analyses.
CHEMISTRY, UNIVERSITY O F CALIFORNIA, L O S ANGELES 24, CALIF.]
Salt Effects and Ion Pairs in Solvolysis and Related Reactions. XVI1.I Induced Common Ion Rate Depression and the Mechanism of the Special Salt BY S. WINSTEIN,PAULE. KLINEDINST, JR.,
AND
G. C. ROBINSON
RECEIVED AUGUST12, 1960 For the mechanism of the special effect of non-common ion salts in acetolysis of certain arenesulfonates, there were previous strong indications against some type of "physical" explanation. The occurrence of induced common ion rate depression by added common ion salts supports a mechanism for the special salt effect involving diversion of a carbonium ion pair to a new ion pair species, thus suppressing ion pair return. The kinetic form of the special salt effect and induced depression is appropriate for this mechanism. Also, it shows that the ion pair exchanges responsible for the special salt effect and induced depression phenomena involve ion pairs of the added salts. The parameters derived from the kinetics of the special salt effect and induced depression in acetolysis of 1-anisyl-2-propyl and 3-anisyl-2-butyl arenesulfonates provide considerable insight into the magnitude of various rate ratios and equilibrium constants connected with the solvolysis scheme. Ingold's interpretation of the nucleophilic substitution and exchange reactions of trityl chloride in benzene is discussed and criticized.
As illustrated in solvolysis scheme I, it is helpful manuscript, where solvolysis product ROS is to arise from solvent-separated ion to distinguish between three varieties of carbonium visualized1a,b,6,7 ion intermediate in acetolysis of various ~ y s t e m s , ~pair , ~ I11 and also the dissociated carbonium ion II' the intimate and solvent-separated ion pairs I1 and when the latter is formed. Return5p6of carbonium 111, respectively, and the dissociated carbonium ion intermediates toward covalent RX may occur ion IV. Scheme I is designed for systems such as from the dissociated carbonium ion stage (external those with which we are concerned in the present ion return) or from ion pairs (ion pair return).6 Ion pair return may be further dissectedfi~~ into (1) Previous papers in this series: (a) X, S. Winstein and A. H. Fainreturn from the intimate ion pair (internal return) berg, THISJOURNAL, 80,459 (1958) ; (b) XI,E. Jenny and S. Winstein, H d u . Chim. A&, 41, 807 (1958); (c) XII, S. Winstein, E. Allred and and return from the solvent-separated ion pair (exP. Klinedinst, Jr., page 48, Foreign Papers a t VIIIth Mendeleev ternal ion pair return). o f Pure and Applied Chemistry, Moscow, U.S.S.R., March, Congress Certain systems in acetolysis respond to the addiXIII, S. Winstein, S. Smith and D. Darwish, THIS JOUR1959; (d) tion of salts like lithium perchlorate with a comNAL, 81,5511 (1859); (e) XIV, S.Winstein, S. Smith and D. Darwish, Tcfrohcdron Lefters, 16,24 (1969); (f) X V , S.Winstein and J. S. Gall, bination of steep ~ p e c i a l ~ ~salt - ~effects ~ ~ , ~at, ~low ibid., 3, 31 (1960); (g) XVI, S. Winstein, J. S. Gall, M. Hojo and S. concentrations of added salt and the more shallow Smith, TEISJOURNAL, 82, 1010 (1960). linear normalldJ salt effects at higher concentra(2) Presented in part at: (a) V I t h Reaction Mechanism Conference, tions of salt. It is clear that the special salt effect is Swarthmore, Pa., Sept. 12, 1956; (b) VIIIth Mendeleev Congress of Pure and Applied Chemistry, Moscow, U.S.S.R.. March, 1959. concerned with prevention of ion pair return.la-cj6-9 (3) Research supported by the National Science Foundation. However, in acetolysis of several systems examined (4) Research sponsored by the Office of Ordnance Research, U. S. in detail, namely, the 3-anisyl-2-butyl,7 2-anisy-I Army. propyl, la 2-ani~yl-l-ethyl'~ and 4 - r n e t h o ~ y - l - p e n t y l ~ ~ (5) (a) W. G. Young, S. Winstein and H. L. Goering THISJ O U R N A L , 73, 1953 (1951); (b) S. Winstein and D. Trifan, ibid., 74,1164 (1952); arenesulfonates, the elimination of ion pair return
(c) S. Winstein and K. C. Schreiber, ibid.. 74, 2165. 2171 (1952). -(6) (a) S. Winstein, E. Clippinger; A. H. Fainberg and G. C. Robinson, ibid., 1 6 , 2587 (1854); (b) S. Winstein, E. Clippinger, A. H. Fainberg and G. C. Robinson, Chemisfry 6.' Z n d u s t y , 664 (1954); (c) S.Winstein, E.Clippinger, A. H. Fainberg, R . Heck and G. C. Robinson. THISJOURNAL, 78,328 (1956).
(7) S. Winstein and G. C. Robinson, ibid., 80, 169 (1958). ( 8 ) (a) A. H. Fainberg and S. Winstein, ibid., 7 8 , 2767 (1966); (b) S. Winstein and E. Clippinger, ibid., 78, 2784 (1958). (9) (a) -4.H.Fainberg and S. Winstein, ibid., 7 8 , 2763 (1956); ( b ) A. H. Fainberg and S. Winstein, ibid., 7 8 , 2780 (1956).
S. WINSTEIN,PAULE. KLINEDINST,JR., AND G. C. ROBINSON
886
Vol. 83
SOLVOLYSIS SCHEME I
I
ionization _I_jc
I1
*
jntimate ion parr
--+
ck-
dissociation
R@X@
Jr
solventseparated ion p a r
e
RCe))X9
t
dissociated ions
ks
k- 1
1
I
117
__I_)
ks
K1
RX
111
----f
k-:
Re
+X 8
4
KS'V
I internal return
ROS
ROS
external ion palr return
external ion return
ion pair return
by the special salt effect is only partial; a discrete fraction of ion pair return is not eliminated. The best explanations-' of the available facts is that external ion pair return is prevented by salts such as lithium perchlorate in the special salt effect, internal return still being permitted. What we shall be concerned with in the present manuscript is the mechanism and kinetic form of the special salt effect. Most of the observations will deal with the 1-anisyl-2-propyl toluenesulfonate systemlb*lO but two others, namely, thre0-3-anisyl-2-butyP and norbornylSbbrornobenzenesulfonate, will be touched on briefly. Mechanism of Special Salt Effect; Induced Common Ion Rate Depression.-For the mechanism of the special salt effect, one may consider either some type of "physical" explanation, or one involving specific chemical reactions between the added salt and ion pair intermediates in solvolysis. There are strong objections7 to the former type of explanation, the most compelling being the unique specificity of the special salt effect which makes a common ion salt ineffective.? Therefore, the latter type of explanation must apply. As regards the nature of the reaction between the special salt and the ion pair intermediate, the most attractive a priori possibility is an exchange reaction which diverts the solvent-separated ion pair, R e / / Xe, to a new species and prevents ion pair r e t ~ r n . ~ ~ JThis . ~ l exchange is formulated in eq. 1 and 2 as involving either an ion pair or only one of the ions of the added salt, respectively. R@IIXG + M e Y e I_ R@I)YG+ M e XB (1)
+
R@lI(Xa Y@,R@/lYe
+ Xe
(2)
The expected reversibility of such exchange reactions suggests a new common ion salt effect which should appear if exchange is indeed the mechanism of the special salt effect. I n exchanges 1 or 2, the ion pair, Me X@, or common ion, Xe, is produced. Addition of the common ion salt, MX, should depress the special salt-enhanced rate, thus giving rise to what may be termed "induced common ion rate depression." The occurrence of such induced depression would be support for the exchange mechanism of the special salt effect. (10) A. H. Fainberg, G C. Robinson and S. Winstein, J . A m . Chcm. Soc., 78, 2777 (1056). (11) S. Winstein, Erpcricnfio Supplcmcnlum ZI, 137 (1955).
Previous work on the acetolysis of l-anisyl-2demonstrated the propyl toluenesulfonate6c-10s12 absence of common ion rate depression due to accumulating toluenesulfonic acid, HOTS. In the present work, the addition of the common ion salt, LiOTs, resulted in a very shallow linear salt effect illustrated in Fig. 1. With this salt the acetolysis kinetics were cleanly first order, the data being summarized in Table I. This table also shows that the experimental points fit the linearg eq. 3 very well, the derived bt value being 4.0. It is clear that addition of LiOTs by itself kt = kto [ l
+ bt (salt)]
(3)
results in no visible rate depression but gives instead the shallow accelerative normal salt effect. Following a previous discussion,6Cthe present work confirms further that dissociation of the solventseparated ion pair I11 may be neglected in acetolysis of 1-anisyl-2-propyl toluenesulfonate. TABLE I NORMAL SALTEFFECT OF LITHIUM ~TOLUENESULFONATE IX ACETOLYSISOF O.OI.OO M 1-P-ANISYL-2-PROPYL TOLUENE SULFONATE AT 50.0" (LiOTs) , 10'M
-1OKkt.
Obsd.
set.-+
Ca1cd.b
Fit
%
of'kt
0 1.17 i 0 . 0 1 " 1.17 0.0 1.99 1.265 rt ,005 1.263 .2 3.99 1.36 rt .01 1.36 .o 5.98 1.45 i .01 1.45 .o Previously'* reported a t 49.i2': 106&t= 1.20 f 0.01; previously60*areported at 50.0': 10Skt = 1.198 i. 0.007. The value, 1.918, in ref. 10 is a typographical error. Calculated from 106kt = 1.17 [l 4- 4.01 (LiOTs)].
The test for induced common ion rate depression in acetolysis of 1-anisyl-2-propyl toluenesulfonate was performed by adding LiOTs to acetolysis solutions containing 0.005 M LiC104. This concentration of special salt is sufficient to raise kt to 2.47 X set.-', a value well above kto, which is 1.17 X sec.-l. Addition of the common ion salt at this lithium perchlorate concentration does indeed cause rate depression. Table I1 contains the results of this probe for induced common ion rate depression. Keeping the lithium perchlorate concentration constant a t 0.005 M and adding more and more LiOTs over the range 0.001-0.04 M results in progressive depression of the lithium perchlorate(12) s. Winstein, M. Brown, K . C. Schreiber and A. II. Schlesinger, THISJOURNAL, 74, 1140 (1052).
887
THEMECHANISM OF THE SPECIAL SALTEFFECT
Feb. 20, 1961
enhanced acetolysis rate constant, good first-order rate constants being obtained in each run (Table TI).
/ i
l o t
TABLE I1 INDUCED COMMON ION RATE DEPRESSIONBY LITHIUMp TOLUBNESULFONATE IN ACETOLYSISOF 0.0100 M ~ - ~ A N I S Y L %PROPYL ~TOLUENESULFONATE AT 50.0' (LiClO,), 10'M
(LiOTs),
10'M
-lOSkt, Obsd.
sec. -+ Fit, Calcd.0 % of k,
2.44 1.2 2.47 i 0.01 4.96 0 2.42 0.4 0.0997 2.41 i .01 5.04 2.37 .9 .300 2.35 f .01 5.04 2.35 .9 .399 2.33 f .01 5.04 2.25 f .02 2.24 .4 5.04 .997 2.14 .5 1.99 2.13 f .02 5.04 2 . 0 7 f .01 2.06 .5 3.99 5.04 0 Calculated from: k t / ( k , , t k t ) = 0.670 [475(LiC104)] /[1 78.48 (LiOTs)] with 105 k,,t = 2.85 [ l 27.4 (LiCIO,) f 4.0 (LiOTs)].
-
+
+
+
h
0
I
- -
I._ 3
6 [ s - l d ! ,
IC'
9
v
Fig. 1.-Effects of lithium perchlorate and toluencsulfonate in acetolysis of 1-p-anisyl-2-propyl toluencsulfonate a t 50".
summarized in Table 111, LiOTs was added by Clippingerlx in acetolysis of norbornyl bromobenzenesulfonate in the presence of sufficient lithium perchlorate (0.05 116)to bring about a considerable rate enhancement. The addition of LiOTs instead of LiOBs permits one to detect not only a depression of initial rate constant but also any downward drift in rate constant due to conversion of alkyl brornobenzenesulfonate to alkyl toluenesulfonate during the acetolysis.6c Actually, neither effect of the added LiOTs was observed, only normal slight increases in rate constant being caused by the extra added salt. The results obtained with the norbornyl system confirm the explanation of the mechanism of inTABLE I11 duced common ion rate depression. Since induced SOME ACETOLYSIS RATESAT 25.0' OF 3-ANISYL-2-BUTYL depression counteracts the special salt effect, and AND CXO-NORBORNYL ~BROMOBENZENESULFONATES the latter is concerned with elimination of ion pair Added Concn., (LiOBs), 106kr, 10skt, return from the solvent-separated ion pair 111, ROBS salt 101 M 1O'M sec. -1 calcd.. neither the special salt effect nor induced depression dl-threo-31.96 i 0.03 1.90 can be expected to occur with a system which does Anisyl-2- LiClO, 0.10 2.68 i .08 3.03 not show external ion pair return in acetolysis. butyl',' LiC10, .50 4.69 f .16 4.62 Norbornyl bromobcnzenesulfonate is thought to be LiCIO, .50 2.50 3.47 i .18 3.65 just such a system.l LiCIO, 1.00 5.70 i .07 5.48 While the occurrence of induced common ion LiClO, 1.00 1.00 5 . 1 4 4 .17 5.05 rate depression represents qualitative support for a LiClOc 1.00 2.50 4.80 i .20 4.67 mechanism of the special salt effect involving LiOAc 1.00 2.83-2.120 diversion of the solvent-separated ion pair TIT to a LiOAc 3.00 3.58-2. 82d new species by exchange, only the kinetics of the exo-Nor9.0 special salt effect will disclose whether the exchange bor n y P 13 2.50' 9 . 8 mechanism involves an ion pair or a dissociated ion 26.70 LiClO, 5.00 of the salt, and whether it accounts quantitatively LiClO, 5.00 2.50' 29.8 i 0.7 a Calculated from: kt/(Ke,t - Kt) 0.60 4- [829. for the special salt effect. However, before treat163 LiOBs)] with lo6 kext 5.05 [l 4-21.6 ing the kinetics of the special salt effect, it is neces(LiCIOd)]/[l (LiCIO,) + 5(LiOBsl. Cn. 0.01 M . 14 - 87% sary to inquire into the state of aggregation of the . _reaction range. - d 20 .- 78% reaction range. 0.01 - 0.03 M . various salts in glacial acetic acid solvent. 0 Interpolated value based on f LiOTs instead of LiOBs. Salts in Acetic Acid Solvent.-The chemical d a t a a t 0.03 and 0.06 M lithium perchlorate. literature contained some information on the naIt is of some interest to examine the combined ture of electrolytes in glacial acetic acid solvent effect of both the common ion salt and the non- when we began the solvolytic studies in this solvent. common ion salt, lithium perchlorate, in acetolysis More information appeared while the solvolytic of a system which does not show the special salt studies were in progress, and this information has effect in acetolysis. exo-Norbornyl bromobenzene- been supplemented with conductivity studies on sulfonate is such a system, lithium perchlorate various salts carried out in these laboratories. l a displaying essentially the normal pattern of salt As can be expected14 for a solvent with a dielectric effects on the titrimetric rate constant.9hJa As (14) H. 5. Harned and B. B. Owen, "The Physical Chemistry of Just as LiOTs fails to give common ion rate depression but does lead to induced depression in acetolysis of 1-anisyl-2-propyl toluenesulfonate, lithium broniobenzenesulfonate, LiOBs, behaves analogously in acetolysis of threo-3-anisyl-2-butyl p-broinobenzenesulfonate.6cJ Several pertinent measurements summarized in Table 111show clearly that induced common ion rate depression can be made important in acetolysis of the 3-anisyl-2butyl system. In fact, the phenomenon of induced common ion rate depression has been observed in these laboratories in acetolysis of every system which displays the special salt effect.
+
-
(13) E. Clippingcr, S. Smlth and P. Klinedinst. Jr., unpublished work.
Electrolytic Solutions," 3rd ed., Reinhold Publishing Corp., New York N. Y.,1968, Chapter 7.
s. I v I N S T E I N , P A U L E.K L I N E D I N S T , JR.,
888
constant of 6, conductivities of salts in glacial acetic acid indicate that they are very largely undissociated. At the lowest salt concentrations, only the equilibrium between ion pairs and dissociated ions shown in eq. 4 needs to be considered, ion pair dissociation constants, K 2 , being of the order of 10-7-10-6. For lithium perchlorate specifically,13 K2 is 1.6 X so that ion pairs are dissociated to the extent of ca. 10% a t a salt concentration of lo-" -If and less than 2y0 a t 5 X l o p 3 M . For other salts, such as lithium toluenesulfonate and acetate and tetrabutylammonium perchlorate and toluenesulfonate, K2-values13are even lower, and ion pair dissociation is more nearly negligible. At higher salt concentrations in acetic acid solvent, ion triplets, quadruplets and higher aggregates become important. In fact, for typical salts in acetic acid, there are minima in the equivalent conductance-concentration plots. At concentra tions above these minima, which usually occur a t ca. lo-, 1M13t15the conductance is due to ion triplets and higher charged aggregates.l4 The association into ioii triplets and quadruplets may be discussed with the aid of triplet and quadruplet dissociation constants, K 3 and Kdr according to eq. 5 and 6, and an equilibrium constant, K Q , for dimerization of ion pairs to quadruplets as in eq. 7. I n eq. 5 and 6 the two different ion triplets M@Ye;lI@and lie M@Y@are assumed to behave alike in dissociation or association. The equilibrium in eq. 7 is related to those in eq. 4-6, so that K Q can be expressed in terms of K,,K3and K, as shown in eq. 8. RZ
ye XI@ Y 3 M e
ye iue ye Ma Ye Me Ye
Rl
(4)
Ye + M e
I V ye ~
+ ye
(5)
M e Y 3 M@$- Y 3
ye M@ye J_ ya ~e ye + h i e K4
AI@
m + ya MB
(6)
KQ
m~ya M@y 3 h i e ye (7) K Q = K2/'(KJ
