ELECTROPHILIC SUBSTITUENT CONSTANTS
Sept. 20, 1958
amino)-or-methylstyrene, b.p. 125-127' a t 10 mm., was obtained. After recrystallization from methanol, the olefin, m.p. 74-75', was obtained in 23% yield (lit.l9 m.p. 74"). The olefin was treated with methyl iodide in benzene to form the quaternary salt, dec. 213', and the product transformed into the chloride with silver chloride. The chloride was dissolved in methanol and precipitated with diethyl ether, dec. 200'. Anal. Calcd. for Cl2HI8NC1: C, 68.07; H, 8.59; N, 6.62. Found: C, 68.30; H, 8.88; N: 6.85. The chlorides were prepared from the corresponding olefin and carbinol by treatment with hydrogen chloride a t room temperature in absolute ethanol. The meta derivative decomposed a t 160'; pura, at 130". Kinetic Measurements.-The procedures used for the kinetic measurements were similar t o those previously described"' with one exception. In the case of the carboxylate, the weighed potassium metal was added to 100 ml. of absolute ethanol under a nitrogen atmosphere to form the potassium ethoxide solution. Sufficient potassium ethoxide was present to neutralize the carboxylate group and to react with the acid developed in the hydrolysis. After the solution was brought to reaction temperature in a constant temperature bath, the weighed chlorides were added to the solution and mixed rapidly by vigorous agitation. After 3 to 5 min., a 5 ml. "zero time" sample was removed and placed in 100 ml. of acetone a t 0". The excess potassium ethoxide was titrated with 0.030 N hydrochloric acid using bromochlorophenol in ethanol solution as the indicator. Typical kinetic runs are summarized in Table V.
Acknowledgment.-We wish to acknowledge the valuable assistance provided by Mr. T . Inukai who synthesized several of the intermediates required in this study,
4979
TABLE V RATE DATAFOR THE SOLVOLYSIS O F t-CUMYL CHLORIDE CONTAINING CHARGED SUBSTITUENTS x , mi. a - x , ml. ki, hr.-1 Time, hr. A.
m-( Potassium carboxylate)-t-cumyl chloride in ethanol a t 25.0' 0 0.250 .417 .583 .916 1.250
5.690 4.270 3.840 3.330 2.910 2.670 2.310
m
1.420 1.850 2.360 2.780 3.020
Average
2.18 1.90 2.05 1.89 1.80
1.96 f 0.09
B. p-(Trimethylammonium chloride)-t-cumyl chloride in 90% aqueous acetone a t 55.0' 0 1.08 1.83 2.42 2.92 4.58 6.75 7.42 18.83
1.790 2.380 2.730 2.995 3.250 3.870 4.580 4.750 6.060 B ,500
m
4.710 4.120 3.770 3.505 3.250 2,630 1.920 1.750 0.440
Average
0.124 .121 .122 .127 .127 .133 .133 ,126 0.127 =k 0.0034
LAFAYETTE, IND.
(19) V. Braun, A n n . , 472, 43 (1929).
[CONTRIBUTION FROM
THE
RICHARD B. WETHERILL LABORATORY O F PURDUE UNIVERSITY ]
Electrophilic Substituent Constants1i2 B Y HERBERT c . BROWNAND 'IT.
oKAMOT03
RECEIVED FEBRUARY 28, 1958 An excellent linear relationship is observed between log (K/RH) for sixteen ntetn substitutcd t-cumyl chlorides and the Hammett meta substituent constants. The reaction constant, -4.54, is used to calculate a total of forty-one electrophilic substituent constants. The degree of constancy exhibited by these and other electrophilic substituent constants in representative aromatic substitutions and electrophilic side-chain reactions is examined. The data are used to estimate substituent constants for additional groups not examined in the t-cumyl system. These substituent constants correlate the available data on electrophilic aromatic substitution and electrophilic side-chain reactions with reasonably good precision.
The solvolysis of substituted t-cumyl chlorides offered a promising route to electrophilic substituent constants?" capable of correlating the available data on electrophilic aromatic substitution5a and side chain reactions.5b With additional constants now available, 4e-h i t appears appropriate to (1) Directive Effects in Aromatic Substitution.
XXX.
(2) This research was supported in part by a grant from The Petroleum Research Fund administered by the American Chemical So-
ciety. Grateful acknowledgment is hereby made to the donors of this fund. (3) Post-doctorate research assistant on a National Science Foundation Grant (G-2752). 1956-1957. (4) (a) H.C. Brown, J. D. Brady, M. Grayson and W. H. Bonner, THISJOURNAL, 79, 1897 (1957); (h) Y.Okamoto and H. C . Brown, ibid., '79, 1903 (1957); (c) H.C . Brown, Y. Okamoto and G. Ham, ibid., 79, 1906 (1957); (d) Y. Okamoto and H. C . Brown, ibid., 79, 1909 (1957); (e) H . C . Brown, Y. Okamoto and T. Inukai, ibid., 80, 4964 (1958); (f) Y. OkamotoandT. Inukai,ibid., 80, 4969 (1958); (6) Y. Okamoto, T. Inukai and H. C. Brown, ibid., 80, 4972 (1958); (h) Y. Okamoto and H. C. Brown, ibid., 80, 4976 (1958). (5) (a) H . C . Brown and Y . Okamoto, ibid., 79, 1913 (1957); (h) Y.Okamoto and H. C . Brown, J . Org. Chem., 22, 485 (1957).
assemble these coiistants and to examine their utility in treating the available data for electrophilic reactions. The Substituent Constants.-The reaction constant, p, for the solvolysis of the t-cumyl chlorides in 90% aqueous acetone a t 25' was previously calculated to be -4.62 by a least squares treatment of log ( k / k ~ )for the solvolysis of several t-cumyl derivatives (m-Me, m-Et, m-F, m-C1, m-Br, m-I and m-NOz) versus the corresponding Hammett substituent constants. With data for many additional groups now available, i t appeared desirable to redetermine the value of the reaction constant. In Fig. 1 is shown a plot of log ( k / k ~ )for the solvolysis versus the available values of the Hammett substituent constank6 Least squares treatment of all sixteen points for which both g m and log k values are available (MeO-, MeS-, Me-, Et-, ( G ) These constants are taken from the compilation by D . H . McDaniel and H. C . Brown, ibid., 28, 420 (1958).
4980
HERBERT C. BROWNAND Y. OKAMOTO 10
0 TEASED
ON
THERMODYNAMIC PK
0
rT
1-BU
FROM O T H E R PKO DATA
00
0 OMe
Me S
-10
+= W
0 A
-20
- 3.0 0
05
Tlr
.
Fig. 1.--Solvolysis of the meta substituted t-cumyl chlorides in 90% aqueous acetone at 25".
Vol. 80
TABLE I ELECTROPHILIC SGBSTITUEXT COSSTANTS Substituent UIll Dimethylamino Anilino Amino -0.16 Hydroxy ,1216 Acetylamino .21 Benzoylamino Methoxy ,ll5* Phenoxy ,2526 hlethylthio .15 Methyl - .06gb - .07 Ethyl Isopropyl t-Butyl - .l0 Phenyl .OF 3,4-CaHa(B-Naphthyl) Carbethoxymeth yl Chloromethyl n Hydrogen Trimethylsilyl - 0.04 Fluoro ,3376 Chloro ,373b Bromo .391b Todo ,359 Carboxy Carbomethoxy Carboethoxy .37 Trifluoro .42 Cyano .. 5 G ~ Nitro ,7105 Carboxylate ( K +) - .1 Trimethylammonium (CI-) 101
urn +
UP
-0.83 (-0.16)O
- .66 - .37b
-
,047
-
-
,158 ,066 .OG4 ,060 ,059 ,109
-
01 .268b ,320' .OO
- ,170' - ,150'
,399
-
,405 ,359 ,322 ,368 ,3BG , ,520 . ,502 .67-i ,028
(-
.6)"
,778 .5) ,604 ,311 ,295
- ,280
-
.14)'
0.011 ,352
.6)a
- ,197b - .O1
.Ol)G
n
((-
- .15Ib
,042'' ((
7 ' ,I
(-1.7)'' ( - 1.4)" (-1.3)" (-0.92)"
,256 ,179
- ,135 (.lOi)n
(-
n
0 -0.07
.Ol)G
0.021 ,073
.06zh ,227'
-
,2326 .18
,150 .1:55 ,521 ,489 ,482 ,012 ,050 . 790 - .0'?:i
,45 ,54 , 6GOh
. 778b .0
,114
t-Bu-, Ph-, MeSSi-, H-, F-, C1-, Br-, I-, Et02C-, ,359 .S8 . 406 FzC-, NC- and OzN-)' leads to a value for the rea btlues in parentheses are cstimateti U-constants from action constant of -4.45. data on electrophilic renctioris other than the solvol! The u-constants are based on benzoic acid ioni- t-cuniyl chlorides. Sce Table 11. Therinodynnmic tl:tt:t zation data of greatly variable quality.6 In some with estimated ut~certaintyof 4~0.02. C T11erniodyri:tinic cases the uncertainty in the pka values is sufficiently datum with estimated uncertainty of f0.05. large as t o introduce an uncertainty in the values which both um and urnG are available, the average of the substituent constants, u, of +0.10 unit. deviation becomes +0.027. Consequently, i t has been proposed that reaction In electrophilic reactions strong resonance interconstants be calculated only from substituent con- actions occur between electron-supplying substants based on thermodynamic dissociation con- stituents and the electron-deficient centers of the stants in water.6 The uncertainties in these values reacting systems. Consequently, i t has long been are considerably smaller, usually no greater than recognized that the Hammett substituent con10.02 unit. Least squares treatment of the eight are not satisfactory for treating electro"thermodynamic" points (MeO-, Me-, H-, F-, stants philic substituent reactions. Indeed, i t has been C1-, Br-, I- and OzN-) leads to a slightly different argued that the resonance interactions of the sub(The stituent will vary so markedly in different reactions value for the reaction constant, -4.54. thermodynamic point for the cyano group was as to render it impossible to represent the effect of omitted because of the disagreement between differ- the substituent by a constant, as required in the ent workers as to the correct thermodyriamic dis- Hammett treatmenL8 Consequently, it is of insociation constant for m-cyanobenzoic acid.6) terest to calculate uf-constants from data from Since the latter reaction constant is based on more different electrophilic reactions in order to compare precise data for the u-constants, this value, p = the range of variation observed in these constants. 4.54, has been adopted for the calculation of the For this comparisongwe have selected the following up +-constants. reactions : (A) solvolysis of the t-cumyl chlorides; In Table I we have summarized all of the avail- (B) uncatalyzed bromination of aromatic derivaable electrophilic substituent constants based on tives in acetic acid a t 25°8a-10; (C) uncatalyzed the solvolysis of the t-cumyl chlorides together with chlorination of aromatic derivatives in acetic acid the corresponding Hammett substituent constants. (8) (a) P. B . D. de la Mare, J . Chem. Soc., 4450 (1954); (b) J. K. I t is of interest to compare the extent of the and G . S. Hammond, THISJOURNAL, 76, 3445 (1953); (c) agreement between the um-constants derived from Kochi N. C . Deno and A. Schriesheim, ibid., 7 7 , 3051 (1955). For more ionization data and the um +-constants calculated favorable views see (d) D. E . Pearson, J. F. Baxter and J. C . Martin, by the above procedure. I n the case of the eight J . Org. Chem., 17, 1511 (195'2); (e) N. N. Lichtin and H. P. Leftin. "thermodynamic" points, the average deviation J . P h y s . Chem., 6 0 , 164 (1933); (f) PI'. C . Deno a n d W. L. Evans, THIS is hO.019. If we include all sixteen points for J O U R N A L , 79, 5804 (1957). (7) Because of t h e wide variation in the U-values for t h e trimethylammonium group and the large discrepancy observed in t h e rm-and u,+-values, this charged group is not included in this treatment. See discussion in ref. 4h.
(9) A similar treatment for a somewhat more limited series of suhstituents and reactions was recently reported b y Den0 and Evans, ref. 8f ( I O ) H. C . Brown and L. M . Stock, THISJ O U R N A L , 7 9 , 1 4 2 1 (1927), and unpublished work.
L'ALUES O F
TABLE I1 ELECTROPHILIC SUBSTITUENT CONSTANTS CALCULATED FROM VARIOUS REACTIONS A Solvolysis of l-cumyl chloride
Substituent
@-Dimethylamino p-Anilino p Amino @-Hydroxy p - Acetylamino p-Benzoylamino p-Methoxy @-Phenoxy $-Methyl p-Ethyl $-Isopropyl p-&Butyl p-Phenyl p-Trimethylsilyl p-Fluoro p-Chloro p-Bromo P-Iodo p-Carboethoxy p-Sitro p-Chloromethyl p-Carboethoxymethyl m-Amino M-Methoxy m-Methyl m-&Butyl nz-Fluoro m-Chloro m-Bromo m-Iodo m-Carboethoxy m-Chloromethyl m-Carboethoxymethyl
4981
ELECTROPHILIC SUBSTITUENT CONSTANTS
Sept. 20, 1958
P =
-4.54
-0.778
-
,311 ,296 ,280 ,256 - .179 .021 - .073 ,114 ,150 ,135 ,482 .790
-
B Uncatalyzed bromination P =
-12.14
C Uncatalyzed chlorination P =
-8.06
D
Reaction E
-1.58
-1.74
-0.969 ,750 ,750 ,826 ,645 ,278
-0.933
-
-
,240 ,262
.017 - .054 .066 - ,061 ,059 - .065 .352 ,399 ,405 ,359 .366
-2.00
-0.801 - .364 ,305 ,306 ,299 ,287 ,275 ,127 ,041 ,028 .028 ,201 ,070 ,231 ,091
-0.736 -0.368
(-
-
-
,323 ,378)
-0.278
-
- ,087 - .097
at 2 5 0 8 a , l l ., (D) nitration of aromatic derivatives by nitric acid in acetic anhydride or in nitromethane a t 0" and 25012; (E) protonolysis of substituted phenyltrimethylsilanes by perchloric acid in aqueous methanol a t 5 1 . 2 0 1 3 ; (F) ionization of diphenylcarbinols in aqueous sulfuric acidsc)f; (G) ionization of triphenylcarbinols in aqueous sulfuric acidsCsf;and (H) infrared absorption frequency for carbonyl stretching band.14 Ideally, the values of the reaction constant should be based on reaction data for meta substituents and the substituent constants based upon thermodynamic dissociation constants of benzoic acids, Unfortunately, the available data deal primarily with the effect of para substituents, with too few rncta substituents examined to utilize this proce(11) H. C. Brown and L. M. Stock, THISJOURNAL, 79, 5175 (1957), and unpublished work. (12) J. D. Roberts, J. K . Sanford, F. L. J. Sixma, H . Cerfontain and R . Zagt, ibid., 7 6 , 4525 (1954); C. K. Ingold, A. Lapworth, E. Rothstein and D . Ward, J . Chcm. SOC.,1959 (1931); C. K. Ingold and F. R . Shaw, ibid., 575 (1949); M . J. S. Dewar, T. Mole, D . S. Urch and E. W. T. Warford, ibid., 3572 (1956); M. J . S . Dewar, T. Mole and E. W. T. Warford, ibid., 3576 (1956). (13) C . Eaborn, ibid., 4858 (1956). (14) R. N. Jones, W.F. Forbes and W. A. Mueller. Can. J . Chent., 36, 504 (1957).
.167 ,012 ,133 .152 .OS5 .490
,100 .055 ,074
,062
G
F
H
ProIonization Ionization 1.r. spectra tonolysis of of of acetophenones Selected Nitration of ArSiMes ArnCHOH AraCOH p = value of P = P = P = P = U+ -6.22 -4.32 -4.74 -3.44 -12.30
-
-
-
-1.55 -1.39 -1.37 -1.14 -0.833 - ,469 - ,419 - ,721
-
,297
.079
-
,479 .477 .346 .337 .137 .01
-
.5
-0.325
-
.os1 ,107 ,101
,081 .162 ,162
.933
,732
,035 .164
-
-1.7 -1.4 -1.3 -0.92 .6 .6
-
.162
-
-
.01 .16 .16
-
.14 .01
- .083
,422
.426
,406 .406
dure. Consequently, we have plotted the results versus the electrophilic substituent constants (Figs. 2-4, 6, 12) and have estimated the reaction constants from a least squares treatment of the data. The individual values of the substituent constants were then estimated from the relationship uf = l / p log ( k l k ~ ) . The results are summarized in Table I1 to facilitate ready comparison. In the case of the p-methyl substituent, data are available for each of the reactions examined in Table 11. The seven reactions, B-H, lead to an average value for this electrophilic substituent constant of -0.308 0.021, as compared to the u+constant of -0.311 derived from the t-cumyl chloride system. Similarly six reactions (C-H) result in an average value for the 9-chloro substituent of 0.108 f 0.039, as compared to the u+-value of 0.114 derived from the t-cumyl chloride solvolysis. The p-methoxy group possesses a large resonance contribution in electrophilic reactions. Consequently, such a group might be expected to exhibit a large variation in the value of the constant calculated in this way from experimental data from reactions of widely different electronic demand.
*
HERBERT c. BROWKAND IT. OKAhlOTO
4982
VOl. 80
TABLE 111 REACTION CONSTANTS FOR ELECTROPHILIC SUBSTITUTIOXS IN BENZEXOID DERIVATIVES Reaction
Reaction constant, p"
sb
ii d
Fig
Ref.
Bromination of monosubstd. benzenes by bromine in HAC a t 2.5"" - 12.14 0.604 0.987 8 Xa, 10 2 Chlorination of monosubstd. benzenes by chlorine in HAC a t 2.5" - 8.06 .190 ,987 9 3 8a, I 1 Nitration of monosubstd. benzenes by nitric acid in nitromethane or acetic anhydride a t 0' or 25" - 6.22 .287 ,980 13 4 12 Bromination of monosubstd benzenes by hypobromous acid and 10 - 5.78 ,076 ,998 7 perchloric acid in 50% aq. dioxane a t 2.5' 5 Protonolysis of aryltrimethylsilanes by perchloric acid in 72% aq. methanol a t 50' 13 - 4.32 .326 .972 12 6 7 Protonolysis of aryltrimethylsilanes by sulfuric acid in HAC - 4.60 .248 .980 14 17 18 Brominolysis of aryltrimethylsilanes in HAC a t 25" - 6.04 .175 .990 14 8 Brominolysis of benzeneboronic acids in 20% HACand 0.40 M XaBr a t 2509 19 - 4.44 ,114 .992 16 9 Bromination of polymethylbenzenes by bromine in nitromethane a t 20 25" 10 - 8.10 .lo1 .995 11 a The reaction constnnt p is based only on u+-values derived from the t-curnyl solvolyses (Table I). * The standard The number of The correlation coefficient. deviation of the experimental measurements from the regression line. compounds involved in the calculation of p. E The para partial rate factor for biphenyl estimated from rate data and an assumed value of 50% substitution in the para position. f The biphenyl point deviates seriously (Fig.3) and was not included in the calculations. 0 The p-methoxy point deviates seriously and was omitted from the calculations.
However, even in the case of this group, the average value (from reactions B, E-G), -0.771 f 0.042, is in excellent agreement with the u+-value of -0.778 derived from the standard reaction. Finally, even in the case of the p-dimethylamino group, where resonance contributions must be very large, the available data from four reactions (B, E-G) exhibit good internal consistency, and lead to an average constant for this group of -1.72 f 0.13. The excellent agreement realized in these comparisonsls between the u+-constants obtained from the solvolysis of the t-cumyl derivatives and the average values calctilated from other reaction data is highly encouraging. Moreover, it suggests that data from these reactions may be used to calculate tentative u+-constants for groups which have not been examined in the t-curnyl system. Accordingly, such values are listed in the last column of Tablc 11 and are given (in parentheses) in the compilation 01ut-values in Table 1. Correlation of Electrophilic Aromatic Substitution Data.-Our primary objective in developing this set of u+-constants has been the quantitative correlation of directive effects in aromatic substitution. Accordingly, with the extended list of such constants now available, it appeared appropriate to te,t their utility for this purpose. In addition to the substitution reactions examined in Table 11, we have examined the utility of the u+constants for the following additional reactions: bromination by hypobromous acid. catalyzed by perchloric acid, in 50% aqueous dioxane a t 5Ool6; protonolysis of aryltrimethylsilanes by sulfuric (13) Only in t h e case of t h e P-acetylamino and t h e p-benzoylamino groups does there appear t o be a major discrepany between t h e u t
constants calculated from t h e bromination d a t a (B) and t h e corresponding constants obtained from t h e ionization of t h e substituted triarylcarbinols in sulfuric acid ( G ) . Possibly protonation of t h e acylamino groups occnrs in t h e latter reaction and modifies their ability t o stabilize t h e carbonium ion. (10) P. R.D. d e l a M a r c a n d J . T . Harvey, J . Chetn. SOL.,30 (19.3;); P. B. D. de l a Mare and J. T. Harvey, i b i d . , 131 (1957); P. B. D. de la Mare a n d hf. Hassan, i b i d . . 3004 (1957).
acid in acetic acid"; brominolysis of aryltrimethylsilanes in acetic acid a t 2 5 O I 8 ; brominolysis of benzeneboronic acids a t 2Sol9; and bromination of substituted polyinethylbenzenes in nitromethane a t 25O.20
Reaction constants for these reactions, obtained by a least squares treatment of the available data in each case, are summarized in Table I11 together with the results of a statistical analysis of the correlation realized in this treatment. Previously, insufficient data were available to permit the independent treatment of bromination and chlorination. Accordingly, data for these similar reactions were treated together as a single halogenation reaction with p = - 11.35.6a With additional data available, both u+-constants and partial rate factors, individual treatment is now feasible. Considering the enormous range of reactivity covered by the bromination reaction, the correlation with the c+-constants is excellent. On the other hand, no correlation is realized with the uconstants (Fig. 2). The situation is less satisfactory in the case of the chlorination results (Fig. 3 ) . The point for biphenyl appears to deviate seriously from the linear relationship.21 The calculated value for p , - S.06, also appears to be out of line with the value - 12.1 estimated for the related bromination reaction. This unexpected value for the chlorination reaction constant arises primarily from the data for the monohalobenzenes, which correlate poorly with the partial rate factors for the alkylbenzenes (Fig. 3). (17) Private communication by Dr. C. Eaborn. (1s) C. Eaborn and D. E . Webnter, J . Chem. SOC.,4449 (1957). (19, H . G. Kuivila and A . R . Hendrikson, THISJ O U R N A L . 74, ,5008 (1952); H. G . Kuivila and C. E. Benjamin, ibid., 77, 4834 (1955). (20) G.Illuininati a n d G. hlarino, i b i d . , 78, 4975 (1950); G. Illuminati, R ~ L ~ Y Sci., L Q'26, 2752 (1956). (21) I t should be mentioned t h a t this point is based on a n assumed value (60%)8& for para substitution. An unusually l a r g e amount of o ~ l l r osiihstitution in t h i s system would improve t h e situation. nr. P. H. I). de la l l a r e i s investigating t h e isomer distribution in this reaction.
ELECTROPHILIC SUBSTITUENT CONSTANTS
Sept. 20, 1958
10
-
4983
& e' plan to examine the chlorination of these monohalobenzenes. Of the simple aromatic substitution reactions, nitration has been the most extensively studied. Unfortunately, the investigators have utilized different solvents and different temperatures, so that the data are not strictly comparable. Considering the possible effects of these minor variations in the experimental conditions, the correlation of the partial rate factors with the u+-constants can be considered satisfactory (Fig. 4). It should be com-
0 p-OMe
e l-02
-2
-0.1
oOp-phOl
02
03
0 4
0,'s
01
0.2
0.3
04
0 m-Ma 10
o
10
. d
B
I
6
'
0
0
4
J
OI) 0
2
m-me
P-Ph
p-1-8u M :\e
-3.01 0
-2
ao
-0 5
-I 0
0-:
Fig. 2.-Bromination
of monosubstituted benzenes in acetic acid.
1.0
-
0.0
-
2
P- F
0
0 -1
3.0
-
p - ~ e
o
t-
-3.0
1
I
J
1.0
rn-f-Bu
O g
P-F
-01
00
Fig. 4.-Nitration of monosubstituted benzenes in nitromethane or acetic anhydride.
m-Me
0 -
p-CI
H
0 P-Bf
-1.0 -
30
:2 0 I
c1
10
J
00
-10
I -04
-02
u- .+
-
/
0
I
-03
-%
0
c
- 2 0-
,2 . 0 l9
P-
-
-10
0
p-Ph
p.t.Bu
0.
Me
\rn-
=
I
I
-03
-02
I -01
I
00
01
02
I 03
et Fig. 3.-Chlorination
of monosubstituted benzenes in acetic acid.
pared with the corresponding plot based on the Uconstants (Fig. 4). The recent results of P. B. D. de la Mare and his co-workers on the bromination of aromatics by hypobromous acid, catalyzed by perchloric acid, yields an excellent correlation with the u+-constants (Fig. 5). The determination of accurate partial rate factors for aromatic substitutions frequently off ers major experimental difficulties. In the case of some substituents, one isomer may be formed in predominant amount, and i t becomes a difficult problem to determine minor amounts of the product accurately. The protonolysis of aryltrimethylsilanes (Figs. 6 and 7), and the brominolysis of aryltrimethylsilanes and benzeneboronic acids (Figs. 8 and 9) offer a convenient means to determine partial rate constants by a direct experimental rate measurement, without the necessity for an accurate analysis for minor components of the reaction product. I n these four cases, the experimental
4984
HERBERT C . BROWNA N D Y . OKAMOTO
Vol.
0-02 I
-01
Q1
00
I
2.0
so
00
a5
v
l
u-Me
0
-
Io
I
,I
p-OMe
0
.
1
u-1 0u
I
P-Ph
10
p-Me
3,4.
C,H,
.c
I
m-Me
m-1-8"
0
0
0
60.0
h'
0
O
00
W
H
0 _I
m-Ph
-10
0
-
- 5.0
eo
W 0
m - M e\m-I-B"
-I
0
00-
-5.0 I
I
-04
-03
I
I
I
01
- 0 2
c
0 I
0 2
-05
03
U?
30
OT
-93
-0'
-92
90
9'
92
.
9'3
Fig. 7.-Protonolysis
,
of phenyltriniethylsila~~es b y sulfuric acid in acetic acid
(r:
-0.3
-02
-01 1
!
02 I
01 I
0 I
03
0 4
I
I
0 p-OMe
p-Me p-t-8"
m-Me
ow m-Me
00
05
00
UT Fig. 5.-Bromination of monosubstituted benzenes in aqueous dioxane by hypobromous acid, catalyzed by perchloric acid.
, t
I
I
I
00
p-Et p-,-pr
0
0 p-ScMe,
,.
0 o
O
p-s,Me,
0 P - P ~
,"
- 1.0
0 p-
0
-10
p-c I u-01
-2.0
0 P-F
0
m - Ph
I
p-8,
O
P-CI
0
p-!
m-CI
0
3 0
6
\,.
2 0
z
z 0
2
yo
U-F
P-SiMe,
m-Ph
1.0
0.0
-I 0 I
K
-0.3
t
Fig. 6.-Protonolysis of phenyltrimethylsilanes by perchloric acid in 727, aqueous methanol.
I
-0.2
I
-0.1
I
0.0
I
01
I
0.2
~ - C I I
h
0 3
cf Fig. S.-Brominolysis I_
of phenyltrimethylsilanes in acetic acid.
ELECTROPHILIC SUBSTITUENT CONSTANTS
Sept. 20, 1958
Reactions.-Recently we examined the utility of the of-constants in correlating the available data from electrophilic side-chain reactions.5b An excellent correlation was realized. With additional constants available, we have extended our examination of the correlation to include the new values and applied them to several representative electrophilic reactions : solvolysis of t-cumyl chlo-
Q: I
I
0
P-OMe
P-Me p-Ph
0
4985
-0.2-0.1
m-Me
I
-C02Ef
0.0 I
I
CI: 0.1
I
0.3
0.2
0.4
I
I
I
m-N02
0 m-Me
0 P-F
H 4.0
t\
p-l
I
P-CI O O p - B rm-l
p-Ph m-Me
P-l
- 2.0
m-C I
I
I -0.5
I
0.0
I
I
0.5
U?
Fig. 9.-Brominolysis
of benzeneboronic acids in 207, acetic acid.
rate data exhibit an excellent correlation with the electrophilic substituent constants.z2 Illuminati and Marino have developed an alternative procedure for determining the partial rate factors by a direct measurement of substitution rates.z0 Comparison of the rates of bromination of the monosubstituted mesitylenes with that of mesitylene itself yields the meta partial rate factors, m f . Similarly, comparison of the monosubstituted durenes with durene itself yields the para partial rate factors, p f . Excellent agreement between these rate factors and the a+-substituent constants is observed (Fig. 10). The fact that the partial rate factors for both methyl and the halogens agree so well with the a+-values suggests that the relatively poor correlation exhibited by the partial rate factors for these groups in the chlorination reaction (Fig. 3) may not be real. In conclusion the available data on aromatic substitution exhibit a gratifying correlation with the electrophilic substituent constants. Additional data of high accuracy will be required to provide a realistic evaluation of the precision which may be realized in predicting rates and isomer distributions in the substitution of aromatics. Correlation of Electrophilic Aromatic Side-chain (22) Only in the case of the P-methoxy group is a serious deviation observed in the brominolysis of benzeneboronic acid. This value is based on a major experimental extrapolation, which might have introduced some experimental complication. Alternatively, the reaction of this derivative may involve a change in the rate-determining stage. It is highly desirable that this exception be carefully re-examined to ascertain whether or not it constitutes a real deviation from the prapqsed correlation.
-0.3
-0.2
-0.1
0.0
0.1
0.2
0.3
0.4
t
R
Fig. 10.-Bromination
of polymethylbenzenes in nitromethane.
rides in ethanol a t 2 5 0 4 g ; solvolysis of triphenylcarbinyl chlorides in 40% ethan01-60% ethyl ether a t 0" 2 3 ; thermal rearrangement of a-phenylethyl chlorocarbonates in dioxane a t 80"2 4 ; acidcatalyzed rearrangement of phenylpropenylcarbinols in 60y0aqueous dioxane a t 30°2j; migration aptitudes in the acid-catalyzed rearrangement of diphenylmethylcarbinyl azidesz6; equilibrium constants for carbonium ion formation from triarylcarbinols in aqueous sulfuric acid a t 25°8c,f;equilibrium constants for the ionization of triphenylcarbinyl chlorides in liquid sulfur dioxide a t 0" 2 7 ; and the frequency shifts in the infrared stretching frequency of the carbonyl groups in substituted (23) A. C . Nixon and G. E. K . Branch, THISJ O U R N A L , 68, 492 (1936). (24) K . B.Wiberg and T . M. Shryne, ibid., 77, 2774 (1955). (25) E . A. Braude and E. S. Stern, J . C h c n . Soc., 1096 (1947). (26) S. N. Ege and K . W. Sherk, THISJOURNAL, 76, 254 (1953). (27) The data represent equilibrium constants for the ionization to form an ion-pair
(CaHs)sCCl-+
SO%,oo
(CsHa)aC+Cl-
The authors wish to express their appreciation to Professor N. Lichtin for his generosity in permitting this examination of his data prior t o the complete publication of his results.
1101. so
HERBERT C. BROWNAND Y . OKAMOTO
49SG
TABLE IV REACTION COXSTANTS FOR ELECTROPHILIC AROMATICSIDECHAINREACTIONS Reaction constant, p a
Reaction
Solvolysis of t-cumyl chlorides in 90% aq. acetone a t 25"
sh
xd
TC
Iicf.
0.154 0.992 8 ,178 .989 16 ,074 ,998 16 ,099 .994 8 ,031 ,990 14
-4.54' -4.45' -4.67 -2.57 -3.01
Solvolysis of t-cumyl chlorides in ethanol a t 25' 4~ Solvolysis of triphenylcarbinyl chlorides in 40% ethano1-60% ethyl ether at 60" 23 Thermal rearrangement of a-phenylethyl chIorocarbonates in dioxane a t 80' 2.1 Acid-catalyzed rearrangement of phenylpropenylcarbinols, XCeH4CH(OH)CHCHCHa, in 60% aqueous acid a t 30" -2.97 ,155 .078 7 25 Migration aptitudes in the acid-catalyzed rearrangement of diphenylmethylcarbinyl azides -2.69 .091 ,997 S 26 Equil. constants for carbonium ion formation from triphenylcarbinols in aq. sulfuric acid a t 25" -3.64 ,512 ,995 12h 8c,f Equil. constants for ionization of triphenylcarbinyl chlorides to ion-pairs in sulfur dioxide a t 0" -3.73 ,234 ,981 15 27 Frequency shifts in the infrared stretching frequency of the carbonyl group in substd. acetophenones -12.30 ,906 ,980 10 14.28 a The reaction constant p is based only on u+-values derived from the t-cuniyl solvolyses (Table I). * The standard deviaThe number of compounds tion of the experimental measurements from the regression line. The correlation coefficient. Obtained from a least squares treatment of the eight nzeta thermodynamic involved in the calculation of p. E This study. * Obtained from a least squares treatment of all meta points (Me*. points (MeO-, Me-, H-, F-, Cl-, Br-, I-, O&-). MeS-, Me-, Et-, t-Bu-, Ph-, MesSi-, H-, F-, C1-, Br-, I-, EtO,C-, FIC-, NC-, OzN). * Values for p-t-Bu and p-i-Pr derivatives omitted since authors report these values are less accurate and they deviate seriously from the linear relationship followed by remaining compounds.
With this one volysis of the benzyl t o ~ y l a t e s . ~ ~ exception, the o+-constants have been quite satisfactory in correlating numerous reactions of wide diversity. E 0.0
-05 rn- I- E
05
P- Ph
u o
!
m-Me O H
P-"2
