March 5, 1955
SOLVOLYSIS O F [CONTRIBUTION FROM THE
1129
METHYL-2-CYCLOHEXENYL ACIDPHTHALATE
DEPARTMENT O F CHEMISTRY,
UNIVERSITY OF WISCONSIN]
Stereochemistry of Allylic Rearrangements. 111. The Solvolysis of cis- and trans-5-Methyl-2-cyclohexenyl Acid Phthalate in Aqueous Acetone' BY HARLANL. GOERINGAND ERNEST F. SILVERSMITH^ RECEIVED AUGUST13, 1954 The polarimetric and titrimetric rates of solvolysis of cis-(Ia) and trans-5-methyl-2-cyclohexenyl acid phthalate ( I l a ) in aqueous acetone have been examined. The polarimetric rate constant (k,) and titrimetric constant ( k t ) are steady during the reaction and k,> kt under all conditions a t which the two were compared. The greater polarimetric rate is due topartial racemization of the reactant, by an intramolecular process, prior to solvolysis. The kinetic behavior is consistent with the previous interpretation for internal return involving the formation of an ion pair intermediate which gives solvolysis product or returns to the racemic reactant. It is suggested that intramolecular (SNi') rearrangements of allylic acid phthalates which have been described in the literature are similar to the racemization involved in the present work and involve internal return from ion pair intermediates.
I n the previous paper in this series3 the kinetics and stereochemistry of the isomeric anionotropic rearrangement of cis-(Ia) and trans-5-methyl-2cyclohexenyl acid phthalate (IIa) in acetonitrile were described. Evidence was presented which indicates that the isomeric rearrangement involves ionization (alkyl-oxygen cleavage) to the symmetrical hybrid 5-methyl-2-cyclohexenylcarbonium ion and hydrogen phthalate ion followed by recombination to form the rearranged ester. The observation that racemization is faster than geometric isomerization for both optically active I a and I I a was attributed t o internal return4 which was interpreted in terms of ion-pair intermediates. I n order to obtain more information cowerning the importance and nature of ion-pair intermediates in the ionization of the isomeric 5-methyl-2cyclohexenyl acid phthalates we have investigated the solvolysis of these compounds in aqueous acetone solutions. In solvents of this type the isomeric rearrangement is completely diverted by the intervention of solvent and the previously developed4 method of comparing polarimetric and titrimetric solvolysis constants can be used t o determine the extent of internal return. x I
CHs
IC'> I \d
I \-7
I
I1 a, X = 02CCeH4C02H b , X = OH
The polarimetric (k,) and titrimetric (kt) firstorder rate constants for the solvolysis of Ia in 50, 60, 80 and 90% aqueous acetone and for IIa in 80% aqueous acetone are shown in Table I. The titrimetric constants (calculated from the integrated expression for a first-order reaction) were determined a t initial concentrations of 0.20 and 0.05 M from the rate of increase in acidity and were steady a t both concentrations over the ranges that the reactions were followed. Each value of k t in Table I is the average value (and mean deviation) of 6 or 7 determinations during the reaction. (1) This work was supported in p a r t by t h e Office of Ordnance Research a n d in part by t h e Research Commitee of t h e Graduate School with funds given by the Wisconsin Alumni Research Foundation. ( 2 ) Shell Fellow 1954; du Pont summer research assistant 1953. (3) H. L. Goering, J. P. Blanchard a n d E. F. Silversmith, THIS JOURNAL, 1 6 , 5409 (1954). (4) S. Winstein a n d R. Heck, ibid., 1 4 , 5584 (1952), and previous p a p e n in this series.
The polarimetric constants (k,) were determined from the rate of loss of optical activity of solutions of (-)-Ia and (+)-IIa. I n all cases the solvolysis of active I a and I I a resulted in the complete loss of optical activity. At concentrations of 0.05 M the constants determined from the relationship In (ao/at) = kat, where a0 and at are observed rotations a t zero time and time t, respectively, did not show any trends during the reaction. Typical experiments illustrating the first-order behavior of the polarimetric and titrimetric rates are shown in Fig. 1. The polarimetric solvolysis constants for
- 1.3
0.0
-1.4
-0.1
-(d -1.5
-0.2
u
-8
04" Q
Y
u
-1.6
-0.3
8
I
1
4
8 12 Time, hours. Fig. 1.-First-order rate of formation of phthalic acid (upper line, left-hand scale) and loss of optical activity (lower line, right-hand scale) during the solvolysis of (-)-cis-5-methyl-2-cyclohexenylacid phthalate (Ia) in 90% aqueous acetone a t 100'.
I a a t concentrations of 0.20 M showed upward trends during the reaction and k, for these experiments are given as ranges in Table I. In order to determine the cause of the increasing polarimetric constants the effect of phthalic acid (a product of the reaction) on the rate was investigated. The polarimetric constant for Ia a t a concentration of 0.20 M in 80% acetone containing 0.20 M phthalic acid did not show any trends during the solvolysis and as shown by Table I was somewhat greater than the final value reached without added phthalic acid. It thus appears that phthalic acid, either by increasing the ionizing power of the medium or by acid catalysis, causes the upward trend in the polarimetric rate. The titrimetric constant, which is steady a t concentrations of 0.20 M , also is increased (from 12.3 to 18.2 in 80% acetone) by 0.20 M phthalic acid. The upward trend of the polarimetric constants is puzzling in view of the fact that
HARLAN L. GOERING AND ERNEST F. SILVERSMITH
1130
VOl. 77
erize under the conditions of the solvolysis and consequently the composition of the isolated mixture of methylcyclohexenols is not informative TABLE I from a mechanistic viewpoint. cis-5-RiIethyl-2A N D TITRIMETRIC RATECONSTAXTS FOR THE POLARIMETRIC cyclohexenol (Ib) is isomerized in 5.5 hours a t 100" SOLVOLYSIS O F Cis- AND tranS-5-METHYL-2-CYCLOHEXENYL in 607, acetone (4.4 half-lives for the solvolysis ACID PHTHALATE IN AQUEOUSACETONEAT 99.95rt 0.05' of Ia) in the presence of 0.05 dl phthalic or chloroSolacetic acid to a mixture that has the same composivent (Acid % vol. phthal- of tion (55y0Ib and 45y0 IIb) as those isolated from yo ate) reacn. ace102 folka kt the solvolysis of each of the isomeric acid phthalates Cmpd. tonea M lowed Method (102 hr.-') (102 h r . 3 in 607, acetone. 5.00 .45 Pol. 5.37 f 0.22 Ia 90 Because of the acid-catalyzed (Ib does not isom5.00 28 Tit. 3.0s =!=O 07 Ia 90 Pol. 6.64 i 0.45 1Ia 90 5.00 61 erize in aqueous acetone at 100" in the absence 5.00 40 Tit. 3.92 zt .13 of acid) isomerization of the solvolysis products i t is IIa 90 20.00 55 POI. 9.7to7.0~ Ia 90 3.83 = .01 not possible to distinguish between acyl-oxygen 20.00 35 Tit. Ia 90 and alkyl-oxygen cleavage by the usual method of 5.00 81 Pol. 16.5 i 1.0 la 80 1 1 . 1 + .1 Tit. Ia 80 5.00 69 comparing the configurations of the acid phthalates 15 t o 18' 20.00 66 Pol. la 80 and the resulting a1cohols.j There is, however, 19.3 =!= . 4 20.00 53 T i t . Ia 80 other evidence which indicates that alkyl-oxygen 24.0 i 0 . 5 20.00 72 Pol. Ia 80' cleavage is involved. We have found that the 18.2 =k . 9 Tit. Ia 80' 20.00 ti8 70.9 i 3.3 5.00 89 Pol. Ia 60 ethanolysis of Ia involves exclusive alkyl-oxygen 54.8 * 1.2 5.00 81 Tit. Ia 60 cleavage (ethyl 5-methyl-2-cyclohexenyl ether is 63 t o 81b 20.00 87 Pol. Ia 60 the exclusive allylic product) and i t seems most 5 9 , 4 + 1.3 20.00 83 Tit. Is 60 likely that the same type of cleavage is involved 134 * 8 5.00 83 Pol. Is 50 Tit. 106 zk 2 Ia 50 5.00 75 in the hydrolysis in aqueous acetone. 161 t u 201' 20.00 95 Pol. Ia 50 The effect of change in solvent composition on 86 Tit. 105 3= 3 20.00 Ia 50 the solvolysis constants is consistent with an S N ~ ~ a Solvent compositions are based on the volumes of t h e mechanism (alkyl-oxygen cleavage). As shown pure liquids before mixing. * This is the range of the conin Fig. 2, when log kt and log k, are plotted against stant which drifted upward during the reaction. C T h e (D - 1 ) / ( 2 D 1)-where D is the dielectric solvent contained 0.2 M phthalic acid. constant of the solvent7-straight lines are obtained As shown in Table I the polarimetric rate exceeds as predicted by the relationship deriveds from the the titrimetric rate under all conditions where the calculations of K i r k w ~ o d . ~The ka and kt plots two were compared. The effect of varying solvent are nearly parallel with a slope of 2.5 which is composition on kt, k, and the ratio ka/kt is sum- similar to the value of 2.9 observed by Kochi and marized in Table 11. These data show that k, Hammond'O for the solvolysis of benzyl and p and kt are quite sensitive to solvent composition methylbenzyl 9-toluenesulfonates in aqueous aceand that the ratio k,/kt decreases as ionizing power tone. This suggests that the acid phthalates and p of the solvent increases. The effect of varying toluenesulfonates undergo the same type of ionizatemperature on k,, kt and k,/kt in 6070 acetone and tion in the rate-determining step. the activation parameters are shown in Table 111. Unsolvolyzed acid phthalate was recovered from partly solvolyzed solutions of I a and I I a and TABLE I1 examined to determine if geometric isomerization RATECONSTANTS FOR THE SOLVOLYSIS OF 0.05 M cis-5occurs during solvolysis. The first-order behavior METHYL-2-CYCLOHEXENYL ACID PHTHALATE I N AQUEOUS of the titrimetric process indicates that geometric ACETONEAT 99.95Z!C 0.05" isomerization does not occur; however, the ha, kt, difference between kt for Ia and I I a is small (Table Acetone, %, hr. hr. - 1 ka/kt I) and considerable isomerization would be required 0.0308 1.74 0.0537 90 to cause a detectable kinetic disturbance. The 1.49 .165 .111 80 acid phthalate recovered from a 0.20 LILTsolution 1.29 .709 ,548 60 of dl-Ia in 80% acetone after 5.8 hours a t 100' 1.06 1.26 1.34 50 was shown to be pure I a by infrared analysis. TABLE 111 By using synthetic mixtures it was found that 1% RATE CONSTANTS AND ACTIVATIONPARAMETERS FOR THE of the geometric isomer can be detected in Ia or
the titrimetric constants remain steady under the same conditions.
+
-1
SOLVOLYSIS OF CiS-5-METHYL-2-CYCLOHEXENYLACIDPHTHALATE IN 60% AQUEOUSACETONE AH*., AS*> 80" looo kcal. e.u.
k a (102hr.-') kt (10' hr.-I) ( k a - &)(102hr.-')
korlkt
8.15 6.30 1.85 1.29
70.9 54.8 16.1
27.6 27.6 27.6
-0 5 -0.6 -1.2
1.29
The solvolysis of I a and I I a results in the formation of a mixture of cis-(Ib) and trans-5-methyl-2cyclohexenol (IIb) together with phthalic acid. Unfortunately the 5-methyl-2-cyclohexenolsisom-
(5) (a) J. Kenyon, S.M. Partridge and H. Phillips, J . Chem. S O L . , 207 (1937); (b) M . P. Balfe, H. W. J. Hills, J. Kenyon, H. Phillips and B. C. P l a t t , i b i d . , 556 (1942). (6) C. K. Ingold, "Structure and Mechanism in Organic Chemistry," Cornel1 University Press, Ithaca, N. Y . , 1953, pp. 345 8. (7) T h e dielectric constants of t h e solutions were calculated from t h e d a t a of H S. Harned and B. B. Owen, "The Physical Chemistry of Electrolytic Solutions," N. Y.,1943,p . 118.
Reinhold Publishing Corp.. New York.
(8) S. Glasstone, K . J. Laidler and H. Eyring, "The Theory of R a t e Processes," McGraw-Hill Book Co.. Inc., New York. N . Y.,1 9 4 1 , PI'
419 ti. (9) J. G.Kirkwood, J . Chem. P h y s . , 2, 351 (1934). (IO) J. K . Kochi and G. S. Hammond, THISJ O U R N A L , 7 5 , 3413 (1953)
March 5 , 1955
SOLVOLYSIS OF METHYL-2-CYCLOHEXENYL ACID PHTHALATE
I I a by this analytical procedure. From the values of k, and kt i t can be determined that a t the time of isolation optically active acid phthalate would be racemized to the extent of 28%. In other words, 14y0 of the isolated acid phthalate was converted to its enantiomorph without simultaneous geometric isomerization. When the isolation experiment was repeated with the trans-acid phthalate I I a using the conditions described above for the cis isomer, the recovered acid phthalate was found to be pure IIa. From these experiments i t is clear that geometric isomerization does not occur during the solvolysis of Ia and I I a and that the loss of activity associated with solvolysis is due to the simultaneous formation of solvolysis products and racemic reactant. The specific first-order rate constant for racemization is ( k , - kt) and the effects of varying solvent composition and temperature on this constant are included in Table I11 and Fig. 2. The available information indicates that the racemization, measured by (k, - kt), which occurs during the solvolysis of I a and IIa, involves internal return as in the case of the solvolysis of exo-norbornyl p-bromobenzenesulfonate in acetic acid" and 3-phenyl-2-butyl-p-toluenesulfonate in ethanol, acetic acid and aqueous acetone.12 The previous i n t e r p r e t a t i ~ n ~ofp internal ~~~ return, which appears to be consistent with the present data, has been applied to the 5-methyl-2-cyclohexenyl system in chart I. According to this scheme, the solvolysis of I a or IIa (X = HOZCC~H~COZ-), involves the formation of the corresponding ion pair 111 or IV, which either is converted to solvolysis products or returns to racemic reactant. Presumably the conversion of the ion pair to solvolysis products involves dissociation of the ion pair followed by the reaction of water with the 5-methyl-2-cyclohexenylcarbonium ion. However, no information concerning this step in the solvolysis was obtained because of the instability of the products under the conditions of the solvolysis. The observed polarimetric specific rate constant (k,) corresponds to the rate of formation of the ion pair which is necessarily optically inactive because of its symmetry. The titrimetric specific rate constant (kt) and the CHARTI
solvolysis products
IIa
'd IV
(11) S. Winstein and D. Trifan, THISJOURNAL, 1 4 , 1154 (1952). (12) (a) S. Winstein and R. C. Schreiber, i b i d . . 1 4 . 2165 (1952): (b) see also D. J. Cram. i b i d . , 1 4 , 2129 (1952).
1131
r
0.25 I 0.00
- 0.25 -0.50 i
2
-0.75
M
-1.00
- 1.25 -1.50 -1.75
~1 0.474
0.470
0.478
(D - 1)/(20
+ 1).
0.482
0.480
Fig. 2.-The effect of varying dielectric constant on the kt) polarimetric k,, titrimetric kt and racemization (k, specific rate constants for the solvolysis of cis-5-methyl-2cyclohexenyl acid phthalate (Ia) in aqueous acetone a t 100": upper line, k,; middle line, k t ; lower line (k, kt).
-
-
rate constant for racemization (ka - K t ) are related to the rate constants in Chart I as shown by equations l and 2. (ka
+
kt = kaks/(kl k,) = kak1/(ki ka)
- kt)
+
(1) (2)
If this interpretation is correct the racemization without simultaneous geometric isomerization demonstrates that different ion pairs result from the isomeric acid phthalate and that each, by internal return, is converted exclusively to that isomeric acid phthalate from which i t was formed. Thus in the present system i t is possible to distinguish between internal and external return13 (recombination of the dissociated 5-methyl-2-cyclohexenyl cation with hydrogen phthalate ion), because the latter would necessarily result in the isomerization of a t least one and probably both of the isomeric acid phthalates. The lack of isomerization shows that the process responsible for the discrepancy between k , and kt is an intramolecular process and does not involve external hydrogen phthalate ion. This is indicated also by the kinetic behavior. It has been pointed out previously12athat the process responsible for a larger polarimetric than titrimetric rate cannot involve the external anion formed in the solvolysis providing both k, and kt are steady throughout the solvolysis. Other mechanisms besides the one shown in chart I can accommodate the known facts. For example, it is possible that the simultaneous processes, racemization and solvolysis, are completely independent and do not involve a common intermediate. The evidence indicates, however, that a common intermediate (an ion pair according to Chart I) is indeed involved in the two processes. This is indicated by the temperature independence of the ratio kor/kt (Table 111) and the small effect of varying solvent on this ratio (Table 11). In order to interpret the present data in terms of independent mechanisms for racemization (k, kt) and solvolysis (kt) the unlikely coincidence is (13) This is t h e same phenomenon referred t o as "Mass Law Effect" by C. K. Ingold, el at., J. Chcm Soc., 979 (1940).
VOl. 77
HARLAN L. GOERING AND ERNESTF. SILVERSMITH
1132
required that the independent processes have features an "internal acid catalysis" and was sugidentical activation energies and are similarly gested to accommodate the fact that salts of allylic affected by changing the ionizing power of the acid phthalate (anions) rearrange less readily than solvent (Fig. 2 ) . It indeed appears that the large the free acids. effect of varying solvent or temperature on both It now appears that the so-called SNi' rearrangeracemization (k, - Kt) and solvolysis ( k t ) and the ments observed by Kenyon and co-workers are of small effect on K,/kt can best be explained in terms the same type as the one involved in the racemizaof a rate-determining formation of a common tion process in the present work and involve the intermediate. formation of ion-pair intermediates (VIII) which The downward trend in k,/kt with increasing are converted to the rearranged ester by internal dielectric constant of the solvent (Table 11) return. This mechanism is consistent with the parallels previous o b s e r v a t i o n ~ ~ ~that a ~ ~ the ~ - ~ ~fact that the acid phthalates rearrange more ratio of internal return to solvolysis decreases as readily than their salts (anions) since the rearrangeionizing power of the solvent increases. I t has ment of the anion would involve a doubly charged been pointed that this behavior is consistent anion in the ion-pair intermediate. with the ion-pair intermediate interpretation since H the magnitude of k,/kt depends on the ratio kl/k. I (equation 3) and the latter would be expected to decrease as ionizing power increases. C&CH CHR
A
kalkt = 1
+ (kdka)
(3)
The conversion of optically active I a or I I a to racemic material by way of the ion-pair intermediate is a n intramolecular allylic rearrangement. Because of the symmetry of the present system the only structural change resulting from this rearrangement is the conversion of one enantiomorph to the other. I n unsymmetrical allylic systems this process would result in an intramolecular isomeric rearrangement and i t indeed appears that such rearrangements have been observed with certain allylic acid phthalates. Kenyon and coworkers have observed that under certain conditions optically active 1-phenyl-3-methylally1 acid phthalate5a (Va, X = OZCC~H~CO~H) and 1phenyl-3-ethylallyl acid phthalate16 (Vb) rearrange to the corresponding optically active 3-phenyl-lalkylallyl ester (VI). Kenyon and co-workersSa originally proposed an SNi'" mechanism involving a CBH~CHCH=CHR---+ C@H6CH=CHCHR
I
I
x v
VI
2
a , R = CHs b, R CzHs
four-membered cyclic transition state and later18 suggested the cyclic mechanism VI1 to account for the intramolecular nature of the rearrangement indicated by the retention of activity during the rearrangement. More recently Catchpole, Hughes and Ing01d'~ suggested a nine-membered cyclic (SNi') mechanism involving the free carboxyl group of the acid phthalate. The latter mechanism CH CsH6cdh/ \ H R
b\C/l
