G. S.HANMOND, C. E. KEEDER,1;. ’l’. FANG AND J. K. Kociir
5GS
have assumed tacitly that solvolysis involves a carbonium ion mechanism, but the observed increase in the iodine effect with increasing electrostatic field strength in transition state would also follow even if considerable emphasis were laid upon nucleophilic attack by solvent. Conversely, the results reported here do not establish limiting
AMES, IOWA
The Solvolysis of Benzyl Tosylates.
s.
so
character of benzyl solvolysis (see also reference 3). Acknowledgment.-We are grateful to the Office of Ordnance Kesearch for the support of this work and to the Institute of Atomic Research for infrared services.
[CONTRIBUTIOS I.’ROJI JIIB CHEMICAL LAIIORAIORE’
BY GEORGE HAMMOSD, CIIARLES
VOl.
E.I L E E D E R ,
OF THE I O W A S r A I E C O L L E G E ]
V.
Some Solvent Effects
FABIAX
‘r. FAKGAND JAY
K.
I(Ocl11
RECEIVED AUGUST14, 1937 The rates of solvolysis of several benzyl tosylates have been studied at three temperatures in a group of acetonc-water and dioxane-water solvents. T h e correlation of the rates as a function of solvent composition is best accomplishcd by a n erripirical function of the molar concentration of water. T h e substituent effects fit neither the Hammett equation nor Brown’s u+-treatment. The variations of substituent effects with medium indicate a systematic variation in a single mechanisin rather than a clear-cut separation of the reactions into two distinct mechanistic types. Interesting variations in the activation entropies also are noted.
In earlier work we found that the rates of solvolysis of substituted benzyl tosylates I in the solvent “55 volume %’’ acetone-water (prepared from 55 volumes of acetone and 43 volumes of water) showed interesting and qualitatively understandable deviations from the Hammett p-u relationship.’
x’/ - 1
‘
CH~OS02CJIaCI-I~-p
Donor substituents in the p-position supplied a larger driving force than would have been anticipated by a normal Hammett relatiomhip. The result is attributed to the resonance interaction between the substituent and an electron deficiency created a t the benzyl carbon atom during the course of the solvolysis. ’Atany other examples of similar deviations from the original Hammett relationship are k n o ~ v n . ~ * ~ One might hope to use the solvolysis of benzyl systems as a probe in the study of fine details of solvolytic processes. Variations in the magnitude of substituent egects might well provide a measure of the relative importance of carbonium ion character in the reaction transition states. If all other factors could be maintained constant the absolute magnitude of the reaction constant, p , could be expected to increase with increasing concentration of positive charge in the functional group.4 I n actual practice “other factors” can never be held constant and one must proceed semi-empirically while keeping a cautious eye on the importance of factors such as reaction molecularity, specific solvation effects and short range dielectric effects on the interaction between polar substituents and the site of the reaction. I n the present work me have chosen to vary the solvolysis media. Several tosylates were solvolyzed in a series of solvent mixtures a t three dif(1) J. K. Kiochi and G. S. H a n i m o n d , TaIs J O U R N A L7,5 , 3145, 3452 (1853). (2) H. C. Brown and I’.Okamoto, ibid., 79, 1013 (1057). (3) Y. Okamoto aud H. C. Brown. J . Org. C h r i ; i , 22, 48.5 (1057). (4) C. G. Swain, THISJ O U R N A L 72, , 4583 (1080).
ferent temperatures. The mixtures were made up by blending water with dioxane or acetone. Two different cosolvents were used in the hope that additional information could be inferred from the correlation of the rate data obtained from the two sets of media. An additional feature of the study has arisen subsequent to the completion of the work reported in this paper. Okamoto and B r ~ u - nhave ~ . ~ suggested that a single set of substituent constants, designated by u+,can be used to correlate the partial rate factors for a variety of aromatic substitution reactions with other electrophilic reactions such as the solvolysis of phenyldimethylcarbinyl chlorides6 and the ionization of triary1carbinoh6 Our original data’ are exceptional in that they do not fit a rather general relation~hip.~Since the existence of a set of unique a+-values is surprising in itself, we were particularly interested in reviewing our data. Results and Discussion Correlation of Rate with Solvent Composition.It is a well known fact that solvolysis rates rise rapidly with an increase in the concentration of hydroxylic constituents, especially water, in solvent mixtures. The reason for the increase is not clear since mater and related compounds should provide both nucleophilic and electrophilic solvation and will often increase the dielectric constants of solvent mixtures. I n an attempt to understand the basis for the observed effects we have sought some function of composition which serves to unify the effects of adding water to both acetone and dioxane. We investigated various functions of composition such as dielectric constant, water activity and mole fraction. I n no case were the results encouraging. The familiar dielectric function,’ (D 1)/(2D l), has been shown t o be a near linear
+
( 5 ) 11. ibid., 79, 1 9 , 1906 (1037). (6) N.
C. Brown. J. D. Brad?, 11. Grayson a n d W. 11. Bonner, 1897 (1057); H. C. Brown, Y.Okamoto and G. H a m , ibid.. (1957): and H. C. Brown and Y. Okamoto. rbid.. 79, 1909 C. Den0 a n d A . Schriesheim, ibid.,7 7 , 3051 (1955).
SOLVENT EFFECTSIN BENZYL TOSYLATE SOLVOLYSIS
Feb. 5, 1958
function of rate for acetone-water mixtures.' The fit to a linear relationship is only approximate and the complete lack of correlation between the two series of solvents is demonstrated b y Fig. 1. One
569
3 2
o ~ IO I1
' 1 ' 1 ' 1 l I ~ I 1I 1 I 1 12 13 14 I5 16 17 18 19 2 0 2 2 2 2 23 2 4 2 5 PARTIAL PRESSURE OF WATER IN MM OF MERCURY.
D-l
Fig. 2.-Benzyl tosylate solvolysis rates as a function of partial pressure of water: 0, acetone-water; 0 , dioxanewater.
m I .
by about a factor of two than those observed in acetone solvents containing the same water concentrations. A surprising feature of the correlation is the fact that the rates are faster in the dioxane solvents despite the fact that they have the of the reasons for choosing the particular solvent lower dielectric constants. mixtures was the availability of complete data for Precise interpretation of these data is not easy. the total and partial vapor pressures of the mixtures of both acetone and dioxane with water. The fu- The rates cannot be expressed by an integral kigacities of water in the dioxane-water mixtures were netic order with respect to water concentration. calculated from the data of Herz and Lorenz,8 and Third-order rate constants, which would be exthe fugacities of the acetone-water system were pected on the basis of the treatments by Swain and calculated from the data of Beare, McVicar and co-workers,10fit no better than fourth order. HowFergu~en.~ No corrections were made for gas im- ever, i t is worth noting that an average value of a perfections in either case. All available data for third-order constant would give calculated values log k for benzyl tosylate is plotted against the par- of reaction rates with deceptively small error except tial pressure of water in Fig. 2. The form of the a t the ends of the concentration range studied. The plot was chosen in preference to others (such as situation is reminiscent of that described by log-log) in order to make the data manageable Hughes, Ingold, Mok and Pocker" in their report since the values of k diverge much more rapidly of the kinetics of methanolysis of triphenylmethyl than those of p . The data show that it will not be chloride in benzene solution. I n the present work possible to develop any empirical function of the the concentrations of the hydroxylic cosolvent water activity which will correlate the data for the (water) are enormously larger than those of the hyseries of solvents. Furthermore, the rates for the droxylic constituent in earlier study. The simiwater-acetone series rise monotonously through a re- larity in the dependence of rate on the concentragion of solvenl composition in which the partial pres- trion of hydroxylic component in the two reactions suggests, but does not prove, a similarity in funcsure of water passes through a nzaximum. A similar treatment of the rate data as a function tion. The fact that the rates in acetone and dioxof the mole fraction of water is considerably more ane mixtures are reasonably correlated by a concenpromising but the best fit to a single, empirical tration function but not by an activity function function is shown by Fig. 3 in which the log k values implies that the participation of water in the transiare plotted against the molar concentration of water. tion states of the tosylate solvolysis does not disThe rates in the dioxane-water mixtures are faster lodge the water molecules from their typical involvement in the liquid structure. I n other words, (7) S. Glasstone, R. J. Laidler and H. Eyring, "The Theory of Rate the process of withdrawing water molecules from Processes," McGraw-Hill Book Co., Inc., New York, N. Y., 1041, p. Fig. 1.-Rate of benzyl tosylate solvolysis as a function of dielectric constant: open circles, acetone-water; solid circles, dioxane-water.
419 ff.
(8) W.Herz and E. Lorenz, Z.p h y s i k . Chem., 8140,406 (1820). (9) W.G. Beare, G. A. McVicar and J. B. Ferguson, J . Phrs. Chem.. 34, 1310 (1930).
(10) C. G. Swain, R. B. Mosely and D. E. Bown, THIS JOURNAL, 11, 3731 (1955). (11) E. D.Hughes, C. K. Ingold, S.F . M o k and Y.Pockcr, J. Chrm. Soc., 1238 (1957).
570
G.s. HAMRIOND, c. E. REEDER,F. T. F A N G
AND
J. K. KOCHI
1701.
80
the solution to make transition states does not resemble the process of vaporizing water from the solution.
40
Other geometric arrangements can be visualized, but the above seems particularly attractive. 36Fine details of timing, such as the arresting of the reaction a t an intimate ion pair stage, l3 are "washed out" of the kinetic picture in solvents of high polar32ity and we can only infer that the conversion of an intimate ion pair to looser, more heavily solvated 28pairs should be a rapid, and probably irreversible process. It is also reasonable to assume that, especially at low water concentrations, the nucleo24philic roles may be assumed by acetone or dioxane. Herein may lie the basis of the superiority of di+ 120oxane as cosolvent. The energetic picture presented by Swain remains only slightly modified, -0 but the specific mechanistic implications of the 16 treatment become totally ambiguous. The data for the rates of p-nitrobenzyl tosylate 12are also plotted in Fig. 3 . Again a single function associates the rates for the two sets of solvent mix08 tures, although the function is different from the one for the unsubstituted compound. I n particular, the rate rises more slowly in the mixtures of high water content. The third plot in Fig. 2 shows 0 4 t the behavior of p-methylbenzyl tosylate. The rates rise slightly more rapidly than is the case with OA 0'2 0'4 ok ole Ib 112 ,I4 the parent compound and the spread between the MH20. dioxane and acetone mixtures is slightly greater. Fig. 3.--Tosylate solvolysis rates as a function of molar There is evidence that the dominant role of water Concentration of water: open points, acetone-water ; solid concentration is not a coincidence unique to benzyl points, dioxane-water; circles, benzyl tosylate, a = 7 ; tosylates. The data for solvolysis of t-butyl brotriangles, p-methglbenzyl tosylate, a = 6 ; squares, p-nitro- mide in acetone-water and dioxane-water mixtures, benzyl tosylate, a = 8. which have been gathered by Fainberg and WinThe translation of such generalizations into a de- stein,I4 show the same effect. In fact, the curves scription of reaction mechanism must involve some obtained by plotting log k versus water molarity arbitrariness. I n one of the solvent mixtures (the for benzyl tosylate and t-butyl bromide are refastest of the water--acetone solvents) the diversion markably close to congruent, especially in the more of an intermediate, presuniably the benzyl cation, water-rich mixtures. However, superficial exby added nitrate and chloride ions was demon- amination is sufficient to show that the relationstrated.' The kinetic form of the effect was that ship does not serve to relate mixtures of water with usually considered as a classic proof of the rate-de- hydroxylic cosolvents to the acetone and dioxane termining formation of carbonium ions. Lye have systems. The values of b log k/d(H20) are larger found no evidence of any mechanistic discontinuity with non-hydroxylic cosolvents, reflecting the fact in any of the work so we would like to describe the that solvolysis proceeds a t a measurable rate i n influence of water in terms of the so-called "polar anhydrous hydroxylic conipouiids. Substituent Effects.-The variations in temperacosolvent effect." The superiority of water over acetone and dioxane is clearly demonstrated and ture and reaction medium effect large variations in there is a strong implication that dielectric effects the absolute reaction rates. The range of values are not of prime importance. I t would be very for benzyl tosylate is approximately 6000; that for easy to extend Swain's general formulation to in- p-methylbenzyl tosylate is 8300 within the measiirclude larger numbers of solvent molecules as has able Iimits (the reaction was too fast to follow in the been suggested by Hudson and Saville." The faster dioxane solvents a t 43"); and the range for electrophilic reactivity of a hydroxyl group would p-nitrobenzyl tosylate is 1300. The divergence of surely be enhanced by hydrogen bonding of the the rates as the reactions brconie faster shows that oxygen to a second hydroxylic molecule and similar sensitivity of the reactions to substituent effects cooperative action could increase the necessary increases with rate. In our earlier work we showed polarization of a nucleophilic solvent species. that p-NO, kind riietn suhstitucnts :aye a. w r y atlcx'I'ransition states containing aggregates such its (13) A . I€. I'ainberg a n d S . \Yinstein, 'l'ii~s J ( I I I I G ~ A I , , 79, IliO!l the following would be involved. 0
0
(12) R. F. Ilndson a n d
n. Savillc, .7.
C h i n Soc., 4121 (1986)
(1!137), and earlier papers in t h e series. (14) A. €1. 1;aitiberg a n d S. Winstein, i n i d . , 79, ltiO2 (1937).
Feb. 5 , 1958
SOLVENT EFFECTSIN BENZYLTOSYLATE SOLVOLYSIS
57 1
quate fit to the Hammett equation.'5 I n our present work we have assumed that the same relationship will hold and have taken the value of log kP--N02 - log k H as defining the value of p. The values so obtained are summarized in Table I. TABLE I HAMMETT p CONSTANTS FOR THE SOLVOLYSIS OF BESZYL TOSYLATES IN AQUEOUS SOLVENT MIXTURES - p-. Volume % of Cosolvent
Acetone aceton e .Icetone hcetone Dioxan Dioxan Dioxan Reference 1.
cosolvent
96.2 90.9 83.3 55.5" 76.9 66.7 55.6
25 0
350
45 0
1.36 1.48 1.60 2.20 1.99 2.15 2.47
1.35 1.43 1.62
1.39
1.85 2.08 2.33
1.95 2.11 2.34
1.44
1.62
There are no significant trends of p as a function of temperature, but a t any given temperature p
increases with a n increase in the absolute values of the reaction rates.l6 Examination of all the data for
Fig. 4.-Plot
of solvolysis rates of benzyl tosy-lates ill 55.6 volume per cent. acetone against u+.
anism can be diagnosed is important as a possible halogen substituents on an expanded scale shows contribution to the general theory of solvolysis. If two distinctly different mechanisms were inthat halogen substituents give systematic deviations from the Hammett relationship," but such volved in the reactions whose rates are plotted in deviations become insignificant in comparison with Fig. 4, we would normally expect that the two reacthe over-all range of effects and points for p-chloro tion types would show different response to changes fall very close t o the line defined by p-NO2 and the in the nature of the reaction medium. Specifically, parent compound although the rates are always one would expect that the ionization rates would faster than would be predicted by the normal u- slow down much more than the rates of a direct displacement if the water content of the solvents were value. The lack of correlation between benzyl tosylate reduced. As a symptom of the effect we have solvolysis and Brown's a+-treatment is shown by looked for an indication that the two straight lines a plot of the data for 55.6% acetone solvent against of Fig. 4 intersect farther to the right when the u+ as in Fig. 4. Brown3 has suggested that the data for slow solvents are plotted. Plots of all the curvature in the smooth curve drawn through the data show no such trend, but we have looked in points reflects a change in the mechanism of sol- more detail for a breakdown of the linear relationvolysis within the series. The two solid lines in ship between log k and u+-constants for benzyl Fig. 4 represent our best effort to resolve the data tosylate and the fast, p-substituted compounds. into two linear plots. If the line drawn through Since the point for hydrogen lies just above the the faster rates defines a series of compounds which intersection in Fig. 4 a change in the relative imreacts by a carbonium ion mechanism there is portance of two mechanisms should be revealed by some question as to the most appropriate way of the transfer of the rate for benzyl tosylate from the treating the rest of the data. Probably the best line of large slope to the line of small slope. The procedure is to follow the assumption that if u+- calculations of Table I1 show that the linear relavalues are not applicable then the Hammett u- tionship among p-CH3, p-F and the H as substituvalues must be if, as Brown suggests, only two ents remains perfect even in the slowest solvents. values of substituent constants are needed to cor- Since u+ correlates the rates for the three comrelate all data. Plots of the relevant data are in- pounds in all the media, and since p-fluorobenzyl cluded in reference 17. A plot such as Fig. 4 of tosylate is always faster than the unsubstituted all the data might lead one to believe that a reason- compound (Hammett u is positive for p-F), we able linear relationships exists among the data for must conclude that a carbonium ion mechanism the slower compounds. However, examination of persists with these compounds. In order t o mainthe data on an expanded scale show that the slopes tain the view that all of the slower compounds go change ~onsiderably.'~Nonetheless, i t is interest- by a different mechanism, we would have to accept ing to proceed with the assumption that a resolu- the view that the displacement process is as sensition into two Hamniett plots can be accomplished, tive (or niore so) to niediuni effects as is the carsiiicc the possibility that a sharp change in niech- bonium ion reaction. Changes in renctioii type must be more subtle and continuous than arc ini(15) I,. P. I I a m m e t t , "Physical Organic Chemistry," McGramplied by the "two mechanism" analysis. 1940, p. 198. Hill Book Co., Inc., New York, N. Y., (16) T h e t r e n d s established in this work lead us t o be1iei.e that thc As a further illustration of the general contiiiuity extrapolation procedure used t o estimate t h e solvolyhis r a t e of p of substituent effects we have made log-log plots methoxybenzyl tosylate in 55.6 volumc per cent. acetone' must have (not included in the text) of all of the data in Table led us t o underestimate t h e value. V. If one of the solvents of intermediate ionizing (17) F. T. F a n g . J. K. Kochi a n d C. S. H a m m o n d , THISJ O U R N A L , 80, 563 (leas). power, such as 83% acetone or 77% dioxane, is
TABLE I1 \'AI.U13S
01'
p IrROhl APPLICATION OF h I O D I P I E D I ~ A R l h l E T T
EQUATION3 TO
REACTIVE COMPOUNDS
96.2 A 90.9 :l 83.3 A 55.6 h 76.9 D 66.7 D
3.6 3.9 4.3 4.8 4.4 4.9 5.1 55.6 D A = acetone; D = dioxane.
3.6 3.8 4.3 4.8 4.3 4.0 5 . '1
chosen as a reference medium a series of apparently adequate straight lines is obtained. Systematic trends away from linearity are best demonstrated if the rates are plotted against the the rates in either the fastest or the slowest medium. The over-all results simply show again that the divergence in rates in the faster solvents is greatest with the most reactive compounds and that a series of unique p constants cannot quite keep up with the trend. If we compare the rate constants for benzyl tosylate solvolysis in .j5.f3yo acetone with those for the solvolysis of phenyldimethylcarbinyl chlorides in 90y0 acetone (used to define u+) we observe that the sensitivity to substituents is greatest among the fastest of the tosylates. However, the lowest sensitivity also comes in the tosylate series among the unreactive compounds. Table 111 illustrates the comparisons. T A ~ L111 E COMPARISOX OF ISDKIDUAL RATER A r I o s BETWEEN BESZYL TOSYLATE ASD P H E X ~ - L D I R I C T I I Y L CHLORIDE C A R ~ ~SOLI~YL
our data are most complete. The enthalpies vary much less than the entropies. The relative invariance of substituent effects with temperature is, of course, a reflection of the dominance of entropy effects. We can offer nothing approaching a complete account of these data. Taken in conjunction with the correlation of the rates with the volume concentration of water the dependence on entropy suggests that rate variations might well be considered in terms of the probability that tosylate molecules will encounter water aggregates of an appropriate size. Since the least reactive compounds have very high activation entropies the indication is that a system such as 9-nitrobenzyl, which is relatively ineffective in internal charge dispersion, has a relatively strong ordering effect on the solvent envelope. Such a eonclusion.may seem anomalous in view of the evidence discussed in the first section above which indicates that fewer water molecules are involved in the solvolysis of the p-nitro compound. The ordering effect on the smaller number of solvent molecules must be quite strong. While this is exactly the effect which would be espected if the compound were reacting by an S N 2 mechanism, the qualitative observation apparently cannot be taken as diagnostic for a definite mechanistic switch. All changes in activation parameters are gradual over the entire series of compounds and there is nothing reseinbling a sharp division into two groups. Acknowledgment.--We gratefully acknowledge the support of this work by the Office of Ordnance Research. Experimental
VOLYSES
Benzyl Alcohols.--Most of the alcohols were prcparcd, as previously described,'g by reduction of the corresponding benzoic acids. p-Fluorobenzyl alcohol was obtained in 3,37 3.53 P-CIlpO 7Syoyield, m.p. 22-23', b.p. 209' (730 mxn.), Z*"D 1.5080. --1.(i6 -3.50 p-NO* Anal. Calcd.: C, 66.60; H, 5.60. Found: C, 66.02; €1, 5.68. Benzyl Tosy1ates.-b'ith the exception of p-nitrobenzyl 'The low sensitivity aniong the unreactive tosylall of the esters were prepared by the reaction of atcs must indicate that a considerable fraction of tosylate purified p-toluenesulfonpl chloride (m.p. 68.5-69") with the the positive charge left by the departing anion sodium benzoxides in dry ether.19 The esters were recrysmust be dissipated to solvent. Such considera- tallized from either ether or benzene-petroleum ether tions were the basis of the hypothesis that the slow (Skellysolve B ) and were stored in ether or benzene solution. in dry solvents the esters arc stable for periods of tosylates react by a direct displacement mech- Stored months. Small portions were crystallizecl for use as needed. anisme3 The preceding arguments indicate that The only new compound in the series, p-fluorobeuzyl tosylthe additional involvement of solvent is not easily ate melted at 54-55' but was not analyzed because of its infitted to a mechanistic discontinuity. As an alter- stability in the solid form. $-h-itrobenzyl tosylate was by the method of Tipson.1Y.20 native we may assume that even though the grozs prepared Solvents, titrant solutions and lithium perchlor t e wcrc reaction mechanism, as indicated by tests for the prepared as described e l s e ~ h e r e . ~ J ~ existence of intermediates or other similar experiKinetic Measurements.-The rates were followed by the ments, remains unchanged the transition states internlittent titration method'z17 using potentiometric titrawith Beckman no. 1180-42 glass electrodes and calomel can change considerably with respect to such tion reference electrodes. Brom phenol blue v a s used in followcharacteristics as free energy of solvation and the ing the slow reactions. Stoichiometric points were located :lumber of solvent molecules involved in the first by potentiometric titration of toluciiesulfoiiic acid with trisolvation sphere. In short, we believe that all of cthylamine in tlie various solvent media. M o s l of the rewere c:rrried out in a cell which was fitted wit11 stirrer, the benzyl tosylatc reactions are SN1 reactions b u t :ictions therrnotiieter well, and inlets for the buret and electrodes. that most or all of them may not be limiting iii the Bccause of soivent evaporation the very slow reactions were sense of the discussion by Winstein, et a,l.I8 r i i n i l l ~ l a s stoppered ; flasks. Activation Parameters.-'Table I V ~ t i ~ ~ i t l i a r i ~ i ' : ;Rate Data: .-A11 rate constants ivcrr calculated from t h r ~ntrgr;ttvrl first-rrrtler r;tic htv. Illrlivitlual ruiis c l i o l v c d the data for eiithalpy and entropy of activ;ition ;is:! avcrogc deviations OF less t h a u 15;). 1)uplic:itc I t111s 114'
function of solvent for the fivc compounds for which
(18) S. Winstein, E.Grunwald a n d H. W. Jones, THISJ o v ~ x a r . 73, ,
2700 (1951).
.
(10) J . K. Kochi and G. S. Hammond, iurd , 76, 3443 (l!)53) (20) J . Tipson, J 0;'s C h i n . , 9 . 230 (1944).
EFFECTS OF PHENYL ON BENZYL TOSYLATE SOLVOLYSIS
Feb. Ei. 1955
573
TABLE IV PARAMETERS IN TOSYLATE SOLVOLYSIS ACTIVATION ----p-N0*-
-Substituent H --
-9-CI-
Solventa
A H
ASt
96.2A 90.9A 83.3 A 76.9 D 66.7D 55.6 D
19.4 19.6 18.3 18.5 19.1 20.4
31.2 23.8 22.1 23.8 20.1 14.4
AHS 18.9 18.9 18.9
ASS
AHt
19.1 19.4 19.8 1s.9 18.5 18.3
21.8 21.5 19.1
19.1 17.8
--p-P-
-p-CIIn-
AH^ 18.9 18.3 19.1 23.1 21.5
AS$
23.8 19.1 15.1 15.4 14.4 12.8
ASS
23.2 21.8 16.1 13.4 16.4
13.7 15.6 Numbers refer to volume per cent. of cosolvent; A indicates acetone and D indicates dioxane.
AH^ 19.1 19.8 20.0 18.9
ASt
18.4 12.4 8.4 9.4 8.4 6.7
18.3*
17.Xb b
Using 45" rate extrap-
olated from slower solvents.
TABLE V FIRST-ORDER RATECONSTANTS FOR SOLVOLYSIS OF BENZYL TOSYLATES IN AQUEOUSSOLVENT MIXTURES Solvent0
-$-CHI25"
96.2A 90.9 A 83.3 A 76.9 D 66.7D 55.6D
5.26 3.54 4.28 4.88 5.56 6.24
35'
45O
4.02 4.78 5.31 5.95 6.65
Number refers to volume
-9-F-y
PH25O
45O
25O
4.35 1.87 2.77 4.48 2.63 3.50 5.23 3.26 4.17 5.78 3.85 4.76 4.40 5.04 5.83 per cent. of cosolvent;
1.61 2.39 2.95 3.57 4.05 4.60 A refers
different workers with entirely new preparations and reconstructed apparatus gave rate constants which agreed to 4% or better after correction of a numerical error. All rate
[CONTRIBUTION FROM
THE
Benzyl Tosylates. VI.
log k,
+8
35O
46'
2.85 3.41 4.02 4.51 5.03
3.52 3.31 3.89 3.47 4.93 5.47
-p-CI-
7----fi-N02--
25'
1.50 2.24 2.78
45'
35'
2.74 3.26
3.76 4.23 4.33 4.76 to acetone and D to dioxane.
250
350
2.44 0.55 3.14 1.24 3.68 1.70 2.02 4.67 2.38 2.87 4.18 2.68 3.21
459
1.42 2.17 2.63 2.90 3.39 3.65
constants in reference 1 should be multiplied by a factor of 2.303. Our new data are summarized in Table V. AMES,IOWA
CHEMICAL LABORATORY OF IOWA STATE COLLEGE]
The Effects of Phenyl as a Substituent
BY GEORGES. HAMMOND AND CHARLES E. REEDER RECEIVEDAUGUST14, 1957 The isomeric m- and p-biphenylcarbinyl tosylates have been prepared and solvolyzed. The results are compared with those of an earlier study of the influences of 8-styryl and phenylethynyl as substituents. There is a strong qualitative similarity between 8-styryl and phenyl, although the former provides the greater resonance effects. Phenylethynyl is a unique group in that the values of u + and U R apparently have different signs. The entropy of activation in the solvolysis of p biphenylcarbinyl tosylate is less negative than usual.
We have continued our application of benzyl tosylate solvolysis rates to the study of the conjugative effects of unsaturated, all-carbon substituents. We previously showed' that the p-P-styryl substituent in I accelerated the solvolysis rate far more than would have been anticipated by consideration of the Hammett u-constant for the group. The ef-
considerations lead to the expectation that phenyl as a substituent should resemble P-styryl rather than phenylethynyl. Consequently, we now have made measurements with compounds 111 and IV. ~
~
C
H
, 111
0 O
T IV
~
CH~OTS
Discussion OCH=CH ~Results CandH ~OT~ The data for the rates of solvolysis of the biphenI
~
C
~
C
~
C
H
.
O
T
I1
fect of the p-phenylethynyl group in I1 was qualitatively of the same sort but was quantitatively much reduced. The reactivity of the corresponding m-substituted compounds was adequately correlated by the unmodified Hammett equation. The different results were attributed to the firmer binding of electrons by carbon atoms in the sp state of hybridization than by spz carbon atoms. Such (1) J. K. Kochi and G . S. Hammond, THIS JOURNAL, 76, 3454 (1953).
ylcarbinyl tosylates are gathered in Table I. Data
~for styryl and phenylethynyl as substituents are
included for comparative purposes. The behavior of phenyl as a substituent is intermediate between p-styryl and phenylethynyl. The solvolysis rate of I V is correlated very well by the a-constant for m-phenyl as determined by the ionization constant of m-biphenylcarboxylic acid4 as was the case with the other two substituents.' The extra driving forces supplied by the three (2) G . S. Hammond, C. E. Reeder, F. T. Fang and J. K. Kochi,
ibid., 80, 568 (1958). (3) J. K.Kochi and G . S. Hammond, i b i d . , 76, 3445 (1953). (4) N. N. Lichtiq end H. P. Leftin, ibid., 74, 4207 (1952).
