Enthalpy and entropy of activation for benzyl chloride solvolysis in various alcohol-water solvent mixtures Can. J. Chem. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/28/17 For personal use only.

H. S. GO LINK IN^

AND

J. B. HYNE

Departtnetlt of Clremistry, Vt2iversity of Calgcrry, Calgary, Alberta Received July 24, 1967 The first order rate constants for the solvolysis of benzyl chloride in a series of mixtures of methyl, ethyl, i-propyl, and t-butyl alcohols with water are reported at 40.05 and 60.50 "C. The AH* and AS* values are calculated using these rate constants and those previously reported at 50.25 "C (1, 2). The dependence of these parameters on solvent composition is discussed. Canadian Journal of Chemistry, 46, 125 (1968)

Introduction We previously reported the solvent dependence of activation volumes for benzyl chloride solvolysis in aqueous binary solvents (1, 2). The negative value of the activation volume was found to pass through a ininimum as solvent composition was varied. It was also found that the position and depth of the extremum varies in a characteristic manner as the nonaqueous component of the solvent is varied. We were interested in determining whether a parallel effect would be observed in the extremum behavior of the activation enthalpy and entropy, as has been demonstrated for other solvolytic reactions (3-10). Arnett et al. (11-13) have demonstrated that the extrema observed in AH* in aqueous ethyl alcohol can arise as the result of variation of solvation of the initial state, the transition state, or both. We have demonstrated that the behavior of A P o for benzyl chloride solvolysis in the solvent systems studied is due mainly to variation of solvation of the initial state (2). Recently, Martin and Robertson (14) demonstrated that AC*, for t-butyl chloride solvolysis in aqueous ethyl alcohol also shows extremum behavior. The question as to which state causes this extremum must await further calorimetric measurements, but by analogy with the AH:': behavior of this reaction (1 1) it is to be expected that the initial state solvation variation will be found responsible. If the extre~nain the various activation parameters are related to the same physical phenomena, it is to be expected that the variation of 1Present address: Division of Pure Chemistry, National Research Council of Canada, Ottawa, Canada.

these extrema with changes in the organic component of the solvent should parallel one another. It is the purpose of this work to investigate the reality of this expectation. Experimental The conductimetric technique for determining the solvolysis rates at 40.05 and 60.50 "C was similar to that reported previously (8).

Results The conductimetric data were analyzed by the method of Guggeilheim (15) using a least squares calculation (1). The observed first order rate constants at 40.05 and 60.50 "C are presented in Table I. Those at 50.25 "C have been reported previously (l,2). Activation enthalpies were calculated fro111 the rate constants in Table I using eq. [I]. Activation entropies

[I.]

AN* =

R ( [In (k2/kl) -

-

were then calculated by eq. [2] using the rate constants [2] AS*

= R[ln

+

k - In (~Tllz) AH* IRT]

at 50.25 "C. The results for both A H and AS* are presented in Table 11. The reported error limits were estimated by the method previously reported for activation volumes (16). The use of only two rates at temperatures 20" apart for the evaluatioil of AH* requires some justification. Robertson and Scott (17) have reported a value of AC*, of -40.4 cal/deg mole for benzyl chloride sol~olysisin water. The Arrhenius plot of In k versus 1/T is therefore

CANADIAN JOURNAL OF

TABLE I Rate constants for the solvolysis of benzyl chloride

Can. J. Chem. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/28/17 For personal use only.

105

Mole fraction

x R (s-1)

40.05 "C

CHEMISTRY. VOL.

TABLE I1 Activation parameters for the solvolysis of benzyl chloride Mole fraction

AS"5~.25 (cal /deg mole)

Methyl alcohol 20.3 1 0 . 3 19.5410.08 19.3 1 0 . 3 19.0 1 0 . 3 20.1 1 0 . 3 Ethvl alcohol

0.000 0.100 0.200 0.300 0.400

Ethvl alcohol

i-Propyl alcohol 18.26k0.02 18.45k0.07 20.8 k 0 . 0 3 20.16+0.07 21.0 1 0 . 2 t-Butyl alcohol 17.610.2 20.0+0.2 22.210.2 21.2k0.3

i-Pronvl alcohol

t-Butyl alcohol 1.630 1 0 . 0 0 1 0.236 1 0 . 0 0 2

~ f I : k 50.28 (kcal /mole)

60.50 "C

Methvl alcohol

0.050 0.100

46, 1968

9.78 1 0 . 0 4 1.80310.008

nonlinear. By taking the slope of the straight line drawn between the two In k values at 40.05 and 60.50 "C as the tangent to the In k versus 1/T plot at 50.28 "C a very good value for AH* at the median temperature is obtained. Although the activation enthalpies are associated with a mean temperature of 50.28 "C, the difference of 0.03" is negligible for calculating the AS" values at 50.25 "C. Indeed, the AC*, value of -40.4 cal/deg mole reported by Robertson and Scott (17) for benzyl chloride solvolysis in water predicts an error of only 0.005 e.u. in AS* due to this temperature difference. Comparison of the AH" and AS" values in Table I1 with those reported by other workers shows good agreement. Robertson and Scott (17) report values of 20 370 cal/mole for AH*50.28 and -12.4 e.u. for AS:r50.25 in water. Hyne, Wills, and Wonkka (8) report AH*51.65values of 20 367 cal/mole in water and 19 555 cal/mole in 0.203 mole fraction ethyl alcohol. Discussion Entlzalpy of Activation The dependence of AH* on the composition and nature of the solvent is illustrated in Fig. 1.

-12.510.9 -16.110.3 -18. 1 1 . -20.1f0.8 -18.210.8

-20.72k0.06 -23.1 1 0 . 2 -19.1 1 1 . -22.3 1 0 . 2 -20.6 1 0 . 5 -24.5i0.7 -20.53~0.5 -16.4k0.7 -20.7k0.8

It can be seen that in each of the four systems a minimum occurs. The observed minimum in aqueous methyl alcohol is contrary to the findings of Tommila et al. (18-20), who observed no minimum in this medium and concluded that both the substrate and the organic component of the solvent had to possess large hydrophobic groups in order that a minimum be observed (21). The appearance of the minimum in aqueous methyl alcohol in our work signifies that this explanation is an oversimplification. Nonetheless the depth of the minimum in AH", relative to pure water, does increase as the hydrophobic character of the alcohol molecule increases. The activation volume for this reaction (2) and the enthalpy of activation for p-methylbenzyl chlobehave in a similar manner. As the ride (22) .. depth of the minimum increases its position is shifted toward the pure water end of the solvent composition scale. Such behavior was predicted by Arnett and McKelvey (23) on the basis of the behavior of heats of solution in aqueous alcohols. These workers found that the heat of solution of sodium tetravhenvlboride has an " endothermic maximum relative to pure water of 16.9 kcal/mole in aqueous t-butyl alcohol as compared with the value of 10.7 kcal/mole in aqueous ethyl alcohol (23). In t-butyl alcohol z

A

127

GOLINKIN AND HYNE: ENTHALPY AND ENTROPY O F ACTIVATION

Can. J. Chem. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/28/17 For personal use only.

the extremum occurs at 0.045 mole fraction alcohol, whereas it was observed at 0.12 mole fraction alcohol in the ethyl alcohol media.

behavior of this parameter parallels that of both AH" and AVao in that a minimum is observed, the depth and position of which are dependent upon the nature of the alcohol in the solvent. However, the positions of the AH* and AS* minima do not appear to coincide in all cases. In the ethyl and i-propyl alcohol systems the entropy minimum occurs at higher alcohol concentrations than the enthalpy minimum. Hyne and Wills (7) also found that coincidence of these minima is not a general phenomenon; only two of the systems studied by these authors coincided.

Mole Fraction Alcohol FIG. 1. Dependence of A H * on solvent composition.

In order to dissect the contributions of the initial and transition states to the behavior of AH* in the manner now well established by Arnett and co-workers (13), values for the heat of solution of benzyl chloride in each of the solvent systems are required. Arnett and McKelvey (13) presented a preliminary dissection of this type for benzyl chloride in ethanol-water, but the limited solubility of the halide in the highly aqueous end of the solvent range renders the heat measurements at least suspect if not unobtainable. Entropy of Activation The variation of AS.* with solvent composition is shown in Fig. 2. It is apparent that the

I

I

0.1

I

0.2

I 03

I 0.4

I

Mole Fraction Alcohol FIG. 2.

Dependence of AS* on solvent composition.

Due to the lack of available data, no dissection of AS* into initial state and transition state contributions is possible. Burris and Laidler (24) have suggested that since AS* and AV* both reflect solvent behavior for ionic reactions, they should parallel one another in sign and magnitude. In fact these authors obtained a linear relation between these ~arametersfor six different reactions in water. ?he question then arises as to the possibility of a similar linear relationship between AS* and AV* for a single reaction in several solvents. The

128

CANADIAN JOURNAL OF

Can. J. Chem. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/28/17 For personal use only.

present data, available for 17 solvents, represent a fair test of this postulate. The results are shown in Fig. 3. The AV* values for 0.05 mole fraction isopropyl and t-butyl alcohols are interpolated (see ref. 2).

FIG.3. Partial linear relationships (full line) between AV"o and AS". Value shown above points is mole fraction of solvent mixture for point.

It is at once apparent that overall linearity is not observed. Instead we find quite conlplex curves similar to the so-called isokinetic relationships (25, 26). Despite the apparent complexity, however, the linearity of the AS" versus plot from pure water to approxin~ately that solvent composition at which the extremum is found in the behavior of the parameters with respect to solvent composition should be noted. These portions of the plots are shown in full line in Fig. 3. Furthermore the slopes of these sections of the plots are identical within the limits of uncertainty of the data. In these solvent composition ranges, therefore, AS* and AV* do indeed parallel one another in both sign and magnitude, suggesting similar solve~ltbehavior in all four cases. It is in this region of initial cosolvent addition to water that the structuremaking effect of the cosolvent is considered to be

CHEMJSTRY. VOL.

46, 1968

effective (27). On a mole fraction composition basis t-butyl alcohol maximizes water structure at lowest mole fraction (circa 0.05) and methyl alcohol at highest mole fraction (circa 0.3). It is, therefore, not surprising that in each binary solvent mixture, while the cosolvent is enhancing the water structure, the entropy and volumes of activation should manifest such parallelism in behavior. Over the different solvent composition ranges the solvent structure is passing through the same change from pure water structure to enhanced water structure brought about by the added cosolvent. The similarity of the AS* versus AV* slope in all cases reflects this common solvent behavior. Beyond the solvent composition corresponding to maximum cosolventinduced structuredness the bulk effect of the cosolvent becomes predominant and the solvent structures begin to diverge toward those characteristic of the pure cosolvent. Parallelism in the behavior of AS* and AV* would therefore not be expected. Comparison of Activation Parameters and Mixing Pammeters Available data on the position of the extremum of the activation and thermodynamic parameters of mixing for the various binary solvent systems discussed in this work are collected in Table 111. It is immediately obvious that the position of the extremum in the heat and entropy of mixing does not always occur at the same solvent composition. The tendency is for the entropy extremum to occur at somewhat higher cosolvent mole fraction. Insofar as both the activation parameters and thermodynamic paranleters of mixing reflect the structural properties of the binary solvent system it is not surprising that this same tendency is shown in the relative positions of the enthalpy and entropy of activation. TABLE I11 Mole fraction of organic cosolvent at which high aqueous region extremum in parameter appears Activation parameters

Binary solvent mixing parameters

Organic cosolvent

AH*

AS*

AHA'

ASE

Methyl alcohol Ethyl alcohol i-Propyl alcohol t-Butylalcohol

0.28 0.12 0.06 0.05

0.3 0.2 0.1 0.05

0.3 0.22 0.10 0.10

0.3 0.31 0.23

GOLINKIN AND HYNE: ENTHALPY AND ENTROPY OF ACTIVATION

Acknowledgment

Can. J. Chem. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/28/17 For personal use only.

T h e a u t h o r s gratefully a c k n o w l e d g e t h e financial assistance o f t h e N a t i o n a l R e s e a r c h Council of Canada. and W. G. LAIDLAW.J. 1. J. B. HYNE,H. S. GOLINKIN, Am. Chern. Soc. 88,2101 (1966). 2. H. S. GOLINKIN, IKCHOON LEE,and J. B. HYNE. J. Am. Chern. Soc. 89.1307 (1967). 3. A. H. FAINBERG and S. \WINSTEIN. J. Am. Chern. SOC.78.2770 (1956). 4. A. H. FAINBERG and S. WINSTEIN.J. Am. Chern. Soc. 79,1597,1602,5937 (1957). 5. E. TOMMILA.Suornen Kernistilehti, B25, 37 (1952). 6. E. TOMMILA and M. MURTO. Acta Chern. Scand. 17, 1947,1957,1985 (1963). 7. J. B. HYNEand R. WILLS. J. Am. Chern. Soc. 85, 3650 (1963). 8. J. B. HYNE.R. WILLS.and R. E. WONKKA. J. Am. Chern. S O C . ' ~2914 ~ , (1962). 9. J. B. HYNE. J. Am. Chern. Soc. 82,5129 (1960). 10. J. B. HYNEand R. E. ROBERTSON.Can. J. Chern. 34, 931 (1956). 11. E. M. ARNETT, P. M. DUGGLEBY, and J. J. BURKE. J. Am. Chern. Soc. 85,1350 (1963). 12. E. M. ARNETT,W. G. BENTRUDE, J. J. BURKE,and P. M. DUGGLEBY.J. Am. Chern. Soc. 87, 1541 (1965).

129

13. E. M. ARNETT and D. R. MCKELVEY.Record Chern. Progr. Kresge-Hooker Sci. Lib. 26,185 (1965). 14. J. G. MARTIN and R. E. ROBERTSON. J. Am. Chern. SOC.88,5353 (1966). 15. E. A. GUGGENHEIM. Phil. Mag. [7], 2, 538 (1926). 16. H. S. GOLINKIN, W. G. LAIDLAW, and J. B. HYNE. Can. J. Chern. 44,2193 (1966). 17. R. E. ROBERTSON and J. M. W. SCOTT. J. Chern. SOC.1596 (1961). 18. E. TOMMILA, A. KOIVISTO, J. P. LYRRA,L. ANTELL, and S. HEIMO. Ann. Acad. Sci. Fennicae, Ser. AII, 47 11 952). ,---,.

19. L: PEKKARINEN and E. TOMMILA.Acta Chern. Scand. 13,1019 (1955). 20. E. TOMMILA and S. MALTAMO.Suornen Kernistilehti. B28.118 (1955). 21. E. T O M M ~ L A and MALTAMO.Suornen Kernistilehti, B28,73 (1955). 22. J. B. HYNEand R. WILLS. Unpublished results. and D. R. MCKELVEY.J. Am. Chern. 23. E. M. ARNETT. Soc. 87,1393 (1965). 24. C. T. BURRISand K. J. LAIDLER.Trans. Faraday Soc. 51,1497 (1955). 25. J. E. LEFFLER.J. Org. Chern. 20,1202 (1955). 26. J. E. LEFFLER and E. GRUNWALD.Rates and equilibria of organic reactions. John Wiley and Sons, Inc., New York. N.Y. 1963. Chao. 9. 27. F. FRANKS and D. J. G . 1;~s. Quart. Rev. London, 20,l (1966).