The Rate of Decomposition of N-Chloracetanilide by Hydrochloric Acid in Dioxan-Water Mixtures.
[Received by the Editor, February 20, 1947; issued separately, April, 1948.]
The isomerisation of N-chloracetanilide to p-chloracetanilide under the catalytic influence of hydrochloric acid has been measured over a range of acid concentration and at different temperatures, using four mixtures of dioxan and water as solvents. Mechanisms involving collisions between anilide molecules and either the ions of hydrochloric acid or ion-pairs of the acid, have been considered and rejected as inconsistent with the high temperature coefficient of the reaction rate. A mechanism demanding the presence of covalent hydrogen-chloride molecules is outlined and suggestions, are made for further work to test this hypothesis.
The decomposition of N-chloracetanilide is catalysed by hydrochloric acid which acts by forming acetanilide as an intermediate and p-chloracetanilide as the final product:
The rate of decomposition is proportional to the first power of the anilide concentration and to the square of the acid concentration. Such kinetics are consistent with attack on the anilide molecule by hydrogen and chloride ions simultaneously in a termolecular collision or alternatively by undissociated molecules of hydrochloric acid in a bimolecular collision with the anilide molecules. It has been shown, however, by Robinson and Smith (1939) that the very high temperature coefficient of the reaction is consistent with these bimolecular collisions involving the undissociated acid molecule but is difficult to understand on the assumption that termolecular collisions occur with the ions. Since the thermodynamic properties of hydrochloric acid in dioxan-water mixtures have been studied extensively in recent years, it is of interest to follow the kinetics of this reaction in such mixtures; in this connection, attention should be drawn to the study of the acid-catalysed hydrolysis of methyl acetate in dioxan-water by Harned and Ross (1941).
The technique of following the rate of decomposition was similar to that adopted by Robinson and Smith; the iodine liberated by the N-chloracetanilide at different intervals of time was determined by
titration with standard sodium-thiosulphate solution. The initial anilide concentration was approximately 0·1M; as solvents dioxan-water mixtures containing 20, 45, 70 and 82 per cent. dioxan were used, these having dielectric constants of 60·8, 38·5, 17·7 and 9·5 respectively at 25°. In all cases the reaction was found to be of the first order with respect to the anilide, i.e.,
—d[A]/dt = kobs [A]
where [A] is the concentration of N-chloracetanilide and kobs is the first order velocity constant, the time being measured in minutes. The two side reactions, shown by Soper and Pryde (1927) to be important in aqueous solution, were also found to occur in these mixtures. The first of these consists of a reaction between the chlorine formed as an intermediate with a p-chloracetanilide molecule:
Since the N-substituted chlorine is reactive to potassium iodide, the rate of reaction is diminshed by this side reaction. This effect was eliminated by the addition of 0·03 M phenol which acts as a chlorine acceptor. To show that the effect was real, velocity constants were determined in solutions containing 20 per cent. dioxan, at 25°, with different catalyst concentrations but without addition of phenol. The following table compares the observed-velocity constants with those interpolated from experiments, to be described later, in which phenol was added to the solutions.
|mHCl||Without Phenol.||With Phenol.|
The second side reaction consists of a slow hydrogen ion catalysis occurring in any acid solution and not peculiar to hydrochloric acid:
This accelerates the rate of decomposition and the effect can be computed from measurements in perchloric-acid solution. This reaction is also of the first order with respect to the anilide and the magnitude of the velocity constant was found in different dioxan-water mixtures as follows;
|Per cent. Dioxan.||mHClO||k × 103||1000k/mHClO1|
The velocity constants of the main reaction, given in Table I below, have all been corrected for this simultaneous hydrogen-ion catalysis. The first two columns of this table give the experimental results, whilst Table II contains the results of reaction-rate measurements at different temperatures.
|mHCl||kobs × 102||γHCl||α||kobs/γ2HClm22HCl||kobs/(1–α)mCl|
|mHCl||kobs × 102||γHCl||α||kobs/γaHClm22HCl||kobs/(1–α)mHCl|
|mHCl||kobs × 102||γHCl||α||kobs/γaHClm22HCl||kobs/(1– [ unclear: ] )mHCl|
|0.1042 MHCl||0.3180 MHCl|
|20°||k= 3.28 × 10−3||k= 8.29 × 10−3|
|E = 21,530 cal.||E = 21,325 cal.|
|0.0750 M HCl|
|15°||k= 3.67 × 10−3|
|E = 21,130 cal.|
|0.00499 M HCl|
|15°||k= 5.39 × 10−3|
|E = 22,100 cal.|
|0.00260 M HCl|
|15°||k= 9.65 × 10−3|
|E = 23,460 cal.|
There are three possible mechanisms for this reaction.
(1) Reaction may occur as a result of a ternary collision of an N-chloracetanilide molecule, a hydrogen ion and a chlorine ion. The rate of reaction will then be determined by the equation:
d [A]/dt = kB [A] γH γCl mH mCl γN: Cl/γX where γH γCl (= γ2HCl) is the product of the ionic activity coefficients of hydrochloric acid, which have been determined in dioxan-water mixtures by Harned et alia (1938); γN;Cl is the activity coefficient of the anilide and γX that of the “critical complex.” The quantity kobs/γ2HCl m2HCl = kB γN: Cl/γX should be almost independent of acid concentration in any one dioxan-water mixture, any small departure from constancy being attributed to the effect of acid concentration on the ratio γN: Cl/γX. In Table I, kobs/γ2HCl m2HCl has been evaluated and it will be seen that it is almost constant except at very low acid concentrations in the 82 per cent. dioxan mixture, where the departures might well be ascribed to impurities in the solution. This evidence is undoubtedly consistent with a mechanism of ternary collisions but, as was shown by Robinson and Smith in the case of aqueous solutions, the high energy of activation, derived from the temperature coefficient of the reaction rate, is difficult to reconcile with ternary collisions. Table II
shows that the energy of activation is of the order of 22,000 cal. In a solution of 82 per cent. dioxan, in which both the anilide and the acid are present at a concentration of 0·01M, 17 × 10−3 mols of anilide per litre, decompose in one minute at 25°; this corresponds to 1·7 × 1016 molecules per c.c. per sec. As the fraction of collisions which are effective is only e−22000/RT, the total number of collisions must be of the order of 1032. Under these conditions the number of binary collisions is calculated to be the order of 1028. The number of ternary collisions must be much less and of these a large fraction must be ineffective because of steric effects. The reaction therefore proceeds much more rapidly than any reasonable estimate of the collision number will permit; this, in our opinion, makes it very difficult to accept the mechanism of ternary collisions.
(2) Reaction may occur by binary collision between anilide molecules and “ion-pairs” of hydrochloric acid. Owen and Waters (1938) have shown that such ion-pairs are present to an appreciable extent in 70 and 82 per cent. dioxan solution. In this case we have: d[A]/dt = k'B [A] mIP γN:Cl/γX where mIP is the concentration of these ion pairs. But if Ka is the dissociation constant of the acid: γH γCl m2HCl/mIP = Ka and therefore: d[A]/dt = k'B [A] γH γCl m2HCl γN:Cl/Kaγ × an equation which cannot be distinguished from the former equation for ternary collisions. In Table I we have given an estimate of α, the degree of dissociation of the acid, and calculated kobs/(1–α)m HCl =kobs/mIP = k'B γN:Cl/γ X. The values are reasonably constant for any one dioxan mixture. Now calculation will show that the degree of dissociation does not vary much with temperature and therefore the energies of activation obtained from Table II can be applied to this postulated bimolecular mechanism. Taking, as before, the case of a solution in 82 per cent. dioxan at 25° with 0 01M anilide and 0·01M acid, in which it is known that the observed rate of reaction corresponds to about 1032 collision per c.c. per see., we find that α = 0·569 and mIP = 0.00569. Calculating the number of binary collisions that can occur, we again obtain a value of the order of 1028. Consequently we are confronted once more with a reaction proceeding considerably more rapidly than the collision number predicts.
(3) The ion-pairs which have been considered in the preceding section are conglomerates of oppositely charged ions held together by electrostatic forces and therefore formed more readily in a solvent of low dielectric constant such as 82 per cent. dioxan. In addition, there may be present true molecules of hydrogen chloride characterised by a covalent bond. Robinson and Smith postulated such molecules to explain their results in aqueous solution. By elimination of two possible mechanisms we are led to adopt the idea of collision between anilide molecules and these covalent hydrogen-chloride mole cules as the cause of this reaction. Such a mechanism will explain the experimental data provided that (a) the proportion of such
covalent molecules increases with the dioxan content and (b) the proportion increases rapidly with the temperature. These two conditions will secure (a) the increased rate of reaction with increasing dioxan content of the solvent and (b) a lower energy of activation to reduce the required number of collision to a reasonable magnitude. Whether such covalent molecules do exist in the required amount must be left to further work, but we may conclude with the suggestion that a valuable contribution would be made by measuring the partial hydrogen-chloride-vapour pressure over solutions in these dioxan-water mixtures, from which information equilibrium constants could be computed for the ionic dissociation of the covalent molecule.
Harned. H. S., and Donelson, J. G., 1938. J. Amer. Chem. Soc., 60, 339, 2128.
Harned, H. S., Donelson, J G., and Calmon, C., 1938. Ibid., 60, 2133.
Harned, H. S., and Ross, A. M., 1941. Ibid., 63, 1993.
Harned, H. S., and Walker, F., 1939. Ibid., 61, 48.
Owen, B. B., Waters, G. W., 1938. Ibid., 60. 2371.
Robinson, R. A., and Smith, G. M., 1939. Trans. Roy. Soc. New Zealand, 69, 41.
Soper., F. G., Pryde, D. R. 1927. J. Chem. Soc., 00, 2761.