Dec., 1947
DECOMPOSITION OF TRICHL,OROACETIC ACID IN FORMAMIDE-WATER [CONTRIBUTION FROM THE CHEMICAL
2987
:LABORATORYOF THE OHIO STATE UNIVERSITY]
The Kinetics of the Decomposition of Trichloroacetic Acid in Formamide-Water Mixtures BY C. NORMAN COCHRAN AND FRANK H. VERHOEK
Studies of the decomposition of trichloroacetic through its trichloromethyl group, adapting the colorimethod devised by Cole' for chloroform analysis. acid' and of trinitrobenzoic acid2 in dioxane- metric Fifteen milliliter samples of solutions approximately water mixtures show that the rates of these de- 0.1 molar in trichloroacetic acid in an appropriate solvent compositions pass through maxima as the propor- mixture were pipetted into twelve individual reaction tion of dioxane in the water is increased. In the flasks equipped with an external ground joint to provide an air lock between the thermostat mil and the actual case of trinitrobenzoic acid the maximum was ex- closure on the flask.0 Six of these were sunk into each of plained as the result of two opposing effects, based two oil-baths thermostatically controlled a t 50 and 60 ", on the assumptions that the decomposing sub- or 60 and 70". Auxiliary heaters were used to compensate stance in the solution is the acid anion and that it quickly for the slight drop in temperature which occurred on immersion of the samples. The first sample was reis less stable the less highly it is solvated. One moved after eight to fifteen minutes, and others a t suitable effect, tending to decrease the apparent rate of intervals thereafter. The samples were cooled in an icedecomposition, results from the decreased dts- bath and diluted by a factor of one hundred with distilled sociation of the acid into ions in solvents of high water. Air was bubbled for four minutes through a 10aliquot of this solution to remove chloroform, tests on dioxane content; the other effect, tending to in- ml. saturated chloroform solutions having shown no coloricrease the rate of decomposition, results from the metric test for the latter after sweeping out the solution progressive desolvation of the ion as the dioxane for this length of time. A 2-ml. sample of the chloroform-free solution was pipetcontent of the solvent is increased. This point of ted into a Pyrex test-tube, and 4 ml. of 20% sodium view was confirmed by measurements of the acid hydroxide and 2 ml. of pyridine, purified by refluxing over dissociation constants in the various solvent mix- and distilling from barium oxide, added. The test-tube tures and calculation of the ion concentrations; was fitted with a rubber stopper and expansion tube, when the rates were referred to the actually exist- shaken for two minutes, and immersed in vigorously boiling water a t the instant a timer was started. After ing ion concentrations, the maximum disappeared shaking for sixty seconds in the water, the test tube was and a continuous increase in rate constant with removed to a water-bath maintained a t 17 * 1" and increasing dioxane concentration was observed. shaken until the timer read ninety seconds. At one hun10 ml. of alcohol was added and the contents Existing data do not permit of such a treatment dred seconds until only one phase was present. A t two hundred for the trichloroacetic acid case, but from the shaken seconds the solution was poured into the rectangular abparallelism between the behavior of the two acids sorption cell of a Cerico Photelometer, Industrial Model observed in other cases,3 it seems likely that the B2, and a reading of the percentage absorption taken a t exactly four minutes through a green filter with a maximum maximum may be explained in a similar fashion. transmission a t 5300 A. A calibration curve was prepared It is then of interest to investigate the decompo- according to the same procedure using solutions of trisition of one of the acids in a solvent pair in which chloroacetic acid in water whose concentrations were the dissociation constant of the acid does not known from titration and dilution data. Strict adherence a time schedule is desirable, since the red color fades on change appreciably with change in the compo- to standing. The reproducibility of the technique used is sition of the solvent; in this case the whole change attested by the fact that the average deviation of the in rate must be due to changes in the degree of points on the calibration curve from the curve itself is solvation. A mixture of water and formamide, 0.94%, for the twelve points taken over the extinction 0.12 to 0.46 (0.0002 to 0.001 m ) . with dielectric constants of the same magnitude* range Formamide was purified by the chemical and distillaform such a pair. Trichloroacetic acid was chosen tion treatment of Verhoek,E omitting the fractional crysfor study, since its dissociation constant has been tallizations. The material used boiled a t 80-82" a t 5 mm. pressure and had a freezing point of 2.3'. Solvent mixestimated in both water5 and formamide.6 The rate of decomposition of trichloroacetic acid cannot be measured conveniently in formamide-water mixtures by titrating with standard base the acid left undecomposed because hydrolysis of the formamide forms ammonium formate. The amount of ammonium formate produced is of no consequence its far as the composition of the solvent is concerned, its concentration amounting to only 0.2!5 molar after ten hours a t 100" for a 50 mole per cent. formamide-water mixture, but it so buffers the soluticm . that the titrations are unreliable. Accordingly, the reaction was followed by analyzing for trichloroacetic acid (1) Salmi and Korte, A n n Acad. Sci. Fcnnicae, A64, No. 1.0 (1940). (2) Trivich and Verhoek, THISJOURNAL, 66, 1919 (1943). (3) Verhoek, ibid., 61, 186 (1939);67, 1062 (1945). (4) "International Critical Tables," Vol. VI, p. 83. (5) Hall, Chcm. Reu., 8, 191 (1931). ( 6 ) Verhoek. Trrrs J o u a N A L , 68. 2677 (1938).
tures were made up by adding measured volumes of distilled water to weighed amounts of formamide. The decomposition products in formamide were shown to be the same as in water by a qualitative test for chloroform from a heated solution of trichloroacetic acid in formamide, and by collecting the gas from the decomposition of a weighed amount of dissolved trichloroacetic acid; 95% of the theoretical amount of carbon dioxide was obtained.
The rate of decomposition was measured a t 60" over the range of solvent composition from 0 to 90% formamide, and a t 50 and 70' over portions of the range in order to estimate the activation energies. The reaction was first order throughout, as shown in representative experiments in Table I. (7) Cole, J. B i d . Chcm., 7%, 173 (1926). ( 8 ) Trivich. Ph.D. Thcaia, the Ohio State University. 1942.
2988
C. NORMAN COCHRAN AND FRANK H. VERHOEK
Vol. 69
mum they found is due to the changing dissociation of the acid as the composition of the solvent 60 m o k per cent. is changed. This explanation of their data was 75 mole per centb formaniide a t 70' formamide a t 50 apparently missed by these authors.le9 Concn. Concn. __ A1Og Time, - A log Time, min. According to the recorded values of the dism./lit er At min. m./liter A! sociation constants,6+6 trichloroacetic acid a t room 10 0.0740 15 0.0845 0.0119 0.00114 temperature is slightly less dissociated in forrna25 .04!30 145 .0600 .0123 .00108 mide than in water. If these $K values are used, 40 .0320 275 .0435 .0120 .00108 with straight line interpolation between the two 57 .0200 405 ,0315 .0119 .00112 extremes, and the data of Table I1 recalculated to 70 ,0140 535 ,0225 .0128 .00111 give velocity constants for the decomposition of 85 ,00!30 693 .0150 the trichloroacetate ion, curves for log k,,n vs. First-order rate constants obtained from graphs solvent composition give lines more nearly straight of log concentration against time are recorded in than those shown in Fig. 1. The change is not Table 11, together with calculated activation great, however, since the calculated kion is only twice the observed velocity constant even a t 90 TABLE I1 mole per cent. formamide. VARIATION OF THE VELOCITY CONSTANT AND ACTIVATION Comparison of the data with those for trinitroENERGY WITH %LVENT COMPOSITION benzoate ion in water-dioxane mixtures2 indicates Mole that, even over the smaller range of solvent comper cent. Initial Activation energy position investigated in that case, the increase in form- concn. Velocity constants (sec.-'XlW) kcal./mole amide (m./liter) 60' 60' 70' 50-60 60-70 velocity constant and the decrease in activation 0 0.348" 1.713' 36.2" energy are much greater than in the present case. 20 0.0886 1.40 This difference is to be attributed to the greater 35.2 20 .0935 6.58 solvation of the decomposing ion by formamide as 40 .lo00 0.992 compared to dioxane-or a t least to the greater 33.6 40 .1050 4.77 difference between the solvation of the ion and 34.6 40 -1050 21.9 that of the activated complex, for formamide as 60 .lo03 2.29 compared to dioxane. Comparison with the data 33.2 60 .0882 10.8 for the decomposition of trichloroacetate ion in 33.1 60 .0%2 46.4 water-alcohol mixtures1° shows that the change in 75 .0905 4.25 19.1 32.1 activation energy is about the same in the two 85 .0923 5.82 26.5 32.4 cases. If the change in the rate of decomposition 90 .08:78 6.86 29.4 31.1 is due mainly to changes in the solvation energy, Data of Yerhoek, THISJOURNAL, 56, 571 (1934), for this would indicate that the difference in solvation sodium trichloroacetate. energy between water and alcohol, and water and formamide, is the same. For each of the three energies, arid a graph of log R against solverit comsolvent pairs, the decrease in activation energy is position is
