NOTES
628
far as we have been able t o tell, the isomers have never been separated heretofore. In isolating the non-reactive component, samples of purified piperylene were vaporized in a stream of hydrogen and the mixture passed through molten maleic anhydride, the small quantity of hydrocarbon remaining with the effluent gas being condensed and examined. Piperylene from both sources yielded a condensate which had the same index of refraction, w k , 1.4321, and exhibited only an extremely slow and incomplete reaction with maleic anhydride. The normal corrected boiling point of the substance, as determined in a micro-Cottrell apparatus, was 43.8' as compared to 41.9' of the original purified piperylene. On the basis of the higher boiling point and index, it would appear that the unreactive form is the cis isomer, although Provost3 has reported the boiling points of the cis and trans forms as 39 and 42.5', respectively, from distillation data. Finally, i t is believed that the observations mentioned here are of interest not only as a means of isolating what is believed to be cis-piperylene, but also from the standpoint of the selectivity of the Diels-Alder condensation in general. The writers wish to acknowledge the cooperation of Mr. A. L. Ward of the United Gas Improvement Company, who originally called their attention to the discrepancies in the analysis for piperylene in cracked Cg fractions and to the possibility of the presence of close boiling cisand trans-configurations. This work has been done on cracked Cg fractions furnished by Mr. Ward, and on pure piperylene samples supplied by Messrs. E. R. Gilliland and H. E. Buc. (3) Provost, Compt. rend., 182, 1475 (1926).
Vol. 63
that which holds for the other three halides. The equation developed by Scatchard3 predicts that the activity coefficients of the alkali metal fluorides should increase with increasing atomic number of the cation, although quantitative agreement with the freezing point results requires a larger radius for the fluoride ion than that obtained from crystallographic data. Scatchard's theory is therefore supported by data a t the freezing point, It is of some interest to ascertain whether this reversal in order is also to be found at higher temperatures, and for this purpose isopiestic measurements have been made a t 25' on sodium and potassium fluoride. Stock solutions were prepared from Baker c. P. material and the concentrations determined by analysis. The solutions were equilibrated with solutions of potassium chloride with the results given in Table I. TABLE I MOLALITIES OF SOLUTIONS OF POTASSIUM CHLORIDE AND SODIUM OR POTASSIUM FLUORIDE WHICH AREISOPIESTIC AT 25 O Sodium Fluoride mKcl
mNaF
mKCl
mNsF
0.1554 .2490 .4047 .5616 .5774
0.1562 .2504 .4097 .5728 .5870
0.6765 .7874 .8518 .8908
0.6896 .8088 ,8734 .9124 .9356
.9106
Potassium Fluoride mKCl
m XF
mKCI
0.1612 .3254 .7544 .9904 1.084 1.383
0.1598 .3221 .7353 .9572 1.049 1.316 1.480 1.815 1.964
2.245 2.484 3.112 3.188 3.761 3.918 3.982 4.580 4.81
1.568 1.941 2.108
mxF
2.086 2.290 2.823 2.873 3.354 3.497 3.526 4.006 4.183
Esso LABORATORIES OF THE STANDARD OIL DEVELOPMENT Co. CHEMICAL DIVISION ELIZABETH, NEWJERSEY RECEIVED DECEMBER 21, 1940
OSMOTIC m
9NaF
YNsF
m
9NnF
YNsF
The Activity Coefficients of Sodium and Potassium Fluorides a t 25' from Isopiestic Vapor Pressure Measurements
0.1 .2 .3
0.923 ,909 ,899
0.764 .708 .675
0.5 .7 1.0
0.888 .880
0.631 .602 .572
m
9 KF
YKF
m
9KF
YKF
0.1 .2 .3 .5 .7
0.929 .920 .917 .917 .922 .934
0.774 .727 .701 .672 ,657 .649
1.5 2.0 2.5 3.0 3.5 4.0
0.958 .987 1.021 1.054
0.649 .663 .684 .713
1.091
.748
1.130
.790
BY R. A. ROBIN SON^
It is known as a result of freezing point determinations2 that the order of the activity coefficients of the alkali metal fluorides is the reverse of (1) Sterling Fellow, Yale University, 1940. (2) Karagunis, Hawkinson and Damkohler, Z. p h y s i k . Chem., 151A 433 (1930).
1.0
AND
TABLEI1 ACTIVITYCOEFFICIENTS OF POTASSIUM FLUORIDE AT 25
(3) Scatchard, Ckem. Rcv., 19, 309 (1936).
SODIUM AND
.873
Fell., 1941
629
NOTES
__
If i t is assumed that the addition of chlorine is From these data the activity coefficients given slow and that the exchange reaction in Table I1 were calculated. The freezing point data have been used4 to calHC1* Cla HC1 ClCl* culate the activity coefficient of potassium fluo- and the chlorination ride a t 25 '. A value of y = 0.848 is computed a t RH Clz +RC1 HC1 0.1 M , which is considerably higher than that of other 1-1 electrolytes a t this concentration (the proceed a t comparable rates, an expression for the value for lithium iodide is 0.811). Referring the distribution of radioactivity may be obtained as freezing point data to Y = 0.774 at 0.1 M and 25", follows. If a is the fraction of CIS exchanging behowever, values of 0.736, 0.676, 0.648 and 0.638 fore reacting with the organic molecule; a, the a t 0.2, 0.5, 1.0 and 2.0 M , respectively, are ob- initial number of moles of C1- as HCI; 60, the intained. Except a t 2.0 M , the agreement with the itial fraction of C1- which is radioactive; and isopiestic results is moderately good. The freez- $(x), the fraction of C1- which is radioactive after ing point data for sodium fluoride cannot be cor- the addition of x moles of Ch, then rected to 2 5 " ; a value of y = 0.752 a t 0.1 M is (e + x)e Jo dx = a eo (1) obtained4 a t 0", compared with 0.764 a t 25". Referred to the latter value at 0.1 M , the freezing where (a x)O and x = a O dx are the acpoint data give y = 0.703 and 0.625 a t 0.2 and tivities, respectively, of the final solution and of 0.5 M , respectively, again in moderate agreement the chlorinated product. Equation (1) may be with the isopiestic results. readily solved, and the ratio of the activity of the The activity coefficient of potassium fluoride is product to the initial activity of the solution is very close to that of sodium chloride, while sodium found to be fluoride has a much lower activity coefficient, x e,/a eo = 1 - [ a / ( a x)la (2) close to that of rubidium iodide. Using the data of Long and Olson2 one finds a = (4) Landolt-Bornstein, "Tabellen," Dritter Ergansungsband, p. l.04. This result is unity within the experimen2148. tal error, indicating, as the authors pointed out, STERLING CHEMISTRY LABORATORY YALEUNIVERSITY complete exchange between C1- and C1, before RECEIVED NOVEMBER 1. 1940 NEWHAVEN,CONN. chlorination. (The experiment of Long and 01son was repeated using different ratios of Clz to C1-. Calculated values of cr were again unity The Velocity of Rapid Chlorinations within experimental error.) I n order to carry out experiments in benzene BY HENRYC. THOMAS solution radioactive hydrogen chloride was inHalfordl has called attention to the possibility troduced in the form of the dry gas, prepared diof studying rapid halogenations by measuring the rectly from sodium chloride which had been bomdistribution of radioactive halide between product barded by deuterons. The chlorinations were and halide ion when the process is carried out in carried out on p-chlorophenol. Chlorine was inthe presence of tagged halide ions. Long and 01- troduced from a bulb of the gas attached to the son2 have shown that the velocity of chlorination flask of solution through a large-bore stopcock. of acetanilide is slow compared to the velocity of The chlorinated product for radioactive analysis interchange between chlorine and chloride ion in was recovered by distilling the benzene-hydrogen aqueous solution. Experiments have been car- chloride solution through a fractionating column ried out to investigate the possibility of measuring and dilution of the residue to an appropriate volthe rate of chlorination of an organic molecule ume. The solutions, before and after chlorinawith respect to the rate of exchange between tion, were analyzed for hydrochloric acid content chlorine and hydrochloric acid in benzene solu- by shaking with water and titration of the mixtion. Benzene was selected as the solvent in the ture. Two chlorinations of p-chlorophenol were hope that the low degree of dissociation of hy- carried out with the following results drochloric acid in this medium would bring about Concn. a moderately slow exchange reaction. 9-CICsH4OH, m a a + x x Bp/a Bo
+
+
+
+
+
+
6" +
1 ) R.S. Halford, THIS JOURNAL, 62, 3233 (1940). 12) F. A. Long and A. R.Olson, ibid., 68, 2214 (1936). I
0.403 .409
0.00960
.01045
0.01325 ,01330
0.290
.220
