620
HAROLD G. VESPER
AND
G. K . ROLLEFSON
1701. 56
the sulfuric acid-malic acid complex. The in- same inhibiting power in the decomposition of hibitor thus decreases the concentration of the some compound, the curve for the velocity of decomplex and thereby slows down the reaction. cornposition would be symmetrical and exhibit a The explanation for the maximum appears maximum at lOOyoHzS04. to be the same as for the other acids mentioned:? 'The sharp flattening out of the curve after a concentration of about 2y0 sulfur trioxide has i. e , there exists an equilibrium, H2S04ITSO3 H20. -4s water is removed the velocity of de- been reached is interesting. More than enough composition is accelerated until a point is reached sulfur trioxide is then present to form a molecular where the inhibiting effect of the increased con- compound with all of the malic acid or malic centration of sulfur trioxide more than offsets acid-sulfuric acid complex present ; hence, the the accelerating effect due to the removal of velocity measured from this point on may be that water. The maximum in the velocity of de- of the disintegration of this new complex. composition of malic acid came a t a molality of The author wishes gratefully to express his apwater of 0,024 and 0.030 a t 35 and 4 5 O , respec- preciation to Dr. Edwin 0. W i g , under whose tively, as compared to citric acid which exhibits a direction this work has been done, for his sincere maximum a t molalities of water of 0.10, 0.14, and friendly interest in the completion of it. 0.20 a t lg5, 25 and 35', r e ~ p e c t i r e l y . ~ The posiSummary tion of the maximum and the slope of the curve on either side of the maximum in both cases are The decomposition of malic acid by sulfuric evidence for the correctness of Wiig 's explanation. acid has been found to proceed in the same manner In each case sulfur trioxide is a stronger inhibitor as formic, oxalic and citric acids. -1maximum in the rate of decomposition has than water (as evidenced by the steeper slope of the curve in the sulfur trioxide region); hence, been found, the rate increasing up to a certain the maximum appears a t a small concentration of point as water was removed, then decreasing. water. The relative effect is not so pronounced The location of this maximum, which has been in the case of malic acid as in citric acid, so that overlooked in a previous investigation, brings the maximum falls nearer to zero concentration malic acid into agreement with the theory for of water. One would expect -from this theory. inhibition in the decomposition of these acids. if siilfur trioxide and water should have the ROCHESTER, N Y. RECEIVED NOVEMBER 28, 1933
+
[ C O N T R I B U T I O S FROM THE CHEMICAL
LABORATORY O F THE
The Equilibrium in the Reaction C1,
vNIVERSITY O F CALIFORNIA]
+ Br,
=
2BrC1
BY HAROLD G. VESPER A N D G. K. ROLLEFSON The existence of bromine chloride has been a subject of dispute for many years. Early investigators' recorded the formation of a yellow or red-brown liquid when chlorine was passed into bromine in a freezing mixture, or bromine into liquid chlorine, and assumed the liquid to contain bromine chloride. That this was other than a solution of bromine and chlorine was questioned2 on the basis of the small heat developed in the process, and from studies of the freezing point-. and boiling point-composition curves However, Andrews and Carlton3noted a marked contraction
in volume upon mixing liquid bromine and liquid chlorine, and workers4 in the field of organic chemistry have for some time reported that mixtures of bromine and chlorine in addition reactions give bromochloro products, as if the substance adding were bromine chloride. Physicochemical evidence for the formation of the compound BrC1 on mixing BrI and Clz in carbon tetrachloride solution was furnished by Barratt and Stein5 and by Gillam and Morton.6 I t was found that the absorption of light by the mixtures, in the visible region, was markedly less
(1) (a) Balard, A n n . chim. phys., [Z] 31, 337 (1826): (b) Schonbein, J. prakl. Chem., [ l ]88, 483 (1863); (c) Bornemann, .Inn., 189, 184 (1877): (d) T h o m a s a n d Dupois, C o m p f . rend., 143, 282 (1906). (2) (a) Berthelot, Compt. rend.;94, 1619 (1882); (b) Lebeau, ibid., 143, 589 (1906); (c) Karsten, 2 . anorg. Chem., 63, 365 (1907). (3) Andrews a n d Carlton, THIS J O U R N A L , 19, 688 (1907).
(4) (a) Simpson, Proc. Roy. SOC.(London), A W , 118, 424 (1879); cb) James, J. Chem. SOL.,43, 37 (1883); (c) Delepine and Ville, Compt. rend., 110, 1390 (1920); (d) Hanson and James, J , Chem. Soc., 1955, 2979 (1928). (5) Barratt and Stein, Proc. Roy. SOC. (I,ondoo), 8112, 582 (1929). (6) Gillam and Morton. ibid., Al24, 604 (1929).
CHLORINE-BROMINE-BROMINE CHLORIDE EQUILIBRIUM
March, 1034
62 1
than that of the bromine content alone. Assum- perature and estimated the heat of reaction to be ing the absorption coefficient of the inter-halogen about 300 calories. In the course of an investigation in this Laboracompound to be negligible (exact wave length not stated), Barrett and Stein5 calculated equi- tory it was necessary to have an accurate value librium constants for the dissociation of BrCI, for the equilibrium constant for the dissociation obtaining a mean value of 0.28. We have re- of BrCl and also values of the abscjrption cocalculated their data making an allowance for efficient. By the use of wave lengths near 0.50 p the absorption of BrCl and find that this value is we have found it possible to obtain very consistent reduced to a value not greater than 0.1s. The results. results of Gillam and Morton were qualitative but they estimated that the bromine chloride was less than 50% dissociated. Gray and Style’ measured the absorption of gaseous mixtures of Clz and Br2, finding a similar decrease of the bromine absorption, upon the addition of chlorine. Assuming that a t 0.579 the absorption of BrCl is negligible, they give for A-in two experiments, the value 0.132 and 0.12.5. Their Table IV follows Initial Initial XIixt.(CliJ (Brz) I 0.0401 0.00530 ,0115 I1 ,0114
Free Free (Clr) (Brzj 0.0352 0.00045 ,0047 ,0048
(BrCIj 0.0008 ,0134
K
Ii‘
0 132 ,125
0.182 ,126
I< is their published figure; K’ is one recalculated from their data, using their assumption. If now bromine chloride absorbs slightly a t 0.579 p, e. g., if the ratio a g r c I , / ~ B r ,is 0.01, the values of K 0btainc.d from their mixtures I and 11, respectively, arc shown as Initial (Brd 0.0401 0.00538 ,0114 ,0115
I iitial
hlixt. I I1
Clii
Free (Clz)
Free (Brz) (BrCI) K 0.0351 0.00084$~ 0.01OOg 0.121 ,0456 ,00466 .01368 ,114
Similar cdculations for other assumed ratios of the absorption coefficients give values of I< which are shown graphically in Fig. 1. The two experiments give identical results a t K = 0.111, and the above ratio 0.0134. Gray and Style also noted the absence of a decrease in pressure upon mixing gaseous bromine and chlorine, which confirms the conclusion that the compound is BrCl, not Br2C12 or BrC13. W. Jost8 has made measurements of the gaseous equilibrium a t various temperatures between 0 and 230’ using assumptions similar to those of Gray and Style. He gives values for the equilibrium constant a t room temperature, ranging from 0.10 to 0.23. D. M. Yostg has obtained a value of 0.12 to 0.14 for t h e equilibrium constant a t room tem(7) Gray atid S t y l e . i ’ v o c Iirl I‘ost. T I T I ,SI ~ I I J I ~ K A I , ,55, 44%) !1933), and private communications.
Ratio aBrCl/aBrZ. Fig. 1.-Dissociation constant of bromine chloride, recalculated froni the data of Gray and Style.
Apparatus and Experimental Procedure The measurement of absorption was made with a 1l;cinig-Martens spectrophotometer ; l o the shape of reaction vessel was thus fixed. A cylinder of Pyrex glass, 3 8 mm. inside diameter and 239 mm. long, with fused-on plane parallel ends, was employed. A heliuni discharge tube11 was arranged to be used end on, with electrodes a t right angles from each end. The center portion was 40 cm. long, 0.8 cm. inside diameter. The tube was operated a t 80 milliamperes, using a 10,000 volt ”4 kw. Thordarson transformer, and gave an intense discharge. The beam passed through a copper chloride-calcium chloride filter solution 1.4 cm. thick (153 g. CaC12.4H20, 16.7 g. CuCL, which removed wave 1 cc. 2.5 N HC1, 178 cc. H Z O ) ’ ~ lengths shorter than 0.485 /.c, and was focused by two lenses upon the entrance slit of the spectrophotometer. The lens system and front portion of the spectrophotometer, including the reaction vessel and empty comparison vessel, was placed in a light-tight box, with the helium (10) Martens and Grunbaurn. Ann. P h y s i k . 12, 1004 (1903). (11) Loaned through t h e courtesy of Dr. 1-105-d T. Jones of t h e ICiectrical Products Corporation, Oakland. (12) Uhler and Wood. “Atlas of Absorption Spectr.i.” Cariiegie Inst. PUbl. VOl. 71 (1907).
622
Vol. 56
HAROLD G. VESPERAND G. K. ROLLEFSON
discharge tube on the optical axis, just outside the end of the box; the filter cell was placed on the end wall of the box. Thus no light of shorter wave length than 0.485 p could enter the box, except by opening a shutter placed on the side of the box near the reaction vessel. This side opening was closed as well by a Corning 9-mm. number 580 filter, transmitting oiily thc 0.305 p group of the mercury arc spcctrum. A Cooper-Hewitt 220-volt quartz mercury arc operating on 149 volts, 4.1 amperes, was placed 40 cm. from the reaction vessel, in a nearly horizontal position, so that by opening the shutter the vessel could be illuminated by the 0.365 p lines of the arc. Between the arc and the shutter a Pyrex glass cell, 7.5 cm. thick, 25 em. long, and 14 cm. high, supplied with running water, was placed. This removed the infra-red radiation from the arc, preventing undue warming of the reaction box. The spectrophotometer ocular slit was opened wide and the objective double slit was opened sufficiently to cause the images of the slit, formed by the three green lines of the helium discharge, to overlap completely in the field but not wide enough to show the hazy edges of the beam. A uniform field was thus secured. The filter weakened the 0.492 p line considerably; the relative intensities of the three lines, as seen in the spectrophotometer, being 1, 10 and 6 for 0.4922 p , 0.5016 IJ. and 0.5048 p, respectively, giving a center of gravity a t 0.5022 p . All three green lines were used because it was found impossible to isolate the strong center line (by narrowing the objective and ocular slits) without reducing the observed intensity below an adequate amount. Since the same light was used throughout the experiments, and since the absorption curve is practically linear over this range of wave length, no appreciable error was thus introduced. The reaction vessel was connected by a 1.6-mm. capillary side tube with the usual oil and mercury pumps and liquid air trap, a sulfuric acid manometer, and storage vessels for reagents. The bromine uscd was Kahlbaum, kept in presence of solid potassium bromide; the chlorine was prepared from anhydrous cupric chloride by heating, and freed from air by freezing out with liquid air and pumping off residual gas; in later experiments the chlorine was also fractionated, with no appreciable effect on the results. I n reading the spectrophotometer, each recorded value of log (tan2 9) is the average of a t least sixteen settings, eight on each side, of the analyzing nicol prism. The individual settings were usually within 0.2’ of the mean. If the settings began to deviate beyond this, resting of the eyes restored accuracy.
yellow; the green readings are a better indication that equilibrium is not yet reached.) When illuminated by the 0.365 CY lines of the mercury arc (about 1000 microwatts falling on the vessel), a steady state was attained after two to five minutes. That this steady state is identical with the thermal equilibrium was proved by: (1) leaving the vessel, after illumination, in the dark for from six to ninety hours more, the absorption remaining constant within the limits of experimental error; (2) illuminating the vessel after attaining thermal equilibrium in the dark, the absorption remaining constant within the limits of experimental error. It was thus possible to attain the true equilibrium by illuminating under the above conditions for a few minutes, with the advantage that the setting of the spectrophotometer (“blank” reading) does not change appreciably in two or three hours, while it does so to a slight extent overnight or over the several days necessary to be certain of equilibrium attained in the dark. The procedure was then adopted in the actual equilibrium measurements of illuminating until a constant reading was obtained. This was usually accomplished in two minutes, although more time was always allowed. At one time during the course of the experiments the reaction vessel was removed from the line, and washed with water, alcohol and ether and dried, before replacing it. The next experiments with bromine and chlorine gave the results shown in Table I.
Run
Initial
45
1.253
(The experimental error is greater than the sinal1 1xtu.ccn the last three readings in the
These observations indicate that immediately after cleaning, the dark reaction proceeded a t
log I e / I (yellow)
1.660
,802 (15 rnin.) ,541 (45 min.)
0 . 4 8 0 (20 min.)
Remarks Complete in dark after about 30 min.
0 ‘2.3 1 9 . 0 23.2 40.5 48 63 3.1 2 . 3 1 1 . 4 7 6 1.393 1.214 1.180 1.163 icalcd.) (ohs.) 0.2!31 11.213 (I 105 0 . 1 0 2 0.069 0 . 0 7 0 0 . 0 7 1
Time, hours log I o / I (green)
47
1.421
0 . 4 8 5 (15 rnin.) ,470 (60 min.)
Incomplete in dark after 45 min., complete in light after 1.5 min. Slowly reacting .777 (2 min.) in d a r k ; corn, 7 7 7 (30 min.) plete in light after 2 rnin. Reacting m o r e slowly in dark ,374 (0.5 min.) t h a n t h e iden,365 (2 min.) tical R u n No. .365 (24 min.) 46. Complete in light after ahout 2 min.
In the dark, equilibrium was attained only after periods ranging from sixteen to over sixty hours. The course of one run is given: (Run 18. Initial Clz 49.5, Br:! 20.5.)
46
’rAnLn I Values of log Io /I (green) I n dark T h e n illuminated
1.492 (15 rnin.) 1 . 3 3 0 (45 min.)
48
1.410
1.017 (22 min.) 0.780 (45 min.) , 6 8 2 (60 min.)
,383 (0.5 rnin.) .377 (1.5 rnin.) ,377 (15 rnin.)
CHLORINE-BROMINE-BROMINE CHLORIDE SQUILIBRIUM
March, 1934
TABLE I1 least thirty times as fast as before, but that in the Final course of several experiments the various condiFinal pressure Initial pressure log IoII C1, (yellow) Br2 BrCl CIz Zi tions were gradually restored. It is quite likely, a B r C l / a B r ~ Bm 0 1 9 . 3 8 67.94 0 . 0 4 2 2.92 3 2 . 9 5 1 . 5 0.139 therefore, that the dark reaction is catalyzed .11i 0.01 2.58 33.6 51.1 .02 2.24 34.3 50.8 .097 by the walls of the vessel. Since the light reaction 0 19.44 70.6 ,041 2.88 33.1 54.0 ,142 did not appear to be speeded up appreciably by ,119 .01 2.54 33.8 53.7 the cleaning, it is probably homogeneous, al.02 2.19 34.5 53.4 .008 0 9.80 76.5 ,012 0 . 8 4 17.9 67 6 .li8 though a rigorous proof of this would require .01 .66 18.3 6 7 . 4 ,133 further experimentation. .02 .47 1 8 . 7 67 2 ,091 These results are not in agreement with those of Experiments were then made using the green W. Jost,I3who observed a half time of less than one minute. However, Jost does not mention lines (with a center of gravity a t 0.5022 p ) . having taken any precautions to exclude light Two types of experiments were performed; in the first type the initial pressure of bromine from his reaction vessel. Dissociation Constant and Extinction Coeffi- was low and that of chlorine high, giving a low final cient of BrC1: Experimental Data and Calcula- bromine pressure, so that as large as possible a portions.-Pressures were recorded in cm. of sulfuric tion of the final absorption would be due to BrCI, acid (1 cnl. = 0.001iST atm.) corrected to a and the absorption due to Brzwould be small. This common temperature, 0’) and the thickness of small Brz absorption was calculated by assuming the absorbing layer was 23.9 cm. Applying a value for the equilibrium constant K = (Br2) (Clz)/BrC1)2 = x(c - b x ) / ( 2 b - 2 ~where ) ~b Beer’s law EO the mixture we have is initial Brz pressure, c is initial C12 pressure and log10lI = I’(Br2) Q(BrC1) Z(C1,) (1) x was thus determined. x is final Brz pressure; for the green lines of the helium discharge, and Then Q = R - P (Brz)/(BrCl). Errors in the log Io/I= Y(Br2) f W(BrC1) (2) for the yellow lines, where P , Q, Z , Y and W are estimation of the equilibrium constant, if x were (0.001787) (23.9) times the corresponding values small, would introduce only small errors into the value of Q. of the absorption coefficients. In the second type of experiment, the initial The term Z (Clz) in Eq. (1) was small in our pressure of bromine was higher (relative to that experiments, and was calculated by a first apof chlorine), so that the final pressure of bromine proximation t o the chlorine pressure of the mixwould be larger, and most of the light absorption ture. The remainder of the expression was by the equilibrium mixture would be due to brocalled R: R = I’ (Br2) Q (BrCI) = l o g l o / l -2 mine, that due to bromine chloride being small. (C12). The approximation obtained for Q from the The values P == 0.153, 2 = 0.00009 and Y = average result of type 1 experiments was then 0.0142 were determined directly, using the pure inserted in the relation already given, namely, gases. Preliminary experiments with bromine and R = P (Br2) Q(BrC1) = P(x) Q(2b - 2x), chlorine, using the yellow lines a t 0.588 p, and whence (Br2) = x = ( R - 2bQ)/(P - 2Q), also x). making the same assumption as Gray and Style, (BrC1) = (2b - 2x) and (Clz) = (c - b and Jost, namely, that the absorption of BrCl This gave a second approximation to K , since is negligible in this region (i. e., that the second errors in Q introduced only small errors in the term on the right in Eq. 2 is equal to zero), gave value of K, if the absorption due to BrCl were the results shown in Table 11. Recalculations of small. the data, assuming various ratios ~ y ~ ~ ~ Using ~ / cthis ~ value ~ ~ of ~ K ,, a second approximation are also shown, and these results have been for Q was obtained from the results of type 1 runs, plotted, giving curves similar to those in Fig. 1. and so on. The final approximations for Q and The curves cross a t K = 0.103 and 0.105) and K are given in full in Tables I11 and IV. Results : ratio C X B ~ C I / L Y B=~ ~ 0.017. These results are K = (Brz) (C12)/(BrCI)2 = 0.107. Q = 0.0164, similar to those obtained by recalculating the data whence a B r C l = 0.384, a t 0.502 p. W. JostR reported that the presence o€ air or of Gray and Stplc, and indicate that the true value other foreign gases (hence presumably of C12 and of K lies ntar 0.11. BrC1) increased the cxtinction cocfficierit o f (13) \\’ J o i t Z p h s s i k Chem , B14, 413 (1911)
+
+
+
+
+
+
+
HAROLD G. VESPER AND G. K. ROLLEFSON TABLEI11 EXPERIMENTAL DATA,RUNSOF TYPE1. FINALAPPROXIMATIONS FOR Q. Initial pressure in cm. of HzSOi b, Brz c , CIz
10.05
76.0 i6.5 i3.9
9 8 0.3 9 .2
i5.6
Final log Io / I
Part due t o c12
Final
Diff.
(Brd
0,403 ,397 .3i1 ,366
0.006 .006 .006 .00ii
0.397 ,391 ,365 ,360
0.582 ,546 .50rj
R
Final (BrCI) = 2 [ b - x)
= x
18.95 18.5 17.6 17.45
.488
36
(ASSUMEDK = 0.107) P a r t of H
due to Brt
due t o BrCl
0.089 .os4 .078 .Oi4
0 308 307 ,287
v 0 0163 0186 ,0163 . 0104
, Mi
___.
Av. 0.01G4
EXPERIMENTAL DATA,RUNSOF TYPE2. 1..iitial ' ' pressure in cm. b, Bra
of HzSOi
19.5 19.5 10.9 11.0 8.25 10.85 12 2 11 0
c , Clr
70.95 69. i j 9.2; 9.23 17 2 13 8 18.1 14.85
TABLEI V FINALAPPROXIMATIONS F O R K . (ASSUMEDQ = 0.01M)
Part Final log Io / I
due to CIZ
DAB.
0.927 .929 ,944 ,958 ,480 .7i7 .SI8 ,769
0.005 ,005
0.922 ,924 ,944 ,958 ,479 .777 .817 ,769
... ...
,001
... .001
...
R - 2bQ
~~
2.36 2.36 4.88 4.96 1.72 3.50 3.46 3.39
(BrCI) =2(b-x)
34.25 34.30 12.07 12.07 13.06 14.68 17.48 15.20
(CIZ)
I
