August, 1946

INDUSTRIAL AND ENGINEERING CHEMISTRY

817

and noncrybtalline components. Chemical methods, such as the one used in this work, permit quantitative estimations t o be made, although it must be kept i n mind that accessibility IS determined by the physical and chemical conditions imposed during the test (1). Some support for the present results may be taken from the data of Lauer and co-workers ( I O ) , who found that both isotropic and highly stretched rayons of many different types give essentially the same specific heat of wetting. If heat of wetting is a meawre of the internal accessible hydroxyls, the conclusion must again be drawn that accessibility (or crystallinity) is not affected by stretching, but rather is determined primarily in the procc.i of coagulation and regeneration.
LITERATURE CITED

Haas, R. H.. and Purves, C. B.. J . -4m. Chcm. Poc.,66, 59-65 (1944). (21 Chnrch. W. H., and Underwood, W. F. (to E. I. du Pont de Xeniours & Co.), U. S. Patent 2,249,745 (July 2 2 , 1911) C31 Conrad, C. C., and Scroggie, A. G., IND.ESG. CHEJI.. 37.
111 Assaf, . . iG.,
I

592-8 (1915). Dorte. C., hIethods of Cellulose Chemistry, p. 212, New York, D. Vin Sostrand Co., Inc., 1933. ( 5 ) Irnns, E., Die Chemie, 56, 113-20, 132-6, especially 13-1. Fig. 25 ( 1 913).
(4)

G r a l h , N., and Samuelson, 0.. Sccnsk P a p p c r s t i d n . , 48; 1-6 (1945). Hermans, P. H., in Rohrs, Staudinger, :and Viewegs Fortschritte der Chemie, Physik, und Technik der makromolekularen Stoffe, Vol. 11, pp. 17-35, Berlin, J. F. Lehmanns Verlag, 1912. Kratky, O., Angew. Chem., 53, 153-62 (1940). Kratky, O., in Rohrs, Staudinger, and Vien-egs Fortschritte der Chemie, Physik, und Technik der makromolekularen Stoffe, Vol. I, pp. 172-91, Berlin, J. F. Lehmanns Verlag, 1939. Lauer, K., Doderlein, R.. Jiickel, C. and Wilde, O., .I. m k m mol. Chem., 1, 76-96 (1943). Z Lauer, K., and Mansch, W., ellzolle 11. Kunstseide, 1, 39-43 (1943). Mark, H., in L. E. Wises Wood Chemistry, p. 129, New York, Reinhold Pub. Corp., 1914. Nickerson, R. F., IXD. ESG.CHEX,34, 1180-5 (1012). Ritter, G. J., and Mitchell, R. L., Paper 7mfe J . , 108, No. 8 , 59-63 (1939). Rose, L., J . SOC.Dyers Colourists. 61, 113--18(1945). Sisson, W.. .in Einil Otts Cellulose and Cellulose Deriva1, tives, p. 226, Yew York, Interscience Publiehers, Inc.. 1943. Spurlin, H. M., Ibid., pp. 931-42.
PRESEVTED Yorthwest Re:ionnl lleeting of the AMERICAN a t the CAEUICAL SOCIETY. Seattle, Wash., October 20, 191.5. Contrihution No. 4 from t h e Central Chemical Laboratory of Rayonier 1 n c o r p o r a . d

The System Acetone-Water1,1,2-Trichloroethane

J

TERNARY LIQUID AND BINARY VAPOR EQUILIBRIA

Robert E. Treybal, Lawrence D. Weber, and Joseph F. Daley2
S e w York Cnirersity, S e w Y o r k

53, S. Y .

T h e solubility, tie lines, density, and refractive index were determined in the ternary system acetone-water1,1,2-trichloroethane at 25 C. Applications of various tie line correlations are considered, and a convenient method is described for estimating the position of the plait point. Densities, refractive indices, and viscosities at 25 C. as well as vapor-liquid equilibria at 755 mm. Hg were determined for the binary system acetone-trichloroethane. Vapor pressures of trichloroethane were measured over the range 73.5 to 113.5C. The data indicate that 1,1,2-trichloroethane would make an acceptable solvent for extracting acetone from its aqueous solutions.

The 1,1,2-trichloroethane, purchased from the Carbide and Carbon Chemicals Corporation, was distilled by a method similar to that used for acetone, and the central portion of the distillate had a normal boiling point of 113.5 C. over its entire range. Published values range from 113.3to 114 C!. (6,9,O ) , with one I listing of 113.5 C. (8). The central fraction had a refractive index nzj = 1.4715, whereas previously published values range from 1.4711 (8) to 1.4706 (9), and a density d = 1.4412, which : agrees with the published value (9). The third component WBB laboratory-distilled water.
TERNARY LIQUID DATA

teristics of liquid-liquid extraction equipment, it was found desirable to have rather complete data on equilibrium and other physical properties for the ternary system acetone-water1,1,2-trichloroethane. These data were accordingly determined, and are reported in this paper. Reagent grade acetone, dried over Drierite, was fractionated in a laboratory column equivalent to approximately 20 theoretical plates, and a central portion of the distillate amounting to about 75% of the still charge was retained for this nork. The fraction had a normal boiling point of 5 6 . 1 C. over its entire range, a density di5 = 0.7840, and a refractive index n22 = 1.3556; these values agree with the accepted characteristics ( I O ) .
I
2

K C O S S E C T I O S n-ith a program for determining the charac-

A t present in t h e United Statpa K a r y . Irwent address, T h e Flintkote Company, E a s t Rutherford, K. .I

Determinations were made a t 25 C. of the limiting solubility and the tie lines in the ternary system, according to the CURtomary procedures. Mixtures of acetone and trichloroethane of known concentration were held in a water bath a t 25.0 * 0.1. C., and nere titrated with water until the appearance of a slight turbidity indicated the limiting solubility concentrations. The refractive index (by -4bbe refractometer) and density (by pycnometer) of the solutions were determined after the turbidity had settled from the solution by standing for a short time in the water bath. Similar measurements n-ere made by titration of acetonewater solutions with trichloroethane. The solubility of trichloroethane in acetone-free water (Table I) is in agreement with the data of van Arkel and Vles (f). Although the solubility of water in trichloroethane has apparently been measured (16), the present determination could not be checked because the published data were unavailable.

818

I N D U S T R I A L A N D E N G I N E E R I N G CHEMISTRY
mol. wt. acetone
R E T O N E

Vol. 38, No. E niol. wt. H,O

~ ~ 1 5 7

1.811

14.58 (mol. wt. C2HICII)(mol. wt. a , t i i r )

Figure 1 . Liquid Equilibria of System AcetoneWater 1 , 1 , 2 Trichloroat hane a t 2.5 C.

Since the slopes of these lines are so nearly uiiit,y, ttit: v d i i i , I +.3X should nearly equal the initial slope of the mole fravtiiiii (list+ hution curve of Figure 2 (17), and this is the case. A convenient method of estimating the position 1 1 1 t l i c b plait

3 hours. Samples of the settled layers were withdranTn, and their refractive indices measured. Reference t o a large scale plot of refractive index against concentration of acetone for saturated solutions gave the analysis of the equilibrium layers. Data for the tie lines are listed in Table 11. Solubility and tie line data are plotted in*Figure 1 on the usual triangular coordinates. The crosses near the centers of the tie lines represent the over-all composition of the two-liwid-phase mixtures used in determining the individual tie lines; it is necessary, if a material balance i s truly established, that the tie lines pass through these points.
TIE LINE CORRELATIONS

another factor which influences the choice of solvent in an extraction process. They point out that in the extraction of a solute such as acetone from an aqueous solution, it is important that the high relative immiscibility of solvent and water be maintained even at, high solute concentrat,ions. They showed thi*

TAB,,E

I,

LIMITING

SoLUBILrTY

CURVE FOH

.~ck~,,.ox k;-

W.ITER-l,1,2-TRICHLOROETHAhE. T 25 4
Conoentration, Weight % CtHaCla ( a ) Water ( b ) Acetone ( c ) 100 XI 100 100 X s 99,89 94.73 90.11 84.85 79,58 75,72 70,M 67.52 64.17 60.06 54.88 48.78 43.88 38.31 31.67 26.39 24.04 23.20 20.71 15.39 10,oo 9.63 6.77 4.35 2.18 1.72 1.17 1.02 0.92 0.78 0.70 0.65 0.62
0.44

c;.
Densit,.

xz

Refractive Index,

ng

diS

A simple curve of the distribution of acetone between t w o immiscible solvents can be obtained by plotting the concentration of acetone in the water-rich layer against its concentration in the trichloroethane-rich layer. This is presented in Figure 2, in both weight and mole fraction units. The acetone favors the trichloroethane-rich layers; consequently this solvent can be recommended (at least from the point of view of liquid equilibrium) for extracting acetone from aqueous solutions. Using the notation previously presented ( 1 7 ) , the weight fractions of components a,b, and c of the ternary mixtures are X i , X,, and Xt, respectively, and the corresponding mole fractions are xi, x2, and x3. Thc weight fraction of acetone c, in the water @)-rich layer is t,hen S , in the trichloroethane (a)-rich layer , X31,etc. Some years ago, Hand (7) showed that, distribution data plotted in this notation result in straight. lines (Figure 3). The same type of plot was later developed in terms of mole fractions, and its significance was discussed ( 1 7 ) . The present data plot well i n this manner, and the equations of the straight lines are

0.11 0.26 0.36 e.59 0.76 0.99 1.43 1.44 1.87 2.11 2.98 4.01 5.00 6.84 9.78 13.35 15.37 16.63 19.31 26.28 34.96 35.38 41.35 48.47 55.97 61.11 66.58

? = 14.58 !! 211

(E)

1.067

x1 1

= 1.841

(2)
1.05:

It has previously been shown that the coefficients in the equations are relat,ed (17). Thus,

71.80 74.54 80.40 84.94 87.63 94.66 99.66

0.00 5.01 9.53 14.76 19.66 23.29 28.21 31.04 33.96 37.83 42. I4 47.21 31.12 54.85 58.55 60.26 60.59 60.17 39. 98 58,33 55.04 54.99 51.88 47,18 41.85 37.17 32.25 27.18 14.54 18.82 14,30 11.72 4.82 0.00

I . 4683
1.4588

....
1,3737 I ,3286 I, 2763 I ,2936 1.2018 1.6110 1.1170 I 0882 1.0592 i . n2m
0 .j i b 4 0.9834
,...

1.4504
1,4416

1,4338 I . 4285 1.4218 I . 4178 1.4139 1.4088 I . 4032 1.3972 1 ,3927 1.3888 1.3829 1.3792 1,3792 I , 3758 1,3696 1,3672 1.3672 1.3662 1.3626 1.3601 1.3578 1.3543 1.3518 1.3501 1.3460 1.3430 1.3412 1.3362 1.3328

....

0,9386 0.9328 0 , 918!)

0,9240

....

0.9223 0.9223 0. 9259 0.9293 0.9368 0.9452 0.9539 0.9608 0.9647 0.9747 0.9797 0.9984

....

August, 1946

INDUSTRIAL A N D ENGINEERING CHEMISTRY

INITIAL SLOPE
8

819
/

graphically by the use of effective concentration curves; the present data are plotted in this manner in Figure 4. It is evident that the curve remains high over great ranges of acetone concentration.
,4DDITIONAL SOLUTION PROPERTIES

14.58

order to make more complete the avaifable d a t a such mixtures, additional densities, refractive indices, and viscosities were determined. Those for tliv hinary system acetone-trichloroethane are listed In Table I11 and plotted in Figure 5 . Viscosities were detemiined with calibrated Fenske-CannonOstwald viscometers. Several densities of the ternary mixtures in the area of complete mutual solubility were also drtcrmined (Table IV). The d a t a were plotted with the densities of solutions along the wlubility curve, with densities of acetone-water solutions ( I O ) , and with densities of acetone-trichloroethane (Figure 6). The densities and tie line relationships are such that, a t 311 parts of the diagram pscept tlie plait point, the Fater-rich layers are less dense tlian the corresponding equilibrium trichloro.rt,ti:inc-ricli layers.
111
8lJll

01 .0 x, , ,X=

0.20

0.30

04 .0

0.50

0.60

C '0

4POR-LIQUID DAT.4 FOR ACETOZIETRICH LOROETHANE

MOLE OR WEIGHT FRACTION ACETONE IN WATER L A Y E R

Figure 2.

( ~ r d e rfor a solvent to be satisfactory for extraction purposes, it must be easily separable from the extracted solute. In this connection vapor-liquid equilibria a t a pressure of one atmosphere (755 mm. Hg) were determined with a standard Othmer still ( 2 1 , 12) fitted with a barostat. The barostat was capiible of holding the pressure constant within 0.5 mm. of nir:rcriry. Samples from the still and cond(%iisate rcwiver were analyzed by refractive index tvitti thv 1i(-Ip of the previouely determined data for . t m . Results are presented in Table V and Figure 7 . I n order to test the thermodynamic consistency, of t,liesc data, it was desired t o plot the activity coeff~cientsas a function of liquid concentration in the manner outlined by Carlson and Colburn (4); in this method vapor pressure data for the two component,s were necessary. Adequate data were available for acetone ( 1 5 ) , but considerable discrepancy was found between the published vapor pressures for trichloroctlinne ( I O , 15) and similar data made available by the mmufacturers (3). Consequently some additional mensurements were made with a Wash!~urn-lteadapparatus (18) fitted to the barostat used
111

Distribution of Acetone betw-een Water and 1,1.2-Trichloroethane at 2<5"C.

x ,
Figure 3.

x32

x,, 722

., 1

\I{I.C:

IT. I~QVILIBRII-N TIE LIXE COSCESTRATIOSSFOR . - \ ~ . ~ : ~ ~ . ~ ~ - \ ~ ~ . ~ T L R - ~ , ~ , ~ - ATR I c H L o R o E T H A N E T 2.5' C.

Weight

Tie Line Correlation and Determination of Plait Point

_ .

70
(c)

CtIIpCis(a) W a t e r ( b ) Acetone
100 -Y,2 100
x2 2

CzH3C13 ( a i
100
.Yll

Weight % Water (6)

100

100

x2 3

x2,

.icetone ( c ) 100 XSl

8.75 10.28 20.78 25.14 27.66 37.06 38.52 39.39 41.67 42.97 48.21 53.95 57.40 60.34

1 0 .
----0-CLo

0.52 0 54 0.68 0.73 0 7!l I.00 1.02

I 04 1.11

93.52 92 95 85.35 82.23 80 16 73.00 72.06 71.33 69.35 67.95 62.67 57.00 50.20 41.70

5.96 6.51 13.97 17.04 19.05 26.00 26.92 27.63 29.54 30.88 Q5,73 40.90 46.05 61.78

90.93 89.32 78 32 i 3 78 71.01 60,85 59.21 28.21 05.48 53.92 47.53 40.00 33.70 26.26
14.6

0.32 0.40 0 90 1.10 1 33 2.09 2.27 2.40 2.85

-c.=a-<,

o \

09 .

O '\

08 .

1.17 1 60 2.10 3.75 6.82

3.11
4.26 6.05 8.90 13.40 27.4

WT.X ACETONE WT.% ACETONE+ WATER " WATER LAYER

Estimated plait paint

58.0

Figure 4.

Effective Concentrations in the Ternary Liquid System

820
15 .

INDUSTRIAL AND ENGINEERING CHEMISTRY

Vol. 38, No. 8

TABLE 111. PROPERTIES O F AiCETOXk?l, 1,2-TRICHI.OROETHh?E SOLUTIOSS 25' C. AT

1.4
Acetone,

Wt.

so

Density,

dZ5
1.4316 1.3212 1.2337 1.1521 1.0771

r'iscosity, Centipoises
1.061 0.910 0.799 0.677 0.517

Refractivt Index, n: ; 1.1687 1.4498 1.1343 1.4210 1.40iC~ 1.3980 1.3871 1.3796 1.3698 1.3630 1.3556

1 . 3

1 . 2

: :
I-

1.1

0.00 10.5 20.1 30.1 40.6 49.3 50.9 60.6 69.1 80.0 89.1 100.0

....

...

1,0140 0.9588 0.9158 0,8643 0,8226 0.7840

0.509 0.466 0.411 0:347 0.319

....

*1 . 0

TABLE . 11'
Weight

DEssITIEs OF ACETONE-\~ATER-~, - T R I ~,~ CHLOROETHASE SOLUTIONS '

z W
a

W
Acetone 59.23 68.70 89.13 61.99 73.23 83.13 65.05 83.78 91.33 Kater 7.88 6.05 2.10 21.25 16,26 10.24 26.95 12.50 6.68

CzHaCla

HzO/CZHJCI:
0 240 0,240 0.240 1.545 1.547 1.544 3.37 3.36 3.36

d' :

0 . 9

08 .

32.89 25.25 8 77 13.76 10.51 6.63 8.00 ' 3.72 1.99

0.9540 0.9106 0.8239 0.9052 0.8762 0.8405 0,8939 0.8374 0.8125

07 . in accordance with the generalizations baaed on hydrogen bonding of Exell, Harrison, and Berg (6). The general shape of the curves in Figure 9, and the fact that the molar volumes of the components are of the same order of magnitude, indicate that they should be capable of fitting the Margules type of equations (4). The best over-all fit that could be obtained required terminal values of activity coefficients equal to 0.437 for acetone and 0.912 for trichloroethane; the corresponding 'curves are shown as dashed lines in Figure 9. The fit is not good, yet the values of vapor concentration calculated from these curves, as listed in Table V, do not differ greatly from observed data. The ratio of terminai values of acetone activity coefficients for the binary systems acetone-water and acetone-trichloroethanr should equal the coefficient of the mole fraction tie line equation, 14.58, when the exponent in this equation is so nearly unity ( 1 7 ) . If the activity coefficient of acetone at infinite dilution in water is 9.3 ( 2 ) , this would require that the terminal acetone activity ,w$Ficient in trichloroethane solutions be 0.639, nlhich falls near rhc observed value. An exact check with t'he Margules equations may be lacking because of slightly incorrect temperatures takcn in the Othmer apparatus; incorrect readings may have beell m d e because, although the thermometers were calibrated, no provision was made for wetting the thermometer with both liquid and vapor. It may also happen that the Rlargules equations do not apply to this system. For the separation of acetone from trichloroethane solutions containing small amounts of water, as produced by an extraction operation, the flowsheet of Othmer and Itatcliffe (13) xould apply. Their conclusions with respect to the similar system acetone-water-chlorobenzene, that the small amounts of water held by saturated solutions \-,-illnot affect the separation of acetone, seem completely applicable here.
CONCLUSION

06 .
WEIGHT PERCENT ACETONE

Figure 5.

Properties of Acetone-l,1,2 -Trichloroethane Solutions at 2 5 " C.

previously on the Othmer equilibrium still. The results are presented in Table VI and in Figure 8 on (log p us. l / T o K.) coordinates. T h e new data and those made available ( 3 ) agree well and are somewhat lower than those previously published. Activity coefficients were calculated and plotted in Figure 9 (4). Since the activity coefficients are largely fractional, the solution^ show negative deviations from Rnoult's law. This is

A ACETONE

The ternary liquid equilibrium data, both for tie lines and effective concentrations, show that acetone is readily extracted from aqueous solutions by 1,1,2-trichloroethar1e. The density differences betneen equilibrium layers in the range of concentration that would ordinarily be encountered are large, and interfacial tension characteristics, although not measured, are s1lc.h

821

0.4 0.5 0 6 07 0.8 . . 09 . MOLE FRACTION ACETONE: X=LIPUIO, Y m VAPOR

01 .

0 . 2

03 .

1 0 .

Figure 7. Vapor-Liquid Equilibria of Acetone2,1,2-Trichloroethane a t 75.5 RIm. Mercury

I n addition the vapor-liquid equilibrium data shorn that the acetone can be readily separated from trichloroethane solutions by distill% tion. This solvent should therefore be acceptable as an extracLion agent for removal of acetone from its aqueous solutions
NOiMENCLATURE

that settling of insoluble phases is rapid and complete.

TEMPERATURE, C.

Figure 8 . Vapor Pressure of 1,1.2-Trichloroethane

a, b = relatively immiscible solvents c = distributed solute

= liquid concentration, weight fraction z = liquid concentration, mole fraction y = vapor concentration, mole fraction Subscripts 1,2,3= components a, b, c, respectively; first subscript refers to component whose property is indicated, the second to the predominant component of the solution. Thus, zal = mole fraction of c in an a-rich solution p = plait point

w V

1.1 10 .

G
W 0

I &

08 .

* e
I-

06 .
/
/

1
TABLE. VAPOR-LIQUIDEQUILIBRIA XCETONE1 FOR 1,1,%TRICHLOROETHt\SE A T 755 hIM. HG
Tpp., C. 57.3 57.5 58.2 58.9 59.8 61.7 69.0 69.0 72.7 78.8 85.8 87.6 89.5 93.7 95.6 100.6 101.9 102.2 104.2 112.0 Mole Fraction Acetone Vapor Liquid z Obsvd. Ealcd. 0.919 0.991 0.991 0.989 0.990 0.908 0.890 0.988 0.987 0.865 0,984 0.984 0,980 0.980 0.828 0.972 0.970 0.764 0.754 0.969 0.971 0.610 0.984 0.933 0.610 0.934 0.933 0.524 0.901 0.902 0.421 0.842 0.842 0.322 0.754 0.750 0.305 0.732 0.732 0.295 0.722 0.719 0.692 0.273 0.689 0.213 0.605 0.598 0.201 0.577 0.563 0.144 0.453 0.439 0.132 0.418 0.405 0.417 0.393 0.125 0.356 0.341 0.105 0.080 0.060 0.022
P.4POR P R E S S U R E O F

62.1

86.8

TABLE \I.
Temp.,

c.

i3.5

80.3 82.4 83 8 83. C 96.2

Vapor Pressure, Mm. Hg 200 258 278 289 312

445

International Critical Tables, New Tork. hIcGraa-Hill Book Co., Inc., 1928. (11) 1,~,~-TRICHLOROETHANE IOthmer, D. F., I K D . E N G . C H E M . , 20,743 (192s). (12) b i d . , 35, 614 (1943). Vapor (13) Othmer, D. F., and Ratcliffe, R. L., I b i d . . 3!5, i 9 S (1943). Temp., Pressure, (14) Othmer, D. F., Khite, R. E., and Treuger. E., I k i d . , C3, 1240 c. l l m . Hg
(10)
104.6

Activity Coefficient .4cetonr CzHaCIa 1.035 0.821 1.036 0.874 1.031 ,0.772 1.031 0.815 1.041 0.768 1.053 0.725 1.051 0.710 1,000 0.764 1.000 0.764 0.959 0.810 0.951 0.848 0.927 0.879 0.928 0.899 0.923 0.897 .0.912 0.907 0.908 0,930 0.8i5 0.921 0.840 0.944 0.820 0.950 0.825 0.938 0.826 0.947 0.732 0.975

Figure 9.

MOLE FRACTION ACETONE IN LIQUID 4ctivity Coefficients i n the Binary SFstem

-icetone-l,l,Z-Trichloroethane

LITERATURE CITED

(2) Brunjes, d.s., and Bogart, hf. J. P., IXD. E.NG. C m m , 35, 255
(1943).

(1) Arkel, A. E., van, and Vles, S.E., Rec. trac. cltim., 55, 4 0 i (1936).

(3) Carbide and Carbon Chemicals Carp., private communication. (4) Carlson, H. C., and Colburn. A. P., ISD.ENG. CHEM., 581 34,
(1942).

(5) Ewell, R. H., Harrison, J. &I., and Berg. L., Ibid., 36, 871
(1944).

(6)

Ewell, R. H., and Welch, L. A I . , J . Am. C6e7n. Soc., 63, 2475

(1941).

(7) Hand, D. B., J . Phys. Chem:, 34, 1961 (1930). (8) Handbook of Chemistry and Physics, 25th ed., Clewland. Chemical Rubber Publishing Co., 1941. (9) Henne, A. L., and Hubbard, D. M . , J . Am. Chem. Soc.. 58, 404
(1936)
I

108.7 109.8 111.3 113.5

582 655 678 708 760

(1941).

(15)

Perry. J. H.. Chemical Engineers Handbook, 2nd ed., Nca York, McGraw-Hill Book Co., Inc., 1941. (18) Staverinan, A . J.. K E C .trnn. chim., 60, 936 (1041l. (17) Treybal, R . E.. ISD. LSG. CHIN., 36, S i 5 (1944). (18) Kashhurn. E. IT., arid Read. J. W., J . Ani. ritem. SOC..41, 729
( I $11 R I .