Vapor-Liquid Equilibria for the Ternary System

Methyl Acetate-Benzene-Cyclohexane
ISAMU NAGATA'
Department of Chemical Engineering, Kyoto University, Kyoto, Japan

S E P A R A T I O N of the benzene-cyclohexane mixtures mercury (average deviation of * 3 mm. Hg.). Such devi-
resulting from hydrogenation of benzene into pure con- ations affect observed x-y data negligibly. N o correction
stituents is industrially important. This is effectively was made on them.
accomplished bv azeotropic or extractive distillation of the
mixtures with an entrainer. Several investigators have EXPERIMENTAL RESULTS
reported vapor-liquid equilibrium data for ternary systems
including benzene and cyclohexane, which are needed to Binary Systems. The liquid phase activity coefficients were
design suitable distillation equipment ( 3 , 4 ,5 , 7. 13, 24, 2 5 ) . calculated using the following equation:
The objectives of this investigation are: to determine
Y8 = ?rY~/X2PI (1)
vapor-liquid equilibrium data on the methyl acetate-
benzene-cyclohexane system and to correlate the experi- The vapor pressure data for methyl acetate were obtained
mental data by two methods; one is an analytical method from the compilation of Timmermans ( 2 7 ) . The Antoine
expressing logarithm of activity coefficients, and the other equations cited by Lange ( 1 4 ) were used to calculate the
is an algebraic one, for comparison. vapor pressures of benzene and cyclohexane. Binary
equilibrium data are shown in Table 111.
The area condition for isobaric binary systems has been
EXPERIMENTAL
shown by Herington (11) and Thijssen (26)to be
Purity of Compounds. First grade (Japanses Industrial
Standards) methyl acetate was purified in accordance with
the procedure by Hurd and Strong ( 1 2 ) . Cyclohexane of
guaranteed reagent grade was used without further purifi-
cation. Special grade (Japanese Industrial Standards) For close-boiling point and some azeotropic systems the
benzene was purified by recrystallization repeated twice. right side of Equation 2 is nearly zero. The benzene-cyclo-
The physical properties of the pure compounds are shown in hexane system is an example of this statement. I n order to
Table I . check thermodynamic consistency for a binary system
Analytical M e t h o d . The binary mixtures were analyzed by showing an appreciable difference between both sides of
measuring refractive indices. T h e ternary mixtures were Equation 2 the heat of mixing data a t boiling points for
analyzed by combining refractive index and density the same system are necessary. Such data for the present
measurements. Refractive index measurements were made systems were not available. So no further calculations using
using a Shimadzu Pulfrich refractometer. Prism tempera- Equation 2 were made. Chao ( 2 ) modified the Redlich-
ture was controlled at 25 i 0.1" C. by circulating water Kister equations for isobaric conditions, accounting for the
from a constant temperature bath. Monochromatic light requirements of Equation 2. The equation used to correlate
was obtained from a sodium lamp. Modified Lipkin-type the experimental data on log y l / y ? in terms of liquid
pycnometers were used for density determination. The compositions is
values of refractive index and density were reproducible
within +0.0001. Figure 1 is a ternary calibration diagram
log ~
Yl
=a + b (x? - x;) + c (6x1x?- 1) (3)
Yl
constructed from analytical data listed in Table 11.
Temperatures were measured by a copper-constantan Constants of the equation are given in Table IV. The
thermocouple which was calibrated against a standard activity coefficients of the separate components are
thermometer and was connected with a Yokogawa P-7 expressed by
potentiometer. Observed temperatures were corrected to
normal boiling points (20) and are believed to be accurate + C(XI- xd] + x ? [ a+ b ( x 2 - xl)+ c(6xlx?- 1)l
log 7 1= xlx?[B (4)
within k0.05" C.
Vapor-liquid equilibrium data on the mixtures were log ~2 = xixi[B + C ( X i - x ? ) ] -XI[. +b ( ~ i
-xi) + C ( ~ X ~ W- ?l ) ] (5)
obtained under atmospheric pressure using a Colburn
vapor-recirculating still as described by Grisword and B and C in Equations 4 and 5 are not independent of b and c
Buford (8). Atmospheric pressure recorded during experi- respectively. They must fulfill the following relations to
mental runs showed small deviations from 760 mm. of satisfy thermodynamic requirements (19).

' Present address, Kanazawa University, Kanazawa, Japan. B = b, C = c - a (6)

Table I. Properties of Pure Compounds

Methyl Acetate Benzene Cyclohexane

Property Exptl. Lit. Exptl. Lit. (28) Exptl. Lit.
Density, 25" C. 0.9273 0.9273(28) 0.8739 0.8737 0.7738 0.7736(3)
Refractive index, 25' C . 1.3589 1.4979 1.4979 1.4236 1.4235(28)
1.3615(20/D) 1.3614(1)
Boiling point. O C., at 760 mm. Hg. 56.8 56.8(1) 80.1 80.1 80.7 80.7 (28)
~~

VOL. 7, No. 4, OCTOBER 1962 46 1

 Table 11. Data for Analysis of the System Methyl Acetate-Benzene-Cyclohexane

XP X?b d? n% XI X? d? n$ XI X? di5 ns
0.923 0.077 0.9218 1.3698 0.761 0 0.8706 1.3730 0.399 0 0.8123 1.3963
0.872 0.128 0.9181 1.3769 0.661 0.131 0.8704 1.3893 0.341 0.145 0.8200 1.4095
0.798 0.202 0.9131 1.3879 0.527 0.308 0.8708 1.4113 0.258 0.353 0.8315 1.4291
0.714 0.286 0.9081 1.3997 0.465 0.388 0.8708 1.4212 0.225 0.435 0.8365 1.4370
0.647 0.353 0.9041 1.4094 0.380 0.500 0.8711 1.4351 0.189 0.525 0.8420 1.4462
0.547 0.453 0.8986 1.4236 0.325 0.573 0.8714 1.4442 0.163 0.593 0.8461 1.4531
0.448 0.552 0.8920 1.4372 0.262 0.655 0.8718 1.4546 0.125 0.686 0.8522 1.4628
0.343 0.657 0.8881 1.4517 0.186 0.755 0.8721 1.4671 0.088 0.781 0.8583 1.4730
0.298 0.702 0.8861 1.4578 0.128 0.832 0.8727 1.4765 0.058 0.855 0.8635 1.4814
0.190 0.810 0.8814 1.4723
0.678 0 0.8546 1.3782 0.316 0 0.8021 1.4020
0 0.078 0.7791 1.4276 0.576 0.151 0.8569 1.3957 0.258 0.185 0.8132 1.4175
0 0.176 0.7860 1.4328 0.457 0.327 0.8602 1.4161 0.220 0.304 0.8208 1.4277
0 0.281 0.7941 1.4389 0.394 0.418 0.8618 1.4272 0.189 0.403 0.8274 1.4368
0 0.379 0.8025 1.4452 0.365 0.462 0.8626 1.4321 0.160 0.495 0.8338 1.4453
0 0.460 0.8100 1.4507 0.275 0.594 0.8652 1.4482 0.086 0.727 0.8512 1.4684
0 0.561 0.8202 1.4583 0.229 0.662 0.8667 1.4565 0.045 0.857 0.8617 1.4823
0 0.638 0.8281 1.4642 0.109 0.840 0.8703 1.4782
0 0.714 0.8367 1.4707 0.268 0 0.7968 1.4051
0 0.795 0.8463 1.4779 0.624 0 0.8455 1.3816 0.221 0.175 0.8078 1.4187
0 0.867 0.8557 1.4847 0.535 0.143 0.8491 1.3975 0.186 0.305 0.8166 1.4298
0.477 0.237 0.8516 1.4079 0.152 0.431 0.8260 1.4407
0.945 0 0.9129 1.3621 0.400 0.360 0.8549 1.4221 0.128 0.522 0.8326 1.4490
0.787 0.167 0.9041 1.3854 0.333 0.467 0.8580 1.4346 0.103 0.616 0.8402 1.4578
0.679 0.281 0.8990 1.4011 0.266 0.574 0.8608 1.4467 0.082 0.694 0.8464 1.4656
0.600 0.365 0.8954 1.4127 0.193 0.690 0.8641 1.4606 0.058 0.784 0.8541 1.4747
0.501 0.470 0.8911 1.4270 0.081 0.870 0.8696 1.4821 0.042 0.844 0.8593 1.4809
0.411 0.565 0.8874 1.4400
0.342 0.637 0.8850 1.4498 0.550 0 0.8334 1.3865 0.154 0 0.7856 1.4132
0.248 0.737 0.8816 1.4631 0.461 0.162 0.8391 1.4034 0.128 0.171 0.7973 1.4246
0.179 0.811 0.8791 1.4728 0.404 0.265 0.8432 1.4144 0.097 0.371 0.8124 1.4394
0.362 0.342 0.8461 1.4227 0.082 0.465 0.8201 1.4470
0.863 0 0.8928 1.3669 0.272 0.505 0.8527 1.4406 0.072 0.532 0.8260 1.4525
0.727 0.157 0.8887 1.3879 0.216 0.608 0.8570 1.4522 0.056 0.638 0.8358 1.4620
0.595 0.310 0.8853 1.4080 0.182 0.669 0.8593 1.4591 0.045 0.708 0.8424 1.4683
0.511 0.407 0.8830 1.4209 0.082 0.851 0.8672 1.4803 0.031 0.799 0.8517 1.4773
0.457 0.471 0.8816 1.4292 0.022 0.860 0.8580 1.4832
0.362 0.580 0.8799 1.4442 0.478 0 0.8230 1.3911
0.300 0.652 0.8788 1.4531 0.402 0.158 0.8299 1.4066 0.076 0 0.7794 1.4185
0.208 0.758 0.8771 1.4670 0.349 0.269 0.8351 1.4179 0.063 0.180 0.7919 1.4293
0.136 0.843 0.8757 1.4779 0.300 0.374 0.8403 1.4286 0.049 0.365 0.8063 1.4416
0.244 0.479 0.8454 1.4395 0.043 0.439 0.8129 1.4471
0.820 0 0.8831 1.3694 0.208 0.565 0.8498 1.4488 0.035 0.548 0.8228 1.4478
0.744 0.093 0.8818 1.3814 0.132 0.723 0.8584 1.4662 0.028 0.640 0.8321 1.4631
0.595 0.274 0.8793 1.4048 0.076 0.841 0.8649 1.4795 0.021 0.719 0.8403 1.4702
0.500 0.390 0.8780 1.4198 0.015 0.801 0.8494 1.4777
0.443 0.460 0.8772 1.4286 0.010 0.863 0.8568 1.4839
0.344 0.581 0.8763 1.4442
0.274 0.660 0.8758 1.4551
0.198 0.759 0.8748 1.4676
0.152 0.815 0.8746 1.4743 Mole fraction of methyl acetate. * Mole fractiQn of benzene.

Table Ill. Binary Vapor-Liquid Equilibrium Data

Methyl Acetate-Cyclohexane Benzene-C yclohexane

t. = c . X? L. Y I Y, t, c. X? %I 1 1 7)
74.3 0.033 0.182 xi30 1.031 79.5 0.101 0..131 1.321 1.003
68.4 0.085 0.350 2.804 1.046 78.9 0.171 0.211 1.278 1.005
64.9 0.142 0.443 2.383 1.073 78.4 0.256 0.293 1.206 1.020
69.7 0.283 0.575 1.843 1.170 77.8 0.343 0.376 1.177 1.038
59.0 0.313 0.594 1.761 1.195 77.5 0.428 0.445 1.127 1.071
57.9 0.373 0.625 1.614 1.288 77.4 0.525 0.529 1,088 1.098
56.8 0.478 0.664 1.389 1.406 77.4 0.571 0.564 1.074 1.122
56.7 0.507 0.673 1.333 1.454 77.6 0.665 0.645 1.048 1.164
56.0 0.616 0.714 1.194 1.674 77.9 0.759 0.728 1.027 1.230
55.8 0.688 0.744 1.120 1.857 78.2 0.810 0.777 1.017 1.268
55.7 0.722 0.759 1.093 1.969 78.6 0.863 0.834 1.012 1.297
55.5 0.781 0.789 1.054 2.204 79.3 0.945 0.926 1.004 1.405
55.55 0.835 0.820 1.027 2.491
55.8 0.940 0.914 1.007 3.244
Mole fraction of methyl acetate. Mole fraction of benzene.

Table IV. Constants Used in Expressing Log r,/y, Experimental and calculated activity coefficients are
d o t t e d in Figure 2. I t is Dractical to comDare calculated
Y

vapor conpositions and boiling points with the experimental

System a,, b,, C,J
Methyl Acetate-Benzene -o,0308 o,1245 -D.0183 data to estimate the accuracy of numerical correlations
Benzene-C yclohexane 0 o,1447 o.0121 (Figures 3 and 4). The calculated vapor compositions
Cyclohexane-Methyl Acetate 0.0257 0,5274 0.0016 closely agree with the observed values, but the calculated
temperatures are higher than the experimental results. The

462 JOURNAL OF CHEMICAL AND ENGINEERING DATA

 0.4

Figure 1. Lines of constant density

0.2

XI

Figure 2. Log yi/y2 V S .

composition diagram
H Methyl ocetate-cyclohexane
0 Benzene-cyclohexane
A Methyl acetate-benzene

CYCLO.
HEXAM

0.2
1
+ [ X I X ? ( X >+ XI) + x X f ] ( C 1 7 + . ) + [ X ? X i ( X ? + X3) - BXiXl]
, ,

(e23 ... ) + [-2x?x1(xI + X?) + xlX,xl](cll + . ,. )

+ c 2 ( x 1 - x,) + Ci(X> - x-.) + ...] + XIX?X3[(XI + Xl)

(Cl - 2c2 + C?) - X?(2Cl - C? - cy) + . .. ] (7)

Cgclic advancement of the subscripts gives log y2 and log y 3 .

G values are expressed by
0 0.2 0.4 0.6 0.8
*I

Figure 3. Log y vs. composition diagram

1 Benzene-cyclohexane
2 Methyl acetate-benzene
3 Methyl acetate-cyclohexane I n the numerical determination of ternary constants from
0 Experimental the data, b was assumed to be equal to B , and the other
- Calculated ternary constants were set to zero, for the sake of sim-
plicity. The value of b = 0.102 was determined in the
vapor-liquid equilibrium data for the benzene-cyclohexane concentration range where the ternary constant b has a
system have been reported by several investigators (3, 6, 20, significant contribution to log y and was retaine4 to give
22, 23, 24). The author's results agreee well with those of the precise fit of the experimental and the calculated data
Donald and Ridgway (6) and of Sieg (23). The data on the as shown in Figure V . The deviation of the calculated vapor
methyl acetate-benzene system were given by the author compositions from the experimental results is 0.007 for
( I 7). The data on the methyl acetate-cyclohexane system methyl acetate, 0.007 for benzene, and 0.005 for cyclo-
are not reported in the literature. This system shows hexane, respectively. The calculated temperatures average
azeotropic behavior. 0.5" C. higher than the observed data. Repeated trail calcu-
Ternary. system. The ternary vapor-liquid equilibrium lations indicated that the benzene-cyclohexane azeotrope
data are listed with the calculated results in Table V. The diminishes completely a t the liquid concentration of methyl
logarithm of activity coefficients of the ternary system are acetate of 0.293 mole fraction. There is no ternary
expressed by an extension of the binary equations (17). azeotrope.

VOL. 7 , No. 4, OCTOBER 1962 463

 mN03t-t-
mriorim
N O O O G

33-13 r i d d i d

m
U
m
z
. . . . .
r i d 4 1 3
-0 U

GE -1Lo?c??
. I-wwc-w
m m m ~ w

N
C
m
z
+
c
U
E
a
-
x
f
?
5
b
+
U
+
O
n
.- 5
L
wmmm*
rccowt-t-
c9""c??
a
.-
- 0 0 0 0 3

3
0-
UJ
0
.-0-
2
O w d m 0
bQ oowwww
9wu?yc?
2 0 0 0 0 0

2-
E
c
>
m
5
t

xmot--
i m c s m m
"""i"
ooooo 0 0 0 3 0

c-mwxt-
@ J i m 0 0
?*?t;q
C O G 0 3

464 JOURNAL OF CHEMICAL AND ENGINEERING DATA

 1.0, I 1 I I I I I I I

Algebraic Method. The algebraic equations enable us to

carry out rapid calculation of vapor-liquid equilibrium data.
The ordinary three-constant Prahl equation involves more
additional terms which improves the fit of the experimental
x-ydata (9, 10. 15,211.
The following equation was used to correlated the binary
x-y data.

x, + x a, + x,x d ,
:=:( x b +xc, 1 (10)

This equation has another advantage: the ternary (or

multicomponent) vapor-liquid equilibrium are predicable
from component binary data as suggested by Lu, Li, and
Ting (15) if b values satisfy the condition of Equation 11
approximately.
bi?.bn,bp = I (11)
A three-index equation based on dquation 10 is expressed
by
y1:y?:.y3=

,i bi

,i
:
bl,

: (*
where CY'S are ternary constants to be determined horn the
experimental data.
Boiling temperatures of the binary and ternary systems
were calculated in accordance with the method proposed by
the author ( 1 6 , 1 8 ) .For the present ternary mixtures

where yl?a is defined by

Similarly for and y31?. Constants of Equations 1 2 and

VOL. 7 , No. 4, OCTOBER 1962 465

 NOMENCLATURE
al,, b,?,el?; b2?,e??; all, bnl, cl! = binary constants of modified
Redlich-Kister equations
b , e l , c?, el; B . C,, C?, Ca = ternary constants of modified
Redlich-Kister equations
a??,bl?, cI1. d,?; a l i , b21, c n i , d u ; a31, bal, r 7 1 , d , = binary constants
of algebraic equations
d: = density at 25O C. with respect to water a t 4” C.
GE = excess free energy
AH^ = heat of mixing
n% = refractive index for sodium D line at 25” C.
p b = vapor pressure of component i
R = gasconstant
T = boiling tempeiature, K.
T I ? ,T,3, T I ? ,TI?? = constants
x = mole fraction in liquid phase
y = mole fraction in vapor phase
yl!,y ~ ?y13,, y123; y19, y21,y13, y231,’ 3’23, y ~y ?, ~y3,12 = variables
L Y ~ (Y?,
, a 3 = ternary constants of algebraic equations
y = liquid phase activity coefficient
T = total pressure

Subscripts
1, 2, 3, i, j = components
0 42 0.4 0.6 0.8 1.0 m = mixture
X

Figure 5. Boiling points vs. composition diagram LITERATURE CITED

1 Benzene-cyclohexane
2 Benzene-methyl acetate (1) Bushmakin, I.N., Kish, I.N., Zhur. Priklud. Khim. 30, 200
3 Methyl acetate-cyclohexane (1957).
0 Experimental (2) Chao, K.C., Ind. Eng. Chem. 51,93 (1959).
- Calculated (Modified Redlich-Kister) (3) Chao, K.C., Hougen, O.A., Chem. Eng. Sci. 7, 246 (1958).
(4) Dakshinamurty, P., Jayarama Rao, G., Acharya, M.V.R.,
____
Calculated (Nagata method) Venkata Rao, C., Ibid., 5, 188 (1956).
( 5 ) Dakshinamurty, P., Venkata Rao. C., J . A p p l . Chem.
13 are listed in Table VI. The average deviations of the (London) 7,655 (1957).
calculated from the experimental results are as follows: (6) Donald, M.B., Rideway,- - K., Ibid.. 8, 403 (1958).
(7) Ibid., p. 408.
Vapor Composition ( 8 ) Griswold, J., Buford, C.B., Id. Eng. Chem. 41, 2347 (1949).
Methyl acetate 0.015 mole fraction (9) Hala.. E... Collection Czech. Chem. Commun. 24, 2453 (1959).
Benzene 0.012 mole fraction (10) Hala, E., private communication, 1961.
Cyclohexane 0.007 mole fraction (11) Herington, E.F.G., J . Inst. Petrol. 37, 457 (1951).
Boiling Temperature 0.5” C. (12) Hurd, C.D., Strong, J.S., Anal. Chern. 23,542 (1951).
(13) Kurmanadha Rao, K.V., Krishnamurty, V.V.G., Venkata
Rao, C., Rec. Trau. Chim.76, 769 (1957).
Table VI. Constants of Ternary Algebraic Equations (14) Lange, N.A., ed., “Handbook of Chemistry,” 9th ed., p. 1424,
Handbook, Sandusky, Ohio, 1956.
(15) Lu, B.C.-Y., Li, J.C.M., Ting, T.-W., I n d . Eng. Chem. 51,
Ul? bl? CI? dn (YI T:? b,?.b?,.bSi 219 (1959).
0.760 0.277 0.629 0 0.724 338.7 0.876 (16) Nagata, Isamu, J. CHEM.ENG.DATA6, 586 (1961).
an b, c 21 du CY? T,i TI?? (17) Ibid., 7, No. 3 (1962).
1.473 1.015 1.406 0 2.311 348.8 360.0 (18) Ibid., 7, No. 4 (1962).
(19) Naphtali, L.M., Chao, K.C., I n d . Eng. Chem. 51, 1318 (1959).
~3~ bSi C?1 d31 as Ti3 (20) Perry, J.H., ed., “Chemical Engineers’ Handbook,” 3rd ed.,
5.009 3.115 6.698 -2.385 0 327.7 p. 293, McGraw-Hill, New York, 1950.
(21) Prahl, W.H., Ind. Eng. Chem. 43, 1767 (1951).
(22) Richards, AS., Hargreaves, E., Ibid., 36, 805 (1944).
(23) Sieg, L., Chem.-Ing.-Tech. 22, 322 (1950).
CONCLUSIONS (24) Thornton, J.D., Garner, F.H., J . A p p l . Chem. (London) 1,
suppl. No. 1, s 61 (1950).
The methyl acetate is suitable as an entrainer for (25) Ibid., p. s 68.
azeotropic distillation of the benzene-cyclohexane mixtures. (26) Thijssen, M.A., Chem. Eng. Sci.4, 75 (1955).
Both the analytical and algebraic methods give the boiling (27) Timmermans, J., “Physico-Chemical Constants of Pure
temperatures of the same accuracy. The former provides Organic Compounds,” Elsevier, New York, 1950.
us with more precise vapor compositions than the latter (28) Weissberger, A., Proskauer, E.S., Riddick, J.A., Toops, E.,
does, although the method is slow. Jr., “Organic Solvents,” 2nd ed., Interscience, New York,
1955.

ACKNOWLEDGMENT
RECEIVED for review March 16, 1962. Accepted June 29, 1962.
The author thanks Fumitake Yoshida who supported this Part of dissertation submitted by Isamu Nagata to Kyoto Univer-
work and Yoshio Tamaki for carrying out part of the sity in partial fulfillment of the requirements for the degree of
experimental measurements. doctor of engineering.

466 JOURNAL OF CHEMICAL AND ENGINEERING DATA