1957
BENZENE-METHANOL MIXTUREEQUILIBRIA
Oct., 1946 [CONTR.IBUTION FROM THE
DEPARTMENT OF CHEMISTRY, MASSACHUSETTS INSTITUTE OF TECHNOLOGY]
Vapor-Liquid Equilibrium. VI. Benzene-Methanol Mixtures1 B Y GEORGE SCATCHARD,
SCOTT
E. WOOD'
A N D JOHN
M.
MOCHEL3
The second paper* of this series concerned the graduated in mm. for a length of 1 cm. near the top. The arm was closed with an outside cap. vapor-liquid equilibrium of chloroform-ethanol mixtures. Th.e increase of the.enthalpy on mixing They were filled by means of a hypodermic syrwas found to be positive in the chloroform-rich inge fitted with a stainless steel cannula of the apsolutions and negative in the ethanol-rich solu- propriate size and length. In order to reduce the tions, the excess increase of the entropy over that systematic error encountered in these measureof an ideal solution of the same composition on ments in the previous work, the solutions were mixing a t constant pressure was positive for the made up in 50-ml. Kimble glass volumetric flasks. ch10roforn1-rich solutions and negative for the The liquids were transferred by means of the hypoethanol-rich solutions, and the excess increase of dermic syringes. On adding the second compothe free energ:)' over that of an ideal solution of the nent the tip of the cannula was held just above same composition on mixing a t constant pressure the surface of the first liquid. In this way the was found to be positive over the entire concen- weight of the vapor displaced by the cannula is tration range. I n Papers 111, IV, and V5 the very small and the vapor displaced on the addition vapor-liquid equilibria were reported for the cycle of the second component is practically pure in the of binary systems with the components : benzene, first component and its weight can be estimated. The v a l m pressure rneasurernents with methacyclohexane, and carbon tetrachloride. The thermodynamic properties of these three systems were nol and with the solutions were made with the much simpler than for the chloroform-ethanol sys- same apparatus as described previously. Helium tem, the increase of the enthalpy on mixing, the was used as the confining gas. The vapor presexcess entrop:y of mixing, and the excess free en- sure of methanol was determined a t 5' intervals ergy of mixing all being positive over the whole from 25 to 63'. The equation fitted to these data range of concentrations. It was of interest then is ( T = 273.16 t ) to compare the properties of binary systems comlog p = 7.98963 - 1499.361/T 76225/T2 (1) posed of these non-polar components and an alThe normal boiling point calculated according to cohol. this equation is 64.51', which is in good agreement Experimental with 64.46' given by Butler, Thomson and Synthetic methanol was purified by distillation McLennan. I n the measurement of the vapor-liquid equilibin the column described in 111. Water vapor was excluded during the distillation by passing dry rium a t a mole fraction of methanol between 0.02 nitrogen through the column a t all times. The and 0.05 a steady state could not be attained. It purified methanol was stored in a dry box pro- was found that the boiling temperature was extected by dry nitrogen. The density of the meth- tremely sensitive to the distribution of the comanol a t 25' was 0.78654 g. per cc., 'compared to ponents between the various regions of liquid and 0.78650 given by Jones and Fornwalt.6 The ben- vapor in the boiling apparatus. This is due to the zene was purified as described in 111. The vapor marked dependence of the boiling point on the pressure of the purified product agreed within the composition of the liquid and also to the large experimental .accuracy with the equation given in difference between the composition of the liquid and vapor. At these concentrations the pressure IV. New pycnometers were used in determining the was adjusted so as to keep the boiling temperature relation between the density and composition. within 0.05" of the desired temperature for at They were single armed pycnometers having a least a n hour. The run was then stopped and capacity of 2.bout 10 cc. made of Jena 16-111 samples taken. The compositions were determined from the glass.' The c:apillary had a 1-mm. bore and was densities a t 25' which are given by the equation (1) Presented before the Division of Physical and Inorganic
+
Chemistry of the American Chemical Society at Atlantic City, N. J., April 10, 1946. (2) Present address: Yale University, New Haven, Connecticut. (3) Present address: Corning Glass Works, Corning, New York. (4) G . Scatchard and C. L. Raymond, THISJOURNAL, 60, 1278
(1938). ( 5 ) G . Scatchard, S. E. Wood and J. hl. Mochel, 111, J . Phys. Chem., 43, 119 (1939); I V , THISJOURNAL, 61, 3206 (1939); V, 62, 712 (1940). (6) G. Jones anci H. J. Fornwalt, 60, 1683 (1938). A summary of the best values is given in this paper. (7) E. R. Smith and M. Wojciechowski, Rocrniki Chem., 16, 104 (1936); L. G . Lonmworth, T H I S JOURNAL, 6% 1483 (1937).
l / d = wi/di
-
+
Ws/dt
+6
(2)
in which dl and dz are the densities of the pure components and w1 and wz are their weight fractions. The subscript 1 indicates methanol, and benzene. Table I gives the weight fractions and mole fractions of methanol, the measured densities, 6, and the measured values of 100V'/Vo. V is the volume of the mixture, Vothat of the unmixed (8) J. A. V. Butler, D. W. Thomson and W. H. McLcnaan, J . Chem. Soc., 674 (1933).
GEORGE SCATCHARD, SCOTTE. RTOODAND JOHN M. MOCHEL
1958
components, and VM = V - VO. It will be noticed that V " / P is quite small and has both positive and negative values. The quantity, 6, used in determining the compositions was read from t: large scale graph of the values given in Table 1. Th.ese density-composition results agree quite well with those of Washhurn and Lightbody3 I" and Sch mid I.. 'rABLE
I
DENSITIESOF BENZENE-METHANOL MIXTURESAT 25 O
n't. fract.
Mole fract. CHIOH
CHrOH
0.0000 .076#1 .13CO
.1998 .2959 .4107 .4126
.5464
.7460
.7443
,8765 1.0000
1.0000
d
1oov "/
6
vo
0.87368 . . . . . . . ..... ,86606 +0.00041 +0.035 ,86103 .00033 ,028 .85461 .00021 ,017 .84598 - .00004 - . O M ,83586 - .00029 - ,024 .83572 - .00032 - .027 ,82412 - .00045 - .037 .80741 - ,00043 - ,035 .78654 . . . . . . . .....
0.0000 .1673 .2670 ,3783 ,5060 .6295 ,8313
+ +
+ +
Measurements of the vapor-liquid equilibria of this system were made a t intervals of approximately one-eighth in the mole fraction of the whole system a t 35 and 55' and for approximately equal molal mixtures a t 25 and 45'. The thermodynamic functions were calculated in terms of the excess increase of these functions over that of an ideal solution of the same concentration on mixing a t constant pressure by means of the equations developed in 11. These arc p i '
=
+
R T I n P y ~ / P l ~ t (PI
and
-
VI)(P
-
Pi) (3)
4-(8: - V 2 ) ( P - P2) (1) in which x is the mole fraction in the liquid phase, p*E =
ii 1'111 I'y2/1'2x2
P the vapor pressure of the solution, P1and PZthe vapor pressures of the pure components, VI and V2their molal volumes as pure liquids, and the limit a t zero pressure of the difference between the molal volume of th'e vapor and that of a perfect gas a t the same temperature and pressure. It follows from these equations with the assumption that 17," is linear with the temperature that
y the mole fraction in the vapor phase,
+ (FL"Sj. - FxE,,O)/20
FxE = SxE =
xipiE
.= XT(x1 111
XI
2
~
2
+ TSXE + xz 111 Q) + Fx"
IIh\f = F ,1:
F,"
~
~
(5) (6) (7) (8)
I n these calculations the vapor pressure of the pure components in mm. at 25,35,45 and 55' are: benzene, 95.98, 148.17, 223.40 and 327.10; methanol, 126.91, 209.61, 334.20 and 516.20. The values of p in cc. a t these temperatures are: benzene, -226.5, -196G, -1724 and -1525; methanol, --963, -848, -754 and -675. That of methanol is the value calculated by the theory of corresponding states from the equation (9) E. R. Washburn 2nd A. Lightbody, J. Phys. Chem., 34, 2701 (1930). (10) G. C. Schmidt, 2. phys. Chcm., 121, 221 (1926).
Vol. 68
of Keyes, Smith and Gerry'l for water vapor. It was not multiplied by 0.9 as for the non-polar liquids because of the resemblance of methanol to water. The molal volumes were calculated from the expansion equations given in the "International Critical Tables."
Smoothed Values N o simple equation, whose coefficients were determined by the method of least squares, was found that did not introduce an inflection in the curve of FxM plotted as a function of the mole fraction. Such an inflection would indicate the formation of two liquid phases since a line could be drawn tangent to the curve a t two points.lZ Consequently the coefficients obtained by the method of least squares were adjusted at each temperature in such a manner that the inflection was removed but that minimum changes were made in FxEfor solutions having a mole fraction of methanol of 0.5 or greater and that the values of HxMand SxEremained consistent. Equations in mole fractions were found preferable t o those in volume fractions. Expressed in symmetrical form,13the equation used is =
4A1112x:xZ
+
6-41122x:~i
+
(9)
with the parameters given by the equations A1119
All22 A1222
+ +
= 167.46 0.463T = 106.54 0.716T = 552.54 - 0.498T
(10) (11) (12)
The experimental measurements are given in TABLE Ir VAPORPRESSURE OD BENZENE-METHAXOL MIXTURES DW. x c l l : , ( l l ~Y < ' ? l j O I l
ill
4
Dev. 11
iii 1'
Dev.
rX'xE in
F ~ E
3.5'
0.0242 0.2733 ,3128 ,0254 ,4858 ,1302 ,3107 ,5304 ,5546 ,4988 ,6571 ,5191 ,6305 ,5790 ,6421 ,7965 ,9197 ,7688
+ + + +
0.0304 0.3019 ,0493 .4051 ,4841 .lo31 ,5540 ,3297 ,5845 .4874 ,4984 ,5858 ,6078 ,6076 ,7898 ,6716 .!)OI.S ,7607
+0.006 465.84
-0.003 ,030 ,018
-
+
203.29 211.10 274 2 s
,005 ,008 ,006 ,001
288.47 292.50 292.70 292.49 ,008 283.58 ,006 255.82 550
+ + + -+ +
,029 ,005 ,001 ,006 ,005 ,001 .001
527.12 597.48 664.24 675.62
.nos
675.99 678.44 664.91 622.?!)
+-0.013
182.70
+ +
-k 1 . 9
44.85
4- 7 . 6
37.86 166.92 274.62 300.65 298.96 274.06 190.04 88.38
- 6.4 - 0.4 -/- 0 . 8
4- 0 . 8 4- 1 . 0
4- 2.1
-
1.5
+ 9.0
+7.9 -0.9 4-2.6 -3.5 +0.6 +1.1
+1.9 4-0.2 -0.6
52.38 63.71 0.0 129.96 - .4 284.47 - .1 304.49 f . 2 304.06 . 2 285.52 - . I 198.91 - 3.1 100.21
+7.3 -6.9 -4.6 -0.8
-
-1.8
+16.4
-1.4 -1.2 -0.5 -1.3 -0.6
2.50
.4742
.T,X13
1 . 1 207.48
4.50
.5234
,5752
-+0.005
451.98
4- 1.1 302.24 +2.1
(11) F. G. Keyes, L. B. Smith and H. T. Gerry, Pvoc. A m . Acad. Arts Sci., 70, 319 (1936). (12) J. Willard Gibbs, "Collected Works," Vol. I, Longmans, Green and Co., New York, N. Y., p. 118. (13) M. Benedict, C. A. Johnson, E. Solomon and L. C. Rubin, Trans. A m . Inst. Chem. Enprs., 41, 371 (1945).
Oct.. 1946
BENZENE-METHANOL MIXTUREEQUILIBRIA
0.50 0.75 Xl pressure us. mole fraction for benzenemethanol a t 35". 0.25
Fig. 1.-Vapor
.
1959
The uncertainty in FxE a t 35 and 55' and half tnole fraction is estimated t o be about 1 cal. This would cause uncertainties of about 0.1 cal. and per deg. in SxEand about 31 to 33 cal. in TSXE ITx". The uncertainties in these quantities for qolutioncl richer in methanol are probably somewhat smaller thau thrse estimate..;. I t i q for thi? reason th,it thr, L ,tlues of tlir 1attc.r t w o quantities are given to 1 cal. iii 'Table I11 and tlie deviations in the excess change of the free energy listed in Table I1 are given to 0.1 cal. The increase of the volume on mixing a t constant pressure is quite small, making the difference between the change of the thermodynamic functions on mixing :it constant volume ant1 o i l mixing at coilst ant pressure very small.I4 These differences are smaller than the uncertainties in FxE, Hxhc and SxE,and consequently can be neglected. The differences between the excess change of the entropy on mixing a t constant volume and a t constant pressure and between the energy of mixing a t constant volume and the heat of mixing a t constant 1)ressure are also given in Table 111. 300
Table I1 in the order of the mole fraction of methanol in the liquid phase, the mole fraction of 200 methanol in the vapor phase, the total vapor presd, sure in mm., and the excess free energy change on mixing in cal. per mole. Adjacent to the last 8 100 three columns the deviations of the corresponding 2 quantities, calculated by equations 3,4,9,10,11and 12, from the observed quantities are also listed. 8 The results a t 35' are shown in Fig. 1in which the 0 total pressure and the partial pressures plotted against the rnole fraction of the liquid phase are represented by circles and the total pressures - 100 plotted against the mole fraction of the vapor phase are represented by flagged circles. The 0.25 0.50 0.75 solid lines give the corresponding calculated values and the dotted lines show the behavior of the system according to Raoult's law. Fig. 2.-Various thermodynamic functions for berizeneThe smooth values of the excess change of the methanol a t 35'. free energy on mixing a t 3 5 O , the heat of mixing, and 308.16 times the excess change of the entropy The smooth values of FxEand TS," a t 35' and on mixing, all expressed in cal. per mole, are given Hxhfare shown in Fig. 2 . The excess change of the in Table 111 free energy on mixing is fairly symmetrical and remains positive over the whole range of compoTABLE I11 sition. The heat of mixing is also positive over SMOOTH VALUESOF THE THERMODYNAMIC FUNCTIONS IN the whole range of composition but it is quite unCAL. symmetrical, the maximum occurring a t approxi308.16 308.16 SxE ( S v E - S p E ) Ev'-HpM H," mately 0.3 mole fractioii of ethanol. These values x1 F X d are somewhat higher than thosc reported by 133 167 34 -3 -3 0.1 Schmidt16 but the unsymmetrical nature of the 222 247 25 . -3 -2 .2 curve is the same. The excess change of the en- 6 -2 -2 2i4 268 .3 tropy on mixing is positive up to about 0.3 mole - 45 -1 -1 299 254 .4 fraction in ethanol and then becomes negative for - 80 0 0 220 300 .5 solutions richer in alcohol. Thcsc results are dis180 - 102 +1 +1 282 .6 m cussed in the following paper. 139 - 105 +2 +Z 244 . I
-
+ +
.8 .9
187 108
99 56
-
-
88
52
+2 +I
+2 +l
(14) G . Scatchard, Tvans. Faradar Soc., 53, 160 (1937). (15) G. C. Schmidt, 2. p h y i i k . Chew., 121, 221 (1926).
1960
GEORGESCATCHARD,
SCOTT
E. W O O D
Summary The vapor-liquid equilibrium pressures and compositions of benzene-methanol mixtures have been measured a t approximately each eighth mole fraction a t 35 and 55' and for approximately equimolal mi.xtures a t 25 and 45'. The densities have been determined a t 25'. The excess,change of the free energy, enthalpy, and entropy have been calculated and expressed
[CONTRIBUTION FROM THE
AND JOHN
M. MOCHEL
Vol. 68
analytically. The excess free energy change is positive over the entire range of composition and is quite symmetrical. The change of enthalpy is positive over the entire range of composition but is quite unsymmetrical, the maximum occurring a t a mole fraction of about 0.3 in methanol. The excess entropy change is positive for solutions hnve a mole fraction up to about 0.3 in methanol and then becomes negative. C.4MBRIDGE,
hf ASS.
RECEIVED MAY31, 1946
DEPARTMENT OF CHEMISTRY, MASSACHUSETTS INSTITUTE OF
TECHNOLOGY]
Vapor-Liquid Equilibrium. VII. Carbon Tetrachloride-Methanol Mixtures1 BY GEORGESCATCHARD,
SCOTT
E. I v 0 G D 2 AND
JOHN
&I. & 1 0 C H E L 3
TABLE I The present paper continues the measurement of the vapor-liquid equilibrium of solutions of DENSITIES OF CARBON TETRACHLORIDE-METHANOL MIXmethanol with the non-polar substances previTURESAT25' ously studied. The purification of the carbon Q't fraet Mule fract. CHIOH CHIOH d 6 l0OV "/vO tetrachloride was carried out as described in paper 0.0000 0.0000 1.58452 ....... ..... IV4 of this series. The methanol was of the same ,0296 ,1276 1.53764 $0.00032 +O. 049 material used for the measurements on the ben.0640 ,2472 1.48731 + .00026 + .039 zene-methanol system, VI.s The vapor pressure .1326 .4233 1.39691 - .00015 - .021 of carbon tetrachloride agreed with the values .l736 .5022 1.34781 - .00035 - .047 calculated by the equation given in IV within the .2579 ,6252 1.25715 - .00076 - .095 experimental error. .2597 ,6274 1.25526 - .00071 - .089 Experimental .7523 .3874 1.13928 - .00144 - .164 The apparatus and procedure for the density.3927 .7564 1.13472 - .00138 - ,156 composition measurements were the same as de.5987 0 98697 - ,00126 - .124 .8775 scribed in VI.5 Due to the high molecular weight 1.0000 1.0000 0.78654 ....... ..... of carbon tetrachloride i t was found necessary to correct the observed weight of the component eighth mole fraction of the whole system a t 35 and first weighed in making up the solutions for the 5 5 O and for approximately equi-molal mixtures a t weight of the vapor displaced on addition of the 25 and 45', using the same apparatus and prosecond component. The results are shown in cedure described in the earlier papers.& As in the Table I in which are given the weight fraction and attain a steady state for those mixtures whose mole fraction of methanol, the observed density, mole fraction of methanol in the whole system was 6, defined by equation VI-2,6 and 100 VM/Vo. 0.125 and 0.25 due to the extreme sensitivity of The absolute values of P/ VO for this system are the boiling- point to the composition of the liquid somewhat larger in the methanol poor solutions and the extreme sensitivity of this composition and about five times larger in the methanol rich to the distribution of methanol in the various parts solutions than for the benzene-methanol system. of the equilibrium still. For these mixtures the The compositions of the vapor and liquid sam- pressure was adjusted to maintain the temperature ples in the vapor pressure measurements were de- within 0.05' of the desired temperature for a t least termined from their densities by Eq. VI-2, 6 a n hour. being read off from a large scale graph of the obThe excess change in the free energy on mixing served 6 quantities versus the density. a t constant pressure over that of an ideal soluThe vapor -liquid equilibrium measurements tion of the same concentration was calculated were made a t intervals of approximately one- from the observed vapor pressure and the equilibrium mole fractions of the vapor and liquid phases (1) Presented before the Division of Physical and Inorganic Chemistry of the American Chemical Society at Atlantic City, New Jerusing the equations given in VI. The vapor pressey, April 10, 1946. sures of carbon tetrachloride in mm. used in these (2) Present address: Yale University, New Haven, Connecticut. calculations are 113.88, 174.47, 258.94 and 373.53 (9) Present address: Corninn Glass Works, Corning. New York. .~. -. a t 25, 35, 45 and 55', respectively. The limits a t (4) G. Scatchard, S. E. Wood and J. M. Mochel, THIS JOURNAL, 61, 3206 (1939). zero pressure of the difference between the molal (5) G. Scatchard, S. E. Wood and J. M. Mochel, ibid., 68, 1957 volume of the vapor and that of a n ideal gas a t (1946). the same temperature and pressure are -2290, (6) Equation numbers beginning with VI refer to the preceding - 1992, - 1749 and - 1549 a t 25, 35, 45 and 55O, paper.
