THE VAPOR PRESSURE O F BINARY SOLUTIONS O F ISOPROPYL

ALCOHOL AXD BENZENE A T 25C.

ALLEN L. OLSEN
D e p a r t m e n t oj ChemTstry, K a n s a s S t a t e College, M a n h a t t a n , K a n s a s
AND

E. ROGER WASHBURN
D e p a r t m e n t o j C h e m i s t r y , L'niversity o j h'ebraska, L i n c o l n , N e b r a s k a

Received J u l y IS, 1936

-4s a part of a comparative study of binary and ternary systems niade

u p of water, a hydrocarbon, and a lower alcohol, the following vapor pres-
sure measurements have been niade on solutions of isopropyl alcohol in
benzene.
MSTERIALS

The methods of purification and the physical constants of these mate-

rials have been discussed and recorded in a preyious article (1).

EXPERIMENTAL

Total pressure
The difference between the vapor pressure of the isopropyl alcohol and
each liquid mixture mas determined by a differential static method.
The apparatus which n-as used in this measurement n-as a slight niodifica-
tion of that described b y Park.. and Schwenk (3). I t is sketched in figure 2.
Since considerable difficulty was encountered in closing the topi: of the
manometer, mercury-sealed stoppers were employed. S o meaburable
leakage through these stoppers was observed in tn-enty-eight hours when
they were subjected to a difference in pressure of 1 atmosphere.
The manometer was placed within a constant-temperature air bath,
thermostatically controlled to &O.lOC. The readings were observed by
means of a cathetometer reading to 0.1 mm., and it was found that one
could reproduce cathetometer readings to = t O . O 5 mm., which is the limit-
ing accuracy of the inqtrument. Each value which is given in the table
of data representi: the average of many independent determinations, and
the agreement is of the order of kO.05 mni. -After the measurement had
been completed, a portion of the mixture was removed and its composition
was determined by means of a refractiye index measurement. The con-
457
458 A L L E S L. O L S E S AND E . ROGER WASHBURN

version to mole fraction was effected by means of a curve n-hich was plotted
from data in table 1.
Analysis of the vapor phase
The composition of the vapor phase in equilibrium with the solutions
a t 25.00"C. was determined. Thi. was accomplished by passing dry,
carbon dioxide-free air through a pair of bubblers (figure l), each contain-
ing 10 cc. of the mixture under consideration. The bubblers were im-
mersed in a constant-temperature water bath. The air thus saturated
with the vapor of the mixture wa. then passed through a freezing-out tube

.I FIG.I
FIG. 1. Air saturator for vapor analysis
FIG. 2 . Differential static manometer
FIG.2

immersed in liquid air. The alcohol-benzene mixture separated as a solid

on the walls of this tube, and when about 2 cc. of distillate had been col-
lected, it was analyzed by means of the refractometer.
T o determine if there were appreciable concentration changes during
the time required for the vapor to freeze, the following procedure was
carried out. Four bubblers were connected in series and a 10-cc. sample,
chosen successirely from nine different liquid mixtures and varying in com-
position from pure alcohol to pure benzene, was placed in each one. I t was
found that there was no measurable change in composition of the 10-cc.
sample in the first bubbler during the time required for the freezing-out
process. It mas also found that the composition in the first bubbler must
 VAPOR PRESSURE O F BINARY SOLUTIOXS 459

change appreciably before a noticeable difference appeared in the second.

It was further noted that all the liquid must completely disappear in the
first bubbler before a change took place in the third and fourth. The
liquids disappeared in the order of their placement in the series. The air
bubble as it left the first bubbler must have been completely saturated with
the liquid in question during the first part of each determination. -1few
of the values which are recorded in table 2 were obtained by means of a
single unit in which 25 cc. was placed. The value' from this procedure
TABLE 1
Refractive i n d e x of benzene-isopropyl alcohol solutions

MOLE FRACTION O F ALCOHOL

, RErRACTIYE INDEX
I

1 000 I 1.37479
0 823 1.39861
I
0 610 1,42162
0 536 I 1 43444
0 130 1,44725
0 330 I
1,15954
0 222 I 1.47159
0 000 I
1.49800

TABLE 2
C o m p o s i t i o n of vapor of benzene-isopropyl alcohol solutions
MOLE FRACTION O F I MOLE FRACTION or MOLE FRACTION O F MOLE F R A C T I O l OF
ALCOHOL I N L I Q U I D ALCOHOL I N VAPOR ALCOHOL IN LIQEJID ~ ALCOHOL I N VAPOR
-
0.000
0.114
i 0
0
000
187
0
0
633
820
0 334
0 450
0.173 0 223 0 861 0 508
I
0.224 0 231 0 910 I 0 598

1
0.330 0 246 I 0 918 0.739
0.430 0 270 1 000 1 000
0.536 0 290

compared favorably with those obtained from the two-tube unit. A

second bubbler, however, positively insured complete saturation.
The vapor composition data are given in table 2 . The values which are
recorded are representative values of thirty-four determinations. The
average error of a single observation froni the mean of the mole fraction of
alcohol in the vapor phase over liquid niixtures of identical mole fraction
was =t0.005.
Calculation of results
Parks' value of 44.0 mni. was taken as the Tapor pressure of the isc-
propyl alcohol ( 2 ) . The theoretical and the observed results appear in
460 A L L E S L. OLSEN AND E. ROGER WASHBURN

table 3, and the observed results are plotted in figure 3. The values for
the composition of the vapors in equilibrium with these mixtures were not
determined directly, but were obtained from a smooth curve plotted from
values in table 2. In this sense they are really experimental, and when
TABLE 3
P a r t i a l a n d total pressures ut 25.0"C. of benzene-isopropyl alcohol s o l u t i o n s

MOLE FRACTION
OF ALCOHOL
PARTIAL PRESSCRE
OF ALCOHOL
TOTAL PRESSURE
OF MIXTURE
I
1
PARTIAL PRESSURE
OF B E N Z E N E

Liquid Vapor Ibserved Ideal Observed Ideal Observed Ideal

~~ _ _ _ ~
0.000 0.000 0 0 94 4 ~ 94.4
0.059 0.123 129 2 6 104.5 91 4 91.7 88.7 4.969 1,049
0.146 0.205 224 6 4 109.0 87 1 , 86.7 80.6 3.487 1.075
0.362 0.255 27 6 15 9 108.4 76 2 80 8 60.2 1.733 1.342
0.521 0.288 105.8 68 1 75 3 45.2 1.331 1,665
0.700 0.365 99.8 59 1 63 4 28.3 1.182 2.239
0.836 0.470 84.0 52 3 44 5 15.5 1.074 2.874
0.924 0.635 66 4 47 8 24 2 7.2 1.038 3.373
1IO00 1,000 440 I 0.0

FIG.3. Observed total and partial pressures of benzene-isopropyl alcohol system

multiplied by the corresponding total pressures for the various solutions,

give the observed partial pressure of isopropyl alcohol. The observed
partial pressure of benzene was determined in a similar manner. The
ideal pressures in all cases were calculated on the assumption of Raoult's
law
P A = NAP::
 VAPOR PRESSURE O F BINARY SOLUTIOKS 461

where p A and N - 4 are respectively the partial pressure and the mole frac-
tion of component d in a given solution and p i is its vapor pressure in the
pure state.
The activity coefficients for alcohol and benzene are recorded in columns
5 and 6, respectively, in table 3. They were calculated by means of the
equation

wherefis the activity coefficient, p i s the partial vapor pressure of the com-
ponent above the solution, p o is the vapor pressure of the pure component,
and 12is the mole fraction of the component in the solution.
DISCUSSION O F RESULTS

The difference between the vapor pressures of the pure components was
taken as the criterion of accuracy of measurement. The measurement on
this difference, applying many technics, was found to result in an error of
2 to 3 per cent. After having performed the usual routine of technic in
removing the visible air, a tiny bubble of air would always remain above
the liquids \Then the mercury reservoir was raised from its lowered position.
Still further evidence of dissolved air was found in the observation of ab-
normal readings a t the upper end of t h e manometer. The usual procedures
will never remove the dissolved air, for a t the point a t which the stoppers
may be removed there will be placed the existing atmospheric pressure on
the excluded gas, thereby forcing it into the liquid. The effectsof dissolved
air were minimized by making pressure measurements well toward t h e
bottom of the manometer.
I n a study of vapor pressure data and its subsequent calculation to ac-
tivity coefficients, it is observed that the alcohol-benzene system deviates
in a positive manner from Raoults law. The wide deviation from ideal
behavior took place because of the great difference in polarity of the two
components. The polar molecules of alcohol have an abnormally great
attraction for each other, producing greater surface tension, cohesion, etc.,
and tend to squeeze out non-polar molecules from their midst. -4s a
result of the tendency of a liquid of high internal pressure to squeeze out
a liquid of low internal pressure, it is expected t h a t the partial pressure
would deviate in a positive manner from Raoults lam. This concept of
association does not admit any definite polymers like double molecules.
The polar affinities act within the liquid to form groups of molecules which
become impregnable to a non-polar molecule like benzene.
I n the study of the depression of the freezing point the deviations were
ascribed to some combination either of alcohol molecules or alcohol-
benzene molecules. There is a small range of concentration near pure
462 ALLEN L. OLSEN AXD E . ROGER WASHBURN

benzene where the observed and the calculated freezing points do not devi-
ate appreciably. There is a small range of concentration near pure benzene
where the partial pressures of benzene do not deviate far from the theo-
retical values. The alcohol molecules are randomly and sparsely dis-
tributed to such an extent that each is without influence on the other.
h decreased value for the internal pressure of the alcohol in the mixture
would tend to LLsqueeze"out a lesser number of molecules of benzene.
STJMMCIARI

1. The differential static method devised by Parks has been adapted

for the measurement of the total and partial vapor pressures of the com-
ponents of the binary solutions of kopropyl alcohol and benzene a t 25C.
The accuracy of the method has been discussed.
2. These data h a r e been used to calculate the actirity coefficientswhich
indicate deviations from Raoult's law.
3. Deviations from ideal behavior hare been discussedand qualitatively
explained on the basis of polarity.
REFERENCES
(1) OLSEN,A. L., AND WASHBURN, E. R.: J. Am. Chem. SOC.67, 303 (1935).
(2) PARKS,G. S., A N D BARTON,
R.: J. Am. Chem. SOC.60,24 (1928).
(3) PARKS,G. S., AND SCHWESK,J. R.: J. Phys. Chem. 28,720 (1924).