T H E SYSTEM WATER-PHENOL* BY JOHN BRIGHT FERGUSON‘
The system watzr-phenol is frequently cited as an example of a system in which two liquid phases occur. Such systems are of unusual interest since they provide a knotty problem for the proponents of the various theories which deal with “Solubility”. Tha data dealing with such systems are unfortunately not very extensive and we have therefore endeavoured to supplement it by a number of calorimetric and vapour pressure measurements. The results of our experimental work together with a discussion of the available data are given in the following paper.
Part 1. Specific Heats The materials employed were distilled water and redistilled, Kahlbaum’s synthetic C. P. phenol, melting point2 40.8”C. The solutions were made up by weight. The calorimeter proper consisted of a 2 j o cc small-mouth Dewar flask. It contained a glass-enclosed constantan-wire electric heater which had a resistan’ce of 2 6 p j ohms, one end of a glass-enclosed, twelve-junction copper-constantan thermel and a rotary stirrer. The top of the flask was fitted with a rubber stopper which in turn fitted tightly the cases of the heater and thermometer and also the long glass bearing of the stirrer. The bearing wv9s sealed and lubricated by a special very heavy rubber lubricant. The flask stood In a well-controlled air bath which in turn was inside a larger air bath. A similar Dewar flask was immersed in a water thermostat which was placed in the larger air bath. This flask contained the cold junctions of the calorimetric thermometer and one end of a reference thermometer of six junctions the other end of which was placed in an ice bath. This cold junction flask was provided with a rubber stopper having four holes. Two of these nere for the cases of the thermometers and the other two for removable glass plugs. Khen these plugs were removed, a pipette was inserted in one hole and the bath water drawn into the flask. The flask contained no stirring device but all the available space was filled with strips of copper foil which extended the length of the flask. The thermels were calibrated at the following points: (a) B. P. water. (b) Tr.3 SrC126H20,61,341’C. (c) Tr.* Sa2S0410H30,32,384’C. * Contribution from the Department of Chemistry, University of Toronto.
’ For preliminary reports see Ferguson: Trans. Roy. SOC. Can., 13), 17, 160 (1923); Ferguson and Funnell: 18, 122-3;Ferguson and Hope: I Z I (1924). * The melting point was determined with a thermel. Richards and Tngve: J. .4m. Chern. SOC., 40, 89 (1918). Dickenson and Mueller: J. Am. Chern. SOC., 29, 1381 i190j).
T 58
JOHN BRIGHT FERGUSON
The heating current was obtained from a set of storage batteries. A precision potentiometer was used both to measure the temperatures and also to check the heating current by determining the drop of potential across a known resistance. IVi th attention to details, the reference temperature bath was found satisfactory. In one case, two hours after changing the mater, the bath temperature fell 0 , 0 0 3 3 ~ Cin ten minutes. The change of temperature was in all cases regular and could be followed by means of the reference thermometer. The actual temperature of the reference bath was adjusted so that in any experiment it was just below the lowest temperaturz attained by the calorimeter. The cooling or heating corrections were from 3 to 4 percent of the total heat supplied in experiments in which the actual heating took place in about I O minutes. The calorimeter was calibrated using distilled water and in two experiments with cooling corrections, values of 14, 76 and 14.96 cal. mere obtained. An experiment in which a heating correction was necessary gave a value of 1 4 ~ 4 7cal. The value of 1 4 , j 3 cal. was used for the temperature interval 70' to 74'C. The heat usually supplied was 320 cal. The results of these experiments' are given in Table I.
TABLE I The Mean Specific Heats of Kater-Phenol Solutions for the Temperature Interval i o o to 74' Composition K t . percent Phenol
Temperature Interval De g. C .
0,oo
1,001
(assumed value)
20104
1,746 1,741
0,9350 0,9396
401 I 5
1,823 1,660
0,8640 0,8648
60~02
2,018
1,833
0,7597 0,7633
2,IOj
0,6545
2$IIj
0,5486
1,782
0,5487
i9,9 IO0,O
* I calorie
Specific Heat calories'
=
4,182joules.
These results show a slight positive deviation from a straight line relationship and the point of maximum deviation is, approximately, at the consolute composit ion. The specific heat of phenol has been previously determined by Schlamp2 who obtained a value of 0,561 for the temperature 93.9OC. I n the summary of this work previously reported, the composition phenol was erroneously given as 23,04 percent. Schlamp: Ber. oberhess. Ges. f. S a t . u Heilk., 31. IOO (1895).
20,04wt.
percent
T H E SYSTEM WATER-PHEKOL
759
Part 11. The Heats of Mixing The apparatus described in the preceding section was in the main used in the following experiments. A new calorimeter was however necessary. The one liquid was placed in a Dewar flask and the other liquid in a glass container which was immersed in the first liquid when in use. Stirring was obtained by rotating the glass container. The mixing occurred when two glass plugs were removed from openings in this container. One opening .ryas
FIG.I The heat absorbed in the formation of water-phenol solutions at 71" C.
a t the bottom centre and the other a t the top on one side. When the plugs were removed the rotation gave rise to complete mixing due in part to centrifugal action. The heat capacity of this calorimeter was 2 6 , o j + 0 , 4 cal. The cooling or heating corrections were about one percent of the total heat quantity in most of the experiments. The volume capacity of the inner container was 2 j cc so that differential measurements were necessary. Two series of experiments were carried out. In one series, the water was added to the phenol or to a solution and in the other the phenol was added to water or a solution. The results of these experiments, which were carried out a t approximately ~ I ' Cand calculated to this temperature, are given in Table I1 and a graph of them is shown in Fig. I .
760
JOHN BRIGHT FERGUSON
TABLEI1 The Heat absorbed in the Formation of one Gram of Solution from the pure Constituents at 71OC. Experimental Series
Composition Wt. Percent Phenol
A
5,I7
1,012
9!73
1,794 1,867
Heat absorbed I j"calories
I0,23
2,818
I9175 28,06
3,420
35j57
3398
41,86
4,240 43499
47,28
B
4?i06 4!850
4,836 4,643 3,934 2,478
Part 111. Vapour Tension Measurements Schreinemakers' has determined the vapour pressures of solutions of phenol and water. His values are unfortunately not very exact. Our own experiments were designed primarily to check certain of his results. The tensimeter consisted of two glass bulbs connected by a U tube which contained mercury. When in use the bulbs were immersed in a water bath and the emergent, glass tubes were heated electrically t o prevent condensation. The chief difficulty encountered was that of freeing the liquids of air. This was accomplished by heating the liquids above the critical temperature between evacuations and shaking violently. We could not remove the air from the phenol rich layer at room temperature by repeated evacuations and agitation in any reasonable time. After each experiment the solutions were analysed using Lloyd's method.* Water was used as an initial reference material and for it a vapour pressure of 289,1 mm was assumed. Our results are given in Table 111. Especial care was taken with the differential measurements of the 23,33 and 46,81 percent solutions. There can be no question but that the vapour tensions of these solutions do not differ by more than 0,2 mm. Our results confirm the earlier work of Schreinemakers though the actual pressure given by him for these solutions is 294 mm. Schreinemakers: Z. physik. Chem., 35, 464 (1900). Lloyd: J. Am. Chem. SOC.,27, 16 (xgoj); Redman and Rhodes: J. Ind. Eng. Chem., 4, 655 (1912). 2
THE SYSTEM WATER-PHENOL
TABLE I11 The Vapour Tensions of Rater-Phenol Solutions a t js°C. Composition TTt percent Phenol
Tapour Tension Millmeters Mercur?
Wt percent Phenol
Vapour Tension Millimeters Mercury
289,I
46J1
290,I
0
Composition
Part IV. Thermodynamicsi The unaccented symbols refer to a liquid phase at a temperature 8 deg. K. and at atmospheric pressure while the accented symbols refer to the same phase at the same pressure but at a new temperature 8’. The specific heats are assunied to he independent of the temperature. z=x-8n
-z ,
- 8’-z
8
+7 8 - 0 ’ -x - c- 1 l(8 - 8’) - 8’10g,
-
and
I )?;’ ~Dx2
e
- 8’ D*z DX2
- 8’ D2X + 8___ ~
8
Dx2
- 8’ - B’log,
I
!,)% D~
e
,
I1
Equation I defines the zeta for one temperature in terms of the zeta and chi for another temperature, the specific heat and the two temperatures. If we restrict this equation to systems of two components, 2 and y, and differentiate twice in respect to 2 , (x+y = I ) , we obtain equation 11. The first term of this equation is positive for all stable phases, negative for all unstable phases and zero for critical phases. For the system water-phenol we can evaluate the first and second terms of equation I1 from the thermal data alone if we restrict ourselves to the consolute composition. The latter is approximately 34 wt. percent phenol and the consolute temperature* for pure phenol and water is 339 deg. K. For term two (344.1 deg. E(), we obtain 0 . 1 0 2 2 and for term one (338.1 and 333.1 deg. K) respectively -0.o08j and -0.0054. The positive sign for the stable phase is certainly established but the negative signs, while obtained, are somewhat uncertain owing to the size of the quantities involved. However if one could find a similar system in which the last two terms have the same numerical sign. the negative sign for the unstable phase would be established without a numerical calculation. For a discussion of symbols, see W.Lash Miller: Chem. R e v , 1, 294 ( 1 9 2 4 - 5 ) . Hill and Malisoff: J. Am. Chem. SOC.,48, 922 (1926).
762
JOHN BRIGHT FERGUSON
From equation II! the tendency for any solution t o split into two layers on changing the teniperature is a functioh of the zeta and chi of the stable solution and of the specific heats. The function is however such that t,his change is unlikely when the zeta curve for the stable solutions at a temperature which is near their respective dissolution temperatures deviates much from a straight line. For the simple case in which the vapour may be considered as a perfect gasecus solution this statement may be interpreted in terms of the total and partial pressures in the following manner. The zeta for a liquid containing one formula weight of the components is given by the equation:
FIG.2 The partial and total vapour pressures of water-phenol solutions at 7.5' C. .rrl e + nl R e log, -+ PI
e log, 3
111 (nl+n2= 1) Pz in which R is the gas constant, 0 the temperature deg. K., x1 and x z the partial vapour pressures and pl and p2 the vapour pressures of the components in the reference states. I n each case the reference state is taken as the satlrated vapour of the component a t B deg. K. and for these states E and 7 are given zero values. If this equation be differentiated, keeping in mind the approximate equation of Duhem-Margulee, we obtain z =
R
n2 R
IV
THE SYSTEM WATER-PHENOL
7 63
from which it follows that for the zeta curve to approximate a straight line the partial pressure curve for the one component must have a nearly horizontal position parallel to the composition axis.' The total pressure will also have a similar curve. These relations are approximately obeyed by the system under discussion as shown by Fig. 2 . The partial pressures were calculated from the results of Scheinemakers and ourselves on the solutions and the work of Kahlbaum* on pure phenol, upon the assumption that the vapours were perfect gases. The consolute composition is approximately nine mols percent phenol. For the very dilute solutions of phenol the partial pressure for water appears to obey Raoult's Law. The maximum deviation occurs at about 40 mols percent phenol and this is the approximate composition showing the maximum heat of mixing. On plotting the differential heats of mixing obtained by the method of intercepts, the curve for water is convex toward the composition axis as it approaches its origin which is in agreement with the partial pressures of the dilute solutions. I n conclusion, we wish to thank Prof. Lash Miller, Prof. W. S. Funnel1 and Mr. H. B. Hope for their assistance. summary
The specific heats of solutions of phenol and water have been determined at i o - 7 4 deg. C. 2. The heats of formation of these solutions have been measured at approximately 71 deg. C. 3. The vapour tensions of a limited number of solutions were observed a t 7 s deg. C. 4. A thermodynamical study of the available data has been made. I.
Toronto, Canada. January IO, 19?27. 1 Although equation I1 is on a one-gram hasis while equation IV is on a one formula weight basis, this does not materially affect the conclusion. * Kahlbaum: Z. physik. Chem., 26, 604 (1898).