2742
JOHN
H. SLYLOR, VICTOR J. RAXTAND PAULhl. GROSS
[CONTRIBUTION FROM THE DEPARTMENT O F CHEMISTRY,
1701. 64
DUKEUNIVERSITY]
Solubility Studies. VII. The Solubilities of Some Isomeric Ketones in Water BY JOHX H. SAYLOR, VICTORJ. B A X TAND ~ PAULM. GROSS
The possible existence of a linear relationship between the heats and entropies of solution of organic compounds has been suggested by several investigator^^,^ and is of considerable interest from the standpoint of the energy relationships involved in the solution process. In a previous communication from this Laboratory' it was shown that a linear relationship between the heats and entropies of solution existed for three of the ketones studied. These ketones, methyl propyl, methyl isopropyl and diethyl, have approximately the same molar volumes a t the temperatures of the measurements. I t was suggested that equality of solute volumes as well as chemical similarity may well be a prerequisite for the existence of such a linear relationship between the heats and entropies of solution. The. present investigation of the solubilities of the three isomeric ketones, dipropyl, diisopropyl and methyl n-amyl, was undertaken with this suggestion in mind. Experimental The saturated solutions were prepared and analyzed by means of a Zeiss combination liquid and gas interferometer as previously described. 4,5 The ketones were obtained from the Carbide and Carbon Chemical Company. They were purified by repeated shdking with ammoniacal silver nitrate until no discoloration appeared on allowing the ketone to stand for thirty minutes in contact with fresh reagent. Then they were TABLE
A F = AH
THE COMPOUNDS 19. 1,. at Previously 760 m m . , 3C. observed b. p., 'C. Kef.
Dipropyl ketone 144.OO-144.10 114.1 Diisopropyl ketone 121.06 * i)0,5 124.0 1 2 3 . 7 Methyl It-amyl ketone 1 31 I H - 13!). :(2 1 .XI . 2 ~
Results 'The observed solubilities are listed in Table 11. J'apor s o l ~ b i l i t i e s , ~C,, . ~ ~were calculated from these and are listed in Table I1 along with the vapor pressures, PL,of the solute. The vapor solubility is defined as that concentration C, in equilibrium with the vapor of the substance under a standard pressure of 100 mm. a t the particular temperature in question. It is calculated by means of Henry's law from the observed solubilities and the vapor pressure of the liquid solute a t the same temperature. Table I1 also lists the mole fraction, N , corresponding to Cs as well as the densities and molar volumes, P ,, a t each temperature. The free energies of hydration as given by the relation AF = RT In (P/,Y) have been calculated from the mole fractions, N , given in Table 11. The corresponding entropy of hydration A S was evaluated graphically from a plot of AF vs. T . The heats of hydration AH were calculated from the values of A F and A S by means of the relation
1
BOILINGPOIXTSOF Substance
washed, dried over calcium sulfate and fractionated in a 50-cm. Widmer still using calibrated thermometers. The boiling points of the fractions used are given in Table I. I t was necessary to determine the densities of diisopropyl and methyl n-amyl ketone in order to calculate the corresponding molar volumes. This was done a t 10, 30, 50 and 75' by means of appropriate specific gravity bottles.
6 7,8 0
.
(1) This paper was taken in part from the thesis submitted by Victor J . Bart to the Graduate School of Duke University in partial fulfillment of the requirements for t h e degree ot Xlaster of A r b , June,
1940. (2) Evans and Polanyi. i ' ~ a s s F . a l e o n y S u r . . 32, 1:133 (1936). ( 3 ) Butler, i b i d . , 33, 229 (1937). (4) Gross, Rintelen and Saylor, J , P h y ~ C'hem., . 43, 197 (1939). ( 5 ) Gross and Saylor, THISJ O U R N A 63, L , 1747 (1931). ((j) Timmermans, Birll. SOC. chim. b e l g . , SO, 62 (1931) ( 7 ) Mailhe, B d 1 . sot. c h i m , , I-l] 5, 620 11900) (R! "International Critical T;ihl i!ii Park ;mi rIoffInut~, Ind kuy.
-
TAS
These quantities are listed in Table 111. Figure 1 is a plot of A S versus A H for dipropyl, diisopropyl and methyl n-amyl ketones as well as for methyl propyl, methyl isopropyl and diethyl ketones taken from the previous investigation. The linear relationship is not as good in the present investigation. Nevertheless, within the limits of experimental error and the assumptions involved the data point to its validity. It is likely that the relation may have limited validity for the longer carbon chains with their many possible configurations. Butler3 has shown that such a linearity exists in a series of alcohols. I t (10) 5a41or
i t i d e y and Groi.
1 H I \ J O L R k 4 L 60, 373 (1968)
THESOLUBILITIES OF SOME ISOMERIC KETONESIN WATER
Dec., 1942
"713
TABLE I1 SOLUBILITIES AND RELATED QUANTITIES t.,
Moles per 1000 g. HxO
C? molality
N x loa
4.02 1.55 0.348 .113 .336
69.1 27.2 6.23 2.03 0.604
0.834011 .8248
75
0.0643 ,0466 .0331 .0288 .0254
Diisopropyl
10 30 40 50 55 65 75
.0587 ,0404 .0392 .0350 .0339 .033 1 .0376
.858 .201 .118 .0668 .0562 .0335 .0255
Methyl n-amyl
10 30 50 60 65 75
.0472 .0355 .0319 ,0308 .0315 .0339
3.35 .668 .216 ,110 .0890
Ketone
O C .
Dipropyl'
0 10 30
50
a
,0610
PL,b m . )
vm,( m l )
.7913 .7702
1 . 612 3.0 9.5 25.5 75.6
136.9 138.4 141.3 144.3 148.2
15.2 3.60 2.12 1.20 1.01 0.603 .459
.8139 .7968 .7869 .7792 .7734 .7645 .7560
6 . 813 20.1 33.2 52.4 65.5 98.8 147.3
140.3 143.3 145.1 146.5 147.6 149.3 151.0
56.8 11.9 3.86 1.97 1.60 1.10
.8245 .8072 .7898 .7809 .7761 .7676
1.411' 5.32 14.8 28.0 35.4 55.6
138.5 141.4 144.6 146.2 147.1 148.7
dt4
.g081
Data a t 0 and 10 O from the data of Gross, Rintelen and Saylor, ref. 4.
TABLE I11 VALUESOF AF, AH
AND
AS
oc.
AF, kcal.
-AS, cat./degree
o 10 30 50 75
3.95 4.62 5.84 6.93 8.31
65. 8 64.5 58.1 54.1 53.0
14.01 13.63 11.76 10.54 10.13
Diisopropyl
10 30 40 50 55 65 75
4.94 6.16 6.70 7.27 7.50 8.07 8.50
62.1 58.1 58.1 55.1 52.5 46.6 41.0
12.63 11.44 11.49 10.47 9.68 7.68 5.81
Methyl n-amyl
10 30 50 60 65
4.20 5.44 6.52 7.17 7.42 i.90
60.2 59.7 56.7 53.3 52.7 50.2
12.84 12.65 11.79 10.58 10.39 9.57
t.,
Ketone
Dipropyl
-(D
appears, however, that the alcohols with the longest carbon chains do not follow the relationship as closely. Figure 2 is a plot of vapor solubility against molar volume. There is a still greater temperature dependency in the solubility of these compounds than that found previously for ketones of lower molecular weight. A t 10' methyl %-amyl
ketone is 3.9 times more soluble than diisopropyl ketone but this ratio diminishes with increase in temperature until a t 750 i t is 2.4. This greater temperature dependency would be predicted from the explanation previously offered.
11
(11) F. K. Beilstein, "Hendbuch der organischen Chemie," 4th ed., Julius Springer, Berlin, Vol. 1, p. 700, appendix 1, Vol. I , p. 359. (12) Rintelen, Saylor and Gross, THIS JOURNAL, 69, 1129 (1937) (13) Aston and Mayberry, ibid., 66, 2682 (1934). (14) Stuckey and Saylor, i b i d . , 62, 2922 (1940). I
Fig. 1.-Relation
12
13 14 15 - A H kcal. between heats and entropies of solution.
According to this, compounds such as methyl namyl ketone with longer unbranched side-chains
GRINNELL JONES
2744 -ro i r )
AND
WENDELL A. RAY
, 500, 30", 10"
Vol. 64 ited number of configurations and, hence, their solubilities are less temperature dependent.
Summary
----.+
The solubilities in water of dipropyl, diisopropyl and , I methyl n-amyl ketones have J 15 26 35 45 55 been determined a t various Vapor solubility, mole fraction X 101 temperatures from 10 t o 7.5' Meet of temperature on solubilities in relation to volume V , methyl amyl and vapor solubilities have
--_
- _
-- -
--...--
t
I
Fig 2
ketone, 0, diisopropyl ketone,
, dipropvl
may be stretched out a t low temperatures so that the CH2 groups are surrounded by the maximum possible number of water molecules, the attractive forces between these groups and water thus adding to the main interaction between the carbonyl dipole and water. When the temperature increases, the molecule may assume other more compact configurations which will be such as to decrease the number of water molecules adjacent to the CH2 groups. On the other hand, molecules such as that of diisopropyl ketone having compact branched structures can only assume a lim-
ketone
been computed. Free energies, heats and entropies of solution have been calculated. Additional evidence was obtained which tends to confirm the previous suggestion that equality of solute volume as well as chemical similarity is necessary for a linear relation to exist between entropies and heats of solution. Large solubility differences and a large temperature dependence of these differences were noted in agreement with those previously found for other isomeric ketones. DURHAM, NORTH CAROLINA RECEIVED SEPTEMBER 2,1942
[CONTRIBUTIONFROM
THE >1ALLINCm~ODT
1
-_
CHEMICAL LABORATORY O F HARVARD UNIVERSITY]
The Surface Tension of Solutions of Electrolytes as a Function of the Concentration. IV. Magnesium Sulfate BY GNNNELLJONES AND WENDELL A. RAY This paper is a continuation of work previously reported' on the surface tensions of aqueous solution of electrolytes relative to that of pure water. These earlier papers should be consulted for a description of the experimental technique and a discussion of the historical and theoretical background of the problem. The data given are "apparent relative surface tensions" in the sense defined in the second paper of this series. Magnesium sulfate was chosen for this work because none of the salts already studied belonged to this valence type. The measurements have been carried out a t 25.00° over a concentration range of 0.0001 up to 2 molar. Analytical reagent grade magnesium sulfate was purified by two recrystallizations from water. The crystals were washed by centrifugation and (1) Grinnell Jones and Wendell A K a y , THISJOURNAL, 69, 187 (1937). 63, 288, 3262 (1941). Grinnell Jones and I, D Frizzell J rhein Phys , 8, 986 11940)
slowly dried in an electric oven. As the salt was hygroscopic each sample before final weighing was kept in an electric furnace at 475' for several hours or until constant weight was assured. At this temperature the anhydrous salt was formed. The samples were weighed in platinum boats in stoppered weighing bottles. A saturated solution showed no red coloration with phenolphthalein. Two 50-ml. Ostwald type pycnometers were used for the density measurements and gave duplicate results differing by not more than a few parts per million. The surface tension measurements were made by the differential capillary rise method as described in the first and second papers of this series. The figures given in Table I are the mean of a t least two independent determinations for each concentration. The results are shown graphically in Figs I and2.
