June, 1039
HYDROCHLORIC ACIDIN ETHANOLAND ISOPROPANOL-WATER MIXTURES
these authors do not align well with those determined in this Laboratory17 by conductance and by the use of thermodynamically exact cells. It seems certain that the data obtained with cell (18) will require some revision. Summary 1. Measurements of the cell with transference Ag-AgCl/KCl(m in H20-D20)/KCl(m in HzO)/AgCl-Ag
have been made at o.l
Over a wide range of deu-
68, 1642 (1936): ref. (17) La Mer a n d Chittum, THISJOURXAL, 5a; Yates, unpublished conductance measurements.
[CONTRIBUTION FROM THE
1491
terium content and a t 5, 25 and 45’. The partial molal free energy, entropy and heat content changes have been evaluated for the process KCl(m in H20-D20) = KC1 ( m in HzO)
which corresponds to the process occurring in the cell when the differential transport of the waters is neglected. 2 . The absolute activity coefficients of- potassium chloride in isotopic mixtures have been calculated referred to potassium chloride a t infinite dilution in HzO as unity. XEWYORK,N. Y.
DEPARTMENT OF
RECEIVED NOVEMBER 14, 1938
CHENISTRY O F Y A L E UNIVERSITY]
The Properties of Electrolytes in Mixtures of Water and Organic Solvents. I. Hydrochloric Acid in Ethanol- and Isopropanol-Water Mixtures of High Dielectric Constant BY HERBERT S. HARNED AND CALVINCALMON In order to investigate further the effects of different solvent mixtures on the thermodynamic properties of a strong electrolyte, we have measured the electromotive forces of the cells HZ 1 HCl(m), Solvent(X), HzO(Y) 1 AgCl-Ag
in mixtures containing 10 and 207, ethyl alcohol and lOyoisopropyl alcohol a t 25’. In these media of high dielectric constant, hydrochloric acid is completely ionized so that considerations regarding the results will not be complicated by the cffccts of ionic association to be found in media o f lower dielectric constant ( D < 50). These rec u l t s along with similar data in dioxane , methancil and glycerol-water mixtures will serve as a n introduction to this complicated subject. Electromotive Force and Density Data l h e electromotive forces werc obtained in the usual manner with silver-silver chloride electrodes made by heating silver oxide, and subsequently electrolyzing in hydrochloric acid solutions. The density determinations were made with pycnometers of about 30-cc. capacity. The alcohols were purified by fractional crystallization. Density determinations of the alcohol-water mixtures agreed to within less than 0.005~,with those in the “International Critical Tables.” Table I contains the electromotive force, density data, and the vapor pressures employed for correcting cell
electromotive forces to one atmosphere hydrogen pressure. The densities have been expressed by the equation d = d Q + a’m - b’m2 + elma . . . (1) the constants of which are contained in Table 11. The ratio of concentration in moles per liter solution, c, to molality, m, may be expressed to within =tO.O3c/o by the useful equation c/m = do - A’m
(2)
The conslant A ‘ is also included in the table. Standard Potentials ’l‘he standard potentials of the cells were obtained hy the use nf the function E’, defined by the equation
+ 0 1183 log - 01 1183 +Ad26 0 1188 log (I + U 002 m AIAj) = b:, + j ( n i )
b’ = b
I(
V’L
111
(3)
where E is observed electromotive force, u the Debye and Huckel constant, A the parameter which involves the mean distance of approach of the ion$ a, and MX, the mean molecular weight of solvent. ( A = Ka and M x y = 1OO/(X/Ml Y/Mz) where a is in Angstroms and MI and kfz are the molecular weights of organic solvent and water, respectively.) These quantities are given in Table TI. The value of “u” employed was 4.3 A. as found by Harned and Ehlers‘ for aqueous solu-
+
(1) Harned and Ehlers, THISJOURNAL,66, 2179 (1933)
HERBERT S. HARNED AND CALVIN CALMON
1492
Vol. 61
TABLE I ELECTROMOTIVE FORCES O F THE CELLS AND DENSITIES O F SOLUTIONS AT 25 HB HCl(m), ROH(X), H20(Y) AgC1-Ag X = % by ireight of ROH; cell reproducibility = *0.05 mv. (1) Ethanol-Water (X = 10) (2) Ethanol-Water (X = 20) (3) Isopropanol-Water (X = 10) V. p. = 29.6 mm. V. p. = 35.6 mm. V. p. = 27.0 mm.
1
m
1
d
E26
0.0 ..... ,006309 0.47933 .007579 ,47030 .010875 .45261 ,01987 ,42326 ,04210 ,38711 .05100 ,37794 ,07085 .36222 .08000 ,35640 ,08846 .35158 .lo911 .34152 ,19903 .31254 .2999 .29240 .5050 ,26568 .7014 .24793 9946 ,22776 1.4987 ,20163 1.9938 ,18140 a Extrapolated. I
d
m
E25
d
0.0 ,001862 ,004019 ,006356 .008616 .008921 .02089 ,03556 .04856 .06685 ,07947
.....
0.48729 ,46200 ,44955 ,41715 ,38183 ,37145 ,35623 ,34976 ,35378 33823 ,28698 ,26410 ,24115 ,22168 ,19516 ,17351
0.96640 ,96649 96655 ,96661 .96677 ,96716 .96733 ,96762 ,96778 ,96795 ,96812 ,97133 ,97392 .97776 ,98208 ,98996 ,99784
.....
.....
0.98122 ,98129 ,98130 .98132 ,98137 ,98138 .98158 ,98183 .98208 ,98242 ,98265 ,98320 .98457 .98638 ,98875 ,99313 ,99600 ,99825"
m
0.98038 ,98050 ,98056 ,98064 ,98078 ,98114 ,98133 ,98170 ,98189 .98199 ,98240 ,98393 .98562 ,98908 ,99238 ,99726 1.00501 1.01180
E26
.....
0.0 ,004703 ,007872 ,010154 .0 1983 ,04150 ,05166 ,07123 ,08161 ,09237 ,10421 ,3078 ,4751 ,7309 1.0216 1.5492 2.079 ......
,11188 .19215 ,2990 .4451 ,6993 ,8863 1.0000"
0,53930 ,50096 ,47830 ,46335 ,46164 ,42025 ,39459 ,37971 .36444 ,33618 ,33983 ,31387 ,29213 ,27189 ,24767 ,23417 .22704
TABLE I1 CONSTANTS OF EQUATIONS (1) AD : (2) Parameters of equation (12): a = 4.3 A.; A = Ka do
a'
b'
Ethanol-water (X = 10)
0.98038
0.0178
0.00094
Ethanol-water (X = 20)
,96640
.0170
,00248
Isopropanol-water (X = 10)
,98122
,0173
Ethanol-water (X = 10) Ethanol-water (X = 20) Isopropanol-water (X = 10)
D 72.8 67.0 71.4
0,5675 ,6428 .5842
tions, and Harned and Thomas2 for methanolwater mixtures. The plots of E' versus m are straight from the lowest concentrations measured to 0.1 M acid. From them the standard potentials E: could be evaluated with a certainty of *0.05 mv. The standard potential of the cell in these mixtures, obtained by equation ( 3 ) involving the molality and denoted E:, differs from the standard htentials, E: and E&, derived using moles per liter of solution and mole fractions, respectively. These latter quantities can be obtained from E: by the relations E: = E,!,, - 0.1183 log do E& = E: 0.1183 log lOOO/M,,
+
where do is the solvent density.
(4)
(5)
The reference
(2) Harned and Thoma.;, THIS J O I J I I N A I . , 68, 761 (193;).
-0.00004 .00080(m .00064(m ,001
,0005 U
A'
C'
A&
1.467 1.529 1.482
< 1) > 1) K 0.2413 .2515 .2437
0.0180(m .0185(m .0191(m .0188(m ,0184
< 1) > 1) < 1) > 1) MXY
19.19 20.52 19.48
state for E& is a solution of unit molality with the activity coefficient, y , of unity in a given solvent; that for E$ is a solution of unit mole fraction of electrolyte, N , and unit rational activity coefficient, f,or a / N ; and that for E,"is unit concentration of moles per liter, and a unit activity coefficient, y, or a/c. For purposes of thoroughness in examining the standard potential as a function of the dielectric constant, we have computed E:, E: and E& for the solvent-water mixtures. The results are given in Table 111 along with similar values for cells containing glycerol-water, methanol-water and dioxane-water mixtures previously determined by Lucasse, Harned and ThomasP4and Harned and Morrison,6 respectively. (3) Lucasse, Z. physrk Chem , 1P1, 254 (1926). (4) Harned and Thomas, THISJOURNAL, 61, 1666 (1935) ( 5 ) Harned and Morrison, rbrd , 68, 1908 (1936); Harned. r h r d 60, 330 (1938).
HYDROCHLORIC ACIDIN ETHANOLAND ISOPROPANOL-WATER MIXTURES
June, 1939
TABLE I11
E = EL(,) - 0.05915 log mamcl
STANDARD POTENTIALS OF THE CELLS Hz HCI(m), Solvent(Nz), HzO(NI) AgC1-Ag NI, NZ = mole fractions Na D EON E: E:
I
1
7 8 . 5 4 0.22239 0.22327 0.42876 .41946 74.0 .21535 .21639 69.2 .21070 ,41052 ,20881 ,41761 72.8 ,21442 .21544 ,20736 ,20911 ,40709 67.0 ,4239 ,2196 ,2191 77 ,2082 .2058 ,4057 .06 72 ,21460 ,41631 .21363 ,0323 7 1 . 4 ,20231 .40052 .20303 ,0487 6 0 . 8
0 0.0588 .1233 .0417 .0891 .01
Methanol-water Ethanol-water Glycerol-water Isopropanol-water Dioxane-water
The dielectric constants of the mixtures were taken from the data of Akerlof,6 and Akerlof and Short.' As typical of the characteristics of these standard potentials as function of the reciprocal of the dielectric constant, the values of E: are plotted against 1/D in the lower part of Fig. 1. The origin of the plots a t the left of the figure represents E: for pure water. None of these ploG are straight lines. Further, they exhibit pronounced individual characteristics. The two straight lines (dashed) represent plots of Born's equation = EL(")
-
1.21
x
Dz
102-
(6)
E = EL - 0.05915 log rnHmC1
1493
- 0.05915 log YHYCI
- 0.05915 log Y g Y &
(7) (8)
In these, E & ( w ) is the standard potential in water, is the activity coefficient in any solution relative to unity at infinite dilution in water, E& is the standard potential in any mixture relative to unit activity coefficient, r$r&,a t infinite dilution in that solvent. Combining these equations, we obtain YHYCl
E:(,,.)
- EL
YHYCl
= 0.05915 log **
(9)
YHYCI
=
Further by utilizing the thermodynamic relationships of the reaction: H + C1HzO H30f C1-, which prevails in the solutions of high water content, equations (7) and (8) may be converted to
+
EL(,") - (E:,
+
+
- 0.05915 log U H ~ O )= 0.05915 log YHaOYCl ~
G O Y F I
(10)
- 0.2200
where D1 is the dielectric constant of water, Dz that of the solvent-water mixture, EL,,, 0.2200 the standard potential in pure water, that in the solvent, and is the sum of the reciprocals of the ionic radii in Angstrom . -0.2000 units. In this case, we have used 1.2 (lower Q 0.2100 1 plot) and 0.9 (upper plot) for 1.2 corresponds roughly to the values of ionic 0.2000 I I 1 I radii derived from crystallographic data,8'9 0.012 0.014 0.016 1 1/D. while 0.9 is the value of ; correspondFig. I.-E!, and Eh 0.05915 log NI versus 1/D in media of ing to an average ionic diameter of 4.3 A. high dielectric constant: Curve (1) glycerol-water; Curve (2) which we for the methanol-water; Curve (3) ethanol-water; Curve (4) isoAlthough agreement with the over-simpliiied propanol-water; curve( 5 ) dioxane-water. theory expressed by Born's equation is not good, we note that the magnitude of the theoretical where a H 2 0 is the activity of the water in any mixprediction is sufficient to account for the results ture. The activity of pure water has been taken with the exception of the glycerol-water mixtures. to be unity. This latter equation suggests that Plots of 23: and E&versus 1/D have similar charac- a plot of (E: - 0.05915 log aH20) versus 1/D teristics and require no further discussion. should be investigated. The partial vapor presThe electromotive force of these cells may be sure data indicated that N,, the mole fraction of represented by two equations water, could be substituted for aH20 without great (6) Akerlof, THISJOURNAL., M, 4125 (1932). sacrifice in accuracy. Consequently, the plots (7) Akerlof and Short, {bid., 68, 1241 (1936). of the function (E: - 0.059 log N l ) are given in ( 8 ) Pauling, ibid., 49, 70.5 (1927). the upper part of Fig. 1. With the exception of (9) Brayg, Phil. M a g . , [el 40, 1G'J (l
