MARIOGOFFREDI AND THEODORE SHEDLOVSKY
2176
Studies of Electrolytic Conductance in Alcohol-Water Mixtures. 111. Sodium Chloride in 1-Propanol-Water Mixtures at 15, 25, and 3501
by Mario Goffredi and Theodore Shedlovsky Rockefeller University, New York, Neu York 10021
(Received January $1, 1967)
Conductance data are presented for dilute solutions of sodium chloride in l-propanolwater mixtures a t 15, 25, and 35". Measurements were made in approximately 10, 20, 40, 60, 80, and 90 wt % 1-propanol as well as in water, but extrapolated values were obtained for pure 1-propanol because of the very low solubility of the salt in this solvent. Measurements of density, viscosity, and dielectric constant are also presented for the various solvent mixtures at the three temperatures. The conductance data were examined in terms of the 1965 theoretical Fuoss-Onsager linearized equation with the aid of a computer program. This equation describes the experimental results with very high accuracy and with reasonable values for the ion-size parameters. Evidence of ion association is found above approximately 60 wt % 1-propanol. Variations in the Walden products with solvent composition and temperature are reported and briefly discussed.
Introduction
Experimental Section Materials. The conductivity water used in our
fluxing the reagent grade over "activated" alumina for many hours under an atmosphere of dry nitrogen and collecting the middle third of a fractional distillation. The specific conductance of this material was less than 1X ohm-' cm-l. At 25" our measurements of the density and dielectric constant were 0.7998 and 20.40, respectively, in good agreement with the corresponding values in the literature for pure l-propanol.2 Reagent grade NaCl was recrystallized three times from conductivity water and dried for several days at 200". Its molecular weight was taken to be 58.443. Viscosity. Viscosity measurements were made with a quartz viscometer of the Ostwald type described by Washburn and Williamsa having negligible drainage and kinetic energy correction. It was used in a range of Reynolds number less than 10 and calibrated with water at 15 A 0.01", density 0.99913 g ml-l, and viscosity q = 0.011381 poise based on the new values of the National Bureau of standard^.^
studies was prepared by distillation in a "High Purity" Barnstead still after passage through mixed-bed ionexchange resins and also carbon cartridges. It was collected and stored in an all Pyrex glass closed system under purified argon. The specific conductance of this water was always under 2 X lo-' ohm-' cm-'. Purification of 1-propanol was carried out by re-
(1) This research was supported by the National Science Foundation through Grant No. GB-3062. (2) K-Y. Chu and A. R. Thompson, J. Chem. Eng. Data, 7 , 358 (1962); T. A. Grover and P. A. Sears, J . Phys. Chem., 60, 330 (1956). (3) E. W. Washburn and G. Y. Williams, J . Am. Chem. SOC.,35, 737 (1913).
Studies of electrolytic conductance in alcohol-water mixtures over the entire composition range of the binary solvent systems have been in progress in this laboratory. The experimental data for the behavior of strong and weak electrolytes a t various temperatures are of interest not only because of the relevance to electrolyte theory but also because of the light they can throw on the structural interactions between water and alcohol, substances which have similar amphiprotic properties. I n the present paper we present data a t 15, 25, and 35" on the densities, viscosities, and dielectric constants of 1-propanoi-water mixtures over the entire composition range and on the electrolytic conductance of dilute sodium chloride solutions in these systems.
The Journal of Physical Chemistry
STUDIES OF ELECTROLYTIC CONDUCTANCE IN ALCOHOL-WATER MIXTURES
Dielectric Conefant. All our measurements were
2177
~
~~~
Table I : Properties of Water-1-Propanol Mixtures
carried out a t 1MHz using the General Radio Type 1610-AM capacitance-measuring assembly by the subwt % KD x n-CaHTOH do D 107 stitution method. The cell used for the measurements 109 and its temperature control to *0.01" were the same 15" The cell was calias described by Lind and FUOSS.~* 0 0.99913 11.381 81.95 1.o brated with acetone ( 0 2 5 0 20.70), with methanoPb 20 0.97201 26.90 68.39 0.8 40 0.93290 52.45 0.7 36.51 (Dl50 34.68; 0 3 5 0 30.72), and with conductivity water' 60 0.89184 38.05 38.60 0.3 ( 0 1 5 0 81.95; 0 2 5 0 78.30; DW 74.82). The true cell 80 0.85126 28.19 0.2 33.10 constant a t 25" was found to be 7.98 and the lead capaci90 0 83051 28.81 24.50 0.1 tance correction 3.42 pf. 100 0.80733 24.95 21.92 0.03 Conductivity. A Pyrex and a quartz cell of the Shed25 " lovsky type' with unplatinized electrodes were used. 0 0.99707 8.903 78.30 1.6 Cell constants were determined a t 25" by measuring 20 0.9672 18.36 64.89 1.3 the conductance of dilute solutions of aqueous KC1 24.97 40 0.9263 49.70 0.9 and using the equation of Lind, Zwolenik, and FUOSS.~ 60 0.8845 26.61 36.61 0.4 0.8434 23.94 80 26.41 0.2 All measurements were made with the cell in a constant21.62 90 0.82% 22.86 0.1 temperature oil bath maintained within 0.005" with the 19.38 100 0.7998 20.40 0.05 aid of a calibrated platinum resistance thermometer and 35" a hlueller bridge assembly. The electrical resistance 0 0.99406 7.194 74.82 3.2 measurements were made with a carefully calibrated 20 13.30 0.9613 62.14 2.8 ac bridgeg provided with a frequency range of 1,5, and 40 17.93 0.9192 47.02 1.4 10 kc which enabled polarization corrections to be made 60 19.40 0.8766 34.49 0.7 in the usual manner. These were never large but neces80 18.02 0.8350 24.70 0.3 sary to consider in work of high precision with the un90 21.34 0.2 16.62 0.8140 100 15.34 0.7914 19.22 0.07 platinized electrodes we used to avoid adsorption and other possible sources of trouble with platinum black in our solvent mixtures. and viscosity values a t 25 and 35" of Mikhail and The general procedure used in the present work is KimelI2 and the density values a t 15" of the "Interessentially the same as in former studies in this laborat ~ r y . ' , ~ All ~ J ~solvent mixtures and solutions were national Critical Tables," but all the other listings in Table I are from our own measurements. Not listed prepared by weight with appropriate vacuum correcin the table are our viscosity measurements at 0" since tions, of course. Stock solutions of electrolyte (about they were not essential in this paper. These were 0.1 M ) were usually freshly prepared. Weight17.865, 57.55, 75.80, 75.91, 72.08, 56.75, 47.81, and buret techniquo was employed without exposure to 37.05 mpoise in 0, 20, 40, 50, 60, 80, 90, and 100 wt the laboratory atmosphere by using purified argon gas % 1-propanol, respectively. I n Figure 1 the variation in the preparation of the stock solutions and in their in viscosity of the solvent mixtures is shown graphically incremental additions to the conductance cell. Five or as a function of mole per cent 1-propanol a t the various six such additions were made in each experiment, covering a concentration range from a fraction to several millimoles per liter in most cases. (4) J. F. Swindells, J. R. Coe, Jr., and T. B. Godfrey, J. Res. Natl. I
Resultu and Discussion The properties of the water-1-propanol solvent mixtures over their entire range of composition and a t the three temperatures used in our studies (15,25, and 35") are listed in Table I. I n this table, as elsewhere in this paper, weight per cent is the weight per cent 1propanol in the mixture of density do, viscosity 10aq (in millipoise), and dielectric constant D. The specific conductances KO are the "solvent correction" values that were subtracted from the measured specific conductances of the solutions. We made use of density
Bur. Std., 48, RP2279 (1952). (5) (a) J. E. Lind and R. M. Fuoss, J. Phys. Chem., 65, 999 (1961); (b) R. G. Bates and R. A. Robinson, Institute of the Symposium of the Electrochemical Society, Toronto, May 1964. (6) C. G. Malberg and A. A. Maryott, J . Res. Natl. BUT.Std., 56, 1 (1956). (7) T. Shedlovsky, J. Am. Chem. SOC., 54, 1411 (1932). (8) J. E. Lind, Jr., J. J. Zwolenik, and R. M. Fuoss, ibid., 81, 1557 (1959). (9) J. G. Janz and G. D. E. McIntyre, J. EEectrochem. SOC., 108, 272 (1961). (10) T. Shedlovsky and R. L. Kay, J. Phys. Chem., 60, 151 (1956). (11) H. 0. Spivey and T. Shedlovsky, ibid., 71, 2165 (1967). (12) S. Z. Mikhail and W. R. Kimel, J. Chem. Eng. Data, 8, 323 (1963).
Volume 71,Number 7 June 1967
MARIOGOFFREDI AND THEODORE SHEDLOVSKY
2178
ance AO and also K A , whereas E depends on A. but is not otherwise “adjustable.” The dimensionless variable r is the ratio of the “Bjerrum” distance to the DebyeHuckel ion atmosphere thickness. The results of the computer analysis are summarized in the form of conductance parameters in Table I11 in which the symbols have already previously been defined13 except for the last column, uA, which lists the standard deviations for the various experiments. It will be noted that in no case does the standard deviation correspond to more than a few hundredths of a per cent difference between the Fuoss-Onsager equation and the actual data, as is shown by AA in Table 11. The analysis also shows that ion pairing appears to be insignificant below 60% 1-propan01.l~ In those cases where the K A , ion-pair association term, in the three-parameter equation was less than about 6, indicating relatively negligible ion pairing, the data were reprocessed without this term and with
I
IO
1
I
20
30
I
I
40
50
I
I
I
80
90
1
70 Mole per cent n-propanol 60
I 100
Figure 1. Viscosity of water-1-propanol mixtures.
temperatures indicated. The maxima in the curves grow flatter with increasing temperature and seem to shift gradually in position from about 20 mole % at 0” to about 30 mole % at 35”. Measured equivalent conductances A at the concentrations C in equivalents per liter for dilute sodium chloride solutions are listed in Table 11. The values in water at 25” are omitted because they had been previously published.’ Those for solvent above 90% 1-propanol were not obtained because the solubility of the salt became altogether too low for good work. The experimental values listed in the table are representative of at least two and often three experimental series that gave essentially the same results (within no more than about 0.02 or 0.03% in most cases). These are not reported for the sake of reasonable brevity. The conductance data were analyzed with the FuossOnsageP 1965 three-parameter equation in the form A = A,
-K
60% 80% 90% 00%
15‘
I
I
2 5’
3 50
Temp.
A C ~ A ~
in which y is the degree of dissociation, K A the association constant, f the activity coefficient, and S is the limiting Onsager slope. The coefficient L is a function of the “ion size” d and is an adjustable parameter from the data as are the limiting equivalent conductThe JOUTWZ of Phyeical Chemistry
40%
Figure 2. NaCl in water-1-propanol mixtures.
- ~!3C”~y”’+ 2ECy In ryl/a+ LCy
20%
(13) R. M. Fuoss, L. Onsager, and J. T. Skinner, J . Phye. Chem., 69, 2581 (1965). We are indebted to Drs. Fuoss, Skinner, and Lind for making their Fortran computer program available to us. (14) The values in Table I11 for AO, L,and d in water (0% 1-prop anol) at 25 and 3 5 O differ a little from those previously reported from thia laboratory.11 It is due to the fact that different values of vie-
cosity and dielectric constant were used in the two papers. We prefer the present values which are based on newer data.
STUDIESOF ELECTROLYTIC CONDUCTANCE IN ALCOHOL-WATER MIXTURES
2179
Table I1 : Conductances of NaCl in Water-1-Propanol Mixtures 104c
AA
A
104c
A
AA
147.755 146.075 144.939 144.259 143.638
-0.003 -0.005 0.014 0.005 -0.011
5.2965 8.747 12.093 17.909 22.249
90.633 90.159 89,801 89.263 88.926
-0.003 -0.003 0.010 0.000 -0.003
40.01% n-CsHtOH 47.903 0.011 47.440 -0.008 47.108 -0.002 -0.006 46.746 46.458 -0.001 46.205 0.007
3.1132 6.7991 13.021 17.674 26.337
64.053 63.416 62.669 62.229 61.545
0.004 -0.004 -0.002 0.001 0.001
2.7841 7.2432 13.749 20.517 27.103 32.712
60.01% wCSH~OH 34.732 -0.003 34.035 0.003 33.329 0.006 32.749 -0.008 32.298 -0.002 31.964 0.003
5.0869 11 130 18.452 24.239 33.823
45.337 44.269 43.352 42.759 41.959
0.002 -0.006 0.005 0.000 0.000
6.1857 20.918 27.546 38.771 49.356
80.01% w C ~ H ~ O H 0.000 26.257 23.997 -0.003 23.328 0.002 22.423 0.001 21.751 -0.001
4.4477 10.986 17.716 24.607 34.212 42.774
35.081 33.205 31.865 30.812 29.652 28.809
0.004 -0.005
1,4379 4.0318 6.6842 8.6972 11.524 14.322
33.229 31.433 30.150 29.379 28.465 27.702
0.002 -0.002 -0.005 0.005 0.005 -0.003
10'C
A
AA
10.359 15.176 19.680 24.285 35.701 47.425
98.843 98.409 98.066 97.739 97.111 96.554
-0.006 0.003 0,008 -0.008 0.009 -0.006
7.688 15.813 24.233 30.481 38.374 48.374
49.447 48.999 48.667 48.461 48.216 47.949
-0.002 -0.006 0.004 0.010 0.000 -0.006
4.8256 7.0817 9.6815 15.963 19.607 23,923
20.00% n-CsH7OH 68.934 -0.002 68.691 0.003 68.450 -0.001 67.993 0.001 67.770 -0.002 67.541 0.001
5.1663 12.120 23.043 31.119 37.665 53.837
34.615 34.179 33.677 33.398 33.213 32.814
0.005 0.011 -0.012 -0.014 -0.002 0.012
8.4391 13.903 18.988 25.290 31.168 36,972
6.9827 15.131 21.872 29.964 40.449 48.830
24.711 24.115 23.741 23.374 22.970 22.700
0.000 0.000 -0.001 0.003 -0.002 0.001
7.5318 15.123 24,757 32.409 43.719
19.245 18.375 17.592 17.104 16.518
0,001 -0.003 0.001 0.001 -0.001
0.00% n-CaH70H
2.3101 4.3826 6.2153 8,0485 9.6691 11.752
18.876 18.286 17.867 17.501 17.232 16.900
0.000 0.002 0.001 -0.007 0.007 -0.001
33.496 56.705 77.398 91.505 105.664
1.762 3.5088 5.9662 7.9047 8.7871
90.00% n-CsH7OH 25,426 0.001 24.587 -0.002 23.703 0.001 23.136 0.000 22.905 -0.001
y = 1, yielding a corresponding two-parameter equation. This had negligible effect on the values of A. but did influence the values of the mean ionic diameters d which became lowered by several tenths of an angstrom a t about 40% 1-propanol. Otherwise, d remains relatively constant at the three temperatures with the reasonable average value of about 3.3 A. Referring to Table I11 again we note that from 60 to 90 wt % the K A values increase progressively at each of the
I
-0,006 0.004 0.009
-0.007
three temperatures. Also, they become a bit higher as the temperature increases. I n Figure 2 the variations in ho for sodium chloride in the various solvent systems are shown to be reasonably linear with temperature. However, we report the temperature coefficients of conductance (lOO/Ao). (AAo/AT) with greater accuracy in Table I V in which the second column refers to the interval 15-25' and the last column to the interval 25-35'. Volume 71. Number 7 June 1987
MARIOGOFFREDI AND THEODORE SHEDLOVSKY
2180
Table III : Conductance Parameters and Constants for NaCl in Water-l-Propanol Mixtures Wt % nCiHiOH
d
Aa
KA
S
E
L
=A
70.04 70.02 36.82 34.02 38.67 51.35 62.01
15.77 15.75 15.26 24.99 40.16 96.22 139.7
138 148 75 66 100 114 110
0.02 0.009 0.007 0.01 0.002 0.003 0.006
89.75 53.68 49.62 55.42 72.57 85.67
20.33 22.18 36.79 67.29 144 208
191 132 86 170 184 303
0.04 0.002 0.008 0.007 0.002 0.002
111.6 111.6 73.41 68.96 76.72 99.4 115.6
25.58 25.62 29.81 51.68 96.34 212 305
237 219 180 123 332 247 534
0.04 0.01 0.007 0.004 0.006 0,009 0.006
15' 00.00 00.00 20.00 40.01 60.01 80.01 90.00 100.00
101.15" f 0.01 101.057 f 0.008 50.487 f 0.006 35.43 f 0 . 0 1 25.894 f 0.005 21.144 f 0.008 20.19 f 0 . 0 2 19.0 extrapolated)
3.15 f 0.05 3.38 f 0.06 2.75 f 0.06 2.50 f 0.07 3.1 f O . 1 3.19 f 0.06 3.2 f 0 . 6
00.00 20.00 40.01 60.01 80.01 90.00 100.00
126.52 1 0 . 0 2 70.124 f 0.002 49.428 f 0.008 35.79 fO.01 28.734 f 0.006 27.093 f 0.005 22.8 (extrapolated)
3.51 f 0.06 3.36 f 0.03 2.43 f 0.05 3.5 f 0 . 4 3.50 & 0.03 3.52 f 0.02
00 * 00
153.84' i 0.03 153.94 fO.02 92.335 f 0.007 65.329f0.003 47.38 f O . O 1 37.96 f 0 . 0 2 35.34 f O . 0 1 27.7 (extrapolated)
3.61 f 0.07 3.34 f 0.03 3.50 f 0.08 2.56 f 0.02 3.2 f 0 . 2 3.66 f 0.08 3.68 f 0.02
0 f l 31 1 70 f 10
*
2 6 O
8*2 39 1 117 3
* *
35O
00.00 20.00 40.01 60.01 80.01 90.00 100.00
13f 1 45 f 1 158 f 3
" G. C. Benson and A. Gordon, J. C h m . Phys., 13,473 (1945)' recalculated data.
Table IV : Temperature Coefficients of Limiting Equivalent Conductance for Sodium Chloride in Water-l-Propanol Mixtures Wt % n-CaHiOH
0.00 20.00 40.01 60.01 80.01 90.00 100.00
2.52 3.89 3.95 3.82 3.59 3.40 2.00
2.17 3.17 3.22 3.24 3.21 3.06 2.15
A plot of & against the solvent composition is shown in Figure 3 for the three temperatures we had used. The extrapolations that appear as the broken lines between 90 and 100% propanol and the corresponding extrapolated values in Table I11 are based on similar The Journal of Physical C h k t r y
extrapolations of conductance data of sodium acetate solutions which we shall report in a subsequent paper dealing with the ionization constant of acetic acid. Sodium acetate has a more respectable solubility in pure l-propanol than does sodium chloride. We had prepared plots of AO differences between these two salts and these appeared to vary quite smoothly with alcohol enrichment, making the extrapolation of b(NaC1) to 100% l-propanol acceptable in our view. The dips in the curves between 90 and 100% parallel our experimental findings for sodium acetate in this region of solvent composition. I n Figure 4, the Walden products b q o are shown for sodium chloride as a function of the dielectric constant D over the total range of solvent composition for the three curves corresponding to 15, 25, and 35". These curves are not linear and cross flatly in the region between D = 42 and 43 with their order reversed. The crossing region corresponds to about 47% at 35",
STUDIESOF ELECTROLYTIC CONDUCTANCE IN ALCOHOL-WATER MIXTURES
2181
A0
I Soy
A I
IO
IO0
0.5
90
I
80
I
I
70
I
I
I
60
D
I
50
I
I
I
40
30
20
Figure 4. NaCl in water-l-propanol mixtures.
50 40
,3n Ot
Wt. per cent n-propanol
Figure 3. NaCl in water-l-propanol mixtures.
51q;b at 2 5 O , and 56y0 at 15" solvent composition, whereas the viscosity maximum lies between 50 and 60% for all three temperatures. I n the inset of Figure 4 we note that the Walden products Aoqo are linear with temperature for a given solvent composition, with a decreasing negative slope from 20 to 90% at which solvent composition the slope becomes zero, but it d e creases again at 1 0 0 ~ l-propanol. o These observations, interesting though they may
seem, do not obscure the fact that no simple generalization based on the bulk viscosity of the solvent mixtures can satisfactorily account €or the variations in A0 or in the corresponding &,TO products found from our measurements. Although we cannot blandly accept various popular contemporary ideas about water and alcohol structures, since no really good theory for these liquids is yet at hand, it does appear that they have at least some qualitative merit.I1 We prefer, however, to reserve discussion of this matter for a subsequent communication that will make it possible to compare the experimental results for several electrolytes in different alcohol-water systems.
Acknowledgment. We are indebted to Mrs. Catherine Wolowodink for her valuable technical assistance throughout the course of the experimental studies and to Mrs. Marion Angel1 for her most skillful and gracious help in the analysis of our data with the 7090 IBM computer.
Volums 71, Number 7 June 1067