718
P. RUETSCHI AND R. F. AMLIE
Solubility of Hydrogen in Potassium Hydroxide and Sulfuric Acid. Salting-out and Hydration
by P. Ruetschi’ and R. F. Amlie The Electric Storage Battery Company, The Carl F. Norberg Research Center, Yardley, Pennsylvania (Received August 17,1966)
While precise data on the solubility of hydrogen in water are the published work on the solubility of hydrogen in HzS04 and KOH solutions is scant and shows considerable discrepancies. Besides the old data of Geffcken6 and Christoff6 there are known only two recent studies. Vertes and Nagy’ have reported hydrogen solubilities at 20, 30, 40, and 50” in 0.1 and 1.0 N HzS04 solutions, using Winkler’s method.* Knaster and A p e l b a ~ m whose ,~ paper was published during the present study, give values for the solubility of hydrogen in KOH up to 10 N . The pertinent results of these investigations are included for comparison with the values obtained in the present investigation.
Apparatus The solubility apparatus (Figure 1) is similar to that of Ben-Naim and Beerlo and comprises a dissolution system, which is contained in the water bath, I, and a pressure-regulated gas (solute) input system which is external to the water bath. The Pyrex dissolution bottle, G, is provided with two opposed reservoir bulbs, f and f’, which are connected to the bottle near its base by capillary side arms g and g’, respectively, and to the bottle neck by 10-mm tubing. The liquid is agitated by means of a magnetic stirrer, J, placed beneath the bottle and bath. The dissolution bottle is connected to the measuring system at ground joint e. The volume of the bottle, calibrated a t 30” to the level and mark designated by h, is 550 ml. The hydrogen uptake is measured with the 10-ml graduated buret, F. The manometer, E, serves both as a pressure reference (null) indicator and as a mercury cutoff between the gas input (or reference) and dissolution systems. The right arm, b, of manometer E is a graduated 1.0-ml capillary tube, and the total input capacity of the system shown is therefore 11 ml. Mercury levels in manometer E and buret F are regulated with leveling bulbs C and D, respectively. The dissolution section is thermostated at 30.0 i 0.05” in the water bath. The input or reference system comprises a gas reservoir, A, and a closed-end manometer, B. StopThe Journal of Physical Chemistry
cocks 1 and 2 are for evacuating and filling the system, respectively. Sulfuric acid and potassium hydroxide solutions were prepared with reagent grade chemicals. Matheson Prepurified grade hydrogen (99.5% minimum purity) was used directly from the tank.
Procedure The bottle was initially filled to slightly above the calibration mark, h, and cooled to about 3” after connection to the apparatus. The system was then slowly evacuated, and the solution was stirred at a rate sufficient to prevent excessive boiling. For evacuation, the mercury in manometer E was below both arms at the approximate level c, and stopcocks 2, 3, 4, and 5 were closed. The samples were degassed for about 3 (1) Leclanch6 S.A., Yverdon, Switzerland. (2) (a) A. Seidell, “Solubilities of Inorganic and Organic Compounds,” Suppl to 3rd ed, D. Van Nostrand Co., Inc., New York, N. Y., 1952, p 236; (b) “International Critical Tables,” Vol. 3, McGraw-Hill Book Co., Inc., New York, N . Y., 1928, p 256. (3) A. E. Markham and K. E. Kobe, C h m . Rev., 28, 519 (1941). (4) D. M. Himmelblau, J . Chem. Eng. Data, 5 , 10 (1960). (5) G . Geffcken, 2. Physik. Chem. (Leiprig), 49, 268 (1904). (6) A. Christoff, ibicl., 55, 622 (1906). (7) G. Vertes and F. Nagy, Magy. Rem. Folyoirat, 65, 450 (1960). (8) L. W. Winkler, Ber., 22, 1764 (1889). (9) M. B. Knaster and L. Y. Apelbaum, Ruse. J . Phys. Chem., 38, 120 (1964). (10) A. Ben-Naim and S. Baer, Trans. Faraday Soc., 59, 2735 (1963).
SOLUBILITY OF Hz IN KOH AND H2S04
719
hr at 3", at which time no gas evolution was evident even with rapid stirring. The dissolution section was then raised to 30.0' with continuous degassing, and the liquid level was determined relative to the calibration mark, h. The level was usually slightly above this mark and was adjusted to it by additional evacuation with stirring.
Solubility Results The data for hydrogen solubility in distilled waterl4,11 taken to test the apparatus, are given in Table I, which demonstrates that excellent accuracy and precision can be attained. The standard per cent deviat,ion is a very low 0.5%. Table I: Solubility of Hydrogen in Water Cm: of Hz (STP)/I. ( a t 30°)
I /I Source
This work Operator A
A B B h
Himmelblau' Seidell*l
Figure 1. Schematic of solubility apparatus.
After leak-testing the system, hydrogen was admitted to the input section, whereby the pressure was set at slightly above 1 atm by adjusting the mercury level in bulb A. Stopcock 6 was then opened to admit the gas into the dissolution section in contact with the unstirred solution. After establishing the desired gas pressure and mercury level in buret F, stopcock 5 was opened and the mercury level in manometer E was raised, sealing off the thermostated section. Preliminary experiments showed that at least 2 hr tvas required for a detectable quantity of gas to be absorbed by the quiescent liquid. Dissolution was started by switching on the stirrer. To take volume readings stopcock 5 was closed, the input pressure (manometer B) was accurately set, and the mercury menisci in arms a and b of manometer E were exactly leveled by adjustment of the mercury level in buret F. Absorption of gas was initiallyrapid. About 2 hr was required with distilled water before the uptake ceased entirely, while almost 5 hr was required for the most viscous solutions studied. Supersaturation did not appear, and no volume change was evident by a reduction in stirring rate after equilibrium had been attained. Acid and base normalities were determined by titration with standard NaOH or HCl solutions after completion of the run.
Dev
17.03 16.90 17.10 16.93
0.04 -0.09 0.11 -0.06
Av 16.99
10.08
-
17.1 17.0
Solubility values of hydrogen in sulfuric acid and potassium hydroxide solutions are given in Tables I1 and I11 and Figures 2 and 3. Each experimental value is an average of either two or three determinations. Table I1 : Solubility of Hz in H2S04Solutions a t 30'
so
IHzSOd, N
Cms of Hz/l. (at 30')
Log -
0.0011 0.100 0.502 1.02 3.04 5.05 6.95 9.67 12.4 15.2
17.09 16.66 16.15 15.17 12.76 10.83 10.01 8.87 8.11 7.68
0 0.0083 0.0221 0.0495 0.1243 0.195 0.230 0.283 0.322 0.345
S
Solubilities are given in cubic centimeters of gas at STP ( O O , 760 mm) per liter of solution at 30".
Discussion "Salting-out" is the decrease in solubility of a nonelectrolyte in ionic solutions. The vast literature on this phenomenon is summarized in several comprehensive reviews.12-'* Empirically, salting-out is well described by the equation of Setchenowl' (11) See ref 2, p 553. (12) J. N. Sugden, J. Chem. Soc., 174 (1926).
Volume 70,Sumber S March 1566
P. RUETSCHI AND R. F. AMLIE
720
log (S2O/S2) =
(1)
kSC8
where Szo and S2 are the solubilities (in cubic centimeters a t STP per liter) of the nonelectrolyte molecules Table 111: Solubility of H2 in KOH solutions at 30” [KOHI, N
Cma of Hz/l. (at 30°)
0.0091. 0.102 0.510 1.03 1.98 3.04 5.00 7.61 10.23
16.68 16.29 14.13 12.13 9.27 6.71 3.65 1.59
0.77
Log
SO S
in the solvent and in the ionic solution, respectively, and C, is the concentration of the electrolyte in equivalents per liter. Theories on salting-out may be grouped into three categories: (a) electrostatic theories, (b) internal pressure theories, (c) hydration theories. Electrostatic Theories. By minimizing the electrostatic free energy, Debye1*>l9derived the expression for the salting-out effect
0.003 0.018
0.080 0.146 0.264 0.404 0.666 1.029 1.344
S2/S2O where u =
=
1 - Zuni
(2)
JI’
4a
1 - e ~ p [ - ( ( R / r ) ~ ] ~ ~ d r (3)
The integration limits are between the radius of the (hydrated) ion ri and the radius rn of the spherical volume at disposal of one ion. The characteristic length l? is defined by (R4
=
and depends on the molecular volumes of solvent, and nonelectrolyte molecules V1 and V2, respectively, and the corresponding dielectric decrements, whereby the concentrations nl and n2 of solvent and nonelectrolyte are given in molecules/cm3. For values of Z u n i close to 1, it follows from (2) In (S2O/S2) = Z a n ,
H2S04 N O R M A L I T Y
Figure 2.
Solubility of Hn in HzSO, solutions.
(5)
Equation 5 is in qualitative agreement, but in quantitative disagreement, with experimental data. 13,16,20-23 The electrostatic the0ryl~r1~ describes salting-out entirely in terms of dielectric effects. Nonelectrolytes with zero dielectric decrement (such as hydrogen) are thus predicted to show no salting-out, which is not in accord with experimental results. ~~
KOH
NORMALITY
Figure 3. Solubility of Hz in KOH solutions.
The Journal of Physical Chemistry
(13) hl. Randall and C. F. Failey, Chem. Rev.,4, 271, 285, 291 (1927). (14) H. 5. Harned and B. B. Owen, “The Physical Chemistry of Electrolytic Solutions,” 3rd ed, Reinhold Publishing Corp., New York, N. Y.,1958,p 531. (15) B.E. Conway and J. O’hl. Bockris, “Modern Aspects of Electrochemistry,” Butterworth and Co. Ltd., London, 1954,p 95. (16) F. A. Long and W. F. hlcDevit, Chem. Rev., 51, 119 (1952). (17)A. Setchenow, 2. Phyaik. Chem. (Leipzig), 4, 117 (1889). (18) P.Debye and J. McAulay, Phyaik. Z . , 2 6 , 22 (1925). (19) P.Debye, Z.Phyaik. Chem. (Leipzig), 130, 55 (1927). (20)G. Scatchard, Trans. Faraday SOC.,2 3 , 454 (1927). (21)G.Scatchard and M. A . Benedict, J . A m . Chem. SOC.,5 8 , 837 (1936). (22)J. B. Hasted, D. M. Ritson, and C. H. Collie, J . Chem. Phys., 16, 1 (1948). (23) M.Givon, Y. Marcus, and M. Shiloh, J . Phya. Chem., 6 7 , 2495 (1963); see also Y.Marcus, Acta Chem. S c a d , 11, 329 (1957).
SOLUBILITY OF Hz IN KOH AND H8O4
Very recently, Givon, et a1.,23 have tried to improve the Debye treatment of salting-out by introducing into eq 4 a term describing the dependence of the dielectric constant on the ionic concentration of the solution. The neglect of dielectric saturation in the theory of Debye can cause serious errors in the distribution function of polar nonelectrolyte and solvent molecules near the central ion. Introduction of appropriate Dl(r) and D2(r)functions might lead to a more meaningful result than the use of an over-all ionic dielectric decrement. Internal Pressure Theories. Based on concepts developed by T~2it,2~ Tammann,25 and Gibson,26 the salting-out effect has been explained by AlcDevit and Long2’ in terms of an “internal pressure,” exerted by the ions on the nonelectrolyte molecules. Values of IC, calculated with the equation of McDevit and Long are 2-3 times larger than experimental data. However, the right order with respect to the relative effects of different ions is predicted. The theory presumes equal compression of all solvent molecules in the system. I n reality, only the water molecules in the hydration shells are “comprezsed” owing t,o the dipole attraction and corresponding electrostriction. Nonelectrolyte molecules of nonpolar character are, however, expelled from the hydration shells, as discussed below. Hydration Theories. Already in 1907, Philipz8pointed out that salting-out may be explained by assuming that some of the water becomes attached to the electrolyte ions as water of hydration and is thereby removed from its role of solvent. This idea wa,s pursued further by Sugden,12 who showed that hydration numbers derived from saltingout measurements were additive properties of the ions. Eucken and HerzbergZ9suggested that molecular hydrogen is an ideally suited test substance for studying ionic hydration by means of salting-out. Hydrogen molecules, being small, inert, unhydrated, nonpolar, and nonpolarizable particles, whose solubility is very small, whose effects on the dielectric properties are negligible, and whose dispersion interaction with ionsmi8l is not perceptible, so that they are completely expelled from the hydration shells,29 should cause a minimum of distortion in the solution to be studied. Salting-out measurements of hydrogen should prove especially useful for the study of highly concentrated solutions, including strong acids and bases. On the basis of the hydration concept, the ratio of the solubllities in the pure solvent and in the solution may be set simply equal to the ratio of the VOlUmeS of “free” water
72 1
szo/s2 = (1 - 1ZniVi) where Vi is the self-volume of the hydrated ion. Taking the logarithm of (6) and expanding the right side, one derives In (Szo/Sz) = ZniVi
+
(BniVi)’/2
+ ( Z ~ ~ i v i ) ~+/ 3. . .
(7) The linear approximation, valid for dilute solutions ni-0
(where Vio is the ionic volume at infinitely dilute solution) represents the Setchenow equation. For a completely dissociated electrolyte m i = vins,where v i is the number of ions of type i produced upon dissociation of one molecule of salt. Thus
(9) The slope of the Setchenow equation is therefore a In (S2O/Sz) = n,Z(viVio)
measure for the ionic self-volumes Vio at infinite dilution. The neglect of the higher terms in (7) appears fortuitously compensated by the decrease in Vi, and the Setchenow plot can remain linear up to relatively very high electrolyte concentrations. If the electrolyte is incompletely dissociated, ni is no longer proportional to n,, and the Setchenow plot is no longer linear. This appears to be the case for sulfuric acid (Figure 4). I n view of (6) and (9) one derives the following interesting empirical relation for the decrease of the hydrated ionic self-volumes with concentration In (SzO/S2) = n,ZviViO = -In [l - n,(ZviVi)] and
(24) P. G. Tait, ref 14, p 379. (25) G. Tammann, “Ueber die Beziehungen zwischen den inneren Kraften und Eigenschaften der Losungen,” Voss, Leipzig, 1907, p 36. (26) R. E. Gibson, J . Am. Chem. SOC.,56, 4 (1934). (27) W. F. McDevit and F. A. Long, ibid., 74, 1773 (1952). (28) W. Philip, Trans Faraday Soc., 3 , 1 (1907); W. Philip and W. Bramley, J. Chem. SOC.,107, 377, 1831 (1915). (29) A. Eucken and G. Herzberg, Z. Physik. Chem. (Leipzig), 195, 1 (1950). (30) J. O’M. Bockris, F. Bowler-Reed, and M, Kitchener, Trans. Faraday sot., 47, 184 (1951). (31) G. Kortum, 2. Elektrochem., 42, 287 (1936).
Volume YO,Number S March 1066
722
P. RUETSCHI AND R. F. AMLIE
0
' 411.2
f
-
I
/
/'
/
I
I
I
I
I\
\ \
f
NORMALITY
Figure 4.
Figure 5 . Schematic illustration of the ionic volume effect.
Setchenow plots for HPsolubility.
With (9) one obtains from the data of Figure 4 for the ionic self-volume of KOH a t infinite dilution
viVi0 = 5.12 X
cm3
which corresponds to two spheres of radius ?h
=
The distribution of nonpolar nonelectrolyte molecules outside the hydration shell is, according to Debye, given by n2 =
3.94 A
The calculated radii are in the range of the internuclear separation distance K+-HzO, obtained by MoelwynHughes,3z and are, as expected, slightly larger than those estimated from ionic tra.nsport processes,33 which pertain only to tight bonds, but not to the diffuse part of the hydration shell.29 Self-volumes, calculated with (9) from the data of ref 17 are listed in Table IV.
n20 exp [
- (R/T)~]
(12)
where nzo would be the concentration far away from the ion, that is, in the absence of electrolyte, and where I? is defined by (4). The number of nonelectrolyte molecules contained in the volume a t the disposal of one ion is then ZZ = nzoJrnexp Th [ - (R/r)4]4~rzdr
(13)
In the absence of electrolyte, the number of nonelectrolyte molecules in the same volume would be
~~
Table IV : Salting-out of Hydrogen and Ionic Hydration
Salt
NaOH
ICOH NaCl KCl LiCl HCl
Setohenow slope, M -1
Self-volume, cma/molecule x 1022
0.140 0.130 0.114 0.102
5.37 4.97 4.35 3.90 2.90
0.076 0.030
1.15
The above hydration theory may be extended to incorporate the Debye effect outside the hydration shell. As illustrated in Figure 5, each ion with crystallographic radius r, carries a hydration shell of radius rh. If the solution contains Zni ions/cm3, the volume a t the disposal of one ion (dotted sphere) is 1 p r r z d r = (4/3)xrn3 = Zni
The Journal of Physical Chemistry
The lower limit of the integral here is zero since the volume previously occupied by the ion is now also available for dissolution. Comparing the numbers of nonelectrolyte molecules in the same volume, before and after the addition of salt, one has
exp[- (R/r)4])4sr2dr} ~~
(32) E. A. Moelwyn-Hughes, "Physical Chemistry," 2nd ed, Pergamon Press Ltd., Oxford, 1961, p 887. (33) E. R. Nightingale, J . Phys. Chem., 63, 1381 (1959).
ROLEO F SF IN
THE PYROLYSIS O F
= (Zni)(fi?rr2dr
- Sd;*r2dr
ir(1
=
Di-t-BUTYL
723
PEROXIDE
trolyte concentrations, eq 15 becomes, in analogy to (5)
-
1
In (S2O/S2) = ZniViO
exp [ - (R/r)4])47rr2dr
1 - ZViDn* - zcrni
(15)
where Vio is the self-volume of the hydrated ion and u an abbreviation, already defined in (3). Since the ratio Z2/Zzo obviously signifies the ratio of solubilities (S2/S20)per unit volume, eq 15 may be compared immediately with eq 2. For small elec-
+ Zmi
(16)
which represents a refinement over eq 8 in the sense of the Debye theory. For zero dielectric decrements (nonpolar molecules), eq 16 reduces to (8). Acknowledgment. The authors wish to express their appreciation to John W. Harman, 111, and Thomas A. Black for their contributions to the experimental part of this study.
The Role of Sulfur Hexafluoride in the Pyrolysis of Di-t-butyl Peroxide: Chemical Sensitization and the Reaction of Methyl Radicals with Sulfur Hexafluoride
by Leslie Batt and Frank R. Cruickshank Chemistry Department, University of Aberdeen, Aberdeen, Scotland
(Received August 17, 1966)
The acceleration of the rate of decomposition of di-t-butyl peroxide by SF6has been shown to be a chemical effect rather than a physical one as has been postulated previously. Detection and quantitative estimation of methyl fluoride gives the following value for the rate of attack of methyl radicals on SFe: logk = 10.3 - 14,100 ca1/4.575T 1. mole-' sec-l. Detection of methyl fluoride also permits the following mechanism to be postulated which accounts satisfactorily for the accelerated rate. Me. SFs + n!kF SF6., SF5. -t SFc F., F a dtBP -t H F dtBP.-H, dtI3Pa-H -t t-BuOi-BuO, Me. SF5. + MeF SF,. It is concluded that SF8 will not act as an inert energy transfer agent in the presence of organic free radicals particularly over 140".
+ +
+
Introduction
+
+
+
+ Me&O +CH, + CHzCOMe Me. + -CH2COMe-+ EtCOMe
Me.
The gas phase pyrolysis of di-t-butyl peroxide (dtBP) has been shown to be an uncomplicated, homogeneous first-order process by many workers.' The decomposition is satisfactorily explained by the mechanism dtBP +2 t - B ~ 0 * t-BuO.
+
+
+ 31 +Me. + iVlezCO + M
(l)
(2)
2CHzCOMe + ( C H Z C O M ~ ) ~ 2Me. 4CzHe
(3) (4) (5) (6)
(1) For a review, see S. W. Benson, "The Foundation of Chemical Kinetics," McGraw-Hill Book Co., Inc., New York, N. Y.,1960,pp 363-370.
Volume 70, Number 3 March 1966