2340 Shrawder, Irving Cowperthwaite: Activity Coefficients of Sulfuric Acid at Temperatures From To

Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

2340 JOSEPH SHRAWDER,

JR., AND IRVINGA. COWPERTHWAITE Vol. 56

the anion. The comparatively few instances of t i ~ n . This


~ is equivalent to an assumption that
the hydration of an ammonium salt, however, may one molecule of ammonia is coordinated with
be taken as illustrations of anionic solvation,* the ammonium ion, and that four molecules, or
since coordination through the ammonium ion is two diammonol molecules, with the trinitride ion.
improbable. Ammonation of ammonium salts, The formula of the pentammonate may on this
especially a t lower temperatures, occurs f a r more basis be written [NHgNH4]+ [N3*2(NH&]-.
frequently than hydration. Unless this am- Summary
monation can be attributed to a coordination of
simple or associated ammonia molecules through Pressure-temperature-concentration relation-
the ammonium ion, it seems inevitable that it ships of the system ammonium trinitride-am-
must be ascribed to solvation of the anion. monia have been investigated over ranges from
In the system ammonium trinitrideammonia 0 to 4000 mm., -90 to +30°, and 0 to 100% am-
the diammonate is believed to exhibit anionic monia. Two solvates, ammonium trinitride di-
solvation. Whether coordination takes place ammonate, NHdN3.2NH3,and nmmonium trinitride
through hydrogen or through nitrogen atoms, pentummonute, NH4Ng5NH2, both of which are
however, is a question to which the final answer white crystalline solids, have been obtained. The
cannot as yet be given. The pentammonate, on diammonate undergoes inversion into anammon-
the other hand, with its inversion temperature ous salt at -9 O ; the pentammonate undergoes
at - 71 O, probably does not show exclusively inversion into the diammonate at -71". The
anionic ammonation, but may rather be identical eutectic is located a t -87" and 76% ammonia.
(9) In all probability liquid ammonia, like water, is associated,
with the diammonate, except that each of the and contains such molecular species as NHs, ammonol, (NHr)t,
three ammonia molecules has undergone dimeriza- diammonol, and (NH3)3, iiiammonol, analogous to hydrol, dihydrol
and trihydrol.
(8) Sidgwick, "The Electronic Theory of Valency," Oxford
University Press, 1929, p. 194. ITHACA, N.Y. Aucus~13, 1934
RECEIVED

[CONTRIBUTIONFROM THE DEPARTMENT


OF CHEMISTRY, COLUMBIA UNIVERSITY]

The Activity Coefficients of Sulfuric Acid at Temperatures from 0 to SOo1


BY JOSEPH SHRAWDER,
JR., AND A. COWPERTHWAITE
IRVING

Galvanic cells involving sulfuric acid have been is well designed to provide the required measure-
studied by a number of investigators.2 With few ments in the dilute range. The solubility of lead
exceptions the investigations have been confined sulfate, 43.5 mg. per liter in pure water at 20°t4is
to the use of the mercury-mercurous sulfate elec- sufficiently low in dilute acid to obviate any large
trode. While this electrode is reversible to sul- correction from this source.
fate ion and functions nicely in relatively concen- Previous attempts to measure the potential of
trated solutions, the solubility of mercurous sul- the above cell have been made by Lewis and
fate is sucha as to prohibit its use in solutions more B r i g h t ~ n ,Vosburgh
~ and Craig,6 and Baum-
dilute than approximately 0.005 molal. For the stark.2e Of these, only Baumstark was able to
evaluation of E it is desirable that measurements secure reproducible results, the other workers re-
be obtained to a lower concentration, porting deviations of the order of 0.01 volt.
The cell In the present work the electromotive force of
Pb-Hg (2 phase), PbSO&), Hg%(m), HZ this cell has been measured from 0 to 50' over
(1) This paper is from a dissertation presented by Joseph Shrsw- the concentration range 0.02 to 0.001 m.
der, Jr., to the Faculty of Pure Saience of Columbia University in The methods which have been commonly em-
partial fultllment of the requirements for the degree of Doctor of
Philosophy. ployed for the extrapolation of electromotive force
(2) (a) Brdnsted, 2. physik. Chem., 68, 693 (1909); (h) Edgar, data to infinite dilution vary from simple graphical
Thesis, Massachusetts Institute of Technology, 1912; (c) Lewis
and Lacy, THISJOURNAL, 56, 804 (1914); (d) Randall and Cush- extrapolations, on the one hand, through a variety
man, ibid., 40, 393 (1918); (e) Baumstark, Dissertation, Catholic (4) Purdum and Rutherford, ibid., 88, 3221 (1983).
University of America, 1932. (5) Lewis and Brighton, ibid., SQ, 1906 (1917).
(3) Hsrned and Sturgis, THISJOUENAL, 47, 945 (18251. (6) Vosburgh and Craig, ibid., 61, 2009 (1928).
Nov., 1934 THEACTIVITY OF SULFURIC
COEFFICIENTS ACID 234 1

of empirical methods, to the functional methods trodes were used in each cell, the average deviation being
based on the Ilebye-Hiickel theory. The com- of the order of 0.02 mv. The hydrogen electrodes were
freshly plated for each cell.
plete solution of the Debye-Hiickel equation Apparatus.-The type K potentiometer, Eppley stand-
through the third approximation for the case of ard cell, and thermometers were calibrated and frequently
unsymmetrical valence type electrolytes has been checked during the course of the investigation. Water
made available by means of tables communi- thermostats were used with temperature control main-
cated by La Mer, Gronwall and Greiff.' How- tained t o r0.02'.
Lead Nitrate.-Baker Analyzed salt was recrystallized
ever, the validity of extensions of the Debye- four times, filtered by suction and dried a t 110'.
Hiickel theory to the cases of unsymmetrical val- Sulfuric Acid.-C. P. sulfuric acid was redistilled three
ence type electrolytes and of electrolytes giving times in a Pyrex still a t atmospheric pressure. The final
ions of unequal size has been brought into question distillate was then diluted with sufficient distilled water
by Halperna8 In addition, the incomplete dis- to make nine liters of approximately half molar solution.
This stock solution, from which all the solutions subse-
sociation of the bisulfate ion contributes to Ihe quently studied were made by weight dilution, was stored
uncertainty involved in any application of the in a large Pyrex bottle and was carefully analyzed a t
Debye-Hiickel theory to sulfuric acid solutions. intervals of two months by precipitation as barium sulfate,
The work of Hamere in measuring the dissocia- and by titration against acid potassium phthalate using
tion constant of the bisulfate ion over a wide range carbonate-free sodium hydroxide solution as an inter-
mediate. Weight burets were employed in this titration.
of temperatures has offered the possibility of ap- The concentration of the stock solution was known to one
plying this parameter in an effort to obtain EQ part in four thousand.
without recourse to postulates other than the Lead Sulfate.-This was prepared by adding a 0.5
limiting law of Debye. This is desirable since molar solution of sulfuric acid to a 10% solution of purified
the conclusions of Halpern confirm the functional lead nitrate. The product was washed by decantation
for two weeks, after which it was filtered and dried at 110'.
nature of the limiting law (i. e., the square root Mercury and hydrogen were purified in accordance with
dependence, if not the constant). standard procedure.
Experimental Methods Lead Amalgam.-The amalgams were prepared elec-
trolytically following the directions of La Mer and Parks,""
Since the electrodes employed require oxygen-free con- and were stored in reservoirs of the type described by
ditions, a cell similar to that employed by Cowperthwaite, Cowperthwaite and La Mer."b Attempts to obtain satis-
La Mer and BarksdalelO has been used. With the exctp- factory amalgams from stick lead and purified mercury
tion of a hydrogen inlet tube added to the base of the were uniformly unsuccessful, due to the difficulty encoun-
central compartment, the cell employed here differs only tered in obtaining lead of sufficient purity.
in dimensions from that used by the above authors. The
technique employed with this type of cell has been pre- Theoretical
sented elsewhere.1° The cell upon which measurements have been
One unique feature of the manipulation consists in the
made may be represented as
use of dried lead sulfate in the lead amalgam-lead sulfate
electrode. Previous workers" have followed the procedure Pb-Hg (2 phase), PbSOl(s), HzSOa(m), HZ (1)
of storing the lead sulfate under water. In the present for which the corresponding cell reaction is
work the lead sulfate was dried, powdered and added to Pb (2 phase amalgam) + HzSOa(rn) =
the amalgam limbs immediately after the amalgam was
introduced. The addition of sufficient solid lead sulfate
PbSOd(s) + Ha (2)
The conventions employed here and in what fol-
to leave a thin film over the surface of the amalgam has
given excellent results. Electrodes prepared in this man- lows are those established by Lewis and Randall,
ner checked among themselves to better than 0.1 mv. whose nomenclature will be followed throughout.
The hydrogen electrodes employed were of two types. With the reactions written in the direction indi-
For the higher concentrations the standard black elec- cated, the cell potential is positive in sign, and
trodes were found to be satisfactory. For the concentra- the substance of variable activity occurs to the
tion 0.002 m and below, better results were obtained using
bright platinum electrodes. These were prepared follow- left of the equality sign. The equation for the
ing the directions given by Lorch.lZ Four hydrogen elec- cell potential becomes, therefore

nF In 4% m + -
+ 3RT 3RT
(7) La Mer, Gronwall and Greiff, J . Phys. Chem., 36, 2245 (193L). E = EO - nF
In f (3)
(8) Halpern, J . Che.w. Phys., 9, 85 (1934).
(9) Hamer, THIS JOURNAL, 16, 860 (1934). as may be verified by reference to the free energy
(10) Cowpertbwaite, La Mer and Barksdale, ibid., 66, 544
(1934). equation and the definition of mi. The function
(11) (a) La Mer and Parks, ibid., 69, 2040 (1931); (b) Cowper-
thwaite and La Mer, ihid., 19, 4333 (1931); (c) Bray, ibid., 49, 2372
Eo' may then be defined as
(1927) EO' = E - 3R- T In 4I/1m
(4)
(12) Lorch, Dissertation, Columbia University, 1932, 12 F
2342 JOSEPH SHRAWDER,
JR., AND A. COWPERTHWAITE
IRVING Vol. 56

so that at 18 and 30". The measurements of these work-


ers extend over the entire range of acid concentra-
tions studied in this work. The desired solubility
In the equations used thus far, no recognition data were obtained by plotting the available data
has been given the fact of incomplete dissociation on a large scale followed by interpolation to the
of the bisulfate ion. If, instead of the total acid required concentrations and extrapolation to the
concentration, the stoichiometric concentration required temperatures.
of ions is employed in equation (4), a new function Table I records the calculated solubilities in
is thereby defined. This will be denoted by the millimoles per liter in sulfuric acid of the indi-
symbol E*. cated molalities.
The quantity Eo* for an incompletely dissoci-
TABLE I
ated electrolyte is thus the analog of the quantity
CALCULATED
SOLUBILITY
OF LEADSULFATEIN SULFURIC
Eo' for the completely ionized electrolytes. For ACIDOF MOLALITY
m (MILLIMOLES
PER LITER)
the case of sulfuric acid, it can be shown that m 00 12.5O 25' 37.59 50'
0.02 0.0099 0.0124 0.0152 0.0183 0.0220
.Ol .0107 .0134 .0166 .0199 ,0238
.005 .0122 .0153 .OB9 .0227 .0273
.002 .0170 ,0213 .0262 .0314 .0377
(7) .001 ,0238 .0299 .0372 ,0448 .0541

Calculation of CY
In the calculation of Eo*, values for CY, the de-
gree of ionization of bisulfate ion, are required.
The data given by Hamerg along with the calcu-
Equations ( 5 ) and (8) may be used to compute lated solubilities of lead sulfate may be em-
activity coefficients for sulfuric acid at each of the ployed in this computation. The expression for
temperatures and concentrations studied. The the second dissociation constant of sulfuric acid is
activity coefficients presented in this paper have
f H + mE*fBOP mP04' =
been computed on the basis of equation (5). f H 8 O r mE8Or-
K2 (11)
These have been called by BrOn~ted'~ "stoichio-
We may assume that the activity coefficients of
metrical activity coefficients."
the hydrogen and bisulfate ions will cancel,
Experimental Data and Calculations since they are activity coefficients of univalent
The correction for the solubility of lead sulfate ions in a solution of low ionic strength. Then if
in the calculation of EO' and Eo* was made by the concentration of bisulfate ion is denoted by
separating the In 'm terms in equations (4) and (7) x, the expression becomes
into two parts, one for the sulfate ion and one for
the hydrogen-ion concentration. Then equations
(4) and (7) become, respectively where ml and m2 have the significance previously
indicated. The activity coefficient of the sulfate
ion has been assumed to be equal to f, for a typi-
Eo* = E + 2RT
3 In mla'/3 + cal bi-bivalent sulfate and has been computed by
use of the extended theory of Gronwall, La Mer
and Sandved employing an ion size of 3.64 A.
where ml is the gross molality of the acid and The choice of this value for "a" was prompted by
m2 = ml + the molal solubility of lead sulfate in the agreement obtained by La Mer and Parks*la
sulfuric acid of molality m. The solubility of and Cowperthwaite and La Mer using this
lead sulfate i s the various dilute solutions of ion size in applying the extended theory to cad-
sulfuric acid has been calculated from the meas- mium sulfate and zinc sulfate, respectively.
urements of Purdum and Rutherford4 at 20", Since equation (12) contains the two mutually
Pleissner14at 18* and Huybrechts and Ramelot16 dependent unknowns fso,- and x, it has been
(13) Brensted, TRISJOURNAL, 42, 761 (1920). necessary to employ a series of approximations
(14) Pleissner, AT^. Kaiser. Ges. A m f .,. 26.. 419 (1907).
. , in its solution. From the calculated values of x,
(15) Huybrechts and Ramelot, Bull. SOC. chim. Belg., 36, 239
(1927). the corresponding values of CY may be obtained
Nov., 1934 THEACTIVITY
COEFFICIENTS
OF SULFURIC
ACID 2343

directly. Table I1 gives the values for a, thus TABLE IV


calculated, along with the dissociation constants t , "C. 0 12.5 25.0 37.5 50.0
from which they are derived. EO 0.32810 0.33919 0.35052 0.36192 0.37376

The activity coefficients given in Table V have


TABLEI1
been obtained by the introduction of the appropri-
OF HS04-
THEDEGREEOF IONIZATION
12.5O ate values for Eo and Eo' into equation ( 5 ) .
,n 00
&
... 25' 37.6' 50'

TABLE V
0.0148 0.0137 0.0120 0.0102 0.00794 STOICHIOMETRIC OF SULFURIC
ACTIVITYCOEFFICIENTS
0 ACID
0.02 0.5509 0.5375 0.5090 0.4799 0.4174 m 00 12.5' 2S0 37.60 50'
.01 .6620 .6487 ,6245 .5944 .5353 0.02 0.516 0.478 0.444 0.416 0.386
.005 .7586 .7470 .7257 .6993 .6469 .01 .609 .574 .543 .515 .485
.002 .8597 .8512 .8355 .8152 .77G8 ,005 .700 .674 .646 .623 .596
.001 .9124 .9062 ,8949 .8799 .8515 .002 .7Y8 .782 .767 .752 .725
.001 .853 ,846 ,837 .828 .807
EO' and Eo* may now be calculated on the basis
Discussion of Results
of equations (9j and (10). Table I11 gives the
observed values of the electromotive force cor- Despite any theoretical objections to the use of
rected to one atmosphere partial pressure of hydro- the La Mer, Gronwall and Greiff extension of the
gen, along with the corresponding values of Eo'
and Eo*.
TABLE
I11
VALUESOF E( OBS.), EO' AND Eo*
m 0' 12.5' 25' 37.5' 50' 0.320
E(obs.) 0.09536 0.09555 0.09589 0,09625 0.09651
0.001 BQ' .32251 .33302 ,34366 .35432 .36483
EO' .32463 .33541 .34648 .35768 .36922
E(obs.) .11725 .11799 .I1895 .I1985 .I2050
0.340
,002 EO' .32012 .33012 .34032 ,35048 .36036
EO' .32361 .33400 .34484 .35581 .36717
E(obs.) .14487 .I4622 .14745 .14890 .I5037
.005 Eo'
EO*
.31549 .32463 .33366 .34291 .35217
,32175 ,33154 .34156 ,35205 .36365
E(obs.) .I6443 .I6588 .I6744 .I6909 .17067
0.360
[k
.01 Eo' .31060 .31873 ,32698 .33532 .34358
EO' .31981 .32881 .33836 ,34834 .35963
E(obs.) .I8301 .18467 .18637 .18828 .19004
.02 EO' .30473 ,31196 .31922 .32671 .33404
EQ* .31772 .32606 .33513 .34458 .35577

The Extrapolation for EO 0.02 0.06 0.10 0.14


m+.
In Fig. 1, EO' has been plotted against ml/'
Fig. 1.
while Fig. 2 gives the function Eo* plotted in a
similar manner. The Debye limiting slopes for Debye-Hiickel theory, it is of interest to inquire as
the various temperatures have been included on to whether or not the extended theory will fit the
both plots. It will be observed that the function data and be of service as a method of extrapola-
Eo* approaches the limiting slope as the concentra- tion. We have therefore carried out the calcula-
tion is decreased. Eohas been evaluated at each tions necessary to compare the theory with the
temperature by taking the values of Eo*which fall experimental data at 25". In this computation
on the limiting slope and adding to them appro- an ion size of 1.75 A. has been employed. The
priate values of 3RTInF In f computed with the results are indicated graphically on Fig. 3 with
aid of the limiting law. The mean of the several circles denoting experimental points and the solid
values for Eo thus obtained has been taken as the line A representing the extended theory. Crosses
E o for that temperature. Values for Eo at five have been added to represent the values for Eo'
temperatures are given in Table IV. obtained by BaumstarkZeat 25". The value for
2344 JOSEPH SHRAWDER, A. COWPERTHWAITE
JR., AND IRVING Vol. 56

Eo used is that obtained using the extrapolation The values for E o thus obtained are probably
of the function Eo* as previously explained, so accurate to 0.1 millivolt. The value for EO at
that the curve drawn represents the best approach 25’ differs from that obtained by Baumstark
0.300
by 1.65 millivolts, which difference may be as-
signed to the difference in methods of extrapola-
0.320
tion since Baumstark extrapolated the Ea’
curve linearly.

0.340

2’
irl
0.306

0.380

0.02 0.06 0.10 0.14 0.340


?nlla.
Fig. 2.
to the experimental data obtainable by adjust-
ment of the ion size parameter. While the agree- 0.350
ment is unsatisfactory, it should be pointed out
that the theory does not provide convergent I
values for In f when only three terms are em- 0.02 0.06 0.10 0.14
ployed in the case of such a small ion size. In ml/z.
its present form, the third term of the series is of Fig. 3.-At 25’.
the same order of magnitude as the first. It is of interest to compare the activity coef-
In addition to the experiment points for EO’ and ficients obtained in this work with those obtained
the extended theory curve, A, Fig. 3 also contains by other methods. Table VI lists the coefficients
the limiting slope, B, and the experimental points calculated from equation (6) (column 1) with those
corresponding to the function Eo*. Through the given by Lewis and Randall, Randall and Scott’6
latter points has been passed a smooth curve, C, from the freezing point measurements, and
which fuses with the limiting slope. While the Baumstark.2e
point of fusion appears to vary somewhat with TABLEVI
temperature (Fig. 2), still the adherence is such as ACTIVITYCOEFFICIENTS
STOICHIOMETRICAL OF SULFURIC
to warrant the use of the limiting law in the ACID
extrapolation for EO. While the function EO’ must m
I XI
Baumstark
111
I.. & R.
IV
R . & S.
also extrapolate to Eo,the extrapolation is compli- 0.02 0.444 0.519 0.553
cated by the nature of the curve. If the Eo’ func- .Ol .543 ,617 .648
tion is to fuse with the limiting law, at low con- .005 .646 0.615 ,696 .734
centrations, it must exhibit a reversal of curvature .002 ,767 .735 .782 .825
at some point. The complete solution of the ,001 ,837 .804 .831 ,876
Debye-Huckel theory predicts just such a form The values obtained for the activity coefficients
of curvature in very dilute solution, but, as poixted are smaller than those given by Randall and Scott
out, it is not possible to obtain a quantitative and are larger than those of Baumstark. The
evaluation of the theory in the case of sulfuric latter difference arises from the difference in E o
acid. It has seemed expedient therefore to use values used. The values obtained by the freezing
the function Eo* and rely upon the limiting law, (16) Randall and Scott, THIS
JOURNAL,49, 647 (1927).
Nov., 1934 PARTIAL MOLALHEAT OF DILUTION
OF SULFURIC ACID 2345

point method are subject to difficulties associated 0.002 and 0.001 m at each of the temperatures,
with extrapolation which axe fully as severe as 0, 12.5, 25, 37.5 and 50'.
with the electromotive force method, and are 2. A new method has been devised for ex-
particularly aggravated in this case of an im- trapolating to infinite dilution, taking into account
completely dissociated unsymmetrical valence the incomplete dissociation of the bisulfate ion,
type electrolyte. by means of which Eohas been obtained for each
temperature.
Summary
3. The activity coefficient of sulfuric acid has
1. The electromotive force of the cell Pb-EIg been evaluated a t each temperature for the five
(2 phase), PbS04(6), HeSO*(m),H:, has been meas- concentrations studied.
ured for the concentrations 0.02, 0.01, 0.005, NEWYORKCITY AUGUST15, 1934
RECEIVED

[CONTRIBUTION FROM THE ]DEPmTMENT OF CHEMISTRY, COLUMBIA UNIVERSITY]

The Partial Molal Heat of Dilution of Sulfuric Acid from Electromotive Force
Measurements
A. COWPERTHWAITE A N D JOSEPH
BY IRVING SHRAWDER, JR.

In a previous electromotive force meets- tions studied, in which EO' is expressed in the
urements at concentrations from 0.001-0.02 rn and following manner
extrapolated values for Eo were given for the cell EO' = A + Bt + Ct4 + Dta (4)
Pb-Hg, PbSOi(s), H&h(M), HI (1) The constants for these equations have been
at 12.5' intervals from 0-50'. The reaction for determined by the method of least squares. This
this cell is method was selected in order to obtain a smooth-
+
Pb (satd. with Hg) H2S04(na)-+ ing of the experimental data with a view to se-
PbS0&) 4-Ef2 (1 atm.) (2) curing reliable derivatives. Equations of the
As the sulfuric acid is the only reactant of vari- quadratic type have been found to fit the experi-
able activity, these data used in connection with mental points almost as well as the cubic type
the Gibbs-Helmholtz equation permit the compu- used; however, the improvement secured by use
tation of the partial molal heat of dilution of sul- of the cubic equations appeared to be sufficient
furic acid. It has been shown by La Mer and to warrant the use of the four constant equations.
as~ociates,~that the introduction of the function The values for the constants of equation (4) for
EO' = E - ( v R T / n F ) In m into the Gibbs- the various concentrations are given in Table I.
Helmholtz equation gives the convenient form
TABLEI
CONSTANTS EO' = A
FOR THE EQUATION: + Bt + Ct* + Dta
m A B x 104 c x io' D x 100
This form of the equation is particularly useful in 0.000 0.32810 8.8717 1.029 8.294
that it may be used in the evaluation of - AHo .001 .32251 8.3111 9.015 -11.905
since EO' extrapolates smoothly to Eo at infinite .002 ,32012 7.8564 14.059 -20.454
dilution. .006 .31549 7.2306 2.068 0.116
In order to obtain the temperature coefficients .01 ,31060 6.4117 7.918 - 8.447
,02 .30473 5.6728 7.683 - 7.729
required in the above equation, an analytical
function has been set up for each of the concentra- By use of the above coefficients,the first deriva-
(1) This paper is from a dissertation submitted by Joseph Shraiv-
der, Jr., to the Faculty of Pure Science of Columbia University in
tive of EO' with respect to the temperature has
partial fulfilment of the requirements for the degree of Doctor of been evaluated at 25' for each of the values for
Philosophy. m. The calculation of - A H and of is then
(2) Shrawder and Cowperthwaite, Tars JOURNAL.66, 2340
(1934). readily carried oGt. The important quantities
(3) La Mer and Parks, ibid., 63,2040 (1931); La Mer and Cow- used in the thermal calculations have been sum-
p e r t h d t c , {bid., MI, 1004 (1933); Cowperthwaite, La Mer and
Barksdsle, ibid., 66, 644 (lea4). marized in Table 11.

You might also like