Temperature Dependence of E° for the Daniel] Cell

Sister Barbara Buckbeei, Ronald E. Surdzial^ and Clyde R. Metz-%

Indiana University

Abstract

EMF measurements for the cell Zn/ZnSOj (1m)//CuS04 (lM)/Cu were made at various
temperatures over the range of 0-50 °C. The equation describing the voltage as a function
of temperature is

E°(volt) = (1.1028 ± 0.0026) — (0.641 ± 0.425)10- 3 t + (0.72 ± 0.87)10- 5 t2

where t is the Celsius temperature. This result compares favorably with electrochemical
measurements reported in the literature for similar cells, but the derived values of ^G°,
AS° and /\H° differ considerably with accepted thermodynamic values. This disagreement
probably results from the presence of a residual voltage and corresponding temperature
coefficient inherent in the cell from incomplete elimination of the liquid junction potential.

The conventional Daniell cell, Zn/Zn- + //Cu2 + /Cu, is often used in

general chemistry courses to demonstrate the calculation of overall cell
voltage by combining half-cell potentials. Using the data commonly
found in textbook tables (3) for the Zn/Zn-+ and Cu/Cu 2 + couples,
-0.763 v and 0.337 v, respectively, the predicted value of E° at 25 °C is
1.100 v. Upon careful construction of the cell using 0.5m solutions of
the nitrates or sulfates of the metals, the observed voltage is somewhat
lower than the predicted value (11).

For a cell containing the sulfates of the metals, the overall cell
potential E(exp) is given by

E(exp)=E° — RT lnQ + E(LJ) [1]

nF
where Q, the thermodynamic activity quotient, is denned as

Q =: a Zn -+ aCu /a Cu2+ a Zn = a2 ZnS0 ^ a Cu /a2 CllS0 ^ a Zn [2]

and E(LJ) is the liquid junction potential arising from the ZnS0 4 -KCl-
CuS0 4 The activities of the metals in Equation [2] are unity
interfaces.
by convention and the product of concentration and mean activity co-
efficient gives the mean activities for the salts. Using the values given
by Robinson and Stokes (8) for the activity coefficients, the emf contri-
bution in Equation [2] for the activity term is negligible for the 1m
concentrations.

As written above, the cell contains a salt bridge to minimize the

liquid junction potential. Maclnnes
(5) states that E(LJ) is negligible
provided the ionic mobilities of the cation and anion in the bridge are
equal. If a saturated KC1 bridge is used, this equality is nearly achieved
at 25° C. Thus any observed voltage should represent E° for the
chemical reaction of interest: Zn + Cu-+ Zn- + -f Cu. =
1
Current address St. Joseph High School, Brooklyn, N.Y.
:

2 Current address: Highland High School, Highland, Indiana.

3 Current address Indiana University-Purdue University at Indianapolis.
:

123
124 Indiana Academy of Science

Experimental
The half cells consisted of strips of Zn and Cu metal (10x1x0.1 cm)
dipping into 1M solutions of ZnS0 4 and CuS0 4 respectively. To reduce ,

heat transfer from the solutions to the air, the electrodes were thermally
insulated using 1-inch thicknesses of plastic foam. The half cells were
joined by a salt bridge consisting of a saturated KC1 solution suspended
on a strip of chromatographic filter paper. To reduce possible interaction
between the metals and KC1, the electrodes and a small quantity of
solution were isolated from the bulk solution by enclosing them within
glass tubing which was drawn to a capillary tip. At each temperature
fresh solutions and a newly-constructed salt bridge were used. All chemi-
cals were of reagent grade quality and standard quantitative procedures
were used in preparation of the solutions.

The complete Daniell cell was placed in a constant temperature bath

and the temperatures of the half cells were monitored by separate
thermometers. Once temperature equilibrium was reached, the tempera-
ture was recorded to the nearest 0.05° C (corrected for stem immersion)
and the open-cell voltage was read from a Honeywell Potentiometric
Voltmeter, Model 852, to the nearest 0.1 mv. At the sensitivity used, the
input impedance was 10 Mil, so negligible current was drawn from the
cell.

Results

The data appear in Figure 1. Using the method of Bennett and Frank-
lin (1), the regression coefficients for a linear and a quadratic dependence
on temperature were determined. An analysis of variance indicated the
quadratic term to be significant and the corresponding equation is

E° (volt) = (1.1028 ± 0.0026) — (0.641 ± 0.425)10 3 t+

(0.72 ± 0.87)10-5 t2 [3]

where t is the Celsius temperature and varies from 0° to 50° C. The

standard deviation is 1.36 mv and the estimated errors in the regression

coefficients are calculated on a 95% probability limit basis.

The maximum random error voltage resulting from tempera-

in the
ture measurement is negligible, ± The error resulting from
0.02 mv.
temperature gradients is estimated as ± 0.5 mv based on observations
made with nonisothermal cell conditions.
The important thermodynamic properties AG , AS° and AH can be
derived from Equation [3]. These are

AG°(kcal/mole) = -nFE° = -50.863 + 0.0296 t — 0.33x10 -3 t-' [4]

AS° (gibbs/mole) = -d(AG°)/dT = -29.6 + 0.66 t [5]

AH°(kcal/mole) = AG° + TAS° = -58.948 + 0.1813 t +

0.33x10-3 t2. [6]

Table 1 contains values of these properties and accepted thermodynamic

values (10). Although the quadratic equation for AG° allows the esti-
mation of AC° p the confidence limits in the regression coefficients are
,

too large to provide reliable values.

 Chemistry 125

% L
10
TEMPERATURE
20 30 40 50
(°C)
Figure 1. Plot of experimental emfs for the cell Zn/ZnSO',( 1m)//CuSOj,( lu) /Cu. The
curve is the least squares equation.

Table 1. Thermodynamic values for the Daniell cell calculated from

Equations [3] -[6].

t (°C) E° (volt) AG° (kcal/mole) AS° (gibbs/mole) AH° (kcal/mole)

1.1028 —50.863 —29.6 —59.948

10 1.0971 —50.600 —23.0 —57.168
20 1.0929 —50.403 —16.4 —55.354
25 1.0913 — 50.329 (- -50.71)* — 13.1(--3.73)* — 54.621(--51.82)*
30 1.0901 —50.272 —9.8 —53.806
40 1.0887 —50.207 —3.2 —52.224

* The values given in parentheses are accepted thermodynamic values

(10).

Discussion

The calculated value of E° at 25° C from Equation [3] of 1.0913 ±

0.0026 v lower than the predicted value of 1.100 v by roughly 9 mv.
is

Reported values for E° at 15° C of 1.09337 by Cohen, Chattaway and

Tombrock (2) for a cell consisting of amalgamated electrodes in satu-
rated solutions and 1.0962 v (average) by Jahn (4) give similar differ-
ences between experimental and calculated values.
126 Indiana Academy of Science

The value of dE°/dT at 25° C calculated from Equation [3] is

—0.289 mv/deg. This value compares favorably to 0.429 mv/deg —
reported by Cohen, Chattaway and Tombrock (2), to 0.2 mv/deg re- —
ported by Rosset (9), and to —
0.182 mv/deg listed by deBethune and
Loud (3) as an experimental determination. These results are consider-
ably more negative than —
0.083 mv/deg as predicted by combining half
cell values of dE°/dT based on thermodynamic values given by deBethune
and Loud (3). The thermodynamic value for AS° in Table 1 corresponds
to —0.0809 mv/deg (10). Considering the similarities of the species in-
volved in the cell, the smaller thermodynamic values appear to better
express AS for the reaction than do the electrochemical values.

The source of the discrepancies, 9 mv and 0.2 mv/deg, probably

results from the last two terms in the expressions for the overall cell
potential, Equation [1], and the corresponding temperature coefficient

dE(exp)
-dT- =
dE°
^-^ R
(lnQ +T
d In

-l^
Q
)+ "^
dE(LJ)
[7]

which are present because of experimental conditions.

As mentioned earlier, the emf contribution of the second term in

Equation [1] is negligible at 25° C and so the 9 mv must be the result
of the two liquid junction potentials for the ZnS0 4 -KCl and KCl-CuS0 4
interfaces. Maclnnes (5) gives the following general equation

soln II

soln I

which must be applied at each interface. Unfortunately Equation [8]

cannot be integrated directly, but Maclnnes (5) describes two approxi-
mate methods for obtaining values of E(LJ).

The first approximation is based on the assumptions that ionic

mobilities are independent of concentration and that activities and
concentrations are equal. This results in

fli/z >^
RT±ln Wz,) (Ci —
-^>
Ci) nC./ii
^
' i i

E(LJ)=f In
f-^f- [9]
n Mi (Ci"— c!) nC ^i

where /m is the ionic mobility. Using data from Milazzo (7) for Equation
[9], one obtains

E(LJ) = E(LJ) + E(LJ) = —1.3 + 1.6 = 0.3 mv

ZnSO, — KC1 KC1— CuSO^
which is considerably less than the 9 mv and has the incorrect sign.

The second technique for obtaining E(LJ) values from Equation [8]
is by graphical integration of values of t-f/Cf plotted against Cf between
the appropriate concentrations. Following the procedure given by
Maclnnes (5) and assuming that the activities of the sulfate ion in 1m
 Chemistry 127

ZnS0 4 and in lM CuS0 4 are equal and that the changes in the activities
-
of Cl~, SOj' and K 2 S0 4 passing from the salt bridge to 1m ZnS0 4 and
lM CuS0 4 to be identical, the total liquid junction potential is given by

sat KC1

Ir
Zn ~ +
E(LJ) = — ~-
RT O
Vj q~
'

~t~
d(C
ZnS0 4
f
ZnSO,
)

/" V ZnS0 4 ZnS0 4

1m ZnS0 4

sat KC1
{
Cu^+
S C f
d (C
CuS0 4 •
CuS0 4 >7
/
CuS0 4 CuS0 4
1m CuS0 4 Llt'J

Using available data (5)-(8) and making the approximation that the
activity coefficients for the components of mixtures to be equal to the
activity coefficients for the pure components at the concentration of the
component, the graphical integration of Equation [10] gives

E(LJ) = —38.7 + 38.3 = —0.4 mv.

Although the magnitude is low, the sign of E(LJ) is correct and if

proper data were available, Equation [10] should predict values more in
line with the 9 mv. It is felt, therefore, that the 9 mv difference can be
accounted for by the value of E(LJ).

From data temperature dependence of the limiting

reflecting the
ionic conductances (7), Equation [9] can be used to estimate dE(LJ)/dT.
For a 2% increase in the ionic mobilities for an increase in temperature
of one degree, the value of E(LJ) increases by 0.1 mv. Thus a significant
portion of the 0.20 mv/deg discrepancy is accounted for by the dE(LJ) /dT
term of Equation [7]. No attempt was made to estimate dE(LJ)/dT
from Equation [10] because of the uncertainty in the assumptions regard-
ing the choice of data.

In conclusion, the values of E°, AG°, AS° and AH given in Table 1

are valid for the experimental Daniell cell which includes, by necessity,
a liquid junction.

Acknowledgment
Two of the authors (SBB and RES) were participants in the Institute
for High School Chemistry Teachers sponsored by the National Science
Foundation at Indiana University.
128 Indiana Academy of Science

Literature Cited
1. Bennett, C. A., and N. L. Franklin. 1954. Statistical Analysis in Chemistry and
the Chemical Industry. John Wiley and Sons, New York. 724 p.

2. Cohen, E., T. D. Chattaway, and W. Tombrock. 1907. Zur thermodynamik der

normalelemente. Z. Physik. Chem. 60:706-727.

3. deBethune, A. J., and N. A. Swendeman Loud. 1964. Standard Aqueous Electrode

Potentials and Temperature Coefficients at 25°C. Clifford A. Hampel, Skokie, 111. 19 p.

4. Jahn, H. 1886. Uber die Beziehung von chemischer Energie und Stromenergie
galvanischer Elemente. Wied. Ann. 28 :21-28.

5. MacInnes, D. A. 1961. The Principles of Electrochemistry. 2nd. ed. Dover Publica-

tions, New York. 478 p.

6. Maron, S. H., and C. R. Prutton. 1958. Principles of Physical Chemistry. The

Macmillan Company, New York. 789 p.

7. Milazzo, G. 1963. Electrochemistry. Elsevier Publishing Company, New York. 708 p.

8. Robinson, R. A., and R. H. Stokes. 1949. Tables of Osmotic and Activity Coefficients
of Electrolytes in Aqueous Solution at 25°C. Trans. Faraday Soc. 45 :612-624.

9. Rosset, G. 1904. Daniell Cell as Standard Cell for Technical Purposes. Electrochem.
Ind. 2:246.

10. Rossini, F. D., D. D. Wagman, W. H. Evans, S. Levine, and D. Jaffe. 1952. Selected
Values of Chemical Thermodynamic Properties. Nat. Bur. Stand. Circ. 500. 1268 p.

11. Young, J. A., and J. G. Malik. 1969. Chemical Queries. J. Chem. Educ. 46:227-228.