Is it worth investigating the effect of temperature on EMF in a galvanic cell? I suspect the change will be very very small and maybe impossible to test precisely in a high school lab.
This article said the variation of E with temperature can be find with E(T2)=E(T1)+dE/dT(T2-T1) where dE/dT=R/nF (lnQ) but how is this different to using the Nernst equation?


1 Answer 1


TL;DR; version:

Testing the EMF dependence on temperature does make sense, because it is not the matter just of the term $\dfrac{RT}{nF} \ln {Q}$. But you cannot use the Nernst equation directly as it does not explicitly enumerate the temperature dependence of standard reduction potentials.

The derivative of $\text{EMF}$ by absolute temperature depends on:

  • the current component thermodynamic activities(concentrations,pressures, fugacities) ($\dfrac{RT}{nF} \ln {Q}$)
  • the reaction entropy ( $\Delta_\text{r} S^{\circ}$)
  • the difference of the heat capacity of products and reactants ($\Delta C_p$)

Detailed evaluation:

\begin{align} \text{EMF}(T) &= \Delta E^{\circ}(T) - \frac{RT}{nF} \ln{Q} =\\ &= - \frac{\Delta_\text{r} G^{\circ}(T)}{nF} - \frac{RT}{nF} \ln{Q} =\\ &= - \frac{\Delta_\text{r} G^{\circ}(T) + RT \ln{Q} }{nF}\\ &= - \frac{\Delta_\text{r} H^{\circ}(T) - T \left( \Delta_\text{r} S^{\circ}(T) - R \ln{Q}\right) }{nF}\\ &\approx - \frac{\Delta_\text{r} H^{\circ}_{T_1} +\Delta C_p(T - T_1) - T \left( \Delta_\text{r} S^{\circ}_{T_1} + \Delta C_p \ln {\frac{T}{T_1}} - R \ln{Q}\right) }{nF} \tag{1} \end{align}

The key influence has the parameter $\Delta C_p$, what is the difference between the total heat capacity of reaction products and reactants. The approximation in the last step assumes constant molar heat capacities of involved components. If such a condition is not met, things would get complicated because of the need of integration of the functions $\int_{T_1}^{T_2}{\Delta C_p(T)\text{d}T}$ and $\int_{T_1}^{T_2}{\dfrac{\Delta C_p (T)}{T}\text{d}T}$

$$\frac{\text{d}(\text{EMF}(T))}{\text{d}T} \approx -\frac 1{nF} \left( \Delta C_p - \left( \Delta_\text{r} S^{\circ}_{T_1} + \Delta C_p \ln {\frac{T}{T_1}} - R \ln{Q}\right) - T \left( \frac{\Delta C_p }{ T} \right) \right)= \\ = \frac 1{nF} \left( \Delta_\text{r} S^{\circ}_{T_1} + \Delta C_p \ln {\frac{T}{T_1}} - R \ln{Q}\right) \tag{2} \label{2}$$

In the case where $\Delta C_p \approx 0$, the equation gets simplified to:

\begin{align} \text{EMF}(T) = - \frac{\Delta_\text{r} H^{\circ}_{T_1} - T \left( \Delta_\text{r} S^{\circ}_{T_1} - R \ln{Q}\right) }{nF} \tag{3} \end{align}

$$\frac{\text{d}(\text{EMF}(T))}{\text{d}T} = \frac 1{nF} \left( \Delta_\text{r} S^{\circ}_{T_1} - R \ln{Q}\right) \tag{4} $$

As the direct conclusion of the equation \eqref{2} comes the TL;DR; answer.

  • $\begingroup$ Would it be valid instead of using the textbook iteration of the Nernst equation I use the quotient of the ionic activity constants instead of the equilibrium constant? I'm thinking of making a linear trend of the theoretical and see the correlation between that and the experimental values. $\endgroup$
    – Mr. Mister
    Commented May 16 at 1:54
  • $\begingroup$ I am not sure what you mean. You can go further and replace the standard reaction Gibbs energy by the -RT ln K, but then you may have trouble to express K = f(T). $\endgroup$
    – Poutnik
    Commented May 16 at 3:50
  • $\begingroup$ Remind me: can you use Le Chatelier's principle to predict how EMF changes with temperature? $\endgroup$
    – unstable
    Commented May 16 at 11:21
  • $\begingroup$ It can be, qualitatively. Quantitatively, you can use van't Hoff equation d(ln K)/dT = Delta_r H/(RT^2). $\endgroup$
    – Poutnik
    Commented May 16 at 11:43
  • $\begingroup$ Is it possible/good for me to calculate the theoretical d(EMF(T))/dT = (ΔH - ΔG)/nFT then compare this with my experimental d(EMF(T))/dT? Is this (home.moravian.edu/public/chem/public_html/daniell%20cell.pdf) an example of the proper literature values of ΔH and ΔG? $\endgroup$
    – Mr. Mister
    Commented May 21 at 5:03

Not the answer you're looking for? Browse other questions tagged or ask your own question.