Skip to main content
Log in

Catalyst Layer Performance in PEM Fuel Cell: Analytical Solutions

  • Published:
Electrocatalysis Aims and scope Submit manuscript

Abstract

We report the most general analytical solutions to a macrohomogeneous model for performance of the cathode catalyst layer (CCL) in PEM fuel cell. Equations for the shapes of local parameters through the CCL thickness and for the CCL polarization curve are derived for the regime of mixed oxygen- and proton-transport limitations. In the limit of ideal oxygen transport, an equation for the CCL polarization curve valid in the whole range of currents is obtained. Good agreement between analytical and exact numerical results is demonstrated.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4

Similar content being viewed by others

Notes

  1. A reasonable estimate for the derivative is \({\partial {\tilde{j}}}/{\partial {\tilde{x}}} \simeq \tilde{j}_0\). With \(\tilde{c} \leq 1\), we see that in Eq. 10, unity under the square root can be neglected if \(\tilde{j}_0 > 1/\varepsilon^2\). In PEMFCs, ε ≃ 103; with the characteristic current density j * ≃ 1 A cm − 2, this means that this approximation works well for current densities above 10 − 3 mA cm − 2.

  2. Uniform over the cell surface, GDL flooding or membrane drying could be taken into account by simple scaling of the parameter D b or R, respectively.

References

  1. Adzic RR et al (2007) Top Catal 46:249

    Article  CAS  Google Scholar 

  2. Bernardi DM, Verbrugge MW (1991) AIChE J 37:1151

    Article  CAS  Google Scholar 

  3. Springer TE, Zawodzinski TA, Gottesfeld S (1991) J Electrochem So 138:2334

    Article  CAS  Google Scholar 

  4. Frumkin AN, Bagotsky VS, Iofa ZA, Kabanov BN (1952) Kinetics of electrode processes. Izd. Mosk. Univ., Moscow

  5. Perry ML, Newman J, Cairns EJ (1998) J Electrochem Soc 145:5

    Article  CAS  Google Scholar 

  6. Eikerling M, Kornyshev AA (1998) J Electroanal Chem 453:89

    Article  CAS  Google Scholar 

  7. Wang Y, Feng X (2008) J Electrochem Soc 155:B1289

    Article  Google Scholar 

  8. Wang Y, Feng X (2009) J Electrochem Soc 156:B403

    Article  Google Scholar 

  9. Kulikovsky AA (2010) Electrochim Acta 55:6391

    Article  CAS  Google Scholar 

  10. Eikerling M (2006) J Electrochem Soc 153:E58

    Article  Google Scholar 

  11. Ioselevich AS, Kornyshev AA (2001) Fuel Cells 1:40

    Article  CAS  Google Scholar 

  12. Kulikovsky AA (2010) Analytical modelling of fuel cells. Elsevier, Amsterdam

    Google Scholar 

  13. Kulikovsky AA (2009) Electrochem Electochem Solid State Lett 12:B53

    Article  Google Scholar 

  14. Kulikovsky AA (2012) J Electroanal Chem 669:28

    Article  CAS  Google Scholar 

  15. Shen J et al (2011) J Power Sources 96:674

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to A. A. Kulikovsky.

Appendices

Appendix

Here, we report the details of the derivation of Eq. 15. Integrating Eq. 13 once, we get

$$ \dfrac{\partial {\tilde{j}}}{\partial {\tilde{x}}} + \dfrac{\tilde{j}^2}{2} - \gamma\tilde{j} = \left.\dfrac{\partial {\tilde{j}}}{\partial {\tilde{x}}} \right|_1 \equiv -\dfrac{\beta^2}{2} $$
(30)

where β is constant. Equation 30 can, in principle, further be integrated; however, this leads to rather cumbersome expressions. Fortunately, smallness of γ allows us to employ asymptotic technique. We write

$$ \tilde{j}\simeq \tilde{j}^0 + \gamma\tilde{j}^1 $$
(31)

where \(\tilde{j}^0\) is the leading-order solution and \(\tilde{j}^1\) is the first-order correction. Substituting Eq. 31 into Eq. 30, collecting the terms with like powers of γ and neglecting the terms with γ 2, we get equations for \(\tilde{j}^0\) and \(\tilde{j}^1\):

$$ \dfrac{\partial {\tilde{j}^0}}{\partial {\tilde{x}}} + \dfrac{\left(\tilde{j}^0\right)^2}{2} = -\dfrac{\beta^2}{2}, \quad \tilde{j}^0(1) = 0 \\ $$
(32)
$$ \dfrac{\partial {\tilde{j}^1}}{\partial {\tilde{x}}} + \tilde{j}^0\tilde{j}^1 = \tilde{j}^0, \quad \tilde{j}^1(1)=0 $$
(33)

The solution to Eq. 32 is

$$ \tilde{j}^0 = \beta\tan\left(\dfrac{\beta}{2}(1-\tilde{x}) \right) \ \mbox{.} $$
(34)

With this \(\tilde{j}^0\), the solution to Eq. 33 reads

$$ \tilde{j}^1 = - \dfrac{1}{2} \left( 1 + \tan^2\left(\dfrac{\beta}{2}\bigl( 1-\tilde{x}\bigr) \right)\right) \bigl( 1 - \cos\left(\beta(1-\tilde{x})\right)\bigr) $$
(35)

The oxygen concentration and parameter γ can now be calculated using the leading-order solution \(\tilde{j}^0\) in equation for \(\tilde{c}\). Substituting \(\tilde{j}=\tilde{j}^0\) into Eq. 8, we get

$$ \tilde{D}\dfrac{\partial {\tilde{c}}}{\partial {\tilde{x}}} = \tilde{j}_0 - \beta\tan\left(\dfrac{\beta}{2}(1-\tilde{x}) \right), \quad \tilde{c}(1)= \tilde{c}_1 $$
(36)

where \(\tilde{c}_1\) is the oxygen concentration at the CCL/GDL interface. Solving this equation, we find

$$ \tilde{c} = \tilde{c}_1 - \dfrac{\tilde{j}_0}{\tilde{D}}\left( 1 - \tilde{x} \right) + \dfrac{1}{\tilde{D}} \ln\left( 1 + \tan^2\left(\dfrac{\beta}{2}\bigl( 1-\tilde{x}\bigr) \right)\right) $$
(37)

Calculating \({\partial {\ln\tilde{c}}}/{\partial {\tilde{x}}}\) and setting \(\tilde{x}=1\) in the resulting expression, we find

$$ \gamma = \dfrac{\tilde{j}_0}{\tilde{D}\tilde{c}_1} $$
(38)

Thus, the analysis above is valid provided that \(\tilde{j}_0/(\tilde{D}\tilde{c}_1) \ll 1\).

The approximate solution to Eq. 13 is given by Eq. 31 with \(\tilde{j}^0\), \(\tilde{j}^1\), and γ given by Eqs. 34, 35, and 38, respectively. Parameter β can be found from the condition \(\tilde{j}(0) = \tilde{j}_0\); with Eq. 31, this leads to the following equation:

$$ \beta\tan\left(\dfrac{\beta}{2}\right) - \dfrac{\tilde{j}_0\bigl( 1 - \cos\beta \bigr)}{2\tilde{D}\tilde{c}_1} \left( 1 + \tan^2\left(\dfrac{\beta}{2}\right)\right)= \tilde{j}_0 $$
(39)

The shape of local overpotential can now be calculated from Eq. 6:

$$ \tilde{\eta} = {\rm{arcsinh}}\left( - \dfrac{\varepsilon^2}{\tilde{c}}\dfrac{\partial {\tilde{j}}}{\partial {\tilde{x}}} \right) \mbox{.} $$
(40)

Using here Eqs. 31, 34, 35, and 37, we come up with Eq. 15.

Nomenclature

\(\tilde{~}\) :

Marks dimensionless variables

b :

Tafel slope (V)

c :

Oxygen molar concentration (mol cm − 3)

c ref :

Reference oxygen molar concentration (mol cm − 3)

D :

Oxygen diffusion coefficient in the CCL (cm2 s − 1)

D * :

Characteristic oxygen diffusivity (cm2 s − 1), Eq. 5

D b :

Oxygen diffusion coefficient in the GDL (cm2 s − 1)

F :

Faraday constant

j 0 :

Cell current density (A cm − 2)

j :

Local proton current density in the CCL (A cm − 2)

j * :

Characteristic current density (A cm − 2), Eq. 5

i * :

Volumetric exchange current density (A cm − 3)

l b :

GDL thickness (cm)

l t :

Catalyst layer thickness (cm)

n :

Number of electrons transferred in the half-cell reaction

x :

Coordinate across the CCL (cm)

Subscripts

0:

Membrane/CCL interface

1:

CCL/GDL interface

b :

GDL

t :

Catalyst layer

*:

Characteristic value

Greek

β :

Dimensionless parameter, solution to Eq. 39, or Eq. 19

γ :

Small parameter, Eq. 14

η :

Overpotential (V)

η 0 :

Overpotential at x = 0, total voltage loss in the CCL (V)

ε :

Dimensionless Newman’s reaction penetration depth, Eq. 9

σ t :

CCL ionic conductivity (Ω  − 1 cm − 1)

Rights and permissions

Reprints and permissions

About this article

Cite this article

Kulikovsky, A.A. Catalyst Layer Performance in PEM Fuel Cell: Analytical Solutions. Electrocatalysis 3, 132–138 (2012). https://doi.org/10.1007/s12678-012-0091-4

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s12678-012-0091-4

Keywords

Navigation