Polymer electrolyte membrane fuel cells (PEMFC) are promising candidates for the replacement of internal combustion engines. Still, alongside the necessary enhancement of the mass activity of cathode catalyst material, the performances of the PEMFC cathode must be improved in the high current densities range where the O₂ reactant mass transport towards the catalytic site is limiting for the reaction rate [1]. High current density performance of the PEMFC strongly depends on the ionomer distribution in the catalytic layer constituting the electrode, which depends on the solvent used for the catalyst ink preparation [2]. From different studies, it has been observed that the nature of the solvent composition modifies the resulting structure and size of the dispersed ionomer agglomerates in the catalyst ink [3–5]. The main objective of our investigation is to resolve in more detail the influence of the ink solvent composition on the PEMFC performances. For this purpose, the agglomerate size distribution of the 1100 EW Nafion^® ionomer dispersed into various solvents was determined at first by dynamic light scattering (DLS) in a systematic fashion and found to be strongly influenced by the kind of alcohol (isopropanol, ethanol and methanol) and its concentration in water. Then, in order to attempt a correlation between the agglomerate size of the ionomer and the MEA performances, cathode catalyst layers were prepared from inks with these alcohols at different concentrations, while keeping all the other parameters in the CCM preparation unchanged. Electrochemical impedance spectroscopy (EIS) and O₂ mass transport resistance measurements (O₂-MRT) were performed in addition to the recording of single cell polarization curves. The latter were modeled following the equation suggested by Larminie et al. [6] to discriminate the kinetic, ohmic and mass transport overvoltages (losses).

The ink-solvent compositional characteristics significantly influence the O₂ reactant mass transport overvoltages. An analog equivalent circuit including a Nernst impedance element to account for the finite O₂ diffusional limitations at the cathode catalyst layer was fitted to the experimental EIS data. The Warburg parameter scales very well with the O₂ transport resistance through the catalytic layer derived from O₂-MRT measurements, this strengthening the robustness of the overall set of experimental data. As the fraction of ionomer agglomerates in the range 50-500nm progressively increases in the catalyst ink up to ca. 37-45%, the losses attributed to the limited O₂ mass transport to the catalytic sites are lowered by down to ca. 70% at 1.76 A∙cm^-2 current load under the given cell testing conditions, independently of the nature of the alcohol and its relative content. For the higher fraction of ionomer agglomerates in this size range, no further change is observed in the O₂ transport loss parameter. These results highlight quantitatively the significant influence of the size distribution of ionomer agglomerates on the cell performances under high current loads and suggest a convolution with other parameters, such as, e.g., the porous structure of the catalyst layer, that has yet to be quantitatively asserted.

References

[1] A. Kongkanand, M.F. Mathias, J Phys. Chem. Lett. 7 (2016) 1127–1137.

[2] A. Orfanidi, P.J. Rheinländer, N. Schulte, H.A. Gasteiger, J. Electrochem. Soc. 165 (2018) F1254-F1263.

[3] T.T. Ngo, T.L. Yu, H.-L. Lin, Journal of Power Sources 225 (2013) 293–303.

[4] S.A. Berlinger, B.D. McCloskey, A.Z. Weber, J. Phys. Chem. B 122 (2018) 7790–7796.

[5] M. Yamaguchi, T. Matsunaga, K. Amemiya, A. Ohira, N. Hasegawa, K. Shinohara, M. Ando, T. Yoshida, J. Phys. Chem. B (2014) 141210091239007.

[6] J. Larminie, A. Dicks, Fuel Cell Systems Explained, Second Edition.