Cerium enhances the durability of polymer electrolyte membrane (PEM) fuel cells by scavenging reactive radical species which are generated during operation. However, during cell fabrication, conditioning, and discharge, Ce migrates through-plane between the PEM and catalyst layers (CLs) due to concentration and potential gradients.^1,2 In addition, we have observed in-plane Ce migration due to water gradients and also identified degradation of Ce-exchanged PEM side chains as another possible mechanism for Ce migration.³ Ce migration is detrimental because (1) its accumulation in the CL ionomer can diminish the electrode’s proton conductivity, which generates performance losses⁴; and (2) its depletion may leave an ionomer region more susceptible to radical attack. Therefore, it is critical to understand the relative influence of different migration mechanisms under a range of operating conditions in order to stabilize Ce in the PEM and localize it to areas of highest radical generation.

To understand the effects of potential gradients and relative humidity (RH) on Ce migration, ex situ experiments were performed using uncatalyzed Nafion® XL PEMs (DuPont) which contain ~6 μg/cm² ion-exchanged Ce. PEM specimens were operated in H₂ pump mode in a standard conductivity cell (BekkTech) at 80°C with 50% and 100% RH H₂. The evolution of Ce profiles was quantified using X-ray fluorescence (XRF). By comparing the resulting profiles at 4 C of charge transfer for the different potential and RH conditions (Figure 1a), we observe decreased Ce transference at low RH. Under these conditions, decreased PEM water content causes a disproportionate reduction in Ce conductivity relative to proton conductivity.

In these experiments, Ce ion mobility induced by a potential gradient leads to a concentration gradient, which, in turn, induces Ce diffusion in the opposite direction. Therefore, the profiles shown in Figure 1a arise from a combination of ion mobility and back-diffusion due to the resulting concentration gradient. In order to decouple these effects, a transient, 1-D model was developed based on Nernst-Einstein ion mobility and Fickian diffusion, in order to solve for the ion mobility and diffusion coefficients. Experimental and model results for the 2 V, 100% RH case are shown in Figure 1b. At 100% RH, both diffusion and ion mobility coefficients were determined to be an order of magnitude higher than at 50% RH.

In addition to migration within the PEM, Ce is stabilized in the cathode CL, likely in the CL ionomer and/or carbon CL supports.³ Ce accumulation in the cathode catalyst layer was measured to degrade MEA performance. Different mechanisms have been proposed for the performance loss, including increased proton resistance within the CL, and a reduction of oxygen reduction reaction kinetics. However, the effect of each mechanism on performance loss is difficult to quantify. The relative influence of Ce poisoning on the cathode CL performance will also be discussed.

Quantifying the different Ce migration mechanisms has provided a better understanding of the effects of Ce migration in the PEM and CL and the associated losses in operating performance. These results demonstrate that stable Ce compounds, which can be localized to areas of highest reactive radical generation, need to be developed to enhance PEFC durability without compromising performance.

Acknowledgements

This research is supported by the U.S. DOE Fuel Cell Technologies Office, through the Fuel Cell Performance and Durability (FC-PAD) Consortium; Fuel Cells Program manager: Dimitrios Papageorgopoulos. Prof. Ajay Prasad and Prof. Suresh Advani also acknowledge support from the University of Delaware’s Fuel Cell Bus Program. This program is funded by the Federal Transit Administration at the Center for Fuel Cell Research at the University of Delaware.

References

1. Stewart, S. M.; Spernjak, D.; Borup, R.; Datye, A.; Garzon, F. ECS Electrochem. Lett. 3, F19–F22 (2014).
2. Baker, A. M.; Mukundan, R.; Spernjak, D.; Advani, S. G.; Prasad, A. K.; Borup, R. L. ECS Trans. 75, 707–714 (2016)
3. Baker, A. M.; Mukundan, R.; Spernjak, D.; Judge, E. J.; Advani, S. G.; Prasad, A. K.; Borup, R. L. J. Electrochem. Soc. 163, F1023–F1031 (2016).
4. Banham, D.; Ye, S. Y.; Cheng, T.; Knights, S.; Stewart, S. M.; Wilson, M.; Garzon, F. J. Electrochem. Soc. 161, F1075–F1080 (2014).