Fuel cell durability has been a research focus towards successful commercialization of hydrogen-powered fuel cell vehicles. Voltage degradation that defines cell durability must be understood to design the fuel cell system for automotive application and to mitigate voltage loss during cell operation. Voltage degradation is irreversible if it results from permanent damage of cell components. Reversible degradation is caused by certain contaminants that deactivate catalyst surface and can be recovered by contaminant removal. Previous studies indicate that adsorption of mobile anions such as sulfate generated by membrane chemical degradation causes reversible voltage degradation;¹ and fixed anions like sulfonate group in the electrode ionomer side chain can also adsorb onto the platinum surface, thus reducing catalyst specific activity at dry operating conditions.² The purpose of this work is two-fold: (1) to further identify reversible degradation mechanisms by design of experiments; and (2) to develop a mathematical model for assisting development of voltage recovery strategies.

We have used a design of experiments (DoE) to quantify the effect of cathode Pt loading, membrane type (and thickness), and Cerium (Ce) level in membrane. A number of 53 cm² cells were built using a membrane-electrode-assembly (MEA) made of different membranes, cathodes with various Pt loadings, and identical anodes with Pt/Gr-C catalyst. Other cell components were kept the same. A standardized test protocol which includes a voltage recovery procedure (at low V, low T, and over 100%RH) was employed. Cell polarization curves were measured before and during voltage recovery. Water was collected at cell outlet to calculate cell water balance and analyze anion species.

Figure 1 compares the reversible voltage losses with different level of Ce used to mitigate membrane chemical degradation, showing a significantly higher reversible voltage loss for MEAs with unmitigated membrane, and nearly the same cell performance is achieved after voltage recovery. The fluoride release rate (FRR) corresponding to Ce level is greatly reduced with a mitigated membrane. However, the related reversible voltage loss does not decrease further as Ce level is doubled from current level, implying that a portion of reversible voltage loss should be due to fixed anion adsorption. The DoE results show clear difference in the reversible voltage loss between Ce-mitigated and non-mitigated cases, suggesting that the reversible voltage degradation could be de-convoluted into two parts: the membrane chemical-degradation-rate-dependent portion (i.e. sulfate adsorption) and the chemical-degradation-rate-independent portion (or the adsorption of sulfonate end group from electrode ionomer).

We have developed a mathematical model to simulate the voltage recovery process considering that sulfate anion and sulfonate end group adsorption onto Pt surface takes place independently and hence effectively reduces the electrochemical active area (ECA). Sulfonate adsorption-desorption kinetics is derived from the most recent experimental study showing continuous decrease in specific activity when an electrode operates at dry conditions,² which is strongly dependent on electrode potential and relative humidity (RH). The kinetics for sulfate anion adsorption-desorption is obtained by utilizing the published (bi)sulfate coverage on a Pt(1,1,1) electrode as a function of electrode potential and sulfate concentration in a solution of HClO₄ and H₂SO₄ mixture,³ with one adjustable parameter to be determined by fitting to the measured reversible voltage loss with respect to recovery time. Sulfate anion transport is governed by the dilute-solution theory, or the Nernst-Planck equation. It is assumed that sulfate anion leaves the electrode only with liquid water during voltage recovery. Initial sulfate concentration and sulfate coverage are evaluated by the correlation of reversible voltage loss in terms of sulfate coverage obtained from experiments. The reduction in ECA due to sulfate and sulfonate adsorption onto the Pt surface has two effects on cell performance: one is the kinetic effect reflecting a lower specific activity; and the other relates to oxygen transport local to Pt/ionomer interface, which shows a higher impact on cell voltage with increasing current density.

Figure 2 shows that the model agrees well with the measured recovered voltage loss during an 8-hour voltage recovery. As cell performance at high load is sensitive to the amount of Pt embedded in the carbon support particles, accurate catalyst morphology information is required for the model input. With further improvement, the model can be used to assist in developing voltage recovery strategies.

References

1. J. Zhang, B. A. Litteer, F. D. Coms, and R. Makharia, J. Electrochem. Soc., 160, F1067 (2013).
2. S. Jomori, K. Komatsubara, N. Nonoyama, M. Kato, and T. Yoshida, J. Electrochem. Soc., 160, F1067 (2013).
3. A. Kolica and A. Wieckowski, J. Phys. Chem. B, 105, 2588 (2001).