The success of recent deployments of automotive and stationary sources of energy utilizing proton exchange membrane fuel cells (PEMFCs) relies on a high performance under a broad range of environmental conditions. However, PEMFCs using Pt-based catalysts as electrode materials were shown to be very sensitive to many fuel and air contaminants, which typically originate from natural as well as anthropogenic emission sources [1]. The knowledge of the electrochemical behavior of impurities and their intermediates on Pt is a principal requirement for understanding their impacts on the oxygen reduction reaction (ORR), fuel cell performance, mitigation and recovery procedures. AC-impedance is a powerful in-situ technique for determination and characterization of different processes in electrochemical systems. In this work, PEMFC electrode reactions occurring under exposure to various airborne contaminants (SO₂, NO₂, VOCs) were studied by spatial AC-impedance, which gives locally resolved data over the electrode area, while a single cell only provides average values of measured parameters.

The experimental work was performed using a GRandalytics test station, a segmented cell system and a commercially available 100 cm² membrane/electrode assembly (MEA) [2]. The MEA was operated under galvanostatic control of the total cell current. Many airborne impurities including inorganic and organic compounds (SO₂, NO₂, propylene, acetylene, acetonitrile, isopropanol, benzene, naphthalene, bromomethane, methyl methacrylate, etc) were studied.

The electrochemical properties of the selected compounds are typically potential dependent. This should be taken into account when considering the ORR which occurs in parallel with contaminant chemical and electrochemical transformations. The effects of the studied pollutants on fuel cell impedance can be separated into three major groups: 1) the impurity does not significantly affect charge and mass transfer resistances, 2) the impurity increases charge and mass transfer resistances and, 3) the impurity leads to the appearance of a low frequency inductance.

The first type of behavior was found when the cathode was poisoned by isopropanol and SO₂ (at a steady state condition) (Fig. 1 a and c). Rotating ring/disk electrode (RRDE) data showed that isopropanol and SO₂ (at a sulfur coverage below 0.3) do not significantly change the ORR mechanism on Pt, which proceeds via a 4-electron pathway [3, 4]. Contaminant adsorption on Pt decreases the electrochemical surface area (ECSA) and the exchange current density (i₀) but it does not noticeably affect intrinsic Pt properties and ORR kinetics. AC-impedance is not sensitive to i₀ variations at high current operation and it does not detect any changes. However, a decrease in ECSA and i₀lead to a performance loss.

An increase in impedance response was observed for the majority of the studied contaminants and it is connected with an increase in the Tafel slope, a decrease in the proton conductivity and possible diffusion limitations due to foreign species adsorption on Pt (Fig. 1 b). RRDE studies clearly demonstrated that these contaminants cause an increase in the Tafel slope for ORR and noticeable H₂O₂ production, indicating a shift in O₂ reduction to combined 4- and 2-electron mechanisms [3]. This observation implies that the adsorbed contaminant not only decreases the ECSA but also alters the Pt intrinsic activity due to strong electronic interactions. The first and the second AC-impedance behavior groups are usually observed at potentials which are favorable to contaminant oxidation via electrochemical and chemical pathways.

A low-frequency inductance in AC-impedance spectra indicates electrochemical reactions involving adsorbed contaminant species and proceeding with the formation of intermediates on the Pt surface. Moreover, such an inductance means that the current signal follows a perturbation with a phase delay due to the slow relaxation of adsorbate coverage compared to the ORR [5]. The inductive loop was detected for PEMFCs exposed to C₂H₂, SO₂, CH₃CN and naphthalene at potentials where the electroreduction of contaminants can occur (Fig. 1 c).

A detailed discussion of AC-impedance data in connection to ORR mechanisms, contaminant reactions and poisoning propagation along the cathode channel length will be presented. In addition, anode and cathode poisoning processes will be compared.

ACKNOWLEDGEMENTS

The author gratefully acknowledges funding from ONR (N00014-12-1-0496) and DOE EERE (DE-EE0000467), and Hawaiian Electric Company support. The author thanks G. Randolf for system operation support and A. Kulikovsky for discussion.

REFERENCES

1. Y. Zhai, J. St-Pierre, M. Angelo, ECS Trans., 50(2), 635 (2012).
2. T.V. Reshetenko, K. Artyushkova, J. St-Pierre, J. Power Sources, 342, 135 (2017).
3. J. St-Pierre et al., in DOE Hydrogen and Fuel Cells Program - 2014 Annual Progress Report, DOE/GO-102014-4504, November 2014, pp. V-133-V-139.
4. Y. Garsany, O. Baturina, K. Swider-Lyons, J. Electrochem. Soc., 154, B670 (2007).
5. D.A. Harrington, B.E. Conway, Electrochim. Acta, 32, 1703 (1987)