Proton exchange membrane fuel cells (PEMFCs) are a promising power source in transportation and other portable devices¹. The low operating temperature allows for quick device utilization unlike high temperature batteries/fuel cells; which have to be heated to higher temperatures before operation is feasible^1,2. Recently, PEMFCs have been used in cars produced by Audi and Toyota.

The activation and concentration polarizations at the two electrodes is a large factor in determining the efficiency of the device. Electrochemical impedance spectroscopy (EIS) is a common technique to analyze kinetic and mass transport limitations at the electrodes in PEMFCs. However, fitting a spectrum to an equivalent circuit is an inverse problem, where multiple equivalent circuits describing vastly different kinetic processes can yield the same spectrum. Two electrode methods measure the impedance from both electrodes, thus the contributions from the anode and cathode are not resolved without deconvolution. In past studies individual electrode contributions have been deduced by comparing EIS scans done on symmetric electrodes (H₂/H₂) with the result obtained under normal fuel cell operating conditions (H₂/O₂)³. A three-electrode device, using a pseudo-reference electrode, can resolve the different electrode contributions. However, current methods utilizing pseudo-reference electrodes, are sensitive to the position of the reference electrode and the alignment of the anode and cathode^4,5.

Using a novel pseudo-reference electrode embedded in the electrolyte the anode and cathode contributions were resolved in EIS scans at open circuit and under a constant resistive load. An equation was derived to correct for the AC current which passed through the load resistor, in the case where a DC current passes through the cell. Summing the three-electrode results showed reasonable agreement with the two-electrode scan.

EIS spectra of the entire cell, anode-reference, and the cathode-reference is observed in figure 1a. The spectrum of the entire cell and the spectrum obtained by summing the anode-reference scan and the cathode-reference scan is observed in figure 1b. The error analysis is observed in figure 1c. The previously mentioned plots were obtained with the cell operating at 40⁰C with a DC current of ~300 mA and a cell potential of ~0.550 V. Comparison of the spectra under different resistive loads showed that both electrodes were kinetically controlled under the conditions tested.

Acknowledgements: This work was supported in part by the U.S. Department of Energy, Office of Basic Energy Sciences under Grant No. DE-FG02-06ER46086 and in part by the National Science Foundation under Grant No. NSF-CBET-1604008.

References:

1. J. Larminie and A. Dicks, Fuel Cell Systems Explained, Second., John Wiley & Sons, West Sussex, (2003).
2. Y. Shao, G. Yin, Z. Wang, and Y. Gao, J. Power Sources, 167, 235–242 (2007).
3. D. Malevich, E. Halliop, B. Peppley, J. Pharoah, and K. Karan, ECS Trans., 16, 1763–1774 (2008).
4. D. Gerteisen, J. Appl. Electrochem., 37, 1447–1454 (2007).
5. Z. Liu, J. S. Wainright, W. Huang, and R. F. Savinell, Electrochimica Acta, 49, 923–935 (2004).