During operational lifetime, proton exchange membrane fuel cells (PEMFCs) suffer from high performance losses. For the year 2025, the U.S. Department of Energy (DoE) set the durability target for PEMFC light-duty transportation applications at 5,000 hours.¹ To meet the automotive target, different aging protocols were established in order to simulate load cycle variations.² It is well known that load or voltage cycling induces substantial catalyst degradation due to an increasing loss of electrochemically active surface area (ECSA), which is known to be a main driver for the resulting performance penalties.³ Thus, an increasing interest exists in developing accelerated stress test (AST) protocols that can sufficiently describe the degradation of the different components of a membrane electrode assembly (MEA) during automotive application using shorter measurement times.

In this study, we report on a strategy to predict the performance degradation of an MEA, making use of a strongly accelerated voltage cycling based ASTs. For this, voltage cycling based ASTs were performed using 5 cm² MEAs with a 0.1 mg_Pt cm_MEA^-2 loaded Pt/C catalyst (TEC10V20E, Tanaka) for both cathode and anode. Voltage cycling was done under H₂/N₂ (200/75 nccm) at 80 °C, 95% RH, and ambient pressure using square wave profiles with a constant lower potential limit (LPL) of 0.6 V and different upper potential limits (UPLs) of 0.85, 0.95, and 1.0 V, with LPL/UPL hold times of 1, 2, or 8 s. Full characterization of the MEA at beginning-of-life and after each set of voltage cycling intervals was performed by: i) measuring H₂/O₂ and H₂/air polarization curves; ii) determining the ECSA by cyclic voltammetry and CO stripping; iii) conducting limiting current measurement to calculate the O₂ transport resistance (R_O2^total); and, iv) conducting electrochemical impedance spectroscopy (EIS) measurements under blocking conditions to quantify the proton conduction resistance in the cathode catalyst layer (R_H+^cath).

When the ECSA loss ranges between ≈20‑85%, it was found that the ECSA decreases approximately linearly when plotted against the logarithm of the number of cycles (see fig. 1), exhibiting higher slopes for procedures with higher UPLs and longer hold times, in agreement with the results from previous studies.^{3, 4} In addition, a direct and AST protocol independent correlation between the loss in total available surface area, i.e., the roughness factor (rf ≡ ECSA × Pt‑loading), and the individual loss contributions in H₂/air performance curves, namely kinetic (mass/specific activity for the oxygen reduction reaction (ORR)) and O₂ transport resistance contributions, was shown. This is due to the fact, that these individual voltage losses are not only highly affected by the cathode rf,^{3, 5} but that the Pt dissolution/redeposition mechanism seems to be identical for all of the here investigated voltage cycling ASTs, despite the largely varying UPLs and hold times. As a result, a universal correlation between the H₂/air performance losses and the rf deterioration was found throughout all ASTs. The important corollary of this finding is that the data from quickly degrading voltage cycling ASTs (i.e., with high UPL) can be used to project the H₂/air performance losses when cycling under less degrading conditions.

References:

1. "Fuel Cell Technologies Program Multi-Year Research, Development, and Demonstration Plan", U.S. Department of Energy, (2005, revision: 2012, accessed: 30/11/2021), https://www.energy.gov/eere/fuelcells/articles/hydrogen-and-fuel-cell-technologies-office-multi-year-research-development.
2. R. Petrone, D. Hissel, M. C. Péra, D. Chamagne, and R. Gouriveau, Int. J. Hydrog. Energy, 40 (36), 12489-12505 (2015).
3. G. S. Harzer, J. N. Schwämmlein, A. M. Damjanović, S. Ghosh, and H. A. Gasteiger, J. Electrochem. Soc., 165 (6), F3118-F3131 (2018).
4. A. Kneer, N. Wagner, C. Sadeler, A.-C. Scherzer, and D. Gerteisen, J. Electrochem. Soc., 165 (10), F805-F812 (2018).
5. T. A. Greszler, D. Caulk, and P. Sinha, J. Electrochem. Soc., 159 (12), F831-F840 (2012).

Acknowledgement:

We gratefully acknowledge funding from the German Federal Ministry for Economic Affairs and Energy (BMWi) under the funding scheme POREForm (funding number 03ET B027C).

Figure 1: Loss of normalized cathode ECSA over the course of voltage cycling. The aging procedure was done by voltage cycling between 0.6 V and different UPL (0.85, 0.95, and 1.0 V) using square wave modulation with varying LPL/UPL hold times (1, 2, and 8 s). The red dashed line describes the minimum ECSA value (≈15% normalized ECSA, corresponding to an rf of ≈12 cm_Pt cm_MEA^‑2) for that a linear trend of ECSA vs. log(cycles) was observed.