Ex-situ study of PEMFC membrane degradation under coupled chemical/mechanical stresses

EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476235 A1114 Page 1/8 A1114 Ex-situ study of PEMFC membrane degradation under coupled chemical/mechanical stresses Assma El Kaddouri, Guillaume Etienne*, Kevin Mozet, Olivier Lottin Université de Lorraine, CNRS, LEMTA, F-54000, Nancy/France; *Contact corresponding authors: www.EFCF.com/ContactRequest Abstract Increasing the operating temperature of proton exchange membrane fuel cells (PEMFC) is an important step in accelerating their large-scale deployment, especially in heavy-duty transport. Higher temperature is expected to improve the reaction rates and, mostly, makes it possible to use smaller heat exchangers. However, higher material degradation rates are also expected, as the operating conditions are different. For these reasons, it is necessary to study the impact of elevated temperature on the properties and durability of the membrane in PEMFC. An ex-situ approach has been employed to study the evolution of the structure and properties before and after degradation. The aim of this approach is to evaluate the membrane state by decoupling the effect of other components. To do so, a home-made bench was developed, allowing us to induce either coupled or uncoupled chemical/mechanical degradation. The sample is placed in a cell through which a vaporphase hydrogen peroxide (H2O2) is introduced, while simultaneously being mechanically stressed to induce coupled chemical/mechanical degradation. After exiting the cell, the vapor flux is condensed for analysis. The bench enables us to control and therefore vary parameters such as the H2O2 concentration, the relative humidity of the vapor flux, the frequency and the amplitude of the compression cycle. Structural characterization with Fourier-transform infrared (FTIR) spectroscopy on the sample and analyses of the degradation solution by UV spectroscopy are performed after the degradation protocol. To begin, several tests have been performed on pretreated sulfonated (Poly ether ether ketone) (sPEEK) with various H2O2 concentrations and no constrain applied. After degradation, sPEEK sample were macroscopically brittle and the presence of structural damages have been observed by FTIR spectroscopy. However, no degradation products have been detected by UV spectroscopy, possibly indicating that the degradation products are trapped in the sample. The next step is the characterization of sPEEK samples after mechanical and coupled chemical/mechanical degradation. Other similar tests on PFSA membrane such as Nafion 211 are incoming to reach a better understanding of the difference between the two kinds of polymer in terms of durability. The aim of this study is to evaluate the impact of a vapor environment at 95°C, coupled or not with mechanical stresses, on the chemical and functional properties of different membranes. EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476235 A1114 Page 2/8 Introduction Proton Exchange Membrane Fuel Cells (PEMFC) are a promising way to decarbonate heavy-duty transportation, where the use of electric batteries is complicated for reasons of weight and compactness [1]. Several projects have already proved the interest of this technologies. For example, diesel trains have been adapted to run on fuel cells in regions where line electrification is complicated by rugged terrain [2],[3]. Similarly, the use of PEMFCs in buses to limit pollution in urban transport is expanding rapidly [4]. Therefore, it is essential to improve PEMFC system compactness. One way to do it is to operate at higher temperature. The higher difference of temperature between the cell and the environment allows the use of smaller heat exchangers, thus making the system more compact. Also, hydrogen is currently produced by vaporeforming, which induce hydrogen impurities, such as CO2, CO H2S, that are poisoning catalyst sites and lowering fuel cell performances. Therefore, increase cell operating temperature would allow a better tolerance to hydrogen impurities. Several research are already interested in increasing cell temperature. Intermediate and High-Temperature (IT and HT-PEMFC) operate respectively at 90-120°C and 140180°C. If some improvements are waited, such as a higher heat and water management as well as a better tolerance to hydrogen impurities [5],[6], some drawbacks have been identified, particularly in terms of durability [7],[8],[9]. There is therefore an important work to perform to ameliorate materials performance and durability. If Accelerated Stress Test (AST) are well developed to assess in-situ durability, there is not a lot of work focusing on component durability, particularly on the ionomer membrane alone at higher temperature. Lot of studies focused on the chemical durability with the well-known Fenton’s test, which consists in immersing the ionomer in a heated H2O2 solution, catalysed by metal cations [10]. To approach closer conditions to in-situ operation, Robert et al. [11] performed coupled mechanical-chemical ex-situ degradation tests to assess ionomer membrane durability at 80°C. Two types of commercials membranes were investigated: a 25 µm Nafion® NR211 - without any chemical stabilizing additive - and a 28 µm Nafion XL ePTFE-supported PFSA membrane doped with cerium as a chemical stabilizer. Increased fluoride emission rates from Nafion XL and NR-211 membranes were observed under a 5 MPa cyclic compressive stress in a serpentine cell exposed to H2O2 or Fenton’s solution. They also observed morphological changes without any changes in the chemical structure of the ionomer repetition unit - meaning the IEC remain constant - and concluded that chemical degradation is an unzipping mechanism. As the cell operation temperature is raised, there is a need to performed similar studies to understand the impact of temperature and water state. Moreover, it is both interesting and important to carry out the study using H₂O₂ in the vapor phase to better replicate the predominant conditions present in fuel cells. In this study, an experimental bench was built to study the coupled chemicalmechanical degradation of the ionomer membrane in the vapor phase. Due to a significant setup and optimization phase, the results obtained and presented here are limited to chemical degradation tests only (without coupled chemical-mechanical degradation) in the vapor phase. Chemical degradation tests were carried out on a perflurosufonic acid-based ionomer (Nafion NR211), and hydrocarbon-based ionomer: sufonated poly(ether-etherketone) (sPEEK). EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476235 A1114 Page 3/8 1. Experimental procedure Experimental bench and Membranes The experimental bench enables the generation of H2O2 in vapor phase (see figure 1), while allowing control of external parameters such as the relative humidity (RH), the cell temperature and mechanical conditions including static or cyclic stress and stress strength. The experimental bench consists of four stages. First, H₂O₂ vapor is generated by injecting a controlled flow of liquid H₂O₂ into a heated tube, with dry air or N₂ as a carrier gas. Simultaneously, a second gas stream is humidified via a water bubbler to adjust the relative humidity of the combined flow. The mixed vapor passes through a temperature-controlled cell where the membrane is subjected to mechanical stress, and the outlet vapor is then condensed to collect and analyse degradation products. Figure 1: schematic representation of the experimental bench A 5x5 cm² stainless-steel cell (figure 2), featuring 50 straight parallel channels, was designed to be identical to a reference cell used for fuel cell operation measurements. Figure 2: straight parallel channels cell and its dimensions This work focuses on the Fumapem® E730 poly(ether ether ketone) (sPEEK) membrane purchased from Fumatech, and pretreated according to a protocol developed in a previous study [12]. The results are compared to those of the PFSA LSC ionomer (NR211). According to the manufacturer, both membranes are unreinforced and free of stabilizing additives. These preliminary tests were carried out to evaluate the chemical degradation of both Straight parallel channels Number of channels 50 Channels depth 0.5 mm Channels width 0.6 mm Rib width 0.4 mm EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476235 A1114 Page 4/8 membranes under H2O2 vapor, without the addition of ferrous cations (i.e., no Fenton test) and at different H2O2 concentrations. 2. Results Degradation tests at high H2O2 concentration for 2h The first approach was to carry out a chemical degradation test at 95°C on NR211 and pre-treated sPEEK. A flow rate of 15 µL/min of a 45% H2O2 solution, generating an approximate concentration of 1000 ppm, was vaporised over a period of two hours. At the macroscopic scale, no visible degradation was observed for the NR211 membrane, whereas the sPEEK membrane showed significant damage and became highly brittle (see figure 3), making it unsuitable for FTIR spectroscopy analysis. A H2O2 concentration of 1000 ppm therefore appears to be excessively high for the sPEEK membrane. Consequently, a second test was performed using a lower concentration of 10% H2O2 solution, reducing the vaporphase concentration from 1000 ppm to approximately 200 ppm. In addition, as the sPEEK membrane was difficult to remove from the cell due to adhesion (‘glued’ membrane), gas diffusion layers (GDL) were inserted between the membrane and the stainless-steel cell - on both sides - in this second test as a precaution. (a) (b) Figure 3: pretretread sPEEK (a) and NR211 (b) after chemical degradation with a H2O2 vapor flux (1000 ppm) This second test also revealed significant macroscopic degradation of the sPEEK membrane, with partial fragmentation. Nevertheless, the membrane was sufficiently intact to permit analysis by FTIR spectroscopy. Following these initial degradation tests, the chemical structure of the NR211 and sPEEK membranes were characterised by IR spectroscopy (see figure 4) by taking two measurements per membrane at different locations (point 1 and point 2) to assess the uniformity of the degradation. The spectra were then normalised using a specific reference bond: CF2 at 1145 cm-1 for NR211 (see figure 3(a)) and C-O-C at 1221 cm-1 for sPEEK (see figure 3(b)), to track changes in peak intensities. sPEEK NR211 EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476235 A1114 Page 5/8 (a) (b) Figure 4: IR spectrum of pristine and degraded membranes for (a) NR211 (H2O2 vapour at 1000 ppm) and (b) sPEEK (H2O2 vapour at 200 ppm) At first, it is observed that for both membranes, the differences between point 1 and point 2 are not very significant, indicating a certain uniformity across the sample. The NR211 membrane shows (see figure 3 (a)) a slight variation in peak intensities at 1057 cm⁻¹ and 970 cm⁻¹; however, these changes are not significant enough to conclude that there has been any deterioration in the chemical structure of the membrane. In contrast, the sPEEK membrane (see figure 3 (b)) exhibits significant variations for several peaks associated with the chemical bonds of both the sulfonic group and the aromatic ring, indicating notable alterations in the membrane’s chemical structure. To push further the analysis of the FTIR spectrum, the desulfonation of the sPEEK ionomer after degradation was quantified using the degree of sulfonation (DS), based on the work of Al Lafi et al. [13]. Al Lafi et al investigated the evolution of the DS of sPEEK by analysing the intensity of various peaks in the FTIR spectrum. They found a linear correlation between the 1490 cm-1 peak, normalized by the 1600 cm-1 peak, which decreases linearly with the increase in DS following the relation: 𝐷𝑆 =(1.84 ± 0.04)− 𝐴1490 (0.0180 ± 0.0006) 𝑤𝑖𝑡ℎ 𝐴1490 =𝐼1490 𝐼1600 Applying this relation to our FTIR measurements revealed a reduction in the DS from 41% to approximately 30% (29.1% at point 1 and 30.3% at point 2), demonstrating the significant impact of chemical degradation on the chemical structure of sPEEK membrane. These findings highlight key differences in the degradation behaviour of the two membranes and help establish the conditions for future testing. Indeed, a vapor concentration of 1000 ppm H2O2 does not significantly degrade the chemical structure of the NR211 membrane. In contrast, 200 ppm H2O2 vapor causes severe degradation and fragmentation of the sPEEK membrane, indicating the need to apply lower concentrations for viable analysis. For NR211, adding Fe²⁺ ions will be necessary to catalyse radical formation and accelerate membrane degradation. Degradation tests on sPEEK membranes at various concentration and degradation time EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476235 A1114 Page 6/8 The effect of exposure time to H2O2 vapor at 95°C, as well as the investigation of two lower concentrations, were studied on the sPEEK membrane at a fixed flow rate of 20 µL/min. In these cases, the GDLs were removed after being observed as unnecessary at these lower concentration levels. The FTIR spectra of membranes exposed to 3% or 5% H2O2 solution (respectively 33 and 116 ppm in vapor phase) for 4 hours, and to 5% H2O2 for 6 hours (see figure 4) show a significant impact on their chemical structure. Figure 5: IR spectrum of pristine and degraded pretreated sPEEK at 3% H2O2 (33 ppm), 5% H2O2 (116 ppm) during 4h and 6h Using the same approach described above, we quantified the degree of sulfonation (DS) of sPEEK by FTIR spectroscopy. A decrease in DS of 6 to 9% was observed in the degraded membranes compared to the non-degraded membrane. On the other hand, surprisingly, the impact of concentration between 3% and 5% is minimal, showing only about a 1% DS difference. Similarly, increasing the exposure time from 4 to 6 hours results in a modest decrease in DS from 34% to 32% (see Table 1). It would be valuable to further investigate the impact of longer exposure times to confirm this trend. t (h) H2O2 (%) DS (%) pristine 41 4 3 35 5 34 6 5 32 Table 1: estimation of sPEEK DS by FTIR spectroscopy after chemical degradation test 4. Conclusions This study highlights significant differences in the chemical degradation behaviour of sPEEK and Nafion NR211 membranes under H₂O₂ vapor exposure. While the NR211 ,2 , , , ,2 2 2 2 cm (cm ) ristine H2O2 pa h H2O2 pa h H2O2 pa h , ,2 , , , , , , , 2 cm (cm ) , ,2 , , , , , , , 2 cm (cm ) EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476235 A1114 Page 7/8 membrane shows remarkable resistance even at high H₂O₂ concentrations (1000 ppm) with no significant chemical degradation, the sPEEK membrane is highly sensitive, undergoing substantial structural damage and desulfonation even at lower concentrations (around 200 ppm). The reduction in the degree of sulfonation (DS) from 41% to approximately 30% for sPEEK confirms the severe chemical impact of H₂O₂ vapor, alongside membrane fragmentation and loss of integrity. Furthermore, exposure of sPEEK membrane to lower concentration of 3% or 5% H2O2 solution for 4 hours also results in a significant decrease in DS. Surprisingly variations in exposure time and H₂O₂ concentration between 3% and 5% produce only minor changes in the DS. The results also indicate that longer exposure times should be investigated to better understand degradation kinetics. For NR211 membranes, accelerated degradation requires the addition of Fe²⁺ ions to catalyse radical formation, highlighting the distinct chemical stabilities of these two membrane types (Nafion and sPEEK). This work provides a foundation for future studies aimed at understanding degradation impact and mechanisms for membranes subjected to coupled chemical and mechanical stresses under vapor phase. References [1] L. ickaël, . Rebecca, G. Olivier, and L. ierre, ‘TranpLHyn: Transports lourds fonctionnant à l’hydrogène’, F Energies nouvelles, 2 22. [2] ‘Les premiers TER hydrogène dans vos gares à partir de 2 2 ’, Groupe SNCF. Accessed: May 26, 2025. [Online]. Available: https://www.groupesncf.com/fr/innovation/decarbonation-trains/ter-hydrogene [3] ‘ ireo lus H – The next generation of hydrogen trains’, Siemens obility Global. Accessed: May 26, 2025. [Online]. Available: https://www.mobility.siemens.com/global/en/portfolio/rolling-stock/commuter-andregional-trains/mireo/mireo-plus-h.html [4] ‘ remiers déploiements de bus électriques à hydrogène en France : Retours d’expérience.’, France Hydrogène. Accessed: May 26, 2025. [Online]. Available: https://www.france-hydrogene.org/premiers-deploiements-de-bus-electriques-ahydrogene-en-france-retours-dexperience/ [5] Y. Devrim, A. Albostan, and H. Devrim, ‘Experimental investigation of CO tolerance in high temperature E fuel cells’, nt. J. Hydrog. Energy, vol. , no. , pp. 2– 18681, Oct. 2018, doi: 10.1016/j.ijhydene.2018.05.085. [6] V. B. K. , G. Varghese, T. V. Joseph, and . Chippar, ‘Spatial analysis of CO poisoning in high temperature polymer electrolyte membrane fuel cells’, nt. J. Hydrog. Energy, vol. 46, no. 11, pp. 8179–8196, Feb. 2021, doi: 10.1016/j.ijhydene.2020.12.001. [7] Y. Nalbant, C. O. Colpan, and Y. Devrim, ‘Energy and exergy performance assessments of a high temperature-proton exchange membrane fuel cell based integrated cogeneration system’, nt. J. Hydrog. Energy, vol. , no. , pp. –3594, Jan. 2020, doi: 10.1016/j.ijhydene.2019.01.252. [8] A. Suzuki et al., ‘Evaluation for sintering of electrocatalysts and its effect on voltage drops in high-temperature proton exchange membrane fuel cells (HTE FC)’, nt. J. Hydrog. Energy, vol. 37, no. 23, pp. 18272–18289, Dec. 2012, doi: 10.1016/j.ijhydene.2012.09.016. [9] A. Chandan et al., ‘High temperature (HT) polymer electrolyte membrane fuel cells (PEMFC) – A review’, J. ower Sources, vol. 2 , pp. 2 –278, Jun. 2013, doi: 10.1016/j.jpowsour.2012.11.126. [10] F. C. Teixeira, A. . S. Teixeira, and C. . Rangel, ‘Chemical stability of new nafion membranes doped with bisphosphonic acids under Fenton oxidative conditions’, nt. J. EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476235 A1114 Page 8/8 Hydrog. Energy, vol. 48, no. 96, pp. 37489–37499, Dec. 2023, doi: 10.1016/j.ijhydene.2023.04.063. [11] . Robert et al., ‘A chemical-mechanical ex-situ aging of perfluorosulfonic-acid membranes for fuel cells: mpact on the structure and the functional properties’, J. Power Sources, vol. 520, p. 230911, Feb. 2022, doi: 10.1016/j.jpowsour.2021.230911. [12] . Daoudi et al., ‘ mpact of sulfonated poly(ether ether ketone) membranes pretreatments on their physicochemical properties and fuel cell performances’, J. olym. Sci., vol. 61, no. 9, pp. 761–776, 2023, doi: 10.1002/pol.20220649. [13] A. G. Al Lafi, ‘The sulfonation of poly(ether ether ketone) as investigated by twodimensional FT R correlation spectroscopy’, J. Appl. olym. Sci., vol. 2, no. 2, Jan. 2015, doi: 10.1002/app.41242. Keywords: EFCF2025, H2, Low-Temp. Fuel Cells & Electrolysers, Fuel Cells, 95°C, Ionomer membrane, Ex-situ Accelerated Stress Test, Chemical degradation, PFSA-LSC and sPEEK Remark: This work is licensed under Creative Commons Attribution 4.0 International