Electrodes and Diaphragms for Alkaline Water Electrolysis above 130°C

EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476294 A1404 Page 1/9 B1404 Electrodes and Diaphragms for Alkaline Water Electrolysis above 130°C Karl Skadell*, Jakob Scholl, Artur Bekisch, Tino Jacob, Mihails Kusnezoff Fraunhofer IKTS, Arnstadt/Germany; *Contact corresponding authors: www.EFCF.com/ContactRequest Abstract Alkaline water electrolysis (AWE) is the leading technology for the production of green hydrogen. It relies heavily on alkali-resistant core components such as bipolar plates, current collectors, sealings, electrodes incl. electrocatalysts, and diaphragms. The latter is primarily based on polyphenylene sulfide (PPS). Zirfon© from Agfa-Gevaert NV, a recognized product, includes additional inorganic fillers. A significant challenge for Zirfon-type diaphragms is the exposure to temperatures (e.g., operation temperatures or hot spots) exceeding the decomposition temperature, typically between 100 °C and 130 °C under AWE conditions. A temperature-stable diaphragm could reduce degradation and – maybe more important – enhance the water electrolysis efficiency. In the industry-lead project AWEC++, Fraunhofer IKTS and partners are developing components and a stack that enables operation temperatures above 130 °C (above 25 bar). The consortium has the capability to manufacture electrodes, the diaphragm, bipolar plates as well as the stack and a demonstration electrolyzer (20 kW). The key component is a scalable, fully inorganic diaphragm which is produced starting from inorganic meshes. It is (i) stable in aqueous 35 % KOH solution at 200 °C and 35 bar for at least 7 days, (ii) has an average pore diameter of 300 nm (as determined via bubble point analysis), and (iii) exhibits in aqueous 35 % KOH an area specific resistance (ASR) of 1.4 Ω cm2 (thickness: 450 μm). In this project phase IKTS will complete a 5-level short stack with an active area of 100 cm2 with a PGM-free current density of 500 mA cm-2 at 2.1 V. Fig. 1: Cross section of a fully inorganic diaphragm Fig. 2: Design of a stack design aiming at 20 kW EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476294 A1404 Page 2/9 Introduction Alkaline water electrolysis (AWE) is the most most widely deployed electrolyzer technology in comparison to PEM and AEM water electrolyzers (AEMWE, PEMWE) as well as solid oxide electrolyzer cells (SOEC). In addition to its advantages, such as scalability, low cost per installed power, and reliance on abundant materials, AWE still lags behind PEMWE in terms of dynamic response and achievable current densities. One promising strategy to overcome this limitation is to increase the operating temperature beyond the current stateof-the-art of 80 °C or 90 °C. The most crucial bottleneck is the diaphragm which is not stable under higher temperature. It primarily based on polyphenylene sulfide (PPS). Zirfon© from Agfa-Gevaert NV, a recognized product, includes additional inorganic fillers. A significant challenge for Zirfon-type diaphragms is the exposure to temperatures (e.g., operation temperatures or hot spots) exceeding the decomposition temperature, typically between 100 °C and 130 °C under AWE conditions. A thermally stable diaphragm would not only reduce degradation but, perhaps more importantly, improve overall electrolysis efficiency. [1] This is precisely where the present project comes in. By applying advanced ceramic technologies and a novel membrane electrode concept, we aim to enable AWE operation at significantly elevated temperatures [2] This allows current densities to be increased. It shall be demonstrated in a 20-kW alkaline water electrolyzer above 25 bar and above 130 °C. 1. Scientific Approach Since classic diaphragms for alkaline water electrolysis (AWE) are based on polyphenylene sulfide (PPS), such as Zirfon© from Agfa-Gevaert NV, the organic polymers are a weak point in operation at higher temperatures. [3] To overcome this limitation, the use of polymers is avoided by employing a fully inorganic porous diaphragm. The patented [4] process for their preparation involves the deposition of ceramic material onto metal meshes by atmospheric plasma spray (APS). The work began with the conceptual design phase, which included material selection, component specification, and significant advances in the stack layout. This phase also contributed to the development of the process and instrumentation diagram (P&ID). Subsequently, the fabrication processes for both the electrodes and the ceramic diaphragm were refined, and electrochemical performance testing along with materials characterization was carried out. This was followed by design-for-manufacturing optimization of the stack components and the mechanical assembly of the full stack. Stack characterization and supporting simulation activities were also performed. The installation of a 20 kW alkaline electrolyzer system at the Fraunhofer IKTS site in Erfurter Kreuz is currently being prepared. 2. Experiments While Fraunhofer IKTS act as a junction within the partners consortium the tasks related to materials characterization, electrochemical analysis, stack design, and water electrolysis testing fall within its scope. The ceramic diaphragms are constructed by applying insulating ceramic layers via atmospheric plasma spraying (APS) on high-porosity metal substrates (e.g. sintered meshes or laminates). The metallic substrates are specifically tailored to provide mechanical robustness and optimized permeability for ionic transport, while supporting the brittle ceramic top-layer. The components were thoroughly analyzed by SEM imaging (scanning electron microscopy), porosimeter, and flow pressure drop analysis. The components were EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476294 A1404 Page 3/9 tested towards their degradation in an accelerated stress test: aqueous 35 % KOH solution at 200 °C and 35 bar for at least 7 days. Ceramic diaphragms are assembled in zero-gap cells and stacks using nickel-based porous electrodes. The electrochemical tests on cell and stack level were carried out at 60 °C and 80 °C, respectively, at atmospheric pressure. The testing protocol included galvanostatic polarization steps, electrochemical impedance spectroscopy as well as constants current measurements. The gas quality during the measurements were investigated by gas chromatography. The P&ID as well as a stack image is given in Fig. 3 and 4. Fig. 3: P&ID of the electrolyzer set up Fig. 4: Image of the used stack The total work is distributed over partners in the consortium of the publicly funded project AWEC++: Paul GmbH & Co KG (PACO) supplies the backbone metal meshes used both as porous electrodes and as the structural support for the new ceramic diaphragms. PACO’s highprecision nickel (and specialty alloys) weaves deliver outstanding electrical conductivity, chemical resistance, and tailored porosity—key for efficient ion transport and mechanical stability in harsh alkaline‐electrolysis environments. MUW Screentec GmbH handles all atmospheric plasma‐spray (APS) processes, depositing both insulating ceramic layers for the diaphragms and catalytic oxide coatings (e.g. LaCoO₃, BSCF) onto PACO’s meshes. Their APS expertise ensures dense, adherent coatings with precisely controlled pore sizes (~300 nm) and thicknesses (~450 µm), crucial for >130 °C operation without polymer degradation. Merkle CAE Solution GmbH provides multiphysics simulations covering mechanics, fluid dynamics, and electrochemistry to optimize cell and stack designs. Their 3D models predict flow distributions, pressure drops, temperature profiles, and current‐density uniformity, guiding experimental parameters and scale-up strategies. SITEC Industrietechnologie GmbH develops automated manufacturing methods for bipolar and end plates, employing laser cutting/joining and electrochemical machining (ECM) to produce precise flow-field patterns. Their automation workflows enable high-throughput, reproducible stack assembly, minimizing manual error and ensuring component alignment at scale. AP-Miniplant GmbH & Co KG integrates the IKTS-designed short stacks, PACO electrodes, MUW-coated diaphragms, and SITEC plates into a 20 kW demo plant. APM’s skid-mounted system includes KOH circulation, thermal management, gas‐separation modules, and a dynamic control system for variable renewable inputs, proving real-world operation at 130– 180 °C and 25–35 bar. EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476294 A1404 Page 4/9 3. Results 3.1 Coating of meshes – preparation of diaphragm and electrodes Different mesh substrates were investigated before the APS coating. 20 different samples were compared to each other for their application as a diaphragm substrate and as an electrode substrate. Different parameters were looked at like wire thickness, total thickness of the mesh, the mesh size, and the mesh typ. Those 20 could be reduced to five samples, which are listed below. Sample Mesh Typ Wire Thickness / µm Total Thickness / µm Mesh Size / µm A-GTD Smooth braid duplex + 310 o D-5L Five-shaft body ++ 510 + E-GTD Smooth braid duplex o 170 oF-KT Buyer interest o146 - T-5L Five-shaft body +++ 940 ++ The air volume flow of the substrates was determined to distinguish their flow characteristics. Therefore, the volume flow was measured perpendicularly and along to the mesh substrates (Fig. 5). The perpendicular and alongside flow is important for the electrode substrate because electrolyte flows through it. Therefore, a high volume flow is desirable because new reactants should be transported as fast as possible to the active sites. Based on the results, the samples D-5L and T-5L are used as electrode substrates. Fig. 5: Volume flow measurement with air at 10 mbar perpendicular and along the mesh and Fig. 6: Point welded 3D electrode substrate after the APS coating. Both substrates were point welded together to create a 3D electrode substrate with different mesh sizes. Before welding, the substrates were calendared to reduce the total thickness by 25 % of both substrates. Thus, the 3D electrode substrate has a thickness of 1.09 mm (Fig. 6). The substrate with the smaller mesh size is coated with the catalyst. The substrate for the diaphragm coating was determined from the samples A-GTD, E-GTD, and F-KT because of their small mesh size. Furthermore, the substrate thickness is also important since it contributes to the ion conductivity by determining the distance of anode and cathode to each other. Thus, the sample A-GTD was not further investigated. For the diaphragm coating, yttrium stabilized zirconia(IV)-oxide (YSZ) powders with different grain distribution are tested. For the electrode coating, Ni-, Al-, and Mo-powder are used EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476294 A1404 Page 5/9 and mixed to Ni-Al (50-50) and Ni-Mo-Al (50-5-45). The anode is coated with Ni-Al and the cathode is coated with Ni-Mo-Al. The diaphragm coating is applied to the front and back of the mesh samples with a thickness of 100 µm each. The catalyst coating for the electrode is applied only on one site with a thickness of 100 µm. After the coating, the diaphragm is ready to use and the electrodes has to be activated by the following typical procedure for removing aluminum. The electrodes were completely immersed in a solution potassium hydroxide (KOH) and potassium sodium tartrate. The electrodes were kept immersed for 24 hours at 80 °C. Afterward, they were rinsed with DI water and put into a container filled with DI water to prevent oxidation in air. The catalyst coatings on the 3D electrode substrates after the activation procedure are shown in Fig. 7. Generally, the rough surface area of the coatings is recognizable, which indicates a high active surface area. The coating thickness of both catalysts are shown in Figure 2 (c) for NiAl and in (f) for NiMoAl. They are approximately at the same range of about 40 µm. Nevertheless, the 3D electrodes are highly electrochemically active. Fig. 7: Catalyst coatings on 3D electrode substrates after the activation procedure (NiAl: (a), (b), and (c); NiMoAl: (d), (e), and (f)). The different YSZ diaphragm coatings at the substrate E-GTD and F-KT show that the smallest grain distribution has the highest ASR of about 3 Ω cm-2 for both substrates. The lowest ASR was achieved by the biggest grain distribution. By comparing the substrates, the sample E-GTD has the lowest ASR of about 1.27 Ω cm-2, which is about 67 % higher than the ASR of Zirfon (0.77 Ω cm-2). The SEM image shows E-GTD coated with YSZ. The wires of the mesh are recognizable, and the diaphragm coating completely encloses them. The diaphragm has in total a thickness of about 320 µm and the median of the bubble point pore size is 327 nm with a standard deviation of 87 nm (Fig. 8 and 9). EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476294 A1404 Page 6/9 Fig. 8: Area specific resistance measurement results of different YSZ diaphragm coatings on the substrate E-GTD and F-KT Fig. 9: SEM depiction of best performing diaphragm EYSZ75 The components did not show any degradation during the accelerated stress test: aqueous 35 % KOH solution at 200 °C and 35 bar for at least 7 days. 3.2 Concept stack validation The validation towards alkaline water electrolysis for the concept stack with an approximate power input of up to 700 W (5 cells, 100 cm2 active area) was performed at 60° and 80°C. In Fig. 10 it is visible that the stack shows a polarization curve for consistent with typical water electrolysis behaviour which highlights the successful demonstration of the developed diaphragms and electrodes. The overall stack efficiency at 500 mA cm-2 is only ca. 56 % which is comparably low. It can be attributed to the larger ohmic resistance from the diaphragm (e.g., Zirfon© has a 0.77 Ω cm-2 lower ASR) and a cell design that requires further optimization. This results in an ASR stack cell resistance of almost 2 Ω cm-2 as shown in Fig. 14. The stack also demonstrates stable performance over a 2-hour operation period, as indicated by the repeatable polarization data in Fig. 10. Furthermore, the voltage distribution across individual cells remains consistent, as shown in Fig. 12, suggesting homogeneous current distribution and uniform electrochemical activity. Although more extensive analysis is required, this points to a production process for diaphragms and electrodes that is stable to some extent. The constant current experiments at 80 mA cm-2 depicted in Fig. 13 reveal negligible voltage drift across all five cells implying homogeneity and stability of the used materials. One major challenge will remain the gas cross over which is depicted in Fig. 11 as the ratio of hydrogen to oxygen at low current densities. Instead of the desired maximum 2 % hydrogen in oxygen the ratio was found to be between 10 % and 30 %. [5] The steep increase observed during the initial 80 °C measurement suggests possible microstructural damage, such as diaphragm ruptures on the micrometer scale. This phenomenon requires further investigation through post-mortem analysis. EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476294 A1404 Page 7/9 Fig. 10: Stack polarization curves (pc). Fist pcs before 2-h constant current (cc) experiment, second pcs after cc. Fig. 11: Ratio of hydrogen to oxygen (HTO) at the anode at different temperatures during the constant current measurements Fig. 12: Polarization curve of a 5-level stack with a single cell voltage detection Fig. 13: Constant current measurements of a 5-level stack with a single cell voltage detection 3.3 Conclusion and next steps In conclusion, the proof of concept for both the electrodes and the diaphragm, stable at pressures above 25 bar and temperatures exceeding 130 °C, has been successfully demonstrated. These robust components were tested under standard alkaline water electrolysis (AWE) conditions and showed promising performance as part of an integrated cell and stack system. The consortium formed within the publicly funded AWEC++ project plays a critical role in the successful implementation of this technology, bringing together complementary expertise across materials development, coating technologies, simulation, system engineering, and demo plants. The next step is the construction and commissioning of a 20 kW demonstration plant, with operation targeted at 25 bar and 130 °C to validate the high-temperature AWE concept under real-world conditions. EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476294 A1404 Page 8/9 Acknowledgement The authors and the consortium express gratitude towards the Federal Ministry for Economic Affairs and Energy (BMWE) for funding the project AWEC++ as well as the project management Jülich for acting as the funding agency and for the project oversight. References [1] Allebrod, F. et al., Alkaline electrolysis cell at high temperature and pressure of 250 °C and 42 bar, Journal of Power Sources, 229, 2013, 22-31 [2] Xu, Z. et al., A catalyst-coated diaphragm assembly to improve performance of alkaline water electrolysers. Commun. Eng. 2025, 4, 9. [3] Schalenbach, M. et al., Hydrogen Diffusivity and Electrolyte Permeability of the Zirfon PERL Separator. J. Electrochem. Soc. 2023, 170, 045004. [4] Skadell, K. et al., Patent: Electrochemical cell and process for the production of hydrogen and oxygen from water, WO2024179759A1, 2024 [5] Haug, P. et al., Influence of process conditions on gas purity in AWE. Int. J. Hydrogen Energy 2022, 47, 5678–5686. Appendix Fig. 14: Electrical impedance spectroscopy at stack level from Fraunhofer IKTS Fig. 15: Meshes from Paul GmbH & Co KG EFCF 2025: Low-Temp. Fuel Cells, Electrolysers & H2 Processing 1 – 4 July 2025, Lucerne Switzerland https://doi.org/10.5281/zenodo.17476294 A1404 Page 9/9 Fig. 16: Coated meshes by MUW Screentec GmbH Fig. 17: Mechanical simulation by Merkle CAE Solution GmbH Fig. 18: Bipolarplate manufacturing by SITEC Industrietechnologie GmbH Fig. 19: Demo plant by APMiniplant GmbH & Co KG Keywords: EFCF2025, H2, LowTemp. Fuel Cells & Electrolysers, medium temperature water electrolysis, diaphragm, atmospheric plasma spray, demo plant, efficiency Remark: This work is licensed under Creative Commons Attribution 4.0 International