Microstructure Evolution in Ultra-Fast High-Temperature Sintering (UHS) of SOEC Electrode Materials: Grain Growth, Defect Formation, and Electrochemical Implications

Le résumé fourni par la source

The fabrication of solid oxide electrolysis cells (SOECs) depends critically on the microstructural integrity of their electrodes and electrolytes, which influence ionic/electronic transport and long-term durability. Conventional sintering techniques involve prolonged thermal exposure at high temperatures, resulting in high energy consumption and extended processing times. This study investigates Ultra-Fast High-Temperature Sintering (UHS) as an alternative approach, focusing on the effects of rapid thermal exposure on grain growth, defect formation, and electrochemical performance in standard SOEC electrodes materials, including NiO-YSZ, GDC, and LSCF. Samples were sintered using Joule-heated carbon felts under controlled argon atmospheres, with temperatures reaching up to 1300 °C within seconds. UHS employs rapid Joule heating of carbon fibers to achieve heating rates exceeding 1000 °C/min, significantly reducing total sintering times while maintaining structural integrity. In this study, fuel electrode (NiO-YSZ), air electrode (LSCF-GDC/ LSCF), and barrier layer (GDC) materials were sintered via UHS under vacuum and argon atmospheres at temperatures up to ~1300°C. Preliminary tests were also conducted on Ba-based proton-conducting membranes (e.g., BaCeₓFeᵧO₃₋δ, BCF) to evaluate the compatibility of UHS with protonic ceramics. The effects of sintering rate, chemical stability and applied current (ranging from 20 A to 31 A) were analyzed through scanning electron microscopy (SEM), X-ray diffraction (XRD), and electrochemical impedance spectroscopy (EIS). UHS enabled significant densification within 120–200 seconds, reaching relative densities of 90–97% for GDC and up to 96% for LSCF. SEM imaging (Figure 1) revealed well-sintered LSCF structures with average grain sizes of ~5 µm, highlighting the accelerated grain growth kinetics induced by rapid thermal exposure. In contrast, conventional sintering required ~20 hours to achieve comparable densification, underscoring the efficiency of UHS. For NiO-YSZ, early tests showed surface nickel precipitation due to direct interaction with carbon fibers; this phenomenon was confirmed via SEM and elemental mapping using EDS (Figure 2), which showed localized Ni enrichment near the pellet surface. The issue was mitigated using 3D-printed NiO spacers, which provided a chemically compatible interface. No secondary phases were detected by XRD in any of the sintered samples. Electrochemical impedance spectroscopy revealed reduced polarization resistance and improved ionic conductivity in GDC-LSCF layers processed via UHS, indicating a strong correlation between grain connectivity and electrochemical performance. While detailed defect characterization is ongoing via XPS and TEM, the observed trends suggest that UHS-induced microstructural changes—particularly grain growth and phase purity—play a pivotal role in the functional properties of SOEC materials. These findings establish UHS as a viable and scalable sintering technique for SOEC manufacturing, offering substantial reductions in energy consumption and processing time while maintaining desirable electrochemical properties. On the other hand, initial results from UHS-processed BCF membranes further suggest the technique's broader applicability to proton-conducting ceramic systems. Further optimization of heating parameters and atmospheric conditions will be essential to fully exploit the benefits of UHS in next-generation SOEC fabrication. Acknowledgements This work was partially supported through federal project number W912CH-24-9-0006, as part of a partnership with the Center for Clean Hydrogen (CCH) at the University of Delaware. The authors would also like to acknowledge Mr. Ethan Ness and Mr. Daniel Schlaffer for experimental support, and Dr. Qiang Wang and Dr. Marcela Redigolo of WVU Shared Research Facilities (SRF) for assistance in characterization and technical input. Figure 1

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.