• NiFe catalysts in RDE show high OER activity without NiOOH formation. • NiOOH formation is not observed for catalyst layers containing above 23 wt% Fe. • The improved OER performance from Fe spiking in RDE is mitigated in the flowcell. • Ni loses performance rapidly in the flowcell; NiFe remains stable for over 100 hours. • Fe is required for good activity but dissolved Fe is lost to the cathode and reservoir. . Understanding how iron (Fe) influences nickel-based (Ni) catalysts in anion exchange membrane water electrolyzers (AEMWE) under industrial operational conditions is essential for enabling predictable and reproducible performance enhancement, yet the role of Fe in AEM flowcell conditions remains unexplored. This study investigates how Fe enhances Ni catalyst performance when it is intentionally incorporated into co-sputtered catalyst layers versus dissolved in the electrolyte. Preliminary rotating disk electrode experiments show that Ni-only catalyst layers initially display lower activity compared to NiFe catalysts although cycling can induce performance gains by Fe uptake towards comparable activity levels. In-situ Raman spectroscopy indicates that bulk NiOOH emerges only at low sputtered Fe contents but is suppressed above 23 wt% Fe, suggesting bulk NiOOH is not required to achieve the high activity observed for co-sputtered NiFe films. In flowcell experiments Fe-free Ni initially displays a high performance but degrades rapidly, while NiFe layers maintain stable performance over 100 hours. In contrast to the RDE tests, Fe-spiked Ni samples show a gradual decrease in performance, partially attributed to Fe depositing in the reservoir and on the cathode as observed by SEM-EDX. For the pure Ni films, it is hypothesized that deviations from linear Tafel behavior at higher current densities are a consequence of diminishing availability of Fe. These findings help to understand Fe’s role in enhancing Ni electrode performance as well as the need to account for Fe deposition phenomena, while showcasing the limitations of RDE testing for predicting real-world electrolyzer behavior. .
Donker et al. (Sun,) studied this question.