The electrolysis of water to produce hydrogen fuel is crucial for renewable energy. However, the oxygen evolution reaction (OER) is kinetically slow, limiting efficiency. Iron plays a pivotal role in enhancing OER activity in alkaline media in Ni and Co (oxy)hydroxides, whether intentionally added or incidentally present. Maintaining high activity requires soluble Fe species in the electrolyte due to dynamic Fe exchange at the catalyst/electrolyte interface. High-activity perovskite oxides also involve active Fe sites. Fe-containing (oxy)hydroxides are ubiquitous in OER-active materials. Ni-Fe and Co-Fe hydroxides have layered structures, with Fe3+ substituting for M2+ sites. Fe substitution shifts the M2+/M3+ redox potential, indicating electronic interactions. The oxidation state of Fe during OER is debated; both Fe3+ and Fe2+ have been observed. Operando Mössbauer studies show oxidation of Fe3+ to Fe2+ during OER, with Fe2+ persisting after the potential decrease. Discrepancies exist regarding the role of Fe2+, possibly due to different Fe cation populations (interior vs. surface). Previous work suggested two types of Fe sites: OER-active surface sites and less-active interior sites. The intrinsic activity of the most active "surface" Fe and the OER mechanism remain unknown.

The literature extensively explores the use of Fe-containing transition metal (oxy)hydroxides as oxygen evolution reaction (OER) catalysts. Studies have highlighted the importance of Fe in enhancing the activity of Ni and Co (oxy)hydroxides, regardless of the method of introduction. The dynamic exchange of Fe at the catalyst/electrolyte interface and the role of soluble Fe species in sustaining high activity have been established. Research on perovskite oxides has also revealed the importance of active Fe sites. The structural characteristics of Ni-Fe and Co-Fe hydroxides, including the substitution of Fe3+ for M2+ sites and the resulting electronic interactions, have been investigated. However, the exact oxidation state of Fe during OER and the specific roles of different Fe sites (surface vs. interior) remain areas of ongoing debate and investigation. Previous studies have suggested the existence of at least two distinct types of Fe sites with different OER activities.

Electrolyte-permeable NiOOH and CoxHy films were electrodeposited on Pt/Ti/glass substrates. The porosity and thinness ensured electrochemical activity. The redox response of Ni and Co cations was used to understand foreign metal incorporation. Two classes of sites were defined: interior "bulk" sites (substitutional incorporation affecting host redox waves) and surface sites (adsorption onto the structure, with minimal effect on redox waves). NiOOH and CoxHy were tested with added Co2+ and Ni2+, respectively, to study incorporation mechanisms. Voltammetry showed that Co2+ initially formed a separate phase on NiOOH, but subsequent cycling led to dispersion into the NiOOH structure. In contrast, Ni2+ remained on the CoxHy surface as a separate phase. Electrochemical quartz crystal microbalance (EQCM) was used to monitor mass changes during voltammetry. NiOOH showed mass gain during oxidation due to K+ and OH- intercalation, while CoxHy showed negligible mass change. EC-AFM provided supplementary structural characterization. Fe3+ was added during chronoamperometry to limit Fe absorption to surface sites in Fe:NiOxHy and Fe:CoxHy. The influence of Fe addition on OER current was studied. Voltammetry after Fe incorporation revealed minimal changes in the host redox response, but significant activity enhancements, indicating minimal electronic interaction between adsorbed surface Fe and host metal. Cycling increased Fe incorporation into the interior sites. The OER turnover frequency (TOF) was calculated based on the total number of Fe sites. TOF for surface Fe was measured, and the effect of Fe loading was studied, comparing CA vs. cycling vs. co-deposition. Operando XAS was used to study the local structure of Fe adsorbed during CA on NiOOH. DFT calculations were performed on model systems (isolated Fe-O and Fe-O-Fe dimer) to understand OER mechanisms. The calculations investigated different reaction pathways (axial, equatorial, and bridge) and compared overpotentials.

The study found that the location of foreign cations (in or on NiOOH and CoxHy) is influenced by structural changes during voltammetry. NiOOH readily incorporates Fe during cycling, while CoxHy does not. The method of Fe incorporation (chronoamperometry vs. cycling) significantly impacted its location. Chronoamperometry under OER conditions led to primarily surface Fe adsorption, while cycling resulted in more homogeneous mixing of Fe within the host structure. The TOF for surface-adsorbed Fe increased linearly with Fe concentration, contrasting previous studies where Fe site location was uncontrolled. High intrinsic TOF (40 ± 2 s−1 at 350 mV overpotential) was observed for Fe:NiOxHy, significantly higher than benchmark co-deposited phases. This high activity was attributed to a cooperative effect between multiple surface Fe sites. The DFT calculations revealed that the cooperative interaction between Fe atoms in an Fe-O-Fe cluster facilitates charge sharing and stabilization of positive charge, contributing to lower OER overpotentials compared to isolated Fe sites. The Tafel slopes for Fe:NiOxHy and Fe:CoxHy decreased with increasing Fe concentration, suggesting a change in the OER mechanism. The TEM/EDX analysis indicated that the Fe-oxo clusters formed on NiOOH were likely of molecular dimensions. XAS analysis showed that the Fe-O bond length decreased upon cycling, consistent with Fe moving from surface to interior sites. The study demonstrated a strong dependence of Fe site OER activity on local configuration, showing higher activity for surface Fe compared to bulk sites. The intrinsic activity of surface Fe sites was also affected by the host material. The study provided evidence for cooperative effects between multiple Fe sites on the surface, with neighboring Fe atoms stabilizing positive charge during the OER.

The findings address the long-standing question of how and where Fe species enhance OER activity in transition metal (oxy)hydroxides. The results demonstrate that surface-confined, cooperative Fe sites are responsible for the exceptional OER activity observed in these materials. The linear increase in TOF with increasing surface Fe concentration highlights the importance of cooperative interactions between these sites in facilitating charge delocalization and stabilizing reaction intermediates. The DFT calculations provide a mechanistic understanding of these cooperative effects, showing that charge sharing within Fe-O-Fe clusters lowers the OER overpotential. The difference in activity between Fe sites on NiOOH and CoxHy suggests that the host material plays a crucial role in modulating the electronic structure and reactivity of the active sites. This work highlights the importance of controlling both the amount and location of Fe in the design of high-performance OER catalysts.

This study reveals that surface-adsorbed Fe sites, particularly those in cooperative FeOx clusters, are responsible for the remarkably high OER activity observed in Fe-modified NiOOH. The linear relationship between TOF and Fe surface concentration, along with DFT calculations, strongly suggests a cooperative mechanism where neighboring Fe atoms share and stabilize charge, lowering the OER overpotential. Future work should focus on exploring larger FeOx clusters and further understanding the specific role of the host material in modulating the electronic properties of the active sites. Investigating other transition metal oxides with this approach would be beneficial in designing more efficient and stable OER catalysts.

The study focused primarily on two transition metal hosts (Ni and Co). The generalizability of the findings to other hosts needs further investigation. While the DFT calculations provided valuable insights into the OER mechanism, the complexity of the real catalytic system might not be fully captured by the simple models. The operando XAS measurements presented some experimental limitations in fully quantifying the Fe incorporated and thus TOF could not be normalized.