Fe–Ni-based alloys: highly active and low-cost alkaline oxygen evolution reaction catalyst

Water electrolysis in alkaline media enables to produce green hydrogen, an endeavor for renewable electricity storage. However, it is hindered by its low conversion efficiency (overvoltage of water electrolysis reactions), the catalysts' abundance, cost and manufacturability, and their durability.
Published in Chemistry and Materials
Like

The oxygen evolution reaction (OER) is the limiting step in water electrolysis. The development of high-performance OER catalysts is the key to high efficiency (i.e. low OER potential). Regarding the second issue, many research projects are focusing on Fe-Ni-based catalysts for OER in alkaline media. Although these elements are abundant, depending on the preparation method, catalysts can be very costly and difficult to synthesize on a large scale. In addition, to ensure cohesion between the active layer and the catalyst support, complex syntheses are sometimes required to achieve good durability.

 Our study focuses on the use of industrial Fe-Ni-based alloys as catalysts for the OER reaction in alkaline media, such alloys meeting the three technical challenges outlined above. Previous publications[1-5] have demonstrated the good performances of (usually) austenitic steels as OER catalysts ((i) good conversion efficiency): a suitable activation process promotes the growth of an OER active layer on the sample surface, whose performance is linked to the atomic Fe/Ni ratio. The studied samples come directly from a production line of steel and Ni-based alloys ((ii) abundant, inexpensive and easy to process materials). Lastly, the active-surface-layer's self-healing ability ensures the (iii) good durability of these catalysts.

Our aim was to provide a more in-depth understanding of the active-surface-layer's growth mechanisms and how it operates as a catalyst (i.e. investigation of the nature of the active sites). To this end, we studied a very wide range of alloys (with atomic Fe/Ni ratios ranging from 0.004 to 7.4, way beyond the compositions of usually-studied austenitic steels) to measure their catalytic performance and correlate it with a precise characterization of the active-surface-layer (chemistry and microstructure studied by various techniques (TEM-EDS, TEM-ASTAR, XRD and XPS)).

 Our measurements in the initial state (after polishing) confirm the direct dependence between catalytic performance and the atomic Fe/Ni ratio. By subjecting the samples to ageing (open-circuit hold in the 0.1 M KOH electrolyte) and activations (alternated high and low potentials), these performances are improved (decrease of the OER potential) and correlated with the evolution of the Fe/Ni atomic ratio (optimum between 0.2 and 0.4) (Figure 1.a).

Figure 1: (a) Correlation of OER performance with the active-surface-layer chemical composition. The OER potential (at 5 mA cm−2) is shown as a function of the atomic Fe/Ni ratio measured by XPS. (b) Ni active-site activity as a function of the atomic Fe/Ni ratio in the surface layer. For each point, the geometric current density (mA cm−2geo) is taken for a potential of 1.56 V versus RHE, corresponding to the OER region. This value is then normalized by the ECSA value to be expressed in mA cm−2ECSA. The ECSA value for W-316L in the initial state being very low, the error on the measurement is large; the corresponding Fe/Ni ratio of 16.0 is therefore not reported on the graph. The initial state is represented by solid symbols; after ageing, by lighter symbols; and after activations, by open symbols. The orange arrows illustrate how the surface composition and OER activity changes upon aging and activation, depending on the initial Fe/Ni ratio.

Figure 1: (a) Correlation of OER performance with the active-surface-layer chemical composition. The OER potential (at 5 mA cm−2) is shown as a function of the atomic Fe/Ni ratio measured by XPS. (b) Ni active-site activity as a function of the atomic Fe/Ni ratio in the surface layer. For each point, the geometric current density (mA cm−2geo) is taken for a potential of 1.56 V versus RHE, corresponding to the OER region. This value is then normalized by the ECSA value to be expressed in mA cm−2ECSA. The ECSA value for W-316L in the initial state being very low, the error on the measurement is large; the corresponding Fe/Ni ratio of 16.0 is therefore not reported on the graph. The initial state is represented by solid symbols; after ageing, by lighter symbols; and after activations, by open symbols. The orange arrows illustrate how the surface composition and OER activity changes upon aging and activation, depending on the initial Fe/Ni ratio.

All alloys converge towards an optimum atomic Fe/Ni ratio and near-similar OER activity. However, the active-surface-layer growth mechanism differs according to the initial alloy composition: (i) dissolution of additional elements (Cr, Mo...) and Ni-surface enrichment for alloys with high initial Fe/Ni ratios and (ii) Fe enrichment by impurities present in the electrolyte for low Fe/Ni ratios. The active-surface-layer thickness (directly linked to the number of active sites (NiIII/NiII capacitive response)) and the active site performance also depend on the Fe/Ni atomic ratio: the higher the initial Fe/Ni atomic ratio, the thinner the developed active-surface-layer (and the lower the NiIII/NiII reaction capacity, Figure 2); this deficit in the number of active sites is however compensated by their better performance (high efficiency for high Fe/Ni ratio, Figure 1.b).

In summary, this means that, after activations, alloys of very different initial composition (regardless of how they were prepared/polished) converge towards near-similar surface composition (Fe/Ni ratio) and apparent OER activity. So, tuning the initial Fe-Ni-based alloy composition/surface preparation does not appear mandatory if one wants to reach long-term OER activity, at least when the KOH electrolyte contains Fe-impurities (which industrial KOH electrolytes do).

Figure 2: TEM images of four samples after activations. The active surface layer is outlined by the blue area on the right side of each image. For the TEM sample preparation using focused ion beam SEM, a protective layer was deposited over the oxide layer: platinum for W-316L and carbon for W-825, W-718 and W-625. The values for active-surface-layer thickness and NiIII/NiII reaction capacity (i.e. number of active sites) are reported for each sample below its TEM image. The order of the atomic Fe/Ni ratio is W-316L > W-825 > W-718 > W-625.
Figure 2: TEM images of four samples after activations. The active surface layer is outlined by the blue area on the right side of each image. For the TEM sample preparation using focused ion beam SEM, a protective layer was deposited over the oxide layer: platinum for W-316L and carbon for W-825, W-718 and W-625. The values for active-surface-layer thickness and NiIII/NiII reaction capacity (i.e. number of active sites) are reported for each sample below its TEM image. The order of the atomic Fe/Ni ratio is W-316L > W-825 > W-718 > W-625. 

Accelerated durability tests run by alternating very high-frequency alternations of high/low potentials, a very aggressive test, reveal that, most Fe-Ni-based alloys exhibit improved activity in these conditions and extensive OER durability (> 20,000 CV cycles between 0.5 and 1.8 V versus RHE at 1 V s-1). This excellent resistance to harsh potential alternations, that could occur during start/stop and transient operation of the alkaline water electrolyser, indicates these could be robust for long-term practical operation in industrial alkaline water electrolyzers coupled to renewable sources. The only alloy tested that shows a slightly different behavior in the 625 (the initially poorest in Fe), because at such high frequency of potential alternation, Fe-enrichment from the electrolyte (which is needed for optimal OER activity) is barely possible. Whatever this bias (which results in slightly depreciated performance versus the fully-activated state), the OER performances are far from bad and do maintain on the long-term (Figure 3).

Figure 3: Compared OER activities of each surface in its pristine state (After polishing), after 11 cycles of activation (After activations) and after 20 000 cycles of AST (After AST). Data are presented as mean value ± standard deviation.
Figure 3: Compared OER activities of each surface in its pristine state (After polishing), after 11 cycles of activation (After activations) and after 20 000 cycles of AST (After AST). Data are presented as mean value ± standard deviation.

To conclude, Fe-Ni-based alloys are a major research topic in the field of OER catalysts. The deeper understanding provided by our study identified that Fe-Ni-based alloy reach high activity after simple electrochemical activation in Fe-containing KOH, owing to the formation of a rough and very OER-active Fe-doped NiOOH surface. The initial alloys surface treatment and bulk composition play little role on the OER activity and durability of performance after the surfaces are activated: any Fe-Ni-based alloy is capable to reach OER activity surpassing IrO2 benchmark, after proper activation. As a result, Fe-Ni-based alloys (in particular austenitic stainless steels) reveal very promising for future industrial alkaline water electrolysis.

 

References

1 Schäfer, H. et al. Stainless steel made to rust: a robust water-splitting catalyst with benchmark characteristics. Energy Environ. Sci. 8, 2685–2697 (2015).

2 Moureaux, F., Stevens, P., Toussaint, G. & Chatenet, M. Development of an oxygen-evolution electrode from 316L stainless steel: application to the oxygen evolution reaction in aqueous lithium–air batteries. J. Power Sources 229, 123–132 (2013).

3 Moureaux, F., Stevens, P., Toussaint, G. & Chatenet, M. Timely-activated 316L stainless steel: a low cost, durable and active electrode for oxygen evolution reaction in concentrated alkaline environments. Appl. Catal. B Environ. 258, 117963 (2019).

4 Todoroki, N. & Wadayama, T. Heterolayered Ni–Fe hydroxide/oxide nanostructures generated on a stainless-steel substrate for efficient alkaline water splitting. ACS Appl. Mater. Interfaces 11, 44161–44169 (2019).

5 Todoroki, N., Shinomiya, A. & Wadayama, T. Nanostructures and oxygen evolution overpotentials of surface catalyst layers synthesized on various austenitic stainless steel electrodes. Electrocatalysis 13, 116–125 (2022).

Please sign in or register for FREE

If you are a registered user on Research Communities by Springer Nature, please sign in

Subscribe to the Topic

Materials for Energy and Catalysis
Physical Sciences > Materials Science > Materials for Energy and Catalysis
Electrochemistry
Physical Sciences > Chemistry > Analytical Chemistry > Electrochemistry
Materials Chemistry
Physical Sciences > Chemistry > Materials Chemistry
Catalytic Materials
Physical Sciences > Chemistry > Materials Chemistry > Catalytic Materials
  • Nature Materials Nature Materials

    A monthly multi-disciplinary journal that brings together cutting-edge research across the entire spectrum of materials science and engineering, including applied and fundamental aspects of the synthesis/processing, structure/composition, properties and performance of materials.