Nickel–iron and cobalt–iron (oxy)hydroxides are state-of-the-art electrocatalysts for oxygen production in alkaline conditions. Now, the addition of high-valent dopants has been demonstrated to further propel the catalytic rate in these materials by an order of magnitude.
Researchers have long aspired to enable widespread hydrogen production without any associated CO2 emissions via the electrolysis of water, thereby decarbonizing large segments of industry and transport. A primary barricade obstructing this technology is the kinetically sluggish oxygen evolution reaction (OER) occurring at the anode of water electrolysers and the lack of active, robust and low-cost electrocatalysts. Increasing the efficiency and reducing the cost of this reaction is therefore vital for widespread adoption of this technology.
There are two predominant water electrolyser design concepts, namely proton exchange membrane (PEM) electrolysis and alkaline exchange membrane (AEM) electrolysis, each exhibiting their own benefits and drawbacks. While PEM electrolysis involves high current density, it typically requires OER catalysts based on scarce and expensive metals such as iridium1. Conversely, while high current densities cannot be accessed, AEM electrolysis has seen more progress using Earth-abundant catalysts, making this process more amenable to large scale industrial application. Hence, discovering new strategies that tailor material properties to boost the OER in base is a challenge of significant importance.
Now, writing in Nature Catalysis, Edward Sargent and co-workers outline one such fundamental strategy to propel large-scale hydrogen production in alkaline electrolysers2. These authors have shown how the OER can be accelerated in NiFe and FeCo (oxy)hydroxides using high-valent transition metals (that is, Nb, Mo, Ta, W and Re) as dopants. This was achieved using an aqueous sol–gel technique3, whereby metal chloride precursors were dissolved in ethanol and subsequently hydrolysed to produce a gel (Fig. 1a). Then, upon drying the gel with supercritical CO2, amorphous (oxy)hydroxides were produced with a homogeneous dispersion of metal elements (Fig. 1b) confirmed via transmission electron microscopy, bypassing the formation of the phase-separated oxide materials. Inspiration for this came from an earlier work from Sargent et al., highlighting the effect of doping FeCo (oxy)hydroxides with W6+ ions using the same synthetic procedure, which has now been generalized to several other high-valent metals4.
Through careful in situ/ex situ soft and hard X-ray absorption spectroscopic analyses, this research has identified that high-valent dopants increase the proportion of Fe2+ ions on the catalyst surface prior to the OER, facilitating a reduction in the oxidation barrier of the first-row transition metals in the catalyst. Because the OER often occurs at a metal site that cycles through oxidation states, reducing the barrier of oxidation at these sites can lead to an increase in the rate of the reaction. To discuss this in a standardized manner, a typical figure of merit for OER electrocatalysts is the overpotential required to deliver a current density of 10 mA cm–2. Sargent et al. show that their doping strategy reduces this overpotential by approximately 50 mV, representing a 20% reduction compared to the dopant-free catalyst. Intriguingly, this overpotential also shows a slight correlation with the amount of surface Fe2+ observed across the doped NiFe and FeCo-based catalysts prior to the OER (Fig. 1c,d).
Led by Luigi Cavallo and his team, computational density functional theory (DFT) studies reported in the same work justified this by predicting the potential at which the Fe2+/Fe3+ oxidation would occur, with a reduction observed between the control and doped catalysts. This represents another excellent example of how DFT simulations have advanced our understanding of the factors limiting the performance of OER electrocatalysts. Many of these insights have built on the existence of a linear relationship, or scaling relation, between the binding energies of reaction intermediates in oxide catalysts5. This relationship essentially means we cannot vary one binding energy without varying another, limiting our ability to design ideal catalysts. Computational studies accompanying the work by Cavallo and colleagues have used these insights to predict the overpotential of the doped (oxy)hydroxides, finding a good agreement with the observed experimental trends.
With their sights firmly set on real-world applications, Sargent et al. tested one NiFeMo catalyst against the commercially used Ni catalyst in an electrolyser setup of an industrial scale, tested over 120 hours. They demonstrated that this ternary (oxy)hydroxide exhibited an order-of-magnitude-higher current density per unit mass of catalyst loading, a reduced overpotential at high current densities against the commercial material and impressive stability. Overall, the remarkably low overpotentials of the catalysts presented in this work indicate that these multi-metallic materials may circumvent the OER scaling relation by following distinct reaction mechanisms. Given the reduced oxidation barrier in these catalysts, one possibility is the occurrence of additional one-electron oxidations prior to O–O bond formation, which has been proposed to explain the impressive activity seen for certain molecular OER catalysts6. Future work should focus on exploring these alternative reaction mechanisms and determining steps, while also outlining the conditions under which these materials segregate into distinct phases. Furthermore, it would be interesting to probe whether recent advances in NiFe-based OER research could be applied to these catalysts. For instance, the stability of a thin NiFe mixed-hydroxide catalyst has been found to increase via potential cycling7, while OER current densities from NiFe oxide catalysts can more than double in the presence of a magnet8. Combining these insights could represent a substantial step towards hydrogen production without greenhouse gases.
The outstanding stability, activity and affordability of the catalysts presented by Sargent and colleagues leads to thought-provoking questions for OER catalyst discovery: can we do any better under alkaline conditions? The relative gains to be made are sure to be small in comparison to the achievements that may lie ahead in the design of acidic OER catalysts, which still remain unstable, inefficient or expensive for PEM electrolysis. Sargent et al. have set a clear direction for explorative OER catalyst design via doping with high-valent transition metals, which poses another interesting question: will this strategy find application in acidic media? A recent work on Sb5+-doped MnOx catalysts showing the critical role of Sb in the stabilization of Mn3+ in rutile MnO2 for acidic OER catalysis would indicate so9. However, identifying novel materials with the appropriate coordination environments to enable high-valency in acid stable oxides is a daunting yet valuable problem to tackle. Another challenge in catalyst discovery that must be addressed is the optimization of many material properties at once — stability, activity, cost and complexity are all essential considerations. We anticipate that innovation from in situ scrutiny of electrocatalytically active sites10 will bolster our efforts, while machine learning models can tackle the issue of multi-objective optimization in catalysis11. One thing is clear: to accelerate the discovery of novel, active and cost-effective OER catalysts, those approaches must aim to work harmoniously with fundamental experimental investigations, such as this exciting work from Sargent and co-workers.
McCrory, C. C. L. et al. J. Am. Chem. Soc. 137, 4347–4357 (2015).
Zhang, B. et al. Nat. Catal. https://doi.org/10.1038/s41929-020-00525-6 (2020).
Augustyn, V. et al. Nat. Mater. 12, 518–522 (2013).
Zhang, B. et al. Science 352, 333–337 (2016).
García-Mota, M. et al. ChemCatChem 3, 1607–1611 (2011).
Craig, M. J. et al. Nat. Commun. 10, 4993 (2019).
Kuai, C. et al. Nat. Catal. 3, 743–753 (2020).
Garcés-Pineda, F. A., Blasco-Ahicart, M., Nieto-Castro, D., López, N. & Galán-Mascarós, J. R. Nat. Energy 4, 519–525 (2019).
Zhou, L. et al. ACS Catal. 8, 10938–10948 (2018).
Pfisterer, J. H. K., Liang, Y., Schneider, O. & Bandarenka, A. S. Nature 549, 74–77 (2017).
Sun, Y. et al. Nat Catal 3, 554–563 (2020).
The authors declare no competing interests.
About this article
Cite this article
Craig, M.J., García-Melchor, M. Faster hydrogen production in alkaline media. Nat Catal 3, 967–968 (2020). https://doi.org/10.1038/s41929-020-00540-7