Nickel is a promising, cost-effective and robust catalyst for alkaline water electrolysis – a key technology for the climate-neutral production of hydrogen. However, the actual structure of the active surface of such nickel electrodes under reaction conditions had not yet been conclusively determined. A research team at Ulm University has now refuted a decades-old assumption: The catalytically active surface does not consist of nickel oxyhydroxide (NiOOH), as was previously widely assumed, but of nickel dioxide (NiO₂). The findings have been published in the prestigious journal ‘Nature Catalysis’.
During the electrochemical splitting of water, hydrogen is produced at the cathode and oxygen at the anode. “Although hydrogen is usually the focus of attention as a climate-neutral energy carrier, the greatest energy losses frequently occur at the oxygen electrode – that is, where the oxygen is generated,” explains PD Dr Albert Engstfeld, who coordinated this study together with Professor Timo Jacob at the Institute of Electrochemistry at Ulm University.
This oxygen evolution reaction therefore requires efficient catalysts: this is where nickel comes into play. This transition metal is of particular interest for this application, as it exhibits high catalytic activity under alkaline conditions. Under reaction conditions, nickel forms oxidised surface structures, the nature of which is decisive for how the oxygen evolution process unfolds in detail.
How, then, can the molecular structure of the catalyst surface be analytically characterised? “To obtain a valid result, it is crucial to observe the material whilst the chemical reaction is taking place – that is, in situ,” emphasises lead author Justus Leist, a PhD student at the Institute of Electrochemistry. Surface-enhanced Raman spectroscopy (SERS) was utilized for this purpose. To perform this specialized analytical measurement technique, a thin layer of the material under investigation is deposited onto a rough gold surface. For this specific analytical measurement technique, a thin layer of the material under investigation is applied to a rough gold surface. This enhances the interaction of the material with an incident laser, and valuable insights into the material’s structure can be gained from the reflected light.
NiO₂ instead of NiOOH: Surprising discovery necessitates a correction
For decades, it had been assumed that the active surface of nickel catalysts during oxygen evolution consisted of nickel oxyhydroxide, chemically denoted as NiOOH. The Ulm researchers tested this assumption using isotope labelling. To do this, they carried out Raman measurements in both normal water (H₂O) and deuterated, or ‘heavy’, water (D₂O). When hydrogen is replaced by deuterium, the vibrational properties of OH or OD groups normally change. “If the nickel surface were indeed composed of NiOOH, the replacement of the hydrogen isotope would have to result in characteristic changes in the Raman spectra,” says Justus Leist. However, this was not the case!
Supplementary electrochemical investigations using cyclic voltammetry (CV), which were also correlated with density functional theory (DFT) calculations, allowed conclusions to be drawn about the actual chemical composition of the surface. Ultimately, the experimentally observed spectra could be convincingly reconstructed. The scientists finally succeeded in reconciling the theoretically postulated and experimentally measured results. The conclusion: under reaction conditions, the surface corresponds to nickel dioxide, i.e. NiO₂, and not NiOOH. “This difference is fundamentally important from a chemical point of view. NiOOH contains hydrogen atoms, whereas NiO₂ does not. This means that the adsorption sites and reaction pathways on the surface, as previously assumed, must now be reassessed,” emphasises Professor Timo Jacob.
Overtones were misinterpreted for years
In addition to re-evaluating the nickel surface, the study also provides important fundamental insights into the interpretation of Raman spectra of catalytic materials. Certain bands in the spectra have previously often been attributed to so-called reaction intermediates – that is, short-lived intermediate stages of the chemical process. However, the Ulm research team was able to demonstrate experimentally that the signals are attributable to a physical phenomenon rarely observed in this class of materials: harmonicsovertones. “Based on simplified theoretical models, such bands are usually physically forbidden and only become accessible through extended models,” said Justus Leist. Based on further measurements of cobalt- and manganese-based oxides, the team posits postulates that these harmonics occur primarily in materials composed of two-dimensional layers.
The research project was carried out as part of the Collaborative Research Centre 1316 ‘Transient Atmospheric Pressure Plasmas – From Plasma to Liquids to Solids’ (Head of Office: Ruhr University Bochum). In their sub-project, the Ulm chemists are conducting research into plasma electrocatalysis, in which significantly higher current densities and voltages can be utilised than in conventional electrocatalysis. This can be used, on the one hand, for the production of catalytic materials or, on the other, directly for catalytic measurements.
Further information:
PD Dr Albert Engstfeld, Institute of Electrochemistry, email: albert.engstfeld(at)uni-ulm.de
Prof. Dr Timo Jacob, Head of the Institute of Electrochemistry, email: timo.jacob(at)uni-ulm.de
Publication reference:
Leist, J., Neufischer, A., Jacob, T. et al. Consequences of overtones in Raman spectra for assigning nickel anode surface structures as NiO₂ during alkaline electrolysis. Nature Catalysis (2026), 11 September 2026, https://doi.org/10.1038/s41929-026-01613-9
Text and media enquiries: Andrea Weber-Tuckermann
Translation: DeepL