Oxygen vacancies are missing oxygen atoms in an oxide lattice, and they are routinely introduced to tune the electronic structure of catalysts for fuel cells, metal–air batteries and green chemical synthesis. Conventionally, scientists count vacancies and assume more is better, or at least that all vacancies behave alike. The new work challenges that assumption at a fundamental level. When an oxygen atom leaves the lattice, it can leave behind electrons that either remain trapped at the vacancy site or are transferred to neighbouring metal cations. These two configurations correspond to distinct colour centres, known as F centres: an electropositive F1 centre, in which the electrons are not localized at the vacancy, and an electroneutral F2 centre, in which two electrons are trapped within the vacancy itself.

The challenge for the team, led by Xiaoyuan Zhang, Jingwen Sun and Junwu Zhu of Nanjing University of Science and Technology, was to isolate the effect of charge state from every other variable. To do this, they devised a template-assisted synthesis in which the atmosphere during preparation was carefully regulated, allowing them to produce zirconia samples, ZrO2−x, that contain predominantly F1 or predominantly F2 centres while keeping the crystal structure, particle morphology and vacancy concentration essentially unchanged. This clean experimental design meant that any difference in catalytic behaviour could be attributed directly to the charge state of the defects rather than to confounding structural differences.

Characterization confirmed the distinction. Electron paramagnetic resonance spectroscopy, which is sensitive to unpaired electrons, revealed the paramagnetic signature of the F1-type vacancies, while complementary measurements of the electronic structure showed the different local environments around zirconium cations adjacent to each vacancy type—F1-Zr4+ versus F2-Zr3+ configurations. X-ray absorption, electron energy-loss spectroscopy and photoluminescence measurements all supported the picture of two electronically distinct but structurally equivalent defect species. Crucially, electrochemical scanning transmission electron microscopy showed that the catalysts remained stable under operating conditions, ruling out structural reconstruction as the source of the differing reactivity.

The electrochemical consequences were striking. When the electroneutral F2 centres dominated, the catalyst favoured the two-electron oxygen reduction pathway, selectively converting oxygen into hydrogen peroxide. When the electropositive F1 centres dominated, the reaction instead proceeded toward full four-electron reduction, cleaving the oxygen–oxygen bond and producing water. Hydrogen peroxide electrosynthesis is a rapidly growing field because the compound is a green oxidant used in water treatment, disinfection and chemical manufacturing, and producing it on-site in an electrochemical cell could replace the energy-intensive anthraquinone process. A catalyst whose selectivity can be switched by defect charge state therefore has immediate practical appeal.

To understand the mechanism, the researchers deployed in situ electrochemical electron paramagnetic resonance, tracking the paramagnetism of the F centres while the reaction ran, together with in situ Raman spectroscopy to follow the evolution of reaction intermediates. The results overturned a common intuition. The electroneutral F2 centre, despite being the site where electrons are trapped, is not the primary adsorption site for oxygen. Instead, it acts through dynamic electronic compensation: it continuously donates and withdraws electron density to and from adjacent zirconium sites, stabilizing the adsorbed *OOH intermediate that is the hallmark of the two-electron pathway. It is this dynamism, rather than direct binding, that makes F2 centres the gatekeepers of peroxide selectivity.

The electropositive F1 centre behaves in an entirely different manner. It binds molecular oxygen directly at the vacancy site, and the interaction is strong enough to cleave the O–O bond, committing the reaction to the four-electron pathway. In the process, the F1 centre itself is quenched, its paramagnetic signature disappearing as the reaction proceeds. Density functional theory calculations reproduced both behaviours, showing favourable adsorption energetics for O2 at F1 sites and for *OOH at sites electronically modulated by neighbouring F2 centres, and the calculated free-energy landscapes matched the experimentally observed selectivity patterns.

Beyond zirconia, the study carries a broad message for the field. Oxygen vacancies have been invoked to explain catalytic behaviour in ceria, perovskites, cobalt and iron oxides, and countless other systems, but the charge state of those vacancies is rarely controlled or even measured. The authors argue that F-centre charge state should be regarded as an independent design lever, alongside vacancy concentration and position. Because the two charge states can be interconverted by atmosphere control during synthesis, and because in situ electron paramagnetic resonance can now monitor them under working conditions, the toolkit exists to rationally design vacancy chemistry rather than accept whatever defects a preparation happens to deliver.

The work also highlights the power of operando spectroscopy to catch catalysts in the act. Static characterization before and after a reaction can miss the transient electronic exchanges that actually govern selectivity; here, the decisive role of the F2 centre only became visible because its paramagnetism and the Raman signatures of intermediates were tracked simultaneously under electrochemical bias. As the energy transition drives demand for selective, precious-metal-free electrocatalysts for peroxide production, water treatment and chemical synthesis, the ability to dial in defect charge states could prove one of the more consequential ideas to emerge from defect engineering in recent years. What was once an invisible nuance of the oxide lattice has become a switch that chemists can now flip at will.

Subject of Research: Charge-state-dependent oxygen vacancy (F-centre) control of electrocatalytic oxygen reduction selectivity in zirconia

Article Title: F-centre charge state and dynamism govern oxide electrocatalytic selectivity

Article References: Zhang, X., Bukhvalov, D., Su, T., Fang, C., Dai, L., San, S., Duan, H., Wang, Y., Liu, K., Cui, J., Hua, Y., Xue, L., Hou, Z., Zhang, W., Xiong, P., Fu, Y., Sun, J., & Zhu, J. (2026). F-centre charge state and dynamism govern oxide electrocatalytic selectivity. Nature Chemistry. https://doi.org/10.1038/s41557-026-02256-w

Image Credits: AI Generated

DOI: 10.1038/s41557-026-02256-w

Keywords: oxygen vacancies, F centres, zirconia, electrocatalysis, oxygen reduction reaction, hydrogen peroxide, selectivity, electron paramagnetic resonance, vacancy engineering, in situ Raman spectroscopy, defect chemistry, 2e- ORR pathway