Where the Vacancy Sits Decides Everything: Iron-Doped Ceria Cracks the Nitrogen Oxide Puzzle

Every year, power plants and industrial boilers release millions of tonnes of nitrogen oxides into the atmosphere, feeding smog, acid rain and respiratory disease. The workhorse technology for scrubbing these gases, selective catalytic reduction with ammonia, has long depended on vanadium-based catalysts that are toxic, expensive and prone to unwanted side reactions. Now a team of Chinese researchers has shown, with unusual clarity, why a vanadium-free alternative built on cerium oxide behaves so differently depending on which metal you add to it — and the answer turns out to hinge not on how many defects a catalyst contains, but on precisely where those defects sit. The study, published in Catalysis Letters, offers a design rule that could reshape how low-temperature de-nitration catalysts are engineered.

The research group, led by Lirui Sun and Zhaoyang Chen of the National Synchrotron Radiation Laboratory at the University of Science and Technology of China, together with Lidong Zhang and colleagues, prepared four catalysts under an identical protocol: pristine cerium dioxide, and the same oxide modified with iron, cobalt or nickel. Ceria is a fascinating material in its own right. Its fluorite crystal lattice can easily shed and reabsorb oxygen atoms, cycling between Ce4+ and Ce3+ oxidation states, which makes it a natural oxygen reservoir. In ammonia-SCR, that mobility is a double-edged sword. Lattice oxygen is needed to activate ammonia and nitric oxide, but if the oxygen is too aggressive it oxidises ammonia all the way to nitrogen dioxide or even nitrous oxide, destroying the selectivity that the reaction depends on.

The baseline results were sobering. Untreated cerium dioxide managed a peak nitrogen oxide conversion of just 14.83 percent at 300 degrees Celsius under a realistic oxygen-rich feed containing 0.5 percent NO, 0.5 percent NH3 and 5 percent O2. Worse, when the temperature rose to 350 degrees Celsius, conversion collapsed to 6.12 percent while the yield of nitrogen dioxide climbed to 24.92 percent. In other words, the bare catalyst was not merely inefficient — it was actively converting the desired reaction into over-oxidation, producing the very species that fast SCR chemistry needs only in carefully balanced amounts. Too much NO2 alongside insufficient NO conversion means the ammonia is wasted on side reactions rather than being stitched onto nitrogen oxide to form harmless nitrogen and water.

Adding iron changed the picture dramatically. The iron-modified ceria, dubbed FeCe, reached 79.8 percent nitrogen oxide conversion at 300 degrees Celsius and still delivered 75.7 percent at 350 degrees Celsius, while keeping nitrogen dioxide yield down to just 8.0 percent. That combination is the holy grail of this field: high activity and high over-oxidation resistance at the same time. Most catalysts trade one against the other — push the conversion up and the unwanted oxidation products follow. The FeCe sample broke that trade-off, sustaining strong performance across the temperature window where industrial flue gases typically sit.

The other two dopants told a more cautionary tale. Nickel modification shifted the activity maximum down to 250 degrees Celsius, where conversion peaked at 54.0 percent, but the catalyst fell off a cliff at higher temperatures, dropping to a mere 0.3 percent conversion at 350 degrees Celsius. Cobalt modification produced only a narrow low-temperature operating window before failing. Perhaps most striking was what the conventional characterisation techniques revealed. By the usual proxies — X-ray diffraction, high-resolution transmission electron microscopy, Raman spectroscopy and X-ray photoelectron spectroscopy — the nickel-modified sample displayed the strongest defect signatures of all three doped materials. By every standard metric of oxygen vacancy abundance, NiCe should have been the star of the show. Instead, it underperformed the iron catalyst above 250 degrees Celsius by a wide margin.

That paradox is the intellectual heart of the study, and the authors resolved it by looking more carefully at dopant speciation — the chemical form each metal actually adopts inside the ceria host. The three transition metals, despite sitting side by side in the periodic table, chose three entirely different fates. Iron dissolved into the ceria lattice, substituting for cerium atoms and generating oxygen vacancies deep in the bulk crystal. Nickel refused to incorporate and instead precipitated as discrete nickel oxide particles, concentrating vacancies at the NiO/CeO2 interface. Cobalt phase-separated into cobalt oxide, a strongly oxidising spinel phase with its own aggressive oxygen chemistry. Three metals, three architectures, three completely different ways of supplying and positioning reactive oxygen.

Density functional theory calculations then explained why location matters more than quantity. The lattice-substituted iron site combined the lowest oxygen vacancy formation energy among the tested configurations with the widest calculated free-energy separation between two competing reaction channels: the selective pathway in which activated ammonia couples with nitric oxide to form nitrogen, and the non-selective pathway in which lattice oxygen strips hydrogen from ammonia, driving over-oxidation. When the vacancy sits in the bulk lattice near an iron centre, the energetic landscape favours N–N coupling and penalises dehydrogenation. When vacancies cluster at an external interface, as in the nickel case, the local oxygen supply becomes too reactive, tipping the balance toward ammonia destruction at elevated temperatures — which is exactly why NiCe, despite its abundant defect signatures, fell apart above 250 degrees Celsius.

The implications reach well beyond this particular reaction. Oxygen vacancies have become the darling of defect engineering across catalysis, energy storage and electronics, and the field has largely operated on the assumption that more vacancies mean better performance. This study delivers a sharp corrective: vacancy concentration is a crude proxy, and vacancy location is the variable that actually governs the activity–selectivity balance. A catalyst bristling with interface vacancies can look superb under spectroscopic scrutiny and still fail in a reactor, while a lattice-doped material with fewer but better-placed defects can dominate. For ceria-based systems, the message is that dopants must be chosen not just for their electronic properties but for their willingness to enter the lattice rather than segregate out of it.

There is also a practical dimension. Vanadium-based SCR catalysts, typically vanadium pentoxide on titania with tungsten or molybdenum promoters, operate effectively only above roughly 300 degrees Celsius and suffer from vanadium toxicity and ammonia slip issues. A ceria–iron system that sustains nearly 80 percent conversion at 300 degrees Celsius with strong resistance to over-oxidation points toward vanadium-free formulations that could work closer to the source of emissions, where flue gases are cooler, reducing reheating costs. The work was supported by the National Natural Science Foundation of China and the Natural Science Foundation of Anhui Province, and the authors included collaborators from the China Automotive Technology and Research Center, suggesting an eye toward mobile-source applications where low-temperature operation is especially valuable.

What makes this study resonate is its methodological discipline. By preparing all four catalysts under one identical protocol and evaluating them under one identical feed, the team eliminated the confounding variables that plague much of the comparative catalysis literature. The combination of structural characterisation, careful activity testing and first-principles energetics produced a self-consistent story in which every observation — the collapse of bare ceria, the triumph of iron, the paradox of nickel, the narrow window of cobalt — follows from a single principle. Oxygen vacancies are not interchangeable widgets to be counted; they are positional players whose placement in the crystal determines whether they supply oxygen gently enough to build nitrogen bonds or aggressively enough to burn the reductant. As researchers pursue next-generation environmental catalysts, that distinction may prove to be the difference between a promising material and a deployable one.

Subject of Research: Dopant speciation and oxygen-vacancy location in Fe-, Co- and Ni-modified CeO2 catalysts for low-temperature NH3-SCR of nitrogen oxides

Article Title: Unraveling the Role of Dopant Speciation in Oxygen-Vacancy Location and Lattice-Oxygen Supply over Fe-, Co- and Ni-Modified CeO2 NH3-SCR Catalysts

Article References: Sun, L., Chen, Z., Xu, W., Zhang, L., & Li, Z. (2026). Unraveling the Role of Dopant Speciation in Oxygen-Vacancy Location and Lattice-Oxygen Supply over Fe-, Co- and Ni-Modified CeO2 NH3-SCR Catalysts. Catalysis Letters, 156(10), Article 288. https://doi.org/10.1007/s10562-026-05532-z

Image Credits: AI Generated

DOI: 10.1007/s10562-026-05532-z

Keywords: NH3-SCR, ceria, oxygen vacancies, dopant speciation, lattice oxygen, iron doping, nickel oxide, cobalt oxide, DFT calculations, nitrogen oxides, over-oxidation resistance, vanadium-free catalysts