Tiny Lanthanum Dose Turns Cheap Cobalt Oxide into a Fuel-Cell-Friendly CO Scrubber

Hydrogen may be the clean fuel of the future, but the gas that leaves an industrial reformer is rarely pure enough for the fuel cells that are meant to use it. Even trace amounts of carbon monoxide cling to the platinum anodes of proton exchange membrane fuel cells, poisoning the electrodes and steadily eroding performance. The most practical remedy is a chemical trick called preferential oxidation, or PROX, in which a catalyst selectively burns the contaminant CO to carbon dioxide while leaving the vastly more abundant hydrogen untouched. A new study published in Catalysis Letters reports that a whisper-thin coating of lanthanum on cobalt oxide can make this trick work better, cheaper, and at exactly the temperatures where fuel cells operate.

The research, carried out by Reineck Mhlaba of the University of Limpopo in South Africa, focuses on cobalt oxide in its spinel form, Co3O4, a material long celebrated as one of the most active non-precious-metal catalysts for carbon monoxide oxidation. Earlier work by the same author showed that pretreating cobalt oxides with hydrazine boosts their activity and selectivity in PROX at the 80 to 120 degrees Celsius window that matches fuel cell operating conditions. The catch has always been durability: cobalt oxide catalysts deactivate over time, and the moisture and carbon dioxide that inevitably accompany real reformer gas accelerate that decline. The new study asks whether a pinch of lanthanum, a rare-earth element prized for its ionic conductivity and its talent for creating oxygen vacancies, can fix those weaknesses.

The answer, according to the data, is a qualified but encouraging yes. Mhlaba prepared the parent catalyst by precipitating cobalt hydroxide from cobalt nitrate with sodium hydroxide, then loaded lanthanum at 0.5, 1.0, and 2.0 weight percent by incipient wetness impregnation. All four samples were soaked in hydrazine solution for fifteen hours and calcined in air at 350 degrees Celsius before testing. When the catalysts were fed a simulated reformate stream of 1 percent CO, 1 percent oxygen, and 50 percent hydrogen balanced in helium, the catalyst carrying 1 weight percent lanthanum clearly outperformed the rest, delivering its best activity between roughly 60 and 120 degrees Celsius and reaching complete CO conversion at 140 degrees Celsius.

Selectivity, the measure of how much of the precious oxygen goes to CO rather than to wasteful hydrogen combustion, is where the doped catalyst truly shone. The 1 weight percent lanthanum formulation maintained 100 percent selectivity toward carbon dioxide across the temperature range, meaning every oxygen molecule consumed went to destroying the poison rather than burning fuel. By contrast, the catalyst overloaded with 2 weight percent lanthanum lost performance in the crucial 60 to 120 degrees Celsius window and converted noticeably more hydrogen, about 4 percent, an outcome the author attributes to excess lanthanum blocking pores and altering the balance of active sites.

Behind that sweet spot lies a story of structural fine-tuning that the study documents with an unusually complete battery of characterization techniques. X-ray diffraction confirmed the spinel Co3O4 phase in every sample and revealed a slight shift in the diffraction peaks upon lanthanum incorporation, along with the complete absence of any separate lanthanum-containing crystalline phase. That combination suggests the lanthanum atoms are either highly dispersed on the surface or tucked into the cobalt oxide lattice, where their larger ionic radius distorts the structure. The distortion matters because, as other researchers have shown, lanthanum ions substituting at octahedral cobalt sites generate oxygen vacancies, defects that activate lattice oxygen and boost oxidation activity.

Temperature-programmed reduction experiments added a second layer of evidence. Pure Co3O4 shows two characteristic reduction peaks, at about 218 and 313 degrees Celsius, corresponding to the stepwise reduction of cobalt from the plus-three to the plus-two state and then to metallic cobalt. Doping with lanthanum shifted these features and introduced a new high-temperature reduction peak near 578 degrees Celsius, which the author interprets as evidence of a strong electronic and structural interaction between lanthanum and the cobalt oxide lattice, likely through La-Co associated species. X-ray photoelectron spectroscopy backed this up, showing shifts in the cobalt core-level binding energies after doping and an increase in the ratio of Co3+ to Co2+ species on the 1 weight percent catalyst, the very redox couple thought to govern PROX performance.

The texture of the catalysts changed too. BET measurements showed the surface area climbing from 26.8 to 35.5 square meters per gram when 1 weight percent lanthanum was added, with the adsorption isotherms retaining the H4 hysteresis loop that signals an intact mesoporous structure. Electron microscopy revealed hexagonal plate-like and cubic morphologies, with lattice fringe spacings of 0.25 and 0.48 nanometers corresponding to the (111) and (311) planes of the spinel. Thermogravimetric analysis confirmed that all the catalysts were thermally stable well beyond their calcination temperature, losing less than 1.3 percent of their mass up to 600 degrees Celsius.

Perhaps the most practically important results came from the stability tests, where the feed was deliberately fouled with the two enemies of PROX catalysts: water and carbon dioxide. When 10 percent moisture was introduced, the 1 weight percent lanthanum catalyst suffered only a slight decline in activity, a resilience the study links to the peculiar electronic structure of lanthanum. Its unfilled 4f orbitals, shielded by a filled 5s2p6 shell, suppress the hydrogen bonding that lets water molecules dissociate on the surface, limiting the formation of hydroxyl groups that would otherwise smother the active sites. XPS of the spent catalyst showed a decrease in surface hydroxyl concentration, consistent with that protective mechanism.

Carbon dioxide proved a tougher adversary. Adding 15 percent CO2 to the feed cut catalytic activity by roughly 30 percent over 21 hours on stream at 130 degrees Celsius. The spent catalyst showed no structural transformation by FTIR or SEM, but its temperature-programmed reduction profile broadened and shifted, and a new carbonate band appeared at 1480 wavenumbers in the infrared spectrum, pointing to carbonate species accumulating on the surface. The XPS data of the spent sample also revealed a drop in oxygen vacancy concentration and reoxidation of Co2+ to Co3+, sketching a picture of gradual surface chemical change rather than catastrophic structural collapse.

The author is candid about the study’s limits: the catalytic experiments were not independently repeated, and future work will add replicate runs and error bars to quantify variability. Even so, the overall message is compelling. A cheap, earth-abundant cobalt oxide, nudged with a single weight percent of lanthanum and pretreated with hydrazine, can scrub carbon monoxide from hydrogen-rich gas with perfect selectivity at fuel cell temperatures while shrugging off moisture. If the stability against carbon dioxide can be extended over the thousands of hours a commercial fuel cell demands, lanthanum-doped Co3O4 could become a serious contender to replace expensive platinum-group PROX catalysts, bringing cleaner, longer-lived hydrogen fuel cells a step closer to the pump.

Subject of Research: Lanthanum-doped cobalt oxide catalysts for the preferential oxidation of carbon monoxide in hydrogen-rich fuel cell feed streams

Article Title: Effect of La on the Structure, Selectivity and Stability of La/Co3O4 Catalysts in Preferential Oxidation of Carbon Monoxide in Hydrogen-Rich Streams

Article References: Effect of La on the Structure, Selectivity and Stability of La/Co3O4 Catalysts in Preferential Oxidation of Carbon Monoxide in Hydrogen-Rich Streams. (n.d.). https://doi.org/10.1007/s10562-026-05547-6

Image Credits: AI Generated

DOI: 10.1007/s10562-026-05547-6

Keywords: PROX, lanthanum doping, Co3O4, carbon monoxide oxidation, hydrogen purification, fuel cells, catalyst stability, oxygen vacancies, XPS, temperature-programmed reduction, selectivity, cobalt oxide spinel