The material at the heart of the study is niobium pentoxide, or Nb2O5, a semiconductor widely produced in Brazil, which holds most of the world’s niobium reserves. Niobium oxide has long attracted attention as a photocatalyst because of its unusual surface chemistry: it hosts both Brønsted and Lewis acid sites that can chemisorb CO2, forming an unstable carboxyl intermediate that light-driven electrons can then reduce into fuels and oxygenated chemicals. Under aqueous conditions, Brønsted acid sites become dominant, anchoring CO2 through its carbon atom and enabling the cascade of reduction steps that ultimately yield products such as methanol, a potential liquid fuel, and acetic acid, an industrial feedstock.

But acidity alone is not the whole story. The team began by synthesizing a highly reactive form of Nb2O5 using the oxidant peroxide method, dissolving a niobium oxalate precursor in water and hydrogen peroxide, heating the mixture to form a gel, and then drying and gently calcining the solid at just 150 degrees Celsius. This mild treatment preserves a disordered, defect-rich structure bristling with reactive peroxo groups, which give the material its characteristic yellow color and high initial activity. The researchers then dispersed the powder in phosphoric acid solutions at concentrations of 0.1, 0.5 and 1.0 mol per liter for 48 hours, washing and drying the resulting phosphatized catalysts, labeled Nb-0.1, Nb-0.5 and Nb-1.0 according to the acid concentration used.

To quantify how phosphatization changed the surfaces, the team measured the concentration of acidic sites through indirect potentiometric titration in alkaline suspensions, a technique that probes the Brønsted acidity of these amphoteric oxides in water. The response to phosphatization was strikingly non-linear. Acidity rose from moderate values at low phosphoric acid concentration to a peak of 0.99 plus or minus 0.07 millimoles per gram at the intermediate treatment, then fell back slightly at the highest concentration. This optimum, the researchers found, reflects a delicate balance: phosphate groups both introduce new Brønsted acid sites and clear away organic residues left over from synthesis, but too much phosphate begins to clog the very pores and sites the reaction depends on.

An extensive characterization campaign using X-ray diffraction, infrared and Raman spectroscopy, X-ray photoelectron spectroscopy, nuclear magnetic resonance, X-ray fluorescence, electron microscopy, atomic force microscopy with infrared detection and nitrogen physisorption painted a consistent picture of what phosphatization actually does. The treatment stripped residual carboxylate and carbonate species from the surface, evidenced by declining carbon content in elemental analysis and changes in the oxygen 1s photoelectron spectra, while covalently anchoring phosphate tetrahedra to the niobia framework through Nb–O–P linkages. X-ray photoelectron spectroscopy revealed phosphorus in mixed plus-five and plus-three oxidation states at 133.2 and 134.1 electronvolts, confirming the formation of a surface niobium phosphate layer rather than a separate bulk phase. Crucially, the amorphous, pseudohexagonal TT structure of the underlying oxide, built from distorted NbO6 octahedra rich in catalytically useful defects, remained intact throughout.

The textural transformation was equally dramatic. The untreated control material showed weak, Type III nitrogen adsorption isotherms, a low surface area of 13.47 square meters per gram, and pores blocked by organic debris. After phosphatization, the isotherms shifted to Type II behavior and the surface area soared more than sixfold, reaching a maximum of 82.15 square meters per gram for the intermediate sample, whose surface roughness also dropped from 154 to 65 nanometers as measured by atomic force microscopy. At the highest acid concentration, however, excess phosphate accumulation drove the surface area back down to 41.62 square meters per gram, underscoring that more is not always better.

When the catalysts were put to work, the differences became unmissable. In a quartz reactor filled with CO2-saturated water and illuminated by ultraviolet lamps at 254 nanometers, the phosphatized catalyst prepared with 0.5 mol per liter phosphoric acid delivered the standout performance. It produced acetic acid at 267 plus or minus 49 micromoles per gram per hour and methanol at 181 plus or minus 33 micromoles per gram per hour, with the two products together accounting for 99 percent of everything detected. Selectivity reached roughly 59 percent for acetic acid and 40 percent for methanol, while carbon monoxide, a common and less useful byproduct that made up about 20 percent of the control’s output, was suppressed to below 0.1 percent. Formic acid and carbon monoxide lingered only as traces of around 0.2 percent, suggesting they act as fleeting intermediates on the modified surface rather than endpoint products.

The mechanistic explanation, the researchers propose, lies in how phosphate-modified Brønsted acid sites handle the reaction intermediates. On the untreated oxide, carbon monoxide formed during reduction is released prematurely, cutting the reaction chain short and limiting yields. On the phosphatized surface, carbonate and bicarbonate species adsorb more effectively, and intermediates such as the carboxyl radical are retained long enough to undergo the additional reduction and even carbon–carbon coupling steps needed to form methanol and the two-carbon acetic acid. Control experiments, including photolysis without catalyst, irradiation under visible light, and nitrogen bubbling in place of CO2, confirmed that the product formation genuinely depended on the photocatalyst, the ultraviolet light, and the presence of carbon dioxide.

Perhaps most importantly, phosphatization solved a chronic durability problem. Peroxo groups inherited from the oxidant peroxide synthesis route are highly reactive but tend to deactivate within the first reaction cycle, and the untreated control lost essentially all activity by its second run, while continuing to emit carbon monoxide. The phosphatized catalysts, by contrast, remained functional across four consecutive four-hour reaction cycles, retaining roughly half of their initial activity even as their yellow peroxo coloration faded to white. Post-mortem analysis showed that the characteristic Raman band of the Nb–O–P bond at 872 wavenumbers and the corresponding infrared phosphate bands survived the reaction, along with carbonate signatures near 2400 to 2500 wavenumbers that pointed to active CO2 adsorption on the spent surface. Although wavelength-dispersive X-ray fluorescence recorded a partial drop in the phosphorus-to-niobium ratio after cycling, from 0.24 to 0.11, the residual phosphate layer continued to supply the Brønsted acidity and structural stability that kept the catalyst alive.

The authors caution that the acidity measured by titration must be interpreted alongside surface composition, textural accessibility and catalytic performance, and that direct confirmation of the proposed surface intermediates will require in situ or operando spectroscopic studies. They also note that the bandgap of the semiconductor, between 3.04 and 3.10 electronvolts, barely changed with treatment, meaning the performance gains came entirely from surface engineering rather than optical tuning. Even so, the message of the work is clear and potentially far-reaching: by pairing an abundant, locally produced oxide with a cheap acid treatment, the team has shown that surface acidity, when combined with accessible texture and phosphate stabilization, is a strategic design parameter for artificial photosynthesis. As the world searches for ways to recycle carbon dioxide into fuels rather than merely capture it, a modest dip in phosphoric acid may prove to be one of the most elegant tricks in the playbook.

Subject of Research: Phosphate-modified niobium pentoxide photocatalysts for the aqueous photoreduction of CO2 into methanol and acetic acid

Article Title: Influence of the Surface Acidity of Niobium Catalysts Modified with Phosphate in the Photocatalysis of CO2 for Conversion into Methanol and Acetic Acid

Article References: Oliveira, E., da Cruz, J. C., Magalhaes, W. L. E., & Ribeiro, C. (2026). Influence of the Surface Acidity of Niobium Catalysts Modified with Phosphate in the Photocatalysis of CO2 for Conversion into Methanol and Acetic Acid. Catalysis Letters, 156(10), Article 277. https://doi.org/10.1007/s10562-026-05495-1

Image Credits: AI Generated

DOI: 10.1007/s10562-026-05495-1

Keywords: CO2 photoreduction, artificial photosynthesis, niobium pentoxide, photocatalysis, surface acidity, phosphatization, methanol, acetic acid, Brønsted acid sites, niobium phosphate, sustainable fuels, carbon dioxide conversion