Order Matters: How Mixing Metals Transforms Cooking Oil into Jet Fuel

It sounds almost too simple to matter: the order in which two metals are added to a catalyst. Yet a team of Brazilian researchers has shown that this seemingly mundane detail of laboratory procedure can completely reshape the chemistry of turning fatty acids into renewable fuels. In a study published in Catalysis Letters, scientists from the Federal University of Rio de Janeiro, the National Institute of Technology, INMETRO, and the State University of Rio de Janeiro prepared three versions of a nickel-rhenium catalyst on an acidic zeolite support and found that each one produced a strikingly different menu of hydrocarbons from the same oleic acid feedstock. One catalyst churned out light aromatic molecules resembling gasoline components, another preserved long-chain paraffins suited to renewable diesel, and none of them managed to do everything at once. The finding carries real weight for the emerging industry that converts vegetable oils and animal fats into drop-in fuels, where the difference between a diesel molecule and an aromatic ring can determine whether a product meets a jet-fuel specification or ends up as unwanted byproduct.

The catalyst at the heart of the study is a bifunctional material, meaning it combines two chemically distinct active sites that must work in concert. Nickel metal, dispersed on the surface, activates hydrogen gas and cleaves carbon-oxygen bonds, stripping oxygen out of fatty acid molecules. The zeolite support, HZSM-5, contributes acidic sites that can crack, isomerize, and aromatize the hydrocarbon intermediates. Rhenium, a rare and expensive metal, acts as a promoter: it improves the reducibility and dispersion of nickel and helps activate the stubborn carboxyl group that makes fatty acids so hard to deoxygenate. The catch is that conventional industrial hydrotreating catalysts, based on sulfided cobalt-molybdenum or nickel-molybdenum, require continuous sulfur dosing and release sulfur-containing effluents, while catalysts built entirely from noble metals such as platinum or palladium are prohibitively expensive. Nickel-based alternatives avoid sulfur entirely, but they are prone to sintering, coking, and runaway secondary reactions that degrade the desired products.

To probe how preparation chemistry governs these trade-offs, the researchers prepared catalysts containing nominally 6 weight percent nickel and 2 weight percent rhenium on HZSM-5 using incipient-wetness impregnation, a technique in which a metal-salt solution is added to the porous support in an amount just sufficient to fill its pores. Three routes were compared: co-impregnation, in which both metal precursors were introduced together; a nickel-first sequence; and a rhenium-first sequence. Each impregnation step was followed by vacuum drying at 120 degrees Celsius and calcination at 500 degrees Celsius. That seemingly repetitive calcination turned out to be decisive. Rhenium oxide species are volatile and mobile under oxidative treatment, and every additional calcination step gave them another chance to escape. The co-impregnated catalyst retained the full 2.0 weight percent rhenium, while the nickel-first sample retained 1.5 percent and the rhenium-first sample, which endured two calcinations with rhenium on board, kept only 1.0 percent.

Those losses cascaded through every subsequent measurement. X-ray diffraction of the reduced catalysts revealed that the co-impregnated material possessed the smallest nickel crystallites of the series, at 9.2 nanometers, along with a subtle shift of the nickel diffraction peak suggesting a modified lattice environment consistent with close nickel-rhenium contact. Transmission electron microscopy, counting roughly 400 particles per sample, painted a complementary picture: the co-impregnated catalyst showed a comparatively uniform metal distribution, the nickel-first sample contained small particles but also external aggregates, and the rhenium-first sample displayed the broadest and coarsest distribution, with a pronounced tail toward large particles and localized regions where rhenium content reached approximately 12 weight percent. Hydrogen chemisorption, interpreted cautiously because rhenium can alter hydrogen adsorption stoichiometry and promote spillover, gave the co-impregnated catalyst the highest apparent dispersion of the series at 2 percent, against 1.5 percent for both sequential variants.

The acidity of the zeolite, arguably as important as the metal itself, was also reshaped by the preparation route. The parent HZSM-5 was overwhelmingly Brønsted-acidic, with a Brønsted-to-Lewis ratio of 9.31 measured by infrared spectroscopy of adsorbed pyridine. Loading the metals barely dented the Brønsted population in the nickel-first catalyst, which preserved about 93 percent of it, but cut it by roughly a third in the other two. Meanwhile, the apparent Lewis acidity, associated with coordinatively unsaturated metal environments rather than framework aluminum, jumped by an order of magnitude in all bimetallic samples. Notably, nuclear magnetic resonance detected no extra-framework aluminum in the co-impregnated and nickel-first samples, meaning their large Lewis populations must arise from pyridine coordinating to nickel species, rhenium oxide species, and metal-support interfacial sites. When pyridine was desorbed at 350 degrees Celsius, the co-impregnated catalyst held onto 77.6 percent of its initial acid sites, the most thermally persistent population of the series.

The moment of truth came in a continuous fixed-bed reactor running oleic acid in cyclohexane at 340 degrees Celsius and 30 bar of hydrogen for 48 hours. All three bimetallic catalysts maintained high oleic acid conversion, but the composition of the liquid products diverged dramatically, especially in the first 10 hours. The co-impregnated catalyst, with its retained rhenium, small nickel domains, and persistent Lewis acidity, directed 63 to 94 percent of the quantified condensable products into alkyl aromatics, predominantly toluene and xylenes in the C7 to C9 range. Producing these molecules from an 18-carbon fatty acid requires extensive carbon-carbon bond scission followed by oligomerization, cyclization, and dehydrogenation, a punishing sequence of secondary chemistry enabled by the close coupling of metal and acid functions. Only C17 hydrocarbons were detected over this catalyst throughout the entire test, indicating that decarbonylation and decarboxylation pathways, which shed one carbon atom, dominated over carbon-conserving hydrodeoxygenation.

The rhenium-first catalyst told the opposite story. With the least retained rhenium, the largest and most heterogeneous metal particles, and the smallest accessible acid population, it suppressed the cracking and aromatization network and instead favored C17 to C18 paraffins and isoparaffins, which made up 33 to 51 percent of the quantified liquid products and included the branched molecule 2,6,10-trimethyltetradecane. Its average C17-to-C18 ratio of 3.5, however, revealed that even this catalyst was shedding carbon through decarboxylation rather than conserving it through true hydrodeoxygenation. Thermogravimetric analysis of the spent catalysts added a final dimension: apparent coke contents followed the sequence nickel-first at 7.58 percent, co-impregnated at 6.32 percent, and rhenium-first at 4.49 percent, with the monometallic nickel reference worst of all at 8.53 percent. Rhenium clearly mitigates carbon deposition, but how much it does so depends entirely on where the preparation route puts it.

Perhaps the most sobering observation came after roughly 24 hours on stream. On every catalyst, stearic acid, the fully hydrogenated but still oxygen-bearing counterpart of oleic acid, began to accumulate even as oleic acid conversion stayed high. This means the hydrogenation function survived while the deoxygenation function quietly failed, likely because carbon deposits covered metallic sites and blocked access to the zeolite micropores. The problem is compounded by molecular size: oleic acid is too bulky to slip easily through the narrow MFI micropore openings, so primary conversion probably occurs at external or pore-mouth sites, precisely where progressive coking does the most damage to the delicate cooperation between metal and acid functions. Sintering may also have contributed, though the authors note that post-reaction particle sizes were not measured, so this remains unconfirmed.

The broader lesson is that catalyst design must be matched to the fuel in question. For renewable diesel, the metallic function should dominate and strong Brønsted acidity should be restrained, through higher silicon-to-aluminum ratios, selective site passivation, or hierarchical pore architectures, in order to preserve C17 to C18 hydrocarbons. Sustainable aviation fuel demands an intermediate balance, converting primary long-chain products into the C8 to C16 range through controlled hydrocracking and isomerization without tipping into over-cracking, aromatization, or coke. Gasoline-oriented catalysts would need even greater accessible Brønsted acidity and deeper cracking toward C5 to C12 products. The study suggests that aviation-fuel catalysts should sit between the co-impregnated and rhenium-first extremes, combining stable, dispersed nickel-rhenium interfaces with moderate, spatially controlled acidity and improved molecular transport. No single impregnation route achieved oxygen removal, hydrocarbon formation, and stability simultaneously, but by demonstrating exactly how the order of a few pipette additions cascades into rhenium retention, metal dispersion, acid-site character, and coke formation, the work gives fuel chemists a powerful and surprisingly accessible tuning knob for building the next generation of renewable-fuel catalysts.

Subject of Research: Effect of impregnation sequence on Ni–Re/HZSM-5 catalyst structure and oleic acid hydrodeoxygenation selectivity

Article Title: Engineering Metal–Acid Interactions in Ni–Re/HZSM-5 for Selective Hydrodeoxygenation of Oleic Acid: The Effect of Impregnation Sequence

Article References: de Oliveira Fontoura, C. R., Dionizio, D. G. B., de Souza Júnior, R. L., Archanjo, B. S., dos Santos, R. D., Henriques, C. A., & Toniolo, F. S. (2026). Engineering Metal–Acid Interactions in Ni–Re/HZSM-5 for Selective Hydrodeoxygenation of Oleic Acid: The Effect of Impregnation Sequence. Catalysis Letters, 156(11), 307. https://doi.org/10.1007/s10562-026-05540-z

Image Credits: AI Generated

DOI: 10.1007/s10562-026-05540-z

Keywords: hydrodeoxygenation, oleic acid, Ni-Re catalyst, HZSM-5, zeolite, impregnation sequence, renewable diesel, sustainable aviation fuel, rhenium retention, coke formation, bifunctional catalysis, acid sites