The study, published in Catalysis Letters, addresses a problem that has plagued furfural hydrogenation for decades. When furfural meets a hydrogenation catalyst, it has two chemically distinct destinations. Hydrogen can add to the carbonyl group of the aldehyde, producing furfuryl alcohol, the desired product, or it can attack the furan ring itself, over-hydrogenating the molecule all the way to tetrahydrofurfuryl alcohol, which requires more hydrogen and more severe conditions. Side reactions such as acetalization with alcohol solvents, decarbonylation to furan, and condensation of furfuryl alcohol on acidic surfaces further erode the yield. The balance of these competing pathways is exquisitely sensitive to the chemistry of the catalyst surface, particularly the density and strength of acid sites on the support and the electronic and geometric state of the metal particles dispersed on it.

Copper has long been favored for this transformation because it preferentially hydrogenates the aldehyde function while leaving the aromatic furan ring untouched. But bare copper catalysts supported on acidic oxides suffer from two chronic weaknesses. First, strongly acidic support sites catalyze the resinification of furfural and furfuryl alcohol, converting valuable product into insoluble humins that poison the catalyst and lower selectivity. Second, copper particles sinter and oxidize under reaction conditions, progressively losing activity until the process must be shut down for regeneration. The Chinese team, comprising Qihang Ye, Zhaoping Zhong, Yuxuan Yang, Wei Wang, You Jia, Qi Xiong, Huanqi Chen, and Xiang Zheng, reasoned that a carbon layer introduced onto the catalyst support could simultaneously moderate the acid strength of the surface and shield the copper species from deactivation.

To test this hypothesis, the researchers prepared a family of carbon-modified, copper-loaded catalysts in which carbon-containing organic precursors were used to build a carbon encapsulation layer on the support before and during the dispersion of the copper active phase. The amount of carbon introduced became a tunable knob: too little carbon left the harsh acidity of the pristine support intact, while too much began to block pores and bury the very active sites the reaction depends on. Through systematic optimization of the carbon loading and of the reaction conditions, the team arrived at a configuration that delivered complete conversion of furfural with a furfuryl alcohol selectivity of 99.17 percent at a reaction temperature of just 160 degrees Celsius over four hours, using isopropanol as the solvent medium.

The mechanistic explanation for this exceptional performance emerges from a battery of characterization techniques the authors deployed, including X-ray diffraction, Brunauer–Emmett–Teller and Barrett–Joyner–Halenda porosimetry, scanning electron microscopy, X-ray photoelectron spectroscopy, inductively coupled plasma analysis, hydrogen temperature-programmed reduction, Fourier-transform infrared spectroscopy, pyridine-adsorbed FTIR, and ammonia temperature-programmed desorption. Together, these measurements revealed that the carbon loading did something chemically subtle: it converted the strong acid sites on the support into medium-strong acid sites. That shift matters because strongly acidic sites promote the condensation and resinification side reactions that destroy furfuryl alcohol, whereas medium-strength sites can participate productively in adsorbing and activating furfural without triggering destructive chemistry.

Equally important was the second consequence of the carbon treatment: an increase in the ratio of Lewis to Brønsted acid sites on the catalyst surface. Lewis acid sites, which are coordinatively unsaturated metal or metal-oxygen centers, are known to coordinate the oxygen atom of the furfural carbonyl group, polarizing the carbon–oxygen double bond and making it more susceptible to hydrogen attack from adjacent copper sites. Brønsted sites, by contrast, donate protons and promote the oligomerization chemistry that generates the carbonaceous deposits known to clog and deactivate hydrogenation catalysts. By raising the Lewis-to-Brønsted ratio, the carbon layer effectively steered the surface population of acid sites toward the geometry that favors aldehyde activation and away from the one that accelerates deactivation. The well-dispersed copper species identified on the carbon-modified surface then supply the hydrogenation function, working in concert with the re-engineered acid sites in a bifunctional arrangement that has become the design paradigm for modern furfural conversion catalysts.

Durability, the quality that separates laboratory curiosities from industrial candidates, was demonstrated through five consecutive catalytic cycles, in which the carbon-encapsulated catalyst maintained its high furfuryl alcohol selectivity with only minor losses in performance. The authors attribute this stability to two reinforcing features: the protective carbon layer, which physically and chemically shields the underlying support from direct contact with reactive intermediates and slows the migration and sintering of copper, and the consistently well-dispersed state of the copper species on the catalyst surface, which preserves the high density of accessible active sites cycle after cycle. Thermogravimetric analysis supported the picture of a catalyst resistant to the carbon deposition that degrades unmodified analogues, a conclusion with direct implications for the economics of continuous biomass upgrading operations where catalyst replacement and regeneration costs dominate operating budgets.

To probe the chemistry at the molecular level, the team turned to density functional theory calculations using the Vienna Ab initio Simulation Package, employing the Perdew–Burke–Ernzerhof generalized gradient approximation with projector augmented-wave potentials and analyzing the density of states and projected density of states of the adsorption configurations. The computations showed that furfural preferentially adsorbs on the catalyst surface through the η¹(O)-aldehyde configuration, meaning the molecule anchors through a single oxygen atom of the carbonyl group rather than lying flat through the furan ring. This adsorption geometry is precisely the one that exposes the carbonyl carbon to hydrogenation while protecting the ring from over-reduction, and its energetic preference on the carbon-modified surface provides a quantum-mechanical explanation for both the selectivity and the promotional role of the carbon layer observed experimentally. In effect, the surface chemistry funnels furfural down the furfuryl alcohol pathway by making the correct orientation of the adsorbed molecule the most stable one.

The broader significance of the work lies in how it reframes catalyst design for biomass valorization. Rather than treating carbon in a catalyst purely as a mechanical coating or an inert dopant, the study demonstrates that carbon loading functions as an electronic and acidic regulator, one that can be introduced with inexpensive organic precursors and tuned to shift the surface acidity, the Lewis-to-Brønsted balance, and the microenvironment of the metal phase all at once. Because furfural is produced at the scale of hundreds of thousands of tons per year and furfuryl alcohol commands a substantial premium over its parent aldehyde, even incremental gains in selectivity translate into meaningful economic and environmental returns, reducing hydrogen consumption, solvent losses, and waste generation. The results also dovetail with a growing body of literature on carbon-encapsulated and carbon-coated metal catalysts for furfural and cinnamaldehyde transformations, suggesting that the carbon-layer strategy discovered here may generalize to other oxygenate upgrading reactions across the platform-chemical landscape.

The research was supported by the National Key Research and Development Program of China, the China Postdoctoral Science Foundation, and the Taizhou Key Science and Technology Programme Projects. Corresponding author Zhaoping Zhong and his colleagues emphasize that data will be made available on request, and the team reports no competing interests. With a near-perfect selectivity achieved at moderate temperature, a catalyst that survives five reaction cycles intact, and a mechanistic narrative that connects carbon engineering to acid-site chemistry and first-principles adsorption energetics, the study offers the furfural industry a concrete blueprint for next-generation hydrogenation catalysts built from abundant copper and a whisper of carbon. As biomass-derived molecules continue their march into the fuel and chemical sectors, the ability to sculpt a catalyst surface with something as humble as a carbon layer may prove to be one of the quiet breakthroughs on which the bioeconomy’s chemical foundations are rebuilt.

Subject of Research: Carbon-layer-modified copper catalysts for the selective hydrogenation of biomass-derived furfural to furfuryl alcohol.

Article Title: Cu-Loaded Catalysts for Efficient Hydrogenation of Furfural to Furfuryl Alcohol: Modulation of Hydrogenation Performance by Introducing Carbon Layer

Article References: Ye, Q., Zhong, Z., Yang, Y., Wang, W., Jia, Y., Xiong, Q., Chen, H., & Zheng, X. (2026). Cu-Loaded Catalysts for Efficient Hydrogenation of Furfural to Furfuryl Alcohol: Modulation of Hydrogenation Performance by Introducing Carbon Layer. Catalysis Letters, 156(10), Article 282. https://doi.org/10.1007/s10562-026-05493-3

Image Credits: AI Generated

DOI: 10.1007/s10562-026-05493-3

Keywords: Furfural, Hydrogenation, Furfuryl alcohol, Copper catalyst, Carbon-encapsulated catalyst, Lewis acid sites, Brønsted acid sites, Biomass valorization, Catalytic stability, DFT calculations, Selective hydrogenation, Platform chemicals

The appeal of waste apples as a fermentation feedstock lies in their biochemical makeup. Spoiled Jonagold apples supplied by Werder Frucht GmbH and prepared at ATB were pressed into a sugar-rich mash containing about 71.6 grams per liter of fructose, 21.5 grams per liter of glucose, and 7.5 grams per liter of sucrose, along with a high moisture content of 88.9 percent. Unlike apple pomace, the fibrous solid residue left after juice extraction that has dominated previous research, the mash retains readily fermentable simple sugars and requires minimal pretreatment. The solid fraction, composed mainly of peel and seeds, was excluded from the study. This compositional simplicity matters because lignocellulosic residues typically demand energy-intensive pretreatment to release their sugars and often generate inhibitory by-products that slow microbial growth.

Before fermentation could begin, the researchers needed to prepare the mash as a workable medium. They tested several commercial enzyme preparations, including Pectinase L40, Cellic CTec3 HS, Dextrozyme GA, and Viscoferm, at the mash’s natural pH of 4.5 and a temperature of 50 degrees Celsius. Surprisingly, none of the enzymes significantly increased the total reducing sugar content compared with an untreated control, a result the team attributes to the absence of starch in ripe apples and the low levels of cellulose and hemicellulose in the dry matter. However, Cellic CTec3 HS produced a marked reduction in viscosity, which is critical for fermentation because thick media impede mass transfer, create gradients of pH, temperature, and nutrients, and complicate downstream separation. The researchers ultimately selected a low-dose combination of Pectinase L40 and Cellic CTec3 HS, each at 0.5 milliliters of enzyme per kilogram of biomass, to liquefy the mash for subsequent microbial screening.

With a liquefied substrate in hand, the team screened microbial candidates for both target products. For lactic acid, five strains of Heyndrickxia coagulans, formerly known as Bacillus coagulans, were drawn from ATB’s internal collection of 700 isolates. Optical density measurements in apple mash medium, with and without yeast extract supplementation, identified strains A35, A138, and A203 as the strongest performers. For succinic acid, the researchers compared Actinobacillus succinogenes DSM 22257 against several strains of Basfia succiniproducens, confirming earlier reports that A. succinogenes converts both pentose and hexose sugars to succinic acid with superior yields. Subsequent bioreactor screening at 0.25 liters showed that all three H. coagulans strains produced lactic acid at similar concentrations, between 71.8 and 75.5 grams per liter, with yields of 0.91 to 0.94 grams per gram of sugar consumed. Strain A203, however, achieved the highest productivity at 3.5 grams per liter per hour and showed no lag phase, making it the clear choice for scale-up.

Nutrient availability emerged as a decisive factor in both fermentations. When yeast extract was omitted, lactic acid production by H. coagulans A203 collapsed to 32.3 grams per liter after 49 hours, with productivity falling to 0.7 grams per liter per hour. The effect was even more dramatic for A. succinogenes, whose succinic acid output dropped from 35.6 to 6.7 grams per liter without supplementation. These results indicate that apple mash, despite its abundant sugars, lacks sufficient nitrogen and growth factors to sustain industrial fermentation performance. The authors note that replacing yeast extract with cheaper agro-industrial nitrogen sources, such as wine lees or tomato pomace, represents a promising avenue for reducing production costs, since nutrient supply is one of the main economic burdens in biobased organic acid manufacturing.

The central innovation of the study lies in what the researchers removed from the process rather than what they added. Conventional bioprocesses typically separate the enzymatic hydrolysis and fermentation stages with centrifugation to remove solids and sterilization, usually at 121 degrees Celsius under pressure, to eliminate contaminants. Each of these steps adds capital cost, energy demand, and processing time. Instead, the team performed enzymatic liquefaction and fermentation sequentially in the same vessel. For lactic acid, no sterilization was needed at all because H. coagulans A203 is thermophilic, growing optimally at 50 degrees Celsius, a temperature that suppresses most contaminating microbes and shortens the fermentation window. For succinic acid, where A. succinogenes prefers a mesophilic 37 degrees Celsius, the researchers inserted a simplified thermal inactivation step, heating the mash to 80 to 85 degrees Celsius for 15 minutes, which denatures the proteins and disrupts the membranes of most spoilage organisms without the energy burden of full autoclaving.

After validating this simplified procedure at 1-liter laboratory scale, where it matched the performance of the conventional approach with no significant differences in concentration, yield, or productivity, the team moved to pilot scale. Lactic acid fermentation was carried out in a 30-liter working volume using 30 kilograms of apple mash in a Biostat UD bioreactor. The process delivered 73.8 grams per liter of lactic acid with a yield of 0.91 grams per gram of sugar consumed and a productivity of 2.7 grams per liter per hour. Remarkably, the final product exhibited an optical purity of 99.7 percent L-lactic acid, a strong indicator that contamination never took hold, since contaminating bacteria typically produce a mixture of L- and D-isomers. The overall product yield reached 0.78 grams of lactic acid per gram of apple mash on a dry basis, and the bacteria also consumed the malic acid naturally present in the fruit.

The succinic acid pilot run, conducted with 20 kilograms of apple mash in a 25-liter working volume, produced 36.8 grams per liter of succinic acid with a yield of 0.69 grams per gram and a productivity of 1.0 gram per liter per hour, alongside acetic and formic acids as natural by-products of A. succinogenes metabolism. Notably, the strain fixed carbon dioxide during succinate formation, an inherent environmental advantage of the process. These figures compare favorably with previous pilot-scale succinic acid fermentations using other food wastes: oat pomace with acid whey yielded only 19.6 grams per liter at 0.27 grams per liter per hour, while industrial candy waste reached 38.99 grams per liter. The lactic acid results similarly outperformed earlier work on apple pomace hydrolysate, which achieved just 40.72 grams per liter at a productivity of 0.58 grams per liter per hour, and mixed food waste, which produced 68.5 grams per liter at a yield of only 0.38 grams per gram of total solids.

Beyond the headline numbers, the study carries significant implications for how biorefineries handle agri-food waste. Retaining solids in the reactor eliminates an intermediate solid-liquid separation step, meaning only a single separation is needed after fermentation, when microbial biomass must be removed regardless of process configuration. Avoiding intermediate autoclaving also prevents heat-induced degradation of sugars and proteins, limiting the formation of unwanted by-products that complicate downstream purification, which is widely regarded as the economic bottleneck of large-scale bioproduction. Economically, waste apples serve as a zero- or low-cost feedstock that displaces commercial sugars and sidesteps the expensive pretreatment required for lignocellulosic biomass. Environmentally, the simplified process reduces energy consumption and carbon dioxide emissions, the high moisture content of the mash minimizes water input, and diverting spoiled fruit from landfill avoids methane generation.

The authors frame their work as a template for integrating agri-food residues into a circular bioeconomy, in line with the European Union’s Bioeconomy Strategy and Waste Framework Directive. They point toward future advances in intelligent biorefineries that use artificial intelligence and machine learning to adapt fermentation conditions to feedstock variability, as well as genetically engineered microbial strains designed to boost fermentation rates and product diversity. For now, the demonstration that a pilot-scale bioreactor can convert 30 kilograms of rotten apples into nearly 74 grams per liter of high-purity lactic acid marks a tangible step toward turning one of agriculture’s most familiar waste streams into a dependable industrial resource.

Subject of Research: Pilot-scale biotechnological production of lactic and succinic acids from waste apples

Article Title: Upcycling waste apples into platform chemicals: pilot-scale production of lactic and succinic acids

Article References: Portugal Rios da Costa Pereira, L., Schneider, R., Olszewska-Widdrat, A., Sturm, B., & Kaetzl, K. (2026). Upcycling waste apples into platform chemicals: pilot-scale production of lactic and succinic acids. Biotechnology for Biofuels and Bioproducts, 19(1), Article 67. https://doi.org/10.1186/s13068-026-02807-w

Image Credits: AI Generated

DOI: 10.1186/s13068-026-02807-w

Keywords: waste apples, lactic acid, succinic acid, fermentation, pilot scale, biorefinery, circular bioeconomy, Heyndrickxia coagulans, Actinobacillus succinogenes, agri-food waste, platform chemicals, bioplastics