The material at the heart of the work is a lithium manganese oxide cathode with the formula Li1.27Mn1.73O4, a composition typical of spent lithium-ion batteries from the lithium-manganese-oxide family. Instead of dismantling the chemistry and rebuilding the catalyst from purified metal salts — the conventional, energy-hungry approach — the team treated the spent cathode powder itself as the precursor. After removing the residual lithium using a deep eutectic solvent system based on choline chloride and lactic acid, they were left with a manganese-rich oxide framework ready to be activated. This lithium-extraction step proved critical, because the removal of lithium ions from the host lattice destabilizes the crystal structure and opens the door to defect formation on a scale that ordinary synthesis routes rarely achieve.

The second ingredient in the recipe is the heating method. Rather than ramping up a muffle furnace over hours, the researchers employed flash Joule heating, a technique in which a large electrical current is passed directly through the sample, generating intense heat within the material itself in seconds. The team systematically varied the applied current, producing a series of catalysts labeled FMn-aA, where the current ranged up to 25 amperes. For comparison, they also prepared catalysts from the same battery-derived material using conventional muffle-furnace calcination, and from analytical-grade metal salts processed both ways. This four-way comparison allowed them to isolate the individual and combined effects of the precursor source and the heating regime.

The results were striking. The optimized catalyst, designated FMn-25A, achieved more than 90 percent conversion of benzene at only 208 degrees Celsius. In the catalytic oxidation of VOCs, the temperature at which 90 percent of the pollutant is converted — known as T90 — is a key benchmark, because lower operating temperatures translate directly into lower energy costs and reduced risk of unwanted byproduct formation. Manganese oxides are well known catalysts for this chemistry, but reaching such low light-off temperatures typically requires careful doping, composite formation or noble metal promotion. Here, the researchers obtained the performance simply by choosing a waste-derived precursor and heating it in the right way.

The reason for the enhanced activity lies in the defect chemistry of the material. Detailed characterization combining X-ray diffraction, electron microscopy, X-ray photoelectron spectroscopy, hydrogen temperature-programmed reduction, oxygen temperature-programmed desorption and electron paramagnetic resonance showed that flash Joule heating combined with lithium-ion removal generated an unusually high density of oxygen vacancies — atomic-scale holes in the oxide lattice where an oxygen atom is missing. These vacancies are not passive imperfections. They are chemically active sites that change the electronic environment of neighboring manganese atoms, making the lattice easier to reduce and reoxidize, and they create anchoring points where gas-phase oxygen can adsorb and be converted into reactive oxygen species.

To trace exactly how benzene is destroyed on these surfaces, the team turned to in-situ diffuse reflectance infrared Fourier transform spectroscopy, or DRIFTS. This technique monitors the chemical species present on a catalyst surface while the actual reaction is running. The spectra revealed that the abundant oxygen vacancies promoted the formation of active oxygen species, which then attacked the benzene ring and its partially oxidized intermediates, accelerating the entire oxidation cascade toward complete conversion to carbon dioxide and water. In essence, the vacancies act as continuous suppliers of the reactive oxygen that drives the chemistry, explaining why the defect-rich battery-derived catalysts outperformed their conventionally prepared counterparts, including the CMn-MS sample made from metal salts in a muffle furnace and the salt-derived catalyst heated by flash Joule heating.

The study also underscores a broader lesson about how synthesis conditions shape catalytic materials. Conventional furnace calcination proceeds slowly and at relatively moderate temperatures, allowing the oxide lattice to relax into well-ordered, defect-poor configurations. Flash Joule heating, by contrast, delivers an extreme thermal shock followed by rapid cooling, freezing in a metastable, defect-rich structure. When applied to a lithium-depleted battery cathode — a lattice already destabilized by the extraction of lithium ions — this thermal shock produces exactly the kind of oxygen-deficient, highly reducible surface that low-temperature oxidation catalysis demands. The finding aligns with a growing body of literature on rapid Joule heating for catalyst preparation, which has shown that ultrafast thermal processing can create single-atom sites, defective graphenes and nanostructured electrocatalysts that are difficult or impossible to obtain through conventional routes.

The environmental logic of the approach is twofold. On the waste side, recycling strategies for spent lithium-ion batteries have so far concentrated on recovering lithium, cobalt and nickel from high-value ternary cathodes, while manganese-based cathodes have attracted less commercial attention and often end up in lower-value applications or landfill. Converting this material directly into environmental catalysts gives it a productive second life without demanding the purity levels required for new battery manufacturing. Earlier work by the same research group and others has already demonstrated that spent battery materials can serve as precursors for ammonia-selective catalytic reduction catalysts and for catalysts that simultaneously remove nitrogen oxides and VOCs. The present study extends that concept to lithium manganese oxide chemistry and, crucially, combines it with ultrafast electrified heating rather than energy-intensive furnace treatment.

On the pollution side, catalytic oxidation is one of the most effective technologies for destroying VOC emissions, but its industrial adoption has been constrained by the high temperatures — often above 300 degrees Celsius — that conventional catalysts require to achieve complete conversion. Every hundred degrees saved is a meaningful cut in operating energy and cost, particularly for dilute emission streams from painting, printing, petrochemical and smelting operations, sectors that are significant VOC sources and the focus of tightening regulation. A catalyst that reaches 90 percent benzene conversion at 208 degrees Celsius, made from waste rather than fresh chemicals and activated without prolonged furnace heating, addresses the economic and sustainability dimensions of the problem simultaneously. The authors describe the work as a sustainable strategy for green industrial VOC abatement, and the research was supported by China’s National Science and Technology Major Project on regional environmental improvement in the Beijing-Tianjin-Hebei area.

Significant steps remain before the concept reaches industrial deployment. Real emission streams contain water vapor, sulfur compounds and mixtures of VOCs that can poison or deactivate catalysts, and the long-term stability of a defect-rich structure under continuous operation must be demonstrated. Scaling flash Joule heating from laboratory powder samples to the throughput of a working air-cleaning unit is another engineering challenge, although the process is inherently fast and electrically driven, which favors energy efficiency and compatibility with renewable power. Nevertheless, the study offers a compelling proof of principle: the defective, highly reactive materials that catalyst designers usually struggle to engineer can emerge naturally — and almost instantly — from the wreckage of yesterday’s batteries. In turning one environmental liability into the cure for another, the work captures the kind of cross-cutting chemistry that may define the next generation of sustainable pollution control.

Subject of Research: Flash Joule heating fabrication of defect-rich manganese oxide catalysts from spent lithium-ion battery cathodes for low-temperature VOC oxidation

Article Title: Flash Joule Heating–Driven Fabrication of Defect-Rich Catalysts Derived from Spent Li1.27Mn1.73O4 Batteries for Low-Temperature VOCs Oxidation

Article References: Wang, X., Zheng, Y., Guo, H. Y., Duan, X. H., Gong, C., & Xue, G. (2026). Flash Joule Heating–Driven Fabrication of Defect-Rich Catalysts Derived from Spent Li1.27Mn1.73O4 Batteries for Low-Temperature VOCs Oxidation. Catalysis Letters, 156(10), Article 283. https://doi.org/10.1007/s10562-026-05527-w

Image Credits: AI Generated

DOI: 10.1007/s10562-026-05527-w

Keywords: flash Joule heating, oxygen vacancies, spent lithium-ion batteries, benzene catalytic oxidation, manganese oxides, volatile organic compounds, low-temperature catalysis, lithium-ion removal, deep eutectic solvents, DRIFTS, catalyst recycling, VOC abatement

Selective hydrogenation of furfural traditionally suffers from a compromise between catalytic activity and product selectivity. Conventional nickel catalysts, known for their high hydrogen dissociation capability, often promote over-hydrogenation and side reactions, leading to lower selectivity and undesired byproducts. The introduction of iron into the nickel matrix fundamentally changes the catalyst’s electronic properties and active site distribution. Fe incorporation effectively dilutes continuous nickel surface sites, preventing agglomeration and limiting undesired reactions. More importantly, iron plays a vital role in activating the carbonyl group of furfural, which is the primary functional group targeted during selective hydrogenation.

The synergistic interactions between Ni and Fe provide distinct active sites that perform complementary catalytic functions. Nickel sites predominantly facilitate the cleavage and dissociation of molecular hydrogen (H₂), a crucial step for the hydrogenation process. Meanwhile, iron sites specialize in the activation of the carbonyl (C=O) moiety of furfural molecules, enhancing substrate adsorption and orienting the molecule favorably for selective transformation. This dual-site mechanism elegantly resolves the classic dilemma where high activity often leads to poor selectivity, by spatially and electronically separating the reaction steps on the catalyst surface.

Extensive characterization techniques including X-ray diffraction, transmission electron microscopy, and in-situ spectroscopic analyses confirm the formation of a uniform bimetallic alloy with finely distributed Ni and Fe atoms. The atomic-level modulation alters the electronic density around nickel atoms, reducing their propensity for excessive hydrogenation while promoting selective conversion pathways. These modifications were corroborated by density functional theory (DFT) calculations which revealed significant charge transfer effects and altered adsorption energies for the furfural reactant and key intermediates.

The catalyst showcases not only exceptional catalytic performance but also commendable stability and recyclability over multiple reaction cycles without noticeable loss in activity or selectivity. This durability positions the NiFe catalyst as a practical candidate for scalable industrial applications, addressing sustainability concerns associated with catalyst deactivation and precious metal usage. The economic and environmental benefits derived from using earth-abundant metals like nickel and iron further elevate this catalyst’s appeal for green chemical manufacturing.

This research paves the way for future exploration of bimetallic and multimetallic catalysts designed with precise atomic tuning to tailor reaction pathways. The principles highlighted by the NiFe system—activity enhancement through synergistic active sites and electronic modulation—can be extended to other biomass-derived transformations and selective hydrogenations, broadening the horizon of sustainable catalysis. Moreover, the findings challenge the conventional perception of catalyst design, emphasizing the need for balancing site-specific functionalities rather than merely maximizing active surface area.

The study’s implications extend beyond academic curiosity; transforming furfural efficiently to furfuryl alcohol with high selectivity significantly impacts industries reliant on biobased chemicals. Such advancements align with global efforts to shift from fossil fuel-derived feedstocks to renewable resources, promoting a circular economy and reducing carbon footprint. This catalyst therefore represents a critical technological leap supporting sustainable industrial processes.

Published in the renowned Chinese Journal of Catalysis, this research embodies extensive collaboration and meticulous experimental and theoretical work. The multidisciplinary approach integrating surface science, materials engineering, and computational modeling sets a benchmark for future catalytic investigations. The journal itself holds a prestigious position in applied chemistry research, underscoring the significance of these findings to the global scientific community.

In summary, the tailored bimetallic NiFe catalyst presents a compelling solution to an entrenched challenge in biomass conversion chemistry. By harnessing the complementary roles of nickel and iron at the atomic scale, the catalyst delivers superior performance marked by exceptional selectivity, activity, and durability. These attributes hold vast promise for manufacturing high-purity furfuryl alcohol and possibly other value-added chemicals from renewable sources, heralding a new era in green catalysis.

This advancement exemplifies how sophisticated catalyst design leveraging electronic structure engineering and atomistic synergy can revolutionize traditional chemical processes. With further optimization and scale-up, the NiFe catalyst is poised to become a game-changer in biorefinery technologies and sustainable chemical manufacturing.

Subject of Research: Catalytic selective hydrogenation of biomass-derived furfural using a bimetallic nickel-iron catalyst.

Article Title: Promotion effect of Fe in the selective hydrogenation of furfural with bimetallic NiFe catalysts

News Publication Date: 30-Mar-2026

Web References:

• Chinese Journal of Catalysis
• DOI: 10.1016/S1872-2067(26)64959-5

Image Credits: Chinese Journal of Catalysis

Keywords

bimetallic catalyst, NiFe catalyst, furfural hydrogenation, selective hydrogenation, biomass conversion, furfuryl alcohol, catalyst design, electronic structure modulation, sustainable catalysis, green chemistry, activity-selectivity trade-off, synergistic active sites

Heterogeneous catalysis lies at the heart of numerous industrial and environmental processes, facilitating chemical reactions by providing active surfaces where reactants can interact. The efficiency of these catalysts hinges critically on the architecture of their porous structures, which govern the accessibility, diffusion, and reaction of molecules. However, the complexity and heterogeneity of these porous networks have long posed formidable challenges to experimental characterization and predictive modeling. Traditional imaging and analytical methods often fall short of capturing the spatial and temporal nuances of reactive transport within these materials.

Enter deep learning—a subset of artificial intelligence that excels at extracting meaningful patterns from high-dimensional data. By applying computer vision models trained to interpret complex images, the research team has devised a method to effectively map and analyze porous architectures with remarkable resolution and detail. This deep learning framework leverages convolutional neural networks (CNNs) capable of discerning subtle features in microscopy images, enabling a more nuanced understanding of pore connectivity and distribution than ever before.

Crucially, the researchers incorporated transfer learning into their approach, a technique where a model pre-trained on one dataset is adapted to a related but distinct task. This strategic employment of transfer learning circumvented the need for vast amounts of annotated catalytic data, a common bottleneck in materials informatics. By fine-tuning models initially trained on large, generic image repositories, they harnessed pre-existing knowledge to accelerate learning and enhance predictive accuracy in analyzing catalytic materials.

The power of this methodology was demonstrated through comprehensive visualization of the nexus between porous architecture and reactive transport pathways. Reactive transport—the movement and interaction of reactants within catalyst pores—is a dynamic process that is difficult to capture experimentally. The study’s models successfully predicted how molecular species traverse these porous networks, highlighting preferential channels and identifying bottlenecks that impact catalytic performance.

This holistic visualization framework offers a powerful tool for rational catalyst design. By revealing the intimate relationship between structural morphology and chemical reactivity, it informs targeted modifications of pore geometry to optimize mass transport and surface reactions. This insight is pivotal for improving catalyst lifetime, selectivity, and overall efficiency, all of which are vital parameters in the development of greener chemical processes.

Moreover, the integration of deep learning models with experimental data facilitates a feedback loop for continuous improvement. The researchers emphasize how iterative training with new imaging inputs can refine model predictions and adapt to diverse catalytic systems, including those with complex materials compositions or non-standard pore shapes. This adaptability signals a versatile platform that could be generalized to a broad spectrum of catalytic materials.

Another noteworthy aspect of the work lies in its potential to accelerate catalyst screening and discovery. Conventional trial-and-error approaches are both time-consuming and resource-intensive. By contrast, the deep learning paradigm allows rapid virtual screening of porous architectures before experimental synthesis, drastically reducing development cycles. This data-driven acceleration aligns well with the goals of sustainable chemistry, aiming to minimize waste and energy consumption.

The fusion of advanced AI techniques with catalysis research also underscores the growing interdisciplinary nature of modern science. The project exemplifies how computational sciences, materials characterization, and chemical engineering can coalesce to tackle longstanding scientific puzzles. Such collaboration is essential for pushing the boundaries of what can be observed and understood at the nanoscale within reacting systems.

On a technical level, the study detailed how spatial resolution in microscopy images was enhanced through multi-scale feature extraction, enabling the capture of both macroscopic pore connectivity and microscopic surface irregularities. The inclusion of reactive transport modeling incorporated principles from reaction-diffusion theory, further enriching the physical realism of predictions. These innovations represent a significant stride in integrating physics-based modeling with data-centric AI approaches.

The researchers also highlighted potential challenges and future directions, noting that extending this methodology to real-time in situ observations under operational catalysis conditions would mark the next frontier. Combining time-resolved spectroscopy and electron microscopy with AI-driven analysis could unravel transient phenomena such as catalyst deactivation or structural evolution during reactions, areas currently elusive due to measurement limitations.

Industry stakeholders stand to benefit immensely from these findings, particularly in sectors like petrochemicals, renewable energy, and environmental remediation. Enhanced catalyst designs driven by AI-enabled insights could lead to more cost-effective processes, reduced greenhouse gas emissions, and improved resource utilization. The emergent paradigm demonstrated by Yu and colleagues reflects a step toward smarter, more sustainable chemical manufacturing.

Beyond direct applications, this study serves as a compelling example of how machine learning methodologies can transform traditional scientific disciplines. As AI continues to mature, its role in decoding complex natural and engineered systems will only expand, rendering previously hidden aspects of materials behavior visible and quantifiable.

The collaboration highlighted in the publication also demonstrates the growing importance of open data and model sharing. By making trained models and datasets accessible, the team paves the way for reproducibility and community-driven innovation, accelerating collective progress in catalysis research and materials science at large.

As catalysts form the backbone of numerous processes integral to modern society—from synthesizing pharmaceuticals to converting biomass—the ability to visualize and optimize their internal architecture with such precision marks a pivotal moment. This convergence of AI, materials characterization, and chemical engineering sets the stage for a new era of catalyst innovation, one defined by insight, efficiency, and sustainability.

In summary, the pioneering work by Yu et al. harnesses the transformative power of deep learning computer vision and transfer learning to illuminate the intricate interplay between porous architecture and reactive transport in heterogeneous catalysis. Their approach extends beyond mere visualization, providing actionable insights that promise to accelerate catalyst design and development in pursuit of more efficient and environmentally conscious chemical processes. As this interdisciplinary methodology matures, it will undoubtedly inspire further breakthroughs at the nexus of AI and materials science, reshaping how researchers understand and engineer catalytic systems.

Subject of Research: Visualization and analysis of porous architecture and reactive transport in heterogeneous catalysis using deep learning computer vision and transfer learning.

Article Title: Visualizing nexus of porous architecture and reactive transport in heterogeneous catalysis by deep learning computer vision and transfer learning.

Article References:
Yu, Y., Wu, B., Wei, R. et al. Visualizing nexus of porous architecture and reactive transport in heterogeneous catalysis by deep learning computer vision and transfer learning. Nat Commun 16, 8107 (2025). https://doi.org/10.1038/s41467-025-63481-4

Image Credits: AI Generated