Green pathways to low-sulfur diesel: advances and challenges in desulfurization technologies

Abstract

Diesel desulfurization is crucial for producing ultra-low-sulfur diesel to meet stringent environmental regulations and reduce harmful emissions. Traditional hydrodesulfurization (HDS) processes, while widely used, face limitations in removing refractory sulfur compounds such as DBT and its derivatives, and require harsh reaction conditions, including high pressure and hydrogen consumption. Consequently, non-HDS technologies have emerged as promising alternatives, offering milder operating conditions and reduced environmental impact. This review systematically compares various desulfurization strategies, both HDS and non-HDS methods, including bio-desulfurization, adsorptive desulfurization, oxidative desulfurization, extractive desulfurization, and coupled systems, with a focus on their reaction mechanisms, advantages, current research progress, and limitations. Finally, the challenges of these methods and achieving industrial viability are addressed, and future directions are proposed for the development of green, efficient, and sustainable desulfurization technologies. This review aims to provide a comprehensive understanding of the state-of-the-art desulfurization technologies and their potential for advancing cleaner fuel production in the context of global environmental goals.

---

Peiwen Wu

Peiwen Wu received his Ph.D. in Power Engineering and Engineering Thermophysics from Jiangsu University in 2018, during which he was a joint Ph.D. student at Oak Ridge National Laboratory (USA) from 2016 to 2018. He is currently an Associate Professor and Ph.D./M.S. supervisor at Jiangsu University. His research focuses on molecular refining in the petrochemical field and the design of advanced materials, including ionic liquids, two-dimensional materials, and high-entropy materials for desulfurization and selective conversion of heavy oil components.

Shaojie Ma

Shaojie Ma is a Master's degree candidate in Wenshuai Zhu's research group at Jiangsu University. His main research focuses on the design and synthesis of deep eutectic solvents and their application in extractive desulfurization and denitrogenation of fuel oils.

Wenshuai Zhu

Wenshuai Zhu is currently a Professor and Ph.D. supervisor at China University of Petroleum (Beijing). He received his Ph.D. from Jiangsu University in 2009 and was a visiting scholar at Oak Ridge National Laboratory (USA) from 2015 to 2016. His research focuses on efficient energy and resource conversion and the design of advanced functional materials, including clean and value-added utilization of oil and gas, hydrogen and CO2 conversion, and the development of catalysts and adsorbents enhanced by artificial intelligence and additive manufacturing.

---

Green foundation

1. We discuss greener diesel desulfurization strategies in line with green chemistry principles, systematically covering hydrodesulfurization (HDS), biodesulfurization, extractive desulfurization, adsorptive desulfurization, oxidative desulfurization, and coupled processes.

2. HDS is mature and effective but requires harsh conditions, high hydrogen consumption, and shows limited efficiency for refractory sulfur species. Non-HDS approaches offer milder operation, lower energy demand, and reduced environmental impact. Biodesulfurization uses microorganisms under ambient conditions but needs higher activity and stability; extractive and adsorptive methods require selective, recyclable solvents or adsorbents; oxidative desulfurization enables deep sulfur removal at mild conditions yet faces catalyst and oxidant sustainability challenges.

3. Future directions include multifunctional catalysts, bio-chemical hybrid systems, recyclable materials, and renewable oxidants to achieve efficient, low-carbon, and economically viable desulfurization for ultra-low-sulfur diesel production.

1. Background

1.1. Energy around the world

Since the Second Industrial Revolution, the consumption of fossil energies, particularly petroleum products, has increased gradually.^1–4 These products are not only essential as energy fuels for diesel engines, ships, spacecraft, and other applications, but they are also important raw materials in the chemical industry.^5–7 Petrochemicals currently represent 90% of the total feedstock demand in chemical production.^8,9 Since the 21st century, with the deterioration of the human living environment, decreasing energy reserves, and rising energy consumption, many countries around the world have proposed different strategies to overcome the disadvantages.^10–12 On the one hand, various green and clean alternative energy sources have been developed to fulfill the energy demand and reduce pollution during the energy development and utilization process, although fossil energies still dominate energy consumption (Fig. 1).^13–16 New energy carriers such as hydrogen (H₂), electricity, and solar energy have been quickly developed,^17–22 but they have lower energy densities,^22,23 making them unsuitable for applications such as marine and aerospace engines that require high energy density and energy per unit mass (or volume), and no alternative energy has yet emerged beside fossil energies at a large scale.^24–28 On the other hand, methods and strategies for the green and efficient use of traditional fossil energy are continuously being developed.^6,29–31 To achieve cleaner and more efficient utilization of fossil energies,^32–34 advanced technologies such as ultra-supercritical power generation and oxy-fuel combustion can significantly improve energy efficiency while reducing emissions. Carbon capture, utilization and storage (CCUS) systems help mitigate greenhouse gas impacts by capturing CO₂ for reuse or secure storage.^35–37 Furthermore, integrating fossil fuels with renewable energy systems and employing smart digital management solutions can optimize overall energy use and minimize environmental footprint.^38–41 These approaches enable fossil fuels to serve as a transitional energy source while supporting the shift toward sustainable energy systems.

Fig. 1 Global primary energy consumption by source, adapted from ref. 13 with permission from Global Change Data Lab¹³ under Creative Commons BY license, copyright 2020.

1.2. The importance of desulfurization of oil products

Crude oil and its refined products typically contain certain amounts of organic sulfur compounds, such as mercaptans, sulfides, thiophenes, dibenzothiophenes (DBT), and their derivates.^42–46 During the combustion of diesel, these sulfur-containing compounds generate sulfur oxides (SO_x), sulfuric acid mist, and fine particulate matter (PM), leading to acid rain, air pollution, and ecological disadvantages, thus posing serious threats to human health and environmental safety.^47–52 To reduce the impact of SO_x emissions on the atmospheric environment, countries around the world have promoted increasingly strict regulations on sulfur contents in fuel oils (Table 1).^53–57 For example, the European Union's Euro VI standard limits the sulfur content of diesel fuel to ≯10 ppm;⁵⁸ China has fully implemented the National VI standard since 2020, limiting the sulfur content of gasoline and diesel to be ≯10 ppm; the International Maritime Organization (IMO) has enforced a global sulfur limit of ≯0.5 wt% for marine fuels since 2020 (with some inland waters not exceeding 0.1 wt%).⁵⁹ The implementation of these regulations has raised higher requirements for product quality and environmental protection technologies in refineries, making efficient desulfurization one of the core tasks in the refining process.

Table 1 Gasoline and diesel sulfur regulations in major countries and regions^60–63

Region/country	Gasoline (ppm)	Diesel (ppm)	Year
European Union	10	10	2009
U.S.	10	10	2017
China	10	10	2020/2023
Japan	10	10	2005
Korean	10	10	2015
Canada	10	15	2005
Australia	10	10	2009

---

In addition, aromatic sulfur compounds in diesels have a significant catalyst poisoning effect on subsequent hydrogenation catalytic processes, such as hydrotreating and catalytic reforming.^64–66 Aromatic heterocyclic sulfur compounds readily undergo strong adsorption reactions with the active centers on the catalyst surface under high-temperature hydrogenation conditions, forming metal sulfides that cover catalytic sites, thereby reducing catalytic efficiency and causing catalyst deactivation. This results in decreased selectivity and conversion efficiency in reactions like aromatization and isomerization, affecting product quality.^67–69 Moreover, sulfur compounds in diesel can cause product color deepening, decreased stability, and increased corrosiveness.^70,71 The trace amounts of DBT derivatives remaining in diesel not only affect combustion stability but also catalyze particulate matter formation in the combustion chamber, posing potential risks to engine lifespan and equipment safety.⁷² Simultaneously, sulfur compounds react with metal surfaces to form corrosive compounds, leading to equipment corrosion and increased maintenance costs.⁷³ High-sulfur diesel requires higher reaction temperatures and H₂ pressures in the hydrotreating unit to achieve desulfurization, which not only increases hydrogen consumption, reaction pressure, and temperature inputs but also significantly raises indirect CO₂ emissions, contrary to the trend of green, low-carbon development.^74–76

Furthermore, the resource utilization of desulfurization byproducts is of great importance. Hydrogen sulfide generated during conventional HDS can be further converted into sulfur, which is used in industries such as fertilizers, rubber, and fine chemicals.^77–81 Some high-value-added organic sulfur compounds are also expected to be selectively removed and resourcefully recovered, providing a new economic growth point for the refining process. Currently, based on the maturity of traditional HDS technology, technologies such as biodesulfurization (BDS), adsorptive desulfurization (ADS), extractive desulfurization (EDS), and ODS are continuously emerging, offering new pathways for achieving greener, more efficient, and low-energy-consuming clean refining.^82–86

1.3. Characteristics of sulfur compounds in diesel fuel

Diesel is an intermediate distillate fuel composed of a complex mixture of hydrocarbons, with its sulfur compounds primarily originating from organic sulfur compounds naturally present in crude oil.^87–89 Due to the wide boiling point range of diesel (typically between 180–370 °C), its components exhibit structural diversity, resulting in a rich variety and complex distribution of sulfur compounds (Fig. 2).^90,91 Under conventional refining conditions, sulfur predominantly exists in organic forms, with only a small fraction found in inorganic sulfur forms. Most of these organic sulfur compounds exist in the form of chemically stable heterocyclic compounds, which possess strong thermal stability and low reactivity, making them a key challenge to be addressed in the preparation of clean diesel.⁴³

Fig. 2 Commonly detected sulfides in diesel oils.

Chemically, the sulfur compounds in diesel mainly include mercaptans, sulfides, thiophenes, benzothiophenes (BT), and DBT.⁹² Among these, thiophenes and their derivatives are the most commonly found sulfur compounds in diesel, especially in catalytic cracking diesel, where the contents of benzothiophene, DBT, and their alkylated derivatives, such as 4,6-dimethyl dibenzothiophene (4,6-DMDBT), are relatively high.⁹³ These aromatic heterocyclic sulfur compounds, due to the presence of conjugated π electron systems in their molecular structure, exhibit higher C–S bond dissociation energies, making them difficult to break under conventional hydrogenation conditions.⁹⁴ Particularly for compounds where the alkyl substituents are located at the 4 and 6 positions of the DBT backbone, the steric hindrance effect is more significant, which greatly reduces the proximity of sulfur atoms to the active sites on the catalyst surface. This makes traditional HDS processes less effective for treating these compounds under mild conditions.⁹⁵

The distribution characteristics of sulfur compounds in diesel are also significantly influenced by the type of crude oil and the refining process.^43,96,97 For example, diesel derived from high-sulfur crude oils, such as those from the Middle East or South America, typically has sulfur content above 0.5 wt% and is mainly composed of poly-substituted aromatic sulfur compounds.^43,98 In contrast, diesel derived from light low-sulfur crude oils (e.g., North Sea crude) generally has lower sulfur content, with a higher proportion of mercaptans and sulfides.^99,100 The impact of different refining units on the types of sulfur compounds in diesel is also notable. Atmospheric distillation diesel has relatively simple components and lower desulfurization difficulty, while catalytic cracking diesel (FCC diesel) and coking diesel contain a higher number of aromatic sulfur heterocyclic compounds, making them more structurally complex and harder to desulfurize.^97,101–103 Moreover, as refining processes evolve towards ultra-deep desulfurization, the residual sulfur content gradually shifts from easily removed components to difficult-to-degrade heterocyclic sulfur compounds, imposing higher demands on catalyst structure design and reaction systems.^104,105

In this review, both hydrodesulfurization (HDS) and non-hydrodesulfurization (non-HDS) technologies for diesel fuel desulfurization are comprehensively examined. While HDS remains the dominant industrial process, its limitations in removing refractory aromatic sulfur compounds and the requirement for high-pressure hydrogen highlight the urgency of developing alternative, greener approaches. Therefore, particular emphasis is placed on non-HDS methods—including bio-, adsorptive, extractive, oxidative-based, and coupled strategies—which enable sulfur removal under milder and more sustainable conditions. The reaction mechanisms, structure–activity relationships, and technical bottlenecks of each method are critically analyzed. In parallel, recent progress in advanced desulfurization materials, such as transition metal oxides, heteropoly acids, metal–organic frameworks (MOFs), functionalized ionic liquids, carbon-based catalysts, and some other extractants, adsorbents, and catalysts, are summarized with respect to their desulfurization performance, regenerability, and environmental compatibility. Additionally, coupled process intensification strategies, including adsorptive coupled with catalytic oxidative desulfurization (ACODS) and extractive coupled with catalytic oxidative desulfurization (ECODS) systems, are highlighted for their potential to simultaneously enhance reaction and separation efficiency. This review aims to provide a holistic understanding of current advances and remaining challenges in diesel desulfurization, and to offer perspectives for the rational design of next-generation green technologies toward ultra-deep sulfur removal and cleaner fuel production.

2. Hydrodesulfurization of diesel fuel

2.1. Mechanism of hydrodesulfurization

HDS is one of the most widely used desulfurization processes in industrial petroleum refining.^106,107 The core principle of HDS is to utilize transition metal catalysts under a high-temperature, high-pressure hydrogen atmosphere to convert the sulfur element in organic sulfur compounds into gaseous hydrogen sulfide (H₂S), thereby achieving complete removal of sulfur from the oil product.^108,109 HDS is not only a key unit process for producing clean fuels (such as ultra-low-sulfur diesel, jet fuel, and automotive gasoline), but also provides essential technological support for complying with environmental regulations, extending catalyst life, and improving product performance.^110–112 This process is typically implemented in the middle and downstream sections of the refining process, mainly treating intermediate distillates such as straight-run diesel, catalytic cracking diesel, and vacuum residue oil.

The basic reaction pathways of HDS involve two main processes: the first is the hydrogenolysis of organic sulfur compounds, where the sulfur compound molecules initially adsorb onto the active sites on the catalyst surface. Subsequently, the C–S bond breaks, releasing sulfur atoms and forming alkane desulfurization products. The second is the reaction of the generated sulfur atoms with hydrogen to form H₂S gas.^113,114 Based on whether the aromatic ring is hydrogenated during the desulfurization reaction, HDS can be classified into Direct Desulfurization (DDS) and Hydrogenation–Desulfurization (HYD).^115–117 The DDS pathway directly breaks the C–S bond without hydrogenating the aromatic ring, leading to mild reaction conditions and fewer side reactions. On the other hand, the HYD pathway requires the aromatic ring to be hydrogenated into a cycloalkane structure before the C–S bond is cleaved, resulting in more stringent reaction conditions but making it suitable for sterically hindered or highly aromatic sulfur compounds (Fig. 3).

Fig. 3 Different hydrodesulfurization reaction pathways including HYD and DDS pathways, reproduced from ref. 80 with permission from American Chemical Society,⁸⁰ copyright 2021.

In industrial production, HDS units typically use fixed-bed hydrogenation reactors, with reaction conditions usually ranging from 300 to 400 °C and hydrogen pressures of 3–10 MPa.¹¹⁸ The catalysts used are primarily sulfurized transition metals such as Mo and W, with Ni, Co, and other promoters, supported on high surface area γ-Al₂O₃ or mesoporous oxide supports.^{108,119–121} In recent years, with increasingly strict environmental protection standards, the traditional HDS process faces the challenge of “deep desulfurization”, especially for sterically hindered aromatic heterocyclic sulfur compounds such as 4,6-DMDBT, whose removal efficiency is significantly limited (Fig. 4). This requires higher activity catalysts and optimized reaction conditions.

Fig. 4 The typical refractory sulfurs in petroleum fractions at various HDS technologies, reproduced from ref. 103 with permission from Elsevier,¹⁰³ copyright 2021.

2.2. Research progress in hydrodesulfurization

With environmental regulations becoming increasingly strict and the growing demand for clean fuels, controlling sulfur content in petroleum products has become an important challenge for the refining industry. HDS is the most mature and widely applied desulfurization technology. Although it has been widely successful in conventional petroleum processing, existing catalysts and processes still face challenges such as low conversion rates, high energy consumption, and poor hydrogen utilization efficiency, especially when dealing with sterically hindered aromatic heterocyclic sulfur compounds (e.g., 4,6-DMDBT). Therefore, research on HDS technology mainly focuses on the design optimization of catalytic materials, deeper understanding of reaction pathways, process structure enhancement, and the development of emerging coupled technologies.

In terms of catalyst systems, traditional Co–Mo/γ-Al₂O₃ and Ni–Mo/γ-Al₂O₃ catalysts are no longer efficient enough for ultra-deep desulfurization. Researchers have been improving catalysts by exploring different metal types, support structures, and the introduction of promoters. On the one hand, active metals have gradually shifted from traditional fourth-period elements such as Mo and W to noble metals (e.g., Pt, Pd, Ru) or multi-metallic synergistic systems. For example, the Ni–W/SiO₂–Al₂O₃ system shows better desulfurization performance for high-aromaticity and sterically hindered sulfur compounds due to its stronger hydrogenation ability.¹²⁶ Liu et al.¹²⁷ proposed a new method to study the formation mechanism of the active phase in HDS catalysts, particularly focusing on crystalline polyoxometalate (POM) precursors. This method induces the crystallization of Ni–Mo–O small-cluster crystals in the impregnation solution through coordination bonds with organic ligands and supramolecular interactions, leading to the formation of POMs. The structural guiding effects of the POM precursors have a significant influence on the structure of the active phase and its HDS performance. The structural guiding effects result in a higher content and better dispersion of the NiMoS active phase, which explains why molybdenum–nickel sulfides exhibit higher HDS reaction activity. Zhu et al.¹²² reported a case study in which they controlled the HDS reaction pathway to achieve ultra-deep desulfurization of diesel. They synthesized a mesoporous Ni₂P/Al₂O₃ catalyst via hydrothermal-programmed reduction impregnation and then introduced noble metals to prepare M-Ni₂P/Al₂O₃ (M = Pt, Pd) catalysts. Characterization results indicated that the type of noble metal significantly affected the catalyst surface acidity and the metal–support interactions. The study found that under industrial conditions (3.4 MPa, 340 °C, liquid space velocity of 4.8 h⁻¹), the Pt-modified catalyst achieved a sulfur conversion rate of 88.5% for 4,6-DMDBT, significantly outperforming those of the Pd-modified catalyst (76.3%) and the unmodified catalyst (58.6%). This improvement was attributed to the excellent hydrogen activation ability of Pt, which not only promoted the formation of Brønsted acid sites but also preferentially activated the isomerization reaction pathway, thereby enhancing the removal of aromatic sulfur compounds (Fig. 5a). This study provided an effective strategy for rationally designing high-performance HDS catalysts through reaction pathway control. Xu et al.¹²⁸ proposed a metal-confined catalyst MoS₂/Pt@TD-6%Ti, which combines the Pt metal confinement effect and the hydrogen spillover effect of Pt noble metal. The modified micropores of Mo/Pt@TD-6%Ti only allow small hydrogen molecules to migrate and dissociate, effectively blocking the entry of sulfur compounds. Due to the synergistic effect between the strong hydrogen dissociation ability of Pt and the desulfurization capability of MoS₂, the MoS₂/Pt@TD-6%Ti catalyst demonstrated higher catalytic activity in the HDS reactions of DBT and 4,6-DMDBT, while also offering lower catalyst cost. On the other hand, some researchers have directly employed noble metals for HDS. For example, Prins et al.¹²⁹ were the first to investigate the HDS performance of noble metal catalysts supported on mesoporous zeolites, with a particular focus on the desulfurization of 4,6-DMDBT. The study revealed that these catalysts demonstrated significantly higher desulfurization efficiency compared with those supported on microporous zeolites and γ-Al₂O₃. Based on the experimental results, a new reaction mechanism for the HDS of 4,6-DM-DBT was proposed. It is anticipated that further catalytic performance improvements can be achieved by tuning the acidity and pore size of the mesoporous zeolites. This study confirms that mesoporous zeolites are ideal catalyst supports for deep HDS reactions. Tuning the structures of HDS catalysts for enhanced desulfurization activities has been proved to be an efficient strategy. Future work in HDS catalyst structure tuning may focus on optimizing the dispersion and electronic properties of noble metals through nanostructure engineering to enhance low-temperature activity and sulfur tolerance. Additionally, the development of high-stability supports to reduce noble metal sintering and sulfur poisoning should be explored. Another important direction is the precise regulation of active sites based on reaction mechanism guidance, aiming to achieve deep desulfurization of sulfur compounds such as 4,6-DMDBT while reducing hydrogen consumption.

Fig. 5 Tuning structures of hydrodesulfurization catalysts and developing new characterization approaches for better understanding the structures of the catalysts. (a) Pt doping in Ni₂P/Al₂O₃ catalysts for promoted desulfurization performance, reproduced from ref. 122 with permission from American Chemical Society,¹²² copyright 2024. (b) Detailed reaction pathways of 2,8-DPPDBT hydrodesulfurization on NiMo/SBA-15 catalysts, reproduced from ref. 123 with permission from John Wiley and Sons,¹²³ copyright 2022. (c) In situ FTIR characterization of hydrodesulfurization catalysts using CO as the probing molecule, reproduced from ref. 124 with permission from Elsevier,¹²⁴ copyright 2021. (d) Transient XAS to detect minute levels of reversible S–O exchange at the active sites of MoS₂-based hydrotreating catalysts, reproduced from ref. 125 with permission from American Chemical Society,¹²⁵ copyright 2021.

On the other hand, support materials have evolved from traditional γ-Al₂O₃ to mesoporous materials with large pore sizes and strong metal–support interactions (e.g., SBA-15, MCM-41), titanium oxide (TiO₂), zirconium oxide (ZrO₂), nitrogen-doped carbon materials, and so on. These new supports offer higher metal dispersion, a richer distribution of acid–base sites, and more stable structural support, significantly improving the low-temperature activity and sintering resistance of the catalysts. Xu et al.¹²³ addressed the key scientific issue of diffusion limitation of large molecular organic sulfur compounds in the catalyst pores. They precisely designed NiMo/SBA-15 catalysts with different pore sizes and combined catalytic reaction kinetics studies of a new heavy oil model molecule, 2,8-di(4-pentylphenyl)dibenzothiophene (2,8-DPPDBT). This study established a quantitative structure–activity relationship between molecular size and catalyst pore size for the first time. The study innovatively proposed an empirical factor F(λ) = (1 − λ)^8.5 to quantitatively describe the effect of the ratio λ between the molecular diameter and pore size on diffusion efficiency. Experimental results showed that when λ < 0.3, the diffusion efficiency of large molecular sulfur compounds could reach over 90% of the ideal state. By optimizing the catalyst pore size to 11.3 nm, the hydrogenation desulfurization conversion rate for 2,8-DPPDBT was increased by 3.2 times compared with that of conventional catalysts with a pore size of 6.2 nm, while maintaining an over 85% selectivity. This research not only theoretically clarified the mechanism of pore size effects on heavy oil HDS but also provided a quantifiable scientific basis for the design of industrial catalysts (Fig. 5b). The catalysts developed based on this principle have been successfully applied in heavy oil hydro-processing, resulting in over a 40% improvement in desulfurization efficiency. J. N. Díaz de León et al.¹³⁰ explored the impact of common supports and binary mixed-oxide materials on the preparation of oxide and sulfidic NiW catalysts. Through experimental characterization and density functional theory (DFT), the intrinsic support interactions between WS₂ and Ni-WS₂ clusters were studied. The research revealed how different oxides influence the activity, dispersion, promotion effect, total acidity, and surface-specific bonding of NiW catalysts. The study found a correlation between the total acidity of the catalyst and the HDS reaction of the diesel model compound DBT. Furthermore, DFT calculations indicated that the promoted clusters would undergo bending when interacting with the support, a phenomenon that was experimentally verified. Bussell et al. explored the deep HDS performance of Ni₂P catalysts supported on boria (xB-Al₂O₃) with varying boron content. The study found that the HDS activity of the catalyst was closely related to the amount of boron loaded. When the boron loading was 0.8 wt% (using a hypophosphite source) and 1.2 wt% (using a phosphate source), the catalyst exhibited optimal HDS activity. At 573 K, the Ni₂P/0.8B-Al₂O₃-hypo catalyst showed 2.5 times higher activity compared with the B-free Ni₂P/Al₂O₃-hypo catalyst, and the Ni₂P/1.2B-Al₂O₃-phos catalyst displayed 8.6 times higher activity than the B-free Ni₂P/Al₂O₃-phos catalyst. This study demonstrates that by introducing boron onto the alumina support and selecting an appropriate phosphorus source, the HDS performance of Ni₂P catalysts can be significantly enhanced. Besides tuning the supports to modulate the structure of HDS catalysts, the structure of supports also affects mass transfer, which also influences the HDS performance. In the future, focusing on the regulation of HDS catalyst supports, it is possible to develop sulfur-resistant supports. By utilizing in situ characterization techniques, the evolution of the metal–support interface during the reaction can be revealed. This will aid in the design of high-temperature-resistant, rigid supports that can suppress sintering and improve the performance and longevity of HDS catalysts.

In terms of reaction mechanisms, as research deepens, HDS studies have gradually shifted from macroscopic performance evaluation to microscopic reaction pathway analysis and structure–activity relationship revelation. Traditional DDS and HYD pathways exhibit significant differences in energy consumption, hydrogen consumption, and selectivity.⁷⁹ The DDS pathway mainly occurs at the edge sites of metal sulfide catalysts, making it more suitable for treating structurally simple, non-sterically hindered sulfur compounds. In contrast, the HYD pathway requires hydrogenation of the aromatic ring to form an intermediate before cleaving the C–S bond, making it more suitable for desulfurization of sterically hindered sulfur compounds. Currently, researchers are using techniques such as in situ Fourier transform infrared spectroscopy (FTIR) (Fig. 5c), Raman spectroscopy, X-ray absorption spectroscopy (XANES, EXAFS) (Fig. 5d), high-resolution transmission electron microscopy (HR-TEM), and density functional theory (DFT) calculations to deeply reveal the adsorption configurations, energy barriers, and transition state structures of different sulfur compounds on the catalyst surface. These studies have clarified the relationship between the S–Mo–S active phase, Ni–Mo–S synergistic sites, and reaction pathways, providing a theoretical basis for the precise design of high-performance catalysts and laying the foundation for constructing a deep desulfurization reaction system with controlled pathways and high selectivity.^124,125,131

In terms of process structure and reaction intensification, HDS technology is also developing towards efficiency, greenness, and low energy consumption. Traditional hydrogenation reactors mostly use three-phase fixed-bed reactors, which suffer from mass transfer limitations, hotspot formation, and H₂S accumulation.¹³² To address these disadvantages, researchers have proposed process control strategies such as multi-stage series, hydrogen gas segmented feeding, recirculation, and directional flow to improve fluid distribution and heat-mass transfer efficiency within reactors. Furthermore, hydrogenation reactors based on membrane separation (e.g., membrane reactors, ceramic membrane-coated reactors) can achieve in situ hydrogen injection and H₂S online separation, significantly reducing hydrogen consumption and inhibiting catalyst poisoning. Microchannel reactors, with their high surface area and rapid heat transfer characteristics, are also considered a strong candidate for next-generation green HDS reactors. In terms of energy consumption control, research is gradually focusing on low-temperature desulfurization pathways, high-pressure hydrogen recovery, and waste heat reuse to improve overall energy efficiency.¹³³

2.3. Challenges of hydrodesulfurization and further perspective

Although the HDS process demonstrates high reaction efficiency in removing most sulfur compounds, such as sulfides, thiols, thiophenes, etc., it still reveals a series of significant limitations when facing the deep purification requirements of modern fuels and the demands for green, low-carbon processing.

First, HDS has relatively limited ability to remove aromatic sulfur compounds.¹³⁴ As crude oil resources become heavier and contain more sulfur components, the proportion of sterically hindered, highly aromatic heterocyclic sulfur compounds, such as DBT and 4,6-DMDBT, in the oil products significantly increases. These sulfur compounds have stable molecular structures, high C–S bond energies, and weak adsorption ability on the catalyst surface (because of the steric hindrance effect). Traditional HDS processes are often unable to efficiently remove these compounds, requiring significantly higher reaction temperatures, pressures, or prolonged reaction times. This not only increases energy consumption and equipment load but also hinders the green development of the process.

Second, HDS relies on a high-pressure hydrogen environment, which consumes a large amount of hydrogen and has limited sources. Hydrogen, as the core reactant in the HDS process, not only increases equipment investment costs and safety risks under high-pressure conditions, but its industrial production mainly depends on fossil fuel-based hydrogen production routes such as natural gas reforming or coal-to-hydrogen processes, indirectly leading to increased carbon emissions. In the context of the future low-carbon transition to clean fuels, the high consumption and low efficiency of hydrogen have become important factors limiting the development of HDS.

Furthermore, HDS has poor selectivity for crude oil components, which can lead to a decline in product quality.^135–137 Under deep desulfurization conditions, especially when sterically hindered sulfur compounds are removed via the HYD pathway, side reactions such as hydrogenation of the aromatic ring and saturation of hydrocarbons often occur. These side reactions cause a decrease in diesel aromaticity, increase in density, increase in cetane number, and increased hydrogen consumption, which are detrimental to product combustion performance and engine compatibility. Additionally, the accumulation of H₂S byproducts can lead to catalyst poisoning and deactivation, increasing the frequency of catalyst regeneration and replacement and raising operational costs.

At the same time, the HDS process is significantly limited by heat and mass transfer restrictions.¹³⁸ Due to the limited liquid–solid contact efficiency in three-phase fixed-bed reactors, the mass transfer resistance for sulfur compounds to reach the active sites of the catalyst is significant, especially when processing heavy oils and straight-run diesel with high viscosity. To overcome this issue, high flow rates and high-temperature operation are often required, which further increase energy consumption and catalyst load, making process intensification and system scaling more challenging.

As a result, HDS faces significant challenges in terms of ultra-deep desulfurization of modern fuels, energy consumption control, catalyst stability, and product quality protection. Therefore, the development of non-hydrodesulfurization (non-HDS) technologies has become a key direction to break through the limitations of traditional HDS and achieve efficient, green desulfurization. Currently, widely researched non-HDS technologies include biodesulfurization,¹³⁹ extractive desulfurization,¹⁴⁰ adsorption desulfurization,¹⁴¹ oxidative desulfurization.¹⁴²

3. Bio-desulfurization of diesel fuel oils

3.1. Mechanism of bio-desulfurization

Bio-desulfurization (BDS) is an environmentally friendly green technology that utilizes microorganisms or their enzyme systems to selectively remove organic sulfur compounds from petroleum and its derivatives under mild conditions.^143–147 Compared with HDS, BDS offers several advantages, including mild operating conditions, no hydrogen consumption, minimal by-products, and high selectivity. It is especially suitable for removing sterically hindered aromatic heterocyclic sulfur compounds (e.g., DBT, 4,6-DMDBT), which are difficult to remove efficiently via HDS.¹⁴⁸

BDS primarily relies on specific desulfurizing bacteria to biologically oxidize or cleave the C–S bonds in organic sulfur compounds present in petroleum.^149,150 These bacteria are typically isolated from naturally oil-contaminated environments or obtained through artificial mutagenesis and screening. Common genera include Rhodococcus, Gordonia, Nocardia, and Pseudomonas.¹⁵¹ Among them, Rhodococcus erythropolis IGTS8 has been extensively studied and applied due to its high efficiency in degrading DBT-type compounds.¹⁵² These microorganisms can take up the target sulfur compounds via extracellular adsorption or intracellular transport and then catalyze the oxidation or cleavage of organic sulfur through specific enzymatic pathways. During the metabolic process, the carbon backbone of the molecules remains intact, making the desulfurization process a “non-carbon-destructive” desulfurization, while the sulfur atoms are converted into inorganic sulfur species, such as sulfite and sulfate, ultimately achieving sulfur removal.¹⁵³

Zhao et al.¹⁴³ were the first to systematically investigate the BDS behavior and mechanism of a hydrophobic Gordonia strain (SC-10) in an oil–water multiphase system. The study found that this strain altered its cell surface hydrophobicity by producing branched mycolic acids with chain lengths of C47 and C58, and the formation of intracellular inclusions was observed for the first time during the desulfurization process. These two unique biological features significantly enhanced the bioavailability of the substrate. Strain SC-10 employed a sulfur-specific metabolic pathway to selectively convert organic sulfur compounds in diesel into hydroxylated products, achieving a desulfurization rate of 88.3%, reducing the sulfur content from 167.7 mg L⁻¹ to 19.7 mg L⁻¹. This study not only revealed the adaptive mechanisms of hydrophobic microorganisms in oil–water multiphase environments but also provided an excellent microbial resource and theoretical foundation for developing green and efficient BDS technologies. Experiments confirmed that strain SC-10 has broad substrate adaptability and can achieve deep desulfurization of diesel under mild conditions, showing great potential for industrial application.

The most extensively studied metabolic pathway for BDS is the “4S pathway” (four-step sulfur-specific pathway) employed by Rhodococcus erythropolis IGTS8. This pathway selectively removes sulfur from DBT and its derivatives without degrading the aromatic carbon skeleton, making it a representative mechanism for achieving highly selective BDS (Fig. 6a).^154,155 The 4S pathway consists of the following four key steps: Initial oxidation: DBT is oxidized by the enzyme DszC to form DBT-5,5-dioxide (DBT sulfone). Hydroxylation and ring cleavage: DBT sulfone is further converted by DszA into 2′-hydroxybiphenyl-2-sulfonic acid (HPBS). C–S bond cleavage: HPBS undergoes cleavage of the C–S bond catalyzed by DszB, releasing inorganic sulfur (in the form of sulfate) and yielding the final product, 2-hydroxybiphenyl (2-HBP). Cofactor regeneration: the reducing equivalents (NADH) required for the desulfurization reaction are regenerated through the DszD-mediated cofactor recycling system.

Fig. 6 Different BDS pathways. (a) “4S pathway” of BDS; (b) “Kodama pathway” of BDS. Reproduced from ref. 154 with permission from Springer Nature,¹⁵⁴ copyright 1969.

The high selectivity of the BDS process arises from the microorganisms’ ability to recognize and remove only sulfur atoms without disrupting the organic carbon framework, thus preserving the fuel's calorific value. Furthermore, the reaction proceeds under mild conditions (typically 30–40 °C and atmospheric pressure), significantly reducing energy consumption and equipment requirements. Moholkar et al.¹⁵⁶ used molecular simulation techniques to explore the mechanistic differences in the 4S pathway between DBT and 4,6-DMDBT desulfurization. By combining molecular docking and molecular dynamics simulations, they found that 4,6-DMDBT and its metabolic intermediates exhibited lower binding energies and smaller inhibition constants with the Dsz enzyme system. These findings elucidated the underlying molecular reasons for the slower desulfurization rate of 4,6-DMDBT compared with DBT: (i) intermediates such as DMHBPS and DMHBP generated during 4,6-DMDBT metabolism exert strong competitive inhibition on key desulfurization enzymes; and (ii) methyl substituents increase the structural rigidity of the substrate–enzyme complex, impeding catalytic efficiency.

In addition to the 4S pathway, certain microorganisms can degrade organic sulfur compounds by breaking down the carbon skeleton through β-oxidation or thiolysis (e.g., the Kodama pathway), thereby releasing sulfur. However, these pathways typically lead to carbon skeleton destruction and are unsuitable for industrial applications where fuel quality must be preserved (Fig. 6b).

3.2. Recent research progress of BDS

Current research is largely focused on enhancing the expression efficiency of the 4S pathway through genetic engineering approaches—such as cloning the dszABC gene cluster, optimizing promoter strength, and enhancing cofactor regeneration—to improve bacterial oil tolerance, desulfurization activity, and stability.¹⁵⁷

A key challenge in microbial desulfurization lies in the construction and optimization of high-performance desulfurizing strains. Since the discovery of Rhodococcus erythropolis IGTS8 in 1993 as a highly effective DBT-desulfurizing bacterium,¹⁵⁸ numerous other BDS-active genera have been isolated, including Gordonia, Nocardia, Mycobacterium, Pseudomonas, and Sphingomonas. These strains vary in their ability to degrade DBT and its derivatives (e.g., 4,6-DMDBT), and in their metabolic mechanisms.

In recent years, to overcome the limitations of natural strains, such as low desulfurization efficiency, strong product (2-HBP) inhibition, and poor environmental tolerance, researchers have applied genetic engineering to improve strains like IGTS8.¹⁵⁹ For example: enhancing transcription levels of the dszABC gene cluster via promoter engineering; increasing enzyme expression through fusion expression or multi-copy plasmid systems; introducing pathways or mechanisms to metabolize or adsorb 2-HBP to mitigate feedback inhibition; constructing engineered strains with enhanced tolerance to high temperature, oil toxicity, and shear forces for industrial applicability.^160,161

In addition, synthetic biology has been introduced into BDS system reengineering. For instance, heterologous hosts such as Escherichia coli or Pseudomonas aeruginosa have been used to express desulfurization enzymes, and combined with substrate transport proteins and metabolic regulation modules, modular and controllable desulfurizing strains have been developed, thereby improving system stability and substrate adaptability.¹⁶²

Mawad et al.¹⁶³ conducted an in-depth investigation into the biodesulfurization (BDS) performance and metabolic mechanism of Klebsiella oxytoca strain SOB-1 toward dibenzothiophene (DBT). Analysis showed that the strain was capable of degrading 90% of DBT within 96 hours; however, a significant substrate inhibition effect was observed when the DBT concentration exceeded 0.75 mM. Gas chromatography–mass spectrometry (GC-MS) analysis identified four major metabolites produced via a unique degradation pathway, among which 2-hydroxybiphenyl (2-HBP) and 2-methoxybiphenyl (2-MBP) were the dominant intermediates, accompanied by the release of sulfate ions. The study also confirmed a nonlinear relationship between desulfurization efficiency and substrate concentration, offering valuable parameters for optimizing industrial BDS processes.

Fundamentally, biodesulfurization is an enzymatic reaction network catalyzed cooperatively by a group of key enzymes. With the advancement of proteomics and structural biology, significant progress has been made in elucidating the catalytic mechanisms, structural characteristics, and substrate specificities of the core desulfurization enzymes DszC, DszA, and DszB. Crystallographic studies have shown that both DszC and DszA are flavin-dependent monooxygenases whose active sites contain essential amino acid residues that facilitate stable binding of FMN and DBT. In contrast, DszB is a highly substrate-specific C–S bond hydrolase. Based on these insights, researchers have employed site-directed mutagenesis, protein engineering, and directed enzyme evolution to develop enzyme variants with improved substrate adaptability, enhanced catalytic efficiency, and reduced inhibition by 2-HBP. Recently, novel enzyme systems—such as cytochrome P450 monooxygenases and FMNH₂-dependent reductases—have also been identified in certain microbial strains, capable of degrading a broader spectrum of sulfur-containing compounds, thus expanding the applicability of BDS.^164,165

From a reaction engineering perspective, extensive efforts have been devoted to the design of BDS reactors, microbial immobilization, scale-up processes, and the stability of continuous operations.^166–168 Common strategies for process intensification include: 1. immobilization of desulfurizing cells or enzymes, using carriers such as alginate, polyacrylamide gels, magnetic particles, and mesoporous materials, to enhance their stability and reusability; 2. development of flow-bed and membrane bioreactors, enabling efficient phase separation and continuous substrate feeding; 3. optimization of culture medium components and induction methods, aimed at reducing costs while improving microbial growth and desulfurization activity; 4. control of operating conditions such as temperature, pH, and agitation, to improve mass transfer and reduce reaction time.

Kumar Bal et al.¹⁶⁹ proposed an innovative hybrid system that couples microbial desulfurization with ultrasound technology. Their study demonstrated that ultrasonic cavitation significantly enhances the degradation efficiency of DBT and related organosulfur compounds by both free and immobilized microbial cells. The desulfurization kinetics were improved by 2–3 orders of magnitude. By integrating experimental data with cavitation bubble dynamics modeling, the study elucidated three major mechanisms by which ultrasound promotes BDS: enhancing substrate mass transfer, increasing cell membrane permeability, and stabilizing the conformation of desulfurization enzymes (e.g., DszABC). This work provides a robust theoretical foundation for developing next-generation green desulfurization technologies. The proposed “ultrasound-bio” synergistic system achieved a DBT removal efficiency of 92% at the pilot scale, demonstrating strong industrial application potential. These findings not only overcome the long-standing limitation of low BDS reaction rates but also offer an innovative technical pathway for clean fuel production aligned with carbon neutrality goals.

3.3. Challenges of BDS and further perspective

BDS has emerged as an environmentally friendly alternative for the removal of refractory sulfur compounds from diesel. While promising in laboratory studies, this approach still faces numerous technical and economic hurdles that hinder its large-scale commercialization. The most prominent limitation is its insufficient desulfurization efficiency and narrow substrate specificity. Most known BDS pathways, such as the 4S pathway, selectively target DBTs but exhibit poor activity against thiophene and its derivatives, making it difficult to achieve the ultra-low sulfur levels (<10 ppm) required by modern fuel regulations. Furthermore, inhibitory metabolic intermediates like 2-hydroxybiphenyl can suppress microbial activity through negative feedback, further reducing the efficiency of sulfur removal.

From a reaction engineering perspective, BDS processes suffer from inherently slow kinetics due to the prolonged microbial growth and enzyme activity cycles. The poor mass transfer between hydrophobic sulfur species in diesel and aqueous microbial media also limits the reaction rate, resulting in long processing times and oversized reactors that are incompatible with the throughput requirements of modern refineries. Additionally, maintaining stable microbial activity under variable feedstock compositions and stringent environmental conditions, such as pH, temperature, dissolved oxygen, etc., poses significant operational challenges, making continuous industrial-scale operation difficult.¹⁷⁰

Economically, BDS remains uncompetitive with HDS due to high costs associated with strain cultivation, expensive media, and downstream separation. Moreover, unselective microbial oxidation may degrade hydrocarbons, causing diesel yield losses and deterioration of fuel quality. Issues such as emulsion formation further complicate oil–water separation, increasing process complexity and cost.

At present, biocatalysts demonstrate sluggish desulfurization rates and inadequate stability, with considerable efficiency deterioration during industrial-scale expansion, resulting in delays to commercial implementation. Despite being largely confined to laboratory and pilot-scale demonstrations, BDS holds long-term potential. Future development can focus on metabolic engineering to broaden substrate scope, immobilized enzyme systems for catalyst reuse, and advanced bioreactor designs to enhance mass transfer. Integration with other desulfurization technologies and development of cost-effective biomass separation and product recovery methods will be crucial to improving the feasibility and scalability of BDS for real diesel fuel applications.

4. Adsorptive desulfurization of diesel fuel

4.1. Mechanism of adsorptive desulfurization

Adsorptive Desulfurization (ADS) has emerged as a promising non-HDS technique in recent years, drawing increasing attention for its green and energy-efficient characteristics. The fundamental principle of ADS lies in the use of adsorbents with high selectivity and strong sulfur affinity to capture organosulfur compounds from diesel fuel under mild conditions. These compounds are enriched onto the surface or within the pores of solid adsorbents through physical or chemical interactions, thereby achieving effective separation and removal.¹⁷¹ Compared with conventional HDS, ADS does not require high temperature, high pressure, or hydrogen input, resulting in lower energy consumption and simpler equipment. Moreover, it exhibits significant advantages in the deep removal of refractory aromatic sulfur species, such as thiophenes, BT, DBT, and their alkyl derivatives, making it to be a promising pathway for low-carbon and environmentally friendly fuel refining.¹⁷²

The selectivity and sulfur affinity of the adsorbent are the key factors determining the overall adsorption efficiency (Fig. 7).¹⁷³ Organosulfur compounds, due to their aromaticity and electron-rich sulfur atoms, can interact with specific metal ions or functional groups on the adsorbent through coordination bonding, electrostatic interactions, or hydrophobic/oleophilic interactions. These interactions form the theoretical basis for selective adsorption. The major adsorption mechanisms reported thus far include van der Waals physical adsorption, π-complexation between sulfur-containing aromatic rings and metal sites, σ-bond coordination between sulfur lone pairs and vacant metal orbitals, hydrogen bonding, and acid–base cooperative interactions.

Fig. 7 Different interactions between adsorbents and sulfides.

Among these, π-complexation is widely recognized as the dominant mechanism for aromatic sulfur adsorption. In this process, the metal cations (e.g., Ag⁺, Cu⁺, Ni²⁺) on the adsorbent surface interact with the π-electron cloud of the sulfur-containing aromatic ring to form stable π-complexes. Meanwhile, σ-complexation involves direct bonding between the sulfur atom's lone pair electrons and the empty orbitals of metal centers, which further enhances adsorption strength and selectivity. These mechanisms work synergistically to enable the selective removal of aromatic sulfur compounds under ambient or near-ambient conditions.

4.2. Main types of adsorbents

The selection of adsorbent type is closely related to its desulfurization performance. An ideal adsorbent should possess a high specific surface area, suitable pore size distribution, remarkable thermal stability, tunable surface chemistry, and high selectivity toward sulfur compounds.¹⁷⁴ The major types of adsorbents used in ADS and their characteristics are as follows.

4.2.1. Zeolite-based adsorbents. Zeolites are crystalline aluminosilicates with well-defined pore structures, widely used in the petrochemical industry.^181,182 Synthetic zeolites with well-developed structures, such as Y-type, ZSM-5, β-type, and MCM-41, are among the earliest materials applied in ADS. Their primary advantages include high specific surface area, controllable pore distribution, and excellent thermal and chemical stability.¹⁸³ When specific metal ions (e.g., Ag⁺, Cu⁺, Zn²⁺) are introduced into the zeolite framework, selective adsorption of sulfur compounds can be achieved via π-complexation between the aromatic rings and the metal centers. For example, Ag⁺ forms stable coordination structures with thiophene-based compounds via conjugation with π-electrons, while Cu⁺ exhibits stronger σ-coordination with the lone-pair electrons on sulfur atoms. Additionally, the narrow channels of zeolites provide a molecular sieving effect, aiding in the separation of structurally similar hydrocarbons and heterocyclic sulfur compounds.

Hossi et al.¹⁷⁵ developed a novel fuel desulfurization technology using faujasite (FAU)-type zeolite synthesized from coal fly ash (Fig. 8a). This FAU zeolite, derived from coal-fired power plant waste, reduced the sulfur content of diesel from 155 ppm to 97 ppm under ambient conditions. Through optimization of key parameters such as adsorption temperature, fuel volume, and chemical composition, the material exhibited remarkable selectivity toward organic sulfur compounds. The zeolite's cost was only one-fifth that of commercial zeolites, offering a sustainable and economically viable approach for clean fuel production. Prasassarakich et al.¹⁷⁶ developed a dual-stage adsorption system based on ion-exchanged Y zeolites, addressing simultaneous desulfurization and denitrogenation of diesel. Na–Y zeolite showed optimal adsorption capacity for 4,6-DMDBT and DBT (up to 35 mg S per g), while La–Y zeolite exhibited high selectivity for nitrogen compounds (>90% removal) (Fig. 8b). The study revealed that nitrogen compounds have 1.8 times higher affinity for adsorption sites than sulfur compounds, prompting the design of a sequential system: La–Y for denitrogenation followed by Na–Y for desulfurization. This configuration improved the breakthrough capacity for 4,6-DMDBT by 40% and reduced S and N levels below 15 ppm and 5 ppm, respectively. The modular design lowered operating costs by 30% and maintained over 90% efficiency after five regeneration cycles, providing a highly effective and economical strategy for ultra-clean diesel production.

Fig. 8 ADS of diesel fuel by different adsorbents. (a) Effects of cationic exchange on desulfurization performances of zeolite, reproduced from ref. 175 with permission from John Wiley and Sons,¹⁷⁵ copyright 2024. (b) Two-stage sequential adsorption system for denitrogenation and desulfurization of diesel oil over ion-exchanged Y zeolites, reproduced from ref. 176 with permission from American Chemical Society,¹⁷⁶ copyright 2023. (c) The adsorption mechanism of Pd/SiO₂@GO for thiophene, reproduced from ref. 177 with permission from Elsevier,¹⁷⁷ copyright 2022. (d) Possible interactions during adsorptive removal of sulfur compounds using nitrogen-modified graphene, reproduced from ref. 178 with permission from Elsevier,¹⁷⁸ copyright 2024. (e) Adsorption energy of DBT on graphene and graphene oxide (G: graphene; G with –OH: graphene with a hydroxyl; G with –O–: graphene with an epoxy; G with =O: graphene with a carbonyl; G with –COOH: graphene with a carboxyl), reproduced from ref. 179 with permission from American Chemical Society,¹⁷⁹ copyright 2019. (f) Adsorption mechanism for DBT of DHPCs, reproduced from ref. 180 with permission from Elsevier,¹⁸⁰ copyright 2021.

Despite zeolites’ high selectivity and thermal stability, their regeneration performance can be compromised by the stability and dispersion of metal ions. Zeolites also suffer from limited adsorption capacity under high sulfur loading and are less effective for non-polar compounds like thiophenes due to competitive adsorption with hydrocarbons in complex fuel matrices.

4.2.2. Activated carbon and porous carbon materials. Activated carbon materials are widely used due to their low cost, high availability, and porous structures, typically exhibiting surface areas over 1000 m² g⁻¹.¹⁸⁴ In ADS processes, the adsorption on activated carbon is mainly driven by van der Waals forces, making it less effective for weakly polar compounds like thiophenes. Surface functionalization is thus essential for enhancing desulfurization performance. Introducing acidic or basic oxygen-containing groups (e.g., hydroxyl, carboxyl, amino) can increase polarity-based interactions, while loading with metal oxides (e.g., CuO, ZnO, CeO₂) improves electron affinity, enhancing DBT and derivative adsorption. Additionally, graphene oxide (GO) and nitrogen-doped porous carbons have drawn increasing attention due to their tunable structures and abundant active sites.^185–189 Compared with zeolites, carbon materials exhibit lower regeneration temperatures but still suffer from selectivity issues due to co-adsorption of aromatics.

Zhang et al.¹⁷⁷ prepared a series of PdO/SiO₂@GO hybrid aerogels with varying GO contents via a sol–gel method combined with ambient pressure drying. The incorporation of graphene oxide (GO) significantly improved the specific surface area and Pd loading efficiency, thereby enhancing the adsorption performance toward sulfur compounds (Fig. 8c). Compared with PdO/SiO₂, GO doping introduced additional π–π stacking interactions between GO layers, alongside the synergistic π-coordination and S–Pd bonding between Pd²⁺ and thiophenic compounds, resulting in higher adsorption capacity. Thermodynamic and kinetic analyses indicated that the adsorption of thiophene onto PdO/SiO₂@GO-1.0 is a spontaneous and exothermic process, following the Freundlich isotherm and pseudo-second-order kinetic model. The adsorption rate is jointly governed by intraparticle diffusion and surface interactions. Moreover, the hydrophobic and aromatic nature of GO contributes to enhanced selectivity toward aromatics, olefins, and basic compounds, while suppressing the competitive adsorption of MTBE and water. The saturated adsorbent can be efficiently regenerated via solvent washing, demonstrating excellent cycling stability and promising potential for industrial applications.

Mahuya De et al.¹⁷⁸ synthesized graphene and its nitrogen-doped derivatives, and systematically evaluated their adsorptive desulfurization performance. Nitrogen doping significantly enhanced the specific surface area (from 257 to 301 m² g⁻¹) and pore volume (from 0.48 to 1.12 cm³ g⁻¹), thereby improving sulfur compound adsorption. The high-nitrogen-content sample, 14-N-GR, achieved removal rates of 97.3%, 92.8%, and 88.4% for thiophene, BT, and DBT, respectively, which is markedly higher than the undoped material. This enhancement is mainly attributed to strengthened π–π interactions from pyrrolic and pyridinic nitrogen, as well as improved pore structure facilitating molecular diffusion (Fig. 8d). Additionally, the material showed good regenerability, with only a 10–14% drop in efficiency after five cycles, indicating excellent stability and application potential.

Yang et al.¹⁷⁹ employed DFT calculation to investigate the interaction between DBT and graphene, and their results have shown that the interaction between DBT and graphene is primarily governed by dispersion forces, exhibiting typical π–π interaction characteristics (Fig. 8e). This interaction mechanism involves several key aspects: (1) regions of graphene adjacent to DBT tend to become electronegative, facilitating electrostatic attraction with partially positively charged atoms in DBT; (2) the introduction of oxygen-containing functional groups disrupts the π-conjugated structure of graphene by inducing uneven surface charge distribution, thereby weakening its interaction with DBT; (3) non-covalent long-range interactions play a dominant role in the adsorption process; and (4) the adsorption strength is closely related to the interatomic distance between DBT and graphene. In summary, the DBT–graphene adsorption process is dominated by dispersion interactions, indicating that optimizing the conjugated structure and surface electronic distribution of graphene is crucial for enhancing its desulfurization performance.

Besides graphene-based adsorbents, other types of carbon-based materials have also been reported. Liu et al.¹⁹⁰ developed a novel ADS material via direct carbonization of ZIF-8, avoiding post-treatment steps. The resulting porous carbon exhibited excellent DBT adsorption performance, with a maximum capacity of 26.7 mg S per g in model oil containing less than 174 ppm sulfur, which is significantly higher than traditional activated carbon. When 10% p-xylene was present, the capacity decreased by only ∼17%, indicating good aromatic tolerance. The simplified synthesis process is well-suited for industrial scale-up.

Li et al.¹⁸⁰ reported a high-performance desulfurization material: defect-rich hierarchical porous carbon (DHPCs) derived from coal tar pitch. A one-pot process involving oxidative pretreatment, melamine grafting, and synchronous carbonization/activation created a highly defective, hierarchically porous structure (Fig. 8f). The DHPC-10 sample achieved a record-breaking DBT adsorption capacity of 86.94 mg S per g, which is 3–4 times higher than conventional activated carbon, and maintained 75.11 mg S per g after five cycles. Even with the presence of 10% p-xylene, a capacity of 72.3 mg S per g was achieved, demonstrating excellent tolerance. This work provides a cost-effective strategy for preparing high-efficiency carbon-based adsorbents.

Carbon-based adsorbents such as activated carbon and graphene possess several advantages for diesel desulfurization, including high specific surface area, abundant pore structures, and tunable surface chemistry, enabling effective adsorption of sulfur compounds. In particular, graphene exhibits high selectivity and adsorption capacity for aromatic sulfur species due to its unique two-dimensional structure and excellent electronic properties. However, activated carbon suffers from a broad pore size distribution and limited selectivity, which may lead to the co-adsorption of other diesel components. Despite its superior performance, graphene faces challenges such as high production costs, difficulties in large-scale synthesis, and potential structural degradation during regeneration.¹⁹¹ Moreover, both materials exhibit limited adsorption capacity for highly polar sulfur compounds, and thus often require surface modification or composite formation with other functional materials to enhance their desulfurization performance. However, both types of carbon-based adsorbents may experience pore structure collapse or loss of surface functional groups during regeneration, particularly under high-temperature or harsh chemical conditions, which compromises their recyclability. Overall, while carbon-based adsorbents hold significant promise for diesel desulfurization, further optimization of their structure and surface properties is essential to enhance their industrial applicability.

4.2.3. Transition metal oxide adsorbents. Transition metal oxides, such as ZnO, CeO₂, TiO₂, Fe₂O₃, MnO₂, etc., exhibit excellent ADS performance due to abundant Lewis acid sites and redox properties.^171,192,193 ZnO is particularly notable for its affinity toward thiophenes, making it suitable for light sulfur removal from gasoline and diesel,¹⁹⁴ and it has also been employed in the S-zorb process for desulfurization of gasoline. CeO₂ offers both metal–sulfur affinity and redox potential via reversible Ce⁴⁺/Ce³⁺ transitions and oxygen vacancies, allowing oxidation of refractory sulfur compounds like 4,6-DMDBT into more adsorbable sulfones or sulfoxides.¹⁹⁵ MnO₂ and Co₃O₄ also serve as oxidative adsorbents with good thermal regeneration compatibility, often used in combination with carbon or zeolites to form oxidation–adsorption hybrid systems.¹⁹⁶

Yang et al.¹⁹⁷ reported that a 20 wt% Zn/Al₂O₃ nano-alumina adsorbent prepared via a xerogel method combined with thermal vacuum treatment exhibited excellent desulfurization performance for a thiophene–n-pentane mixture at room temperature, with a saturated adsorption capacity of up to 2.8 mg S per g. Characterization results revealed the formation of a ZnAl₂O₄ spinel structure during thermal treatment, with vacuum conditions promoting a more defective and less crystalline spinel phase featuring smaller particle size and higher specific surface area. These structural features enhance the interaction between Zn²⁺ and thiophene molecules. The adsorbent not only offers a controllable structure and superior performance, but can also be efficiently regenerated through a simple procedure, demonstrating good cycling stability and promising potential for practical applications.

Generally, compared with other adsorbents, fewer reports have been focused on pure metal oxide adsorbents, because of the low specific surface areas, poor exposure of adsorption active sites, and so on. Thusly, the metal oxides are generally dispersed on supports with high specific surface areas. Yao et al.¹⁹⁸ systematically studied the ADS mechanism of fourth-period transition metal oxides based on ion covalency parameters (ICP). Modified activated carbons with various oxides (ZnO, Cr₂O₃, MnO₂, Co₃O₄, Fe₂O₃, CuO, NiO) showed a strong negative correlation (R² = 0.87) between ICP values and sulfur removal efficiency, superior to traditional Pearson hardness theory, highlighting ICP's ability to reflect Lewis acidity. Notably, NiO-modified carbon exhibited excellent selectivity for sterically hindered sulfur compounds and followed plug-flow dispersion models, showing industrial application potential.

Moreover, modifying the molecular structure of aromatic sulfur compounds through oxidation has been shown to enhance adsorption performance. Based on this concept, researchers have proposed a novel reactive adsorption strategy that integrates oxidation to strengthen the overall desulfurization efficiency. Song et al.¹⁹⁹ developed a novel Ti–Ce composite oxide (Ti_0.9Ce_0.1O₂) for air-promoted ADS. XANES analysis showed the material could oxidize sulfur compounds into sulfones/sulfoxides in the presence of air. Adsorption selectivity followed the order: DBT sulfone > DBT ≈ BT > 4-MDBT > 4,6-DMDBT > phenanthrene > methylnaphthalene > fluorene > naphthalene. DFT calculations indicated Ti-O-SR₂ interactions were stronger than Ti-SR₂, a key factor in adsorption enhancement (Fig. 9). The material could be regenerated via air oxidation, offering new pathways for reusable deep desulfurization materials.

Fig. 9 Adsorption configurations of BT and its derivatives over a Ce-doped TiO₂(001) surface. (a) BT, (b) DBT, (c) DBT sulfone, (d) 4-MDBT, (e) 4,6-DMDBT; (f) DBTO; Ce, yellow; C, dark gray; H, white; Ti, light gray; O, red; S, dark yellow, reproduced from ref. 199 with permission from John Wiley and Sons,¹⁹⁹ copyright 2014.

Metal oxide adsorbents have demonstrated unique value in the field of diesel adsorptive desulfurization, primarily due to their specific affinity toward thiophenic sulfur compounds, good thermal stability, and relatively low production cost. These materials selectively remove sulfur through strong interactions between sulfur atoms and surface oxygen vacancies. However, inherent limitations remain, including low specific surface area and limited pore structure, which hinder the adsorption of bulky sulfur molecules. Additionally, their performance can be compromised by competitive adsorption from other fuel components in complex diesel matrices, and sintering during high-temperature regeneration can further reduce their service life.

To overcome these challenges, future research can focus on three key strategies: precise control of nanostructures, functional design of composite supports, and optimization of regeneration processes. Constructing hierarchical porous structures and incorporating carbon-based composite supports can significantly enhance surface area and mass transfer efficiency. Furthermore, doping with transition metals or modifying with rare-earth elements allows for fine-tuning of surface electronic states, thereby improving selectivity toward sulfur compounds. On the process level, the development of mild regeneration techniques and exploration of adsorption–catalysis synergistic mechanisms offer promising pathways to improve material recyclability and long-term stability. Continued advancement in these areas is expected to accelerate the practical application of metal oxide adsorbents in deep desulfurization technologies.

4.2.4. Metal–organic frameworks (MOFs). MOFs are crystalline porous materials composed of metal ions/clusters and organic linkers.^200–204 Structures such as HKUST-1 (Cu-BTC), MIL-101(Cr), and UiO-66(Zr) feature extremely high surface areas and tunable pores, making them promising for ADS.²⁰⁵ Their adsorption mechanisms include: (1) σ-coordination between metal nodes and sulfur atoms; (2) π–π stacking between aromatic frameworks and sulfur compounds; (3) pore matching for molecular sieving.²⁰⁶ Functionalized MOFs (e.g., Ag@MOFs, Ce-MOFs) exhibit high selectivity for DBT derivatives, and composite MOFs incorporating ILs, oxides, or carbon further enhance stability and regeneration.

Liao et al.²⁰⁷ optimized the desulfurization performance of the MOF material NH₂-MIL-125 by tuning the drying temperature. BET analysis revealed that the sample treated at 150 °C exhibited the highest specific surface area and pore volume. Adsorption experiments confirmed that NH₂-MIL-125@150 °C demonstrated the best performance in removing DBT, achieving a saturated adsorption capacity of 112 mg g⁻¹. The adsorption affinity for other sulfur compounds followed the order: DBT > 4,6-DMDBT > BT. Further DFT calculations indicated that hydrogen bonding was the dominant interaction, and the theoretical adsorption trend was consistent with the experimental results (Fig. 10a). This study presents a simple and effective thermal treatment strategy to enhance the desulfurization performance of MOF materials, offering both practical applicability and research significance.

Fig. 10 ADS by different MOF-based adsorbents. (a) Simulation diagram and adsorption energy of three sulfides over NH₂-MIL-125, reproduced from ref. 207 with permission from Elsevier,²⁰⁷ copyright 2023. (b) Ni-based MOF with improved adsorptive desulfurization activity, reproduced from ref. 208 with permission from Elsevier,²⁰⁸ copyright 2021. (c) Possible schematic of the increased ADS activity over the Py/MOF, reproduced from ref. 209 with permission from American Chemical Society,²⁰⁹ copyright 2019. (d) BEs of PAF-56 fragment with DBT and H₂O after geometry optimization, reproduced from ref. 210 with permission from Elsevier,²¹⁰ copyright 2022.

Sun et al.²⁰⁴ systematically studied the desulfurization performance of MOFs against thiophene, BT, and 4,6-DMDBT. Among MOF-5, HKUST-1, MIL-53(Fe), MIL-53(Cr), and MIL-101(Cr), MIL-53(Cr) exhibited the best ADS capacity. IR and TPD analysis revealed that adsorbate–adsorbent interactions were the key factor in performance. MIL-101(Cr), despite high porosity, showed poor desulfurization due to weak interactions. These findings emphasize that enhancing adsorbate–adsorbent affinity is crucial for MOF design.

Hajjar et al.²¹¹ examined five MOFs, including MIL-53(Cr, Al, Fe), Cu-BDC, and HKUST-1, against four aromatic sulfur compounds. MIL-53(Cr) had the best performance, with removal rates of 29% (thiophene), 39% (BT), 61% (DBT), and 88% (4,6-DMDBT). The adsorption followed pseudo-second-order kinetics and Langmuir isotherms, indicating monolayer chemisorption. These results provide valuable insights for developing efficient MOF-based desulfurization materials.

In addition, to further enhance the adsorption performance of MOFs toward aromatic sulfur compounds, researchers have employed strategies such as defect engineering, modification, and functionalization based on the characteristic interactions between MOFs and these sulfur species. To improve the performance of MOF materials in liquid fuel adsorptive desulfurization, Liberty L. Mguni et al.²⁰⁸ developed a Ni-BDC adsorbent via an acid modulation synthesis approach, using formic acid as a modulator to control the crystal structure. Although Ni doping itself reduced crystal integrity and particle size, the introduction of formic acid significantly increased crystallinity and grain size, while effectively suppressing the interpenetration phenomenon commonly observed in MOF-5. Adsorption tests demonstrated that the acid-modulated Ni-BDC nearly doubled its desulfurization capacity for thiophene compounds, with the optimal sample exhibiting an adsorption capacity of 4.14 mg g⁻¹ for thiophene and a partition coefficient of 0.053 mg g⁻¹ ppm⁻¹, outperforming DBT and 4,6-DMDBT (Fig. 10b). Formic acid played a triple role in accelerating nucleation, regulating crystal growth, and controlling crystallinity during synthesis. This study highlights that acid modulation is an effective strategy to tailor MOF structures and significantly enhance their potential in adsorptive desulfurization applications.

Song et al.²⁰⁹ fabricated a structurally stable and highly active composite adsorbent (Py/MOF) by immobilizing the functional ionic liquid [Hnmp][H₂PO₄] onto the surface of MOF-199. After modification, the material exhibited a flower-like stacked morphology while retaining the well-defined octahedral shape of MOF-199 and forming enlarged mesoporous channels. The introduction of IL nanoparticles not only increased the number of acidic sites, enhancing the interaction between sulfur compounds and the adsorbent, but also improved selectivity toward weakly basic sulfur species (Fig. 10c). The adsorption sequence for various sulfur compounds followed the order: DBT > BT > 3-MTP > TP > 2,5-DMTP, and the resistance to interference was in the order: cyclohexene > toluene > water > ethanol. After four regeneration cycles, the removal efficiency for DBT remained at 93.8%, indicating excellent cycling stability. This study clearly demonstrates the significant enhancement in MOF-based desulfurization performance achieved through ionic liquid modification.

To address the poor water resistance of MOFs, which often leads to structural collapse, Hu et al.²¹⁰ developed a composite material, Cu(I)-BTC@PAF30, by crosslinking a hydrophobic and organophilic PAF-56 framework onto the surface of Cu(I)-BTC. This design introduced hydrophobic selective channels, effectively mitigating the performance degradation caused by moisture. The resulting material exhibited strong affinity toward DBT (binding energy of −90.52 kJ mol⁻¹) and excellent water repellency (−14.72 kJ mol⁻¹) (Fig. 10d). In model oil containing 0.5% water, the adsorption capacity for DBT remained as high as 50.8 mg S per g, representing only a 17.6% decrease compared with anhydrous conditions. In contrast, unmodified Cu(I)-BTC showed a dramatic 70.9% reduction under the same conditions. Furthermore, the PAF-56 coating effectively prevented the oxidation of Cu(I), enabling the material to retain over 91% DBT removal efficiency after 20 adsorption–desorption cycles.

MOFs exhibit unique advantages in diesel adsorptive desulfurization due to their ultrahigh specific surface area and highly tunable pore structures, which allow for abundant exposure of active adsorption sites.^211–213 These materials show excellent adsorption performance, particularly for sterically hindered sulfur compounds such as 4,6-DMDBT. The presence of abundant unsaturated metal sites and organic linkers on MOF surfaces enables selective capture of sulfur species through mechanisms such as π–π interactions and acid–base interactions.

However, several significant challenges still hinder the practical application of MOFs. Their poor hydrothermal stability leads to structural collapse in the presence of water-containing fuels; narrow pore size distributions, while beneficial for selectivity, limit the diffusion of larger sulfur molecules; high synthesis costs and difficulties in large-scale production further restrict industrial deployment. Additionally, regeneration by solvent washing often suffers from low efficiency, while thermal treatment may compromise the crystalline structure.

Future improvements in MOF-based adsorbents for desulfurization should take a comprehensive approach that integrates multiple strategies. Priority should be given to the development of hydrophobic and structurally stable MOFs, particularly through mixed-ligand engineering. Constructing hierarchical mesoporous–microporous frameworks can significantly enhance mass transfer efficiency, while sulfur affinity can be improved via metal doping or the introduction of structural defects. At the same time, the adoption of green and cost-effective synthesis methods, such as mechanochemical approaches, will be crucial for sustainable development. In terms of practical application, the advancement of mild and efficient regeneration techniques, including microwave-assisted processes, is essential. Moreover, combining MOFs with functional materials such as carbon-based supports or ionic liquids holds great potential for boosting both their performance and real-world applicability.

4.2.5. Ionic liquid-functionalized adsorbents. Ionic liquids (ILs) exhibit high thermal stability, low vapor pressure, tunable structures, and strong sulfur solubility, making them ideal for constructing novel adsorbents.²¹⁴ ILs immobilized on silica, activated carbon, MOFs, or polymers can create “liquid–solid interface” systems that enhance sulfur selectivity.^215,216 Imidazolium/pyridinium cations provide π–π interactions, while acidic anions engage in hydrogen bonding or electrostatics with sulfur atoms. Some ILs also exhibit oxidative properties, enabling integrated oxidation–adsorption mechanisms.²¹⁷ Key challenges remain in ensuring stable IL immobilization and minimizing leaching during repeated cycles.

Two-dimensional (2D) materials, with their high specific surface area and abundant surface functional groups, provide an ideal platform for the uniform dispersion of ILs. The synergistic effect between the two components significantly enhances π–π interactions and electrostatic adsorption toward aromatic sulfur compounds. Additionally, the layered structure of 2D materials facilitates the diffusion of sulfur molecules, while the tunable polarity of ILs allows for improved selectivity, resulting in highly efficient composite adsorbents. Babak Mokhtarani et al.²¹⁸ reported the non-covalent functionalization of graphene and graphene oxide with 1-methyl-3-octyl-imidazolium-based ionic liquids, including [Omim][PF₆], [Omim][BF₄], and [Omim][SCN]. Comprehensive structural characterization confirmed the successful modification. The IL-functionalized materials were then applied for DBT adsorption from a n-decane model fuel. Among them, graphene modified with [Omim][PF₆] exhibited the highest adsorption capacity (6.5 mg g⁻¹), significantly outperforming unmodified graphene (3.2 mg g⁻¹) and graphene oxide (2.25 mg g⁻¹) (Fig. 11a).

Fig. 11 ADS performance of IL-functionalized adsorbents. (a) DBT adsorption capacities for different synthesized adsorbents, reproduced from ref. 218 with permission from Springer Nature,²¹⁸ copyright 2016. (b) IGMH analysis between MIL-53(Al) clusters and sulfides (TH, BT, and DBT) (gradient isosurfaces (s = 0.01 au) for the most stable configuration), reproduced from ref. 219 with permission from Elsevier,²¹⁹ copyright 2023. (c) The surface plots and counter plots showing the effect of three parameters of interest on the percent of sulfur removal at sorbent mass of 0.836 g, time of 24.7 min, and temperature of 23.2 °C, reproduced from ref. 220 with permission from Springer Nature,²²⁰ copyright 2016. Conditions in 11a: 0.02 g adsorbent, 2 ml DBT solution (500 pm), T = 298 K, t = 60 min.

The ultrahigh porosity and tunable pore size of MOFs complement the properties of ILs, facilitating greater exposure of IL active sites. Additionally, the synergistic interaction between the metal nodes in MOFs and the functional groups in ILs can enhance the chemical adsorption of sulfur compounds. This composite system integrates the structural precision of MOFs with the high sulfur affinity of ILs, making it particularly well-suited for the deep removal of trace sulfur compounds from complex fuel matrices. Ren et al.²¹⁹ developed dual-function porous liquids by combining N-heterocyclic ILs, such as dicationic imidazolium (DBU-guanidinium), with MIL-53(Al) through electrostatic interaction. The system achieved removal rates of 67.67%, 85.5%, and 84.34% for thiophene, BT, and DBT, respectively, from model oil (5000 ppm S). The process involves initial IL-mediated extraction via π–π and van der Waals forces, followed by pore-size-controlled diffusion into MIL-53(Al), where sulfur compounds are strongly adsorbed through electrostatic, Lewis acid, and π–π effects (Fig. 11b). This study reveals the molecular mechanism of porous liquid desulfurization and offers a promising solvent-based platform.

The well-defined pore channels and acidic sites of zeolites provide a stable anchoring environment for ILs, enabling selective adsorption of sulfur compounds while facilitating their mass transfer and diffusion. This combination leverages the size-selective adsorption of zeolites and the high sulfur affinity of ILs. Moreover, the rigid framework of zeolites contributes to improved cycling stability of the composite material, making it a promising system for efficient and reusable desulfurization. Semnani et al.²²⁰ developed a reusable magnetic NaY zeolite by immobilizing [bmim]Cl/FeCl₃ IL for DBT removal in n-hexane. Characterization confirmed the successful synthesis and magnetic recyclability (19.5 emu g⁻¹). In particular, they employed a response surface methodology to optimize the ADS conditions, and the optimal conditions (23.2 °C, 24.7 min, 0.836 g) yielded 97.9 ± 0.5% DBT removal (Fig. 11c). The adsorption conformed to both Langmuir and Freundlich models with a max capacity of 2.957 mg g⁻¹. Notably, the application of response surface methodology in their study allowed for a highly efficient optimization of the adsorption parameters. Magnetic separation allows easy recovery, offering a novel solution for fuel desulfurization.

Supported ionic liquid adsorbents exhibit unique advantages in diesel desulfurization, yet they also face several critical challenges. Their core strength lies in the tunable nature of ionic liquids, which allows precise matching with the polarity of various sulfur compounds, enabling efficient and selective adsorption of thiophenic species. Meanwhile, the solid supports offer high surface area and structural stability, mitigating issues associated with the high viscosity and diffusion resistance of pure ionic liquids. The synergistic interaction between the support and ionic liquid further enhances adsorption capacity and kinetics. However, these materials still encounter key bottlenecks, including potential IL leaching during long-term use, high synthesis costs of certain functionalized ILs, and limited tolerance to competitive adsorption in complex real-fuel matrices. Future developments may focus on designing covalently bonded IL systems, constructing self-healing composite supports, and engineering hybrid materials with multiple interaction sites to achieve an optimal balance between selectivity, stability, and cost-effectiveness.

The ADS process typically operates at room temperature to 150 °C without hydrogen, making it ideal for removing alkylated thiophenes or as a polishing step after HDS. After contact with adsorbents, sulfur compounds are separated from hydrocarbons via surface or pore adsorption. Regeneration methods include heating, gas purging, solvent washing, or in situ oxidation. However, challenges such as active site deactivation, metal leaching, and framework collapse during regeneration may hinder long-term reuse and industrial deployment.

In industrial settings, ADS systems can be implemented in fixed-, moving-, fluidized-bed, or membrane reactors, depending on feedstock and desulfurization requirements. Current applications include ultra-deep desulfurization of light oils, polishing of jet/diesel fuels, and selective thiophene removal from gasoline. ADS can also be integrated as a pretreatment or coupling step with oxidative or biological desulfurization to achieve process synergy and performance complementarity.

4.2.6. Cu(I)-based adsorbents. Cu(I)-based adsorbents exhibit strong π-complexation with sulfur-containing compounds (thiophene, benzothiophene, dibenzothiophene, etc.), high desulfurization capacity, and tunable selectivity, leading to their widespread application in ADS in recent years. Immobilized on zeolites or constructed as novel complexes, these adsorbents can be optimized via synthesis process regulation (one-step reduction, one-pot pyrolysis) to enhance Cu(I) dispersion and active site yield. The synergistic interaction between Cu(I) and the support enhances sulphur adsorption while improving resistance to aromatic interference. This approach ensures the stability, equilibrium capacity, selectivity, and regenerability of the Cu(I) system.

Jiang et al.²²¹ prepared Cu(I)–Y and Cu(I)–Y(III)–Y by reducing Cu(II)–Y/Cu(II)–Y(III)–Y (from NaY ion exchange with Cu(NO₃)₂/Y(NO₃)₃) under H₂. Cu(I) was critical for high adsorptive desulfurization (ADS) capacity: it enabled π-complexation with thiophene (TP)/benzothiophene (BT), making Cu(I)–Y exhibit higher sulfur capacity than Y(III)–Y. Cu(I) also synergized with Y(III) to form a new strong S–M (S: sulfur, M: metal) active site in Cu(I)–Y(III)–Y, combining Cu(I)–Y's high capacity and Y(III)–Y's selectivity—its TP/BT breakthrough loadings exceeded both Cu(I)–Y and Y(III)–Y. FT-IR/TG-DTA confirmed Cu(I)-mediated π-complexation (evidenced by 339 °C weight loss for TP on Cu(I)–Y) as a core ADS mechanism, though Cu(I)–Y was more vulnerable to xylene competition than Cu(I)–Y(III)–Y (lacking Y(III)-enhanced sites).

Subhan et al.²²² developed a facile strategy for fabricating Cu₂O-functionalized 3D ordered KIT-6: via ammonia-assisted deposition–precipitation, Cu(NO₃)₂ precursor was directly introduced into the microenvironment between template P123 and silica walls of as-synthesized (template-occluded) KIT-6, followed by one-step N₂ reduction at 700 °C. This single reduction step simultaneously achieved Cu precursor decomposition to CuO, template removal, and CuO conversion to Cu(I), simplifying synthesis compared with conventional multi-step methods. Among the samples, CuAK-20 (20 wt% Cu loading) exhibited optimal performance: it had a high Cu(I) yield (61.84%), well-dispersed Cu species (particle size <5.5 nm), and more weak/medium acid sites (6.9 mmol NH₃ per g) than CuCK-20 (synthesized from calcined KIT-6). Its thiophene adsorption capacity reached 0.28 mmol g⁻¹, significantly exceeding CuCK-20 (0.142 mmol g⁻¹) and reported Cu-based adsorbents; benzothiophene adsorption was lower due to steric hindrance. The desulfurization mechanism relied on π-complexation between Cu(I) (active sites) and thiophene. CuAK-20 also showed excellent regenerability—over 85% capacity remained after 3 cycles. This strategy efficiently promotes Cu dispersion and Cu(I) formation, making CuAK-20 a promising adsorbent for deep desulfurization.

Subhan et al.²²³ fabricated Cu(I)-functionalized SBA-15 via ammonia-driven incipient wetness impregnation and one-step N₂ reduction. The ammonia-driven incipient wetness impregnation introduced Cu(NO₃)₂ into the microenvironment between template P123 and silica walls of as-synthesized SBA-15, while the one-step N₂ reduction simultaneously removed the template, decomposed the Cu precursor to CuO, and converted CuO to Cu(I). Among samples with 10–30 wt% Cu loading, 20Cu/SBA-15 showed optimal performance. It had a 56% Cu(I) yield, well-dispersed Cu species, and retained the ordered 2D hexagonal structure of SBA-15. Its thiophene adsorption capacity reached 13 mg g⁻¹, higher than 10Cu/SBA-15 (11.33 mg g⁻¹) and 30Cu/SBA-15 (9.6 mg g⁻¹), and the experimental data fitted the Langmuir isotherm model best. The desulfurization process relied on π-complexation between Cu(I) and thiophene. 20Cu/SBA-15 also exhibited good regenerability, with an adsorption capacity of approximately 12.3–12.5 mg g⁻¹ after 2 cycles, showing promising potential for deep desulfurization.

Subhan et al.²²² developed a one-pot thermal treatment strategy to fabricate Cu₂O-based adsorbents for deep desulfurization. They directly introduced Cu(NO₃)₂ precursor into the confined space between silica walls and template P123 of as-synthesized SBA-15, then completed three processes in a single thermal treatment under N₂: decomposing Cu(NO₃)₂ to CuO, removing template P123, and reducing CuO to Cu₂O, which is simpler than conventional multi-step calcination. Among samples, CuAS-3 (with 4.5 mmol g⁻¹ Cu) performed best: its Cu(I) yield reached 73.3%, significantly higher than CuCS-3 (53.3%, prepared from template-free SBA-15). Its thiophene adsorption capacity was 0.35 mmol g⁻¹, exceeding CuCS-3's 0.27 mmol g⁻¹. Desulfurization relied on π-complexation between Cu(I) and thiophene. CuAS-3 also showed excellent regenerability; over 90% of its adsorption capacity remained after three cycles. This facile, high-performance strategy makes Cu₂O/SBA-15 promising for deep desulfurization.

Ma et al.²²⁴ systematically studied the adsorptive desulfurization (ADS) performance of a novel triangular Cu(I)-based adsorbent Cu₃pz₃ (copper(I) 4-nitro-3,5-di(trifluoromethyl)pyrazolate) against dibenzothiophene (DBT) and 4,6-dimethyldibenzothiophene (DMDBT), with comparison with the structurally similar Ag₃pz₃. Under the conditions of 298 K and adsorbent dose of 3.65 g L⁻¹, in toluene–iso-octane (v : v = 15 : 85) model oil, Cu₃pz₃ reduced the DBT sulfur content from 100 mg S per L to ∼9 mg S per L (sulfur removal rate, SR% ∼91%), which was much better than Ag₃pz₃ that only reduced it to ∼56 mg S per L (SR% ∼44%). When aromatic competitors (e.g., toluene, naphthalene) existed, Cu₃pz₃ showed significantly higher selectivity for DBT than Ag₃pz₃; after 6 cycles of regeneration by dichloromethane washing, the DBT desulfurization rate of Cu₃pz₃ still remained ∼92%, and Cu(I) was stable without oxidation (confirmed by XPS). Molecular-level studies (PXRD, SEM, EDS, and binding energy analysis) revealed that the ADS selectivity was correlated with the strength of Cu(I)–S bonding: the higher orbital interaction energy (ΔE_orb) from Cu–S covalent interaction determined the high selectivity of Cu₃pz₃, rather than the total binding energy (ΔE); the adsorption process was achieved by forming Cu₃pz₃-HASC adducts instead of surface adsorption. These findings emphasize that enhancing M–S bonding, weakening M–C bonding, or reducing the number of M–C contacts is crucial for the design of metal-functionalized ADS adsorbents with high selectivity.

In industrial settings, Cu(I)-based adsorbents may be deployed within fixed-bed or fluidized-bed ADS systems for ultra-deep desulfurization of light oils. They may also be integrated as a pre-treatment step prior to oxidative desulfurization, leveraging π-complexation selectivity to augment oxidation efficiency. Recent advances encompass streamlined synthesis (one-step template removal-reduction) and high-performance complexes. Future work must address Cu(I) oxidation during cycling and adaptability to complex fuel matrices to further unlock their industrial application potential.

4.3. Challenges of ADS and further perspective

Although ADS offers a promising route for reducing sulfur content in diesel under mild conditions, it still faces several significant limitations. Firstly, most adsorbents exhibit relatively low sulfur capacities, necessitating frequent regeneration or replacement. This not only increases operational costs and process complexity but also limits the overall efficiency of the desulfurization system. Moreover, the selectivity of current adsorbents is often insufficient, and their performance is easily compromised by competitive components such as aromatics and olefins commonly present in diesel fuel. As a result, the efficiency of sulfur removal significantly decreases, particularly when treating diesel with high aromatic content or complex compositions.

In addition, the regeneration of spent adsorbents remains a considerable challenge. Thermal or chemical regeneration can lead to structural degradation and loss of active sites, resulting in severe performance deterioration after multiple cycles and compromising long-term operational stability. From both technical and economic perspectives, the preparation of high-performance adsorbents is often costly, while their regeneration typically involves high energy consumption. Some adsorbents also require operation under high-temperature and high-pressure conditions, which increases equipment investment and operational risk. Furthermore, deactivated adsorbents often contain residual sulfur and heavy metals, necessitating additional environmental treatment and disposal measures, thereby further increasing the overall cost.

Technologically, ADS still struggles to meet the stringent ultra-low-sulfur diesel standards (e.g., <10 ppm sulfur). Its removal efficiency for refractory compounds such as thiophenes is relatively poor, often requiring multi-stage adsorption to achieve deep desulfurization. The composition of diesel fuel, particularly high aromatic or heavy hydrocarbon content, can also adversely affect the adsorbent's lifespan and performance. During industrial-scale operation, issues such as mass transfer limitations and bed plugging may occur, compromising process reliability and scalability. Furthermore, the regeneration process for adsorbents is energy-intensive, inefficient, and prone to causing the loss of active components, leading to increased replacement frequency and significantly driving up long-term operational costs.

To overcome these challenges, future research should focus on the development of novel adsorbents with high sulfur capacities and improved selectivity, such as modified activated carbons, MOFs, and hybrid materials. Additionally, optimization of regeneration processes and integration with other technologies (e.g., HDS) may offer synergistic benefits, enhancing both economic viability and practical applicability of ADS in the production of ultra-clean fuels.

5. Extractive desulfurization

5.1. Mechanisms of extractive desulfurization

Extractive desulfurization (EDS) is a liquid–liquid separation technique that relies on the difference in intermolecular interactions between organic sulfur compounds and polar solvents to achieve selective transfer of sulfur species from the oil phase to the extractant phase. The key lies in leveraging chemical or physical interactions between sulfur compounds and the extractant to disrupt the solubility equilibrium in the original oil matrix, thereby enabling directional enrichment and efficient removal of sulfur species (Fig. 12).²²⁵ This process typically operates under ambient conditions without the need for hydrogen, offering advantages such as low energy consumption, simple operation, and minimal equipment requirements. EDS is especially effective for the deep removal of refractory aromatic sulfur heterocycles, such as DBT and its alkylated derivatives, which are poorly removed by conventional HDS.^140,226 The extraction mechanism in EDS involves the synergistic contribution of several interaction modes:

Fig. 12 Schematic diagram of the basic mechanism of diesel extractive desulfurization.

5.1.1. Polarity-driven phase transfer. Aromatic sulfur compounds generally exhibit moderate polarity, which results in poor solubility in non-polar hydrocarbon matrices. When a moderately or strongly polar extractant (e.g., acetonitrile, dimethyl sulfoxide, ethanolamine, acetic acid) is introduced, sulfur compounds preferentially migrate toward the polar phase at the liquid–liquid interface, driven by polarity differences. Stronger polarity can increase the driving force for phase transfer; however, extractant polarity must be carefully matched with the sulfur species and oil composition to avoid undesired co-extraction of hydrocarbon components.²²⁷

5.1.2. π–π interactions and hydrogen bonding. Sulfur compounds such as DBT and BT, which possess aromatic rings, have high π-electron densities. When extractants with aromatic moieties (e.g., phenol, pyridine derivatives, or functionalized ionic liquids) are used, they can engage in π–π stacking interactions with these sulfur species, forming stable non-covalent complexes that enhance extraction selectivity. Additionally, extractants containing hydrogen-bond donor groups (e.g., –OH, –NH₂, –COOH) can form hydrogen bonds with the lone pair electrons on sulfur atoms, further stabilizing the sulfur–solvent complex and improving affinity.^228,229

5.1.3. Metal–sulfur coordination. In functionalized EDS systems containing metal salts or metal-based ionic liquids, coordination interactions may occur between sulfur atoms and metal centers. Transition metal ions such as Cu⁺ and Ag⁺ can form σ-type coordination complexes with sulfur atoms, significantly enhancing the extraction efficiency of DBT and its derivatives. These systems often exhibit high single-stage extraction efficiency due to strong chemical selectivity. However, challenges such as incomplete decomposition of metal–sulfur complexes during regeneration and potential metal ion leaching can hinder their practical application.²³⁰

Moreover, advanced EDS systems using ILs or deep eutectic solvents (DESs) offer more complex and tunable interaction mechanisms. In these systems, IL cations can interact with aromatic sulfur compounds via π–π stacking, while anions act as Lewis acid/base sites that engage in electrostatic or hydrogen-bonding interactions with sulfur atoms. Such multifunctional molecular recognition enables selective sulfur enrichment through synergistic binding at multiple interaction sites. Owing to their structural tunability, high selectivity, and reusability, IL- and DES-based systems have emerged as promising candidates for next-generation extractive desulfurization technologies.^134,231

Nevertheless, effective EDS not only requires high selectivity for sulfur compounds but also must minimize co-extraction of valuable hydrocarbons such as aromatics and alkanes to preserve fuel quality. Therefore, extractant design, compatibility evaluation, and recyclability are critical factors determining the industrial viability of this approach. To enhance the overall desulfurization performance, extractants are often combined with oxidants (e.g., H₂O₂, peracetic acid) in oxidative–extractive desulfurization schemes. By first oxidizing inert sulfur species into more polar sulfones, the subsequent extraction efficiency can be significantly improved, thereby expanding the scope of target sulfur compounds and enabling deeper desulfurization.²³²

5.2. Common types of extractants

In EDS systems, the design and selection of extractants are critical to achieving efficient sulfur removal. To realize deep and selective removal of refractory aromatic sulfur compounds in diesel, such as DBT and 4,6-DMDBT, extensive research has been devoted to tuning extractant molecular structures, enhancing molecular recognition capabilities, improving thermal stability, and ensuring environmental compatibility.

5.2.1. Conventional organic solvents. Traditional polar organic solvents, including acetonitrile, dimethyl sulfoxide (DMSO), ethanolamines, and pyridine derivatives, have been widely employed in early diesel EDS studies. These solvents leverage their strong polarity and good solubility toward aromatic sulfur compounds to form hydrogen bonds, dipole–dipole, or van der Waals interactions, enabling selective extraction of aromatic sulfides.²³³ However, their limited selectivity, significant co-extraction of hydrocarbon matrix components, high toxicity and volatility, as well as difficulties in solvent recovery, restrict their practical application. Consequently, attention has shifted towards structurally tunable and property-designable functionalized extractants.

Generally, due to the high cost, potential environmental pollution, and partial miscibility with fuel, the use of single organic solvents as extractants for desulfurization has been relatively limited. To address these issues, Azizian et al.²³⁴ proposed using polyethylene glycol (PEG) as a green, low-cost, efficient, non-toxic, non-corrosive, and recyclable molecular solvent for the extractive desulfurization of liquid fuels. PEG achieved a DBT extraction efficiency of 76% within just 90 seconds. Through three successive extraction cycles, the DBT concentration was reduced from 512 ppmw to 10 ppmw, yielding a desulfurization rate of 98%, which represents one of the best performances reported in terms of time and cycle number. The study also investigated the influence of initial sulfur concentration, PEG dosage, extraction time, and temperature on desulfurization efficiency. PEG showed extraction selectivity in the order of DBT > BT > DMDBT, and equilibrium was reached within 3 minutes in all cases. Its low viscosity likely facilitates rapid extraction, while its high efficiency may be attributed to its ability to form hydrogen bonds with sulfur-containing compounds. After five extraction–regeneration cycles using an adsorption-based regeneration method, the recovered PEG still retained a DBT extraction efficiency of 60%.

In addition, Qi et al.²³⁵ proposed a novel strategy for tuning the extractive desulfurization performance of tetrabutylphosphonium bromide (TBPB) through water regulation. To gain deeper insight into the underlying mechanism, the micro-level interactions among TBPB, water, and DBT were systematically analyzed. The results revealed that in the presence of a small amount of water, significant CH–π interactions occur between the TBP⁺ cation and DBT, while the Br⁻ anion forms hydrogen bonds with DBT, jointly driving the migration of DBT into the aqueous TBPB phase and thereby achieving efficient desulfurization. However, as the water content increases further, the enhanced ion hydration weakens the interactions between TBPB and DBT, leading to DBT aggregation and a decline in extraction capacity, while simultaneously promoting TBPB recovery. This study highlights the dual role played by water regulation in enhancing desulfurization efficiency and facilitating extractant regeneration, providing theoretical guidance for the design of optimized aqueous-phase extractive desulfurization systems.

Organic solvent EDS technology offers several advantages, including simple operation, mild conditions, and low energy consumption. However, this technique also faces significant limitations. It generally exhibits relatively poor removal efficiency. In addition, co-extraction of aromatic hydrocarbons from diesel can lead to fuel loss and quality degradation. The regeneration of organic solvents often involves high energy input and may result in volatile emissions, raising environmental concerns. Due to these drawbacks, the direct use of organic solvents for diesel desulfurization has been relatively underexplored in recent research.

5.2.2. Ionic liquid extractants. Ionic liquids (ILs) have emerged as one of the most representative and promising extractants due to their negligible vapor pressure, excellent thermal stability, and highly tunable structures.^{214,236–241} Typical ILs comprise organic cations, such as imidazolium, pyridinium, quaternary ammonium, phosphonium, etc., paired with inorganic or organic anions, such as [BF₄]⁻, [PF₆]⁻, [NTf₂]⁻, [HSO₄]⁻, etc. ILs can interact with aromatic sulfur compounds through π–π interactions between cation aromatic rings and sulfur-containing aromatics, as well as via hydrogen bonding or Lewis acid–base interactions between anions and sulfur atoms, thereby enhancing selectivity.²⁴² For example, [Bmim]BF₄ exhibits high extraction efficiency for DBT through combined hydrogen bonding and π–π stacking, whereas [Bmim]HSO₄ displays synergistic effects due to its acidic anion's oxidative assistance. To further improve performance, metal-functionalized ILs such as [Bmim]⁺-Cu⁺ systems have been developed, wherein metal d-orbitals form stable complexes with sulfur atoms, significantly enhancing selective recognition of inert sulfur compounds like DBT and 4,6-DMDBT. In parallel, novel ILs including magnetic ILs, polyionic liquids, and thermo-responsive ILs have been synthesized to overcome drawbacks such as high viscosity and challenging recycling, thereby advancing their applicability in green EDS processes.

The most common ionic liquid system for extractive desulfurization based on π–π interactions involves imidazolium-based ionic liquids. This is attributed to the π-electrons of the imidazole ring and the aromatic sulfur compounds, which facilitate strong π–π stacking interactions, thereby enabling efficient sulfur extraction. Imidazolium-based ionic liquids (ILs) are among the most widely studied extractants for desulfurization, owing to their green, biodegradable, and low-toxicity properties. To elucidate the extraction mechanism of these ILs, Wu et al.²⁴³ employed DFT to investigate the interactions between DBT and a series of ILs: N-butyl-N-methylimidazolium tetrafluoroborate ([BMIM][BF₄]), N-butyl-N-methylmorpholinium tetrafluoroborate ([BMmorpholinium][BF₄]), N-butyl-N-methylpiperidinium tetrafluoroborate ([BMPiper][BF₄]), N-butyl-N-methylpyrrolidinium tetrafluoroborate ([BMPyrro][BF₄]), and N-butylpyridinium tetrafluoroborate ([BPY][BF₄]). The results revealed the widespread presence of hydrogen bonding and van der Waals interactions across all IL-DBT systems, along with ion–π interactions between DBT and either the cation or anion. Notably, π⁺–π interactions were observed in the [BMIM][BF₄]-DBT and [BPY][BF₄]-DBT systems, indicating stronger interactions when aromatic cations are involved (Fig. 13a). These findings suggest that ILs containing aromatic cations exhibit higher interaction energies with DBT, thereby enhancing their performance in extractive desulfurization. This study provides theoretical insight for the rational design and structural optimization of ILs for more efficient sulfur removal.

Fig. 13 EDS of diesel by different IL extractants. (a) The sign (λ₂)ρ vs. RDG (left) and the gradient isosurfaces (right) for (a1) [BMIM][BF₄]-DBT, (a2) [BMmorpholinum][BF₄]-DBT, (a3) [BMPiper][BF₄]-DBT, (a4) [BMPyrro][BF₄]-DBT, and (a5) [BPY][BF₄]-DBT. In the isosurfaces, the blue and green regions indicate attraction, the yellow and red regions indicate steric effect, reproduced from ref. 243 with permission from Springer Nature,²⁴³ copyright 2017. (b) Proposed interactions between DBT and N-triethylene glycol ether imidazolium IL, reproduced from ref. 244 with permission from Elsevier,²⁴⁴ copyright 2020. (c) Coordination type model between BT and [BMIM]⁺[FeCl₄]⁻, and the interaction energy and selectivity is linearly related to the amount of charge transfer, reproduced from ref. 245 with permission from Elsevier,²⁴⁵ copyright 2015.

In addition to π–π interactions, researchers have identified other intermolecular forces involved in imidazolium-based ionic liquids (ILs). Fiksdahl et al.²⁴⁴ synthesized a series of 36 bifunctional N-(poly)ethylene glycol-substituted N-allyl/benzyl imidazolium and benzimidazolium ILs to investigate their extractive desulfurization performance toward DBT and 4,6-DMDBT in model oils. The benzimidazolium-based N-allyl ILs exhibited significantly higher desulfurization efficiencies compared with their imidazolium analogues (DBT removal rates of 69% vs. 59%; 4,6-DMDBT removal rates of 52% vs. 29%). The desulfurization performance was strongly influenced by the length of the N-ethylene glycol chain and the type of N-allyl/benzyl substitution, while the anion type (NTf₂⁻, N(CN)₂⁻, SCN⁻, BF₄⁻) had a comparatively minor effect on sulfur adsorption capacity. Further results demonstrate that benzimidazolium ILs interact with aromatic sulfur compounds not only through strong π–π interactions but also via hydrogen bonding and van der Waals forces (Fig. 13b). These multiple, synergistic non-covalent interactions collectively enhance adsorption capability and are key to the superior desulfurization performance observed.

Notably, not all ionic liquid anions have a negligible effect on extractive desulfurization performance. A notable exception are the metal-containing ionic liquids (MILs), which represent a class of functionalized materials incorporating metal elements into the structure of ionic liquids, either as part of the cation, anion, or through coordination complexes. The key feature of MILs lies in their ability to combine the tunable nature of ionic liquids with the unique properties of metal centers, such as coordination ability and redox activity, to create highly active media. In extractive desulfurization applications, MILs typically feature metal ions integrated into the anionic component. These anions often possess coordinatively unsaturated metal centers that can form strong M–S interactions with sulfur-containing compounds, such as thiophenes and dibenzothiophenes. This interaction facilitates efficient sulfur extraction from fuel oils. The enhanced desulfurization performance of MILs is thus primarily attributed to the specific coordination interactions between the metal centers and the target sulfur species.

Maschmeyer et al.²⁴⁶ systematically investigated the extractive desulfurization mechanisms of aromatic sulfur compounds using a combination of experimental techniques and DFT calculations. A comparative study of various commonly used ILs at equimolar loading revealed that the extraction efficiency is strongly influenced by the structural characteristics of both the cation and anion components. The cationic performance followed the trend: pyridinium > imidazolium > pyrrolidinium, while the anionic order was: [NTf₂]⁻ > [OTf]⁻ > [PF₆]⁻ > [BF₄]⁻. Both thiophene and DBT exhibited similar extraction behavior across different ILs. Further DFT calculations using the APFD method were conducted to quantify the complexation energies and dispersion interactions between thiophene and various cations and anions. The computational results were in excellent agreement with the experimental findings. The combined analysis demonstrated that the extractive desulfurization process is predominantly driven by dispersion interactions, and that non-local van der Waals forces between the IL and sulfur compounds play a critical role in determining the extraction efficiency.

Cheong et al.²²⁶ reported that iron(III)-based ionic liquids (ILs), synthesized by reacting anhydrous ferric chloride (FeCl₃) with imidazolium chlorides ([imidazolium]Cl), were effectively applied for the extractive desulfurization of model oils containing DBT. Experimental results showed that increasing the molar ratio of FeCl₃ to [imidazolium]Cl significantly enhanced the DBT extraction efficiency. The superior desulfurization performance of these ILs was primarily attributed to the synergistic effect of their Lewis acidity and favorable fluidity.

Li et al.²⁴⁵ employed DFT calculations to systematically investigate the interaction mechanisms between a series of aromatic sulfur compounds and the ionic liquid [BMIM]⁺[FeCl₄]⁻. The study revealed that coordination structures play a crucial role in the extractive desulfurization process. The interaction energies and extraction selectivity for different sulfur compounds followed the order: TH < DBT ≈ BT. Alkyl substitution on TH or BT, such as in 3-methylthiophene and 3-methylbenzothiophene, was found to enhance their interactions with the ionic liquid. However, in the case of sterically hindered alkyl derivatives like 2,7-dimethylbenzothiophene, the interaction strength was diminished. The primary driving force for the extractive desulfurization process was attributed to charge-transfer effects, wherein electrons are transferred from the aromatic sulfur compounds to the Lewis acidic center of the [FeCl₄]⁻ anion. The authors emphasized that the Lewis acidity of Fe-based ionic liquids should be evaluated from the perspective of the entire complex structure formed between [FeCl₄]⁻ and the sulfur compound, rather than by focusing solely on the individual Fe or S atoms.

Liu et al.²⁴⁷ systematically investigated the interactions between N-methylpyrrolidone–ferric chloride (NMP–FeCl₃) and thiophene (TH) in simulated oil systems, including 1-octene, toluene, cyclohexane, and cyclohexene, using DFT. The study revealed the presence of multiple non-covalent and coordination interactions in the system. These include π–π stacking, hydrogen bonding, C–H⋯π interactions, and sulfur–iron coordination. Among them, the S–Fe coordination interaction, formed between the lone pair electrons of the sulfur atom and the vacant orbitals of the iron center, was found to be the most prominent. This interaction was identified as the dominant mechanism for the efficient extraction of thiophene from fuel by the NMP–FeCl₃ system.

In addition, researchers have investigated reactor design to enhance the scale-up of IL-based EDS processes. Zhao et al.²⁴⁸ introduced a microreactor to intensify the IL-based EDS, employing a slug flow regime (capillary number <0.01) with model diesel as the continuous phase and [BMIM][BF₄] as the dispersed phase to achieve efficient extraction of DBT. Results demonstrated that the extraction equilibrium time was drastically reduced from over 15 minutes in conventional batch reactors to just 120 seconds within the microreactor. A two-film model was applied to describe the mass transfer in the slug flow units, indicating that the primary mass transfer resistance resided in the liquid film surrounding the IL droplets. Further analysis revealed that increasing temperature or the IL-to-MD flow rate ratio enhanced desulfurization performance, attributed to significant improvements in diffusion coefficients and expansion of interfacial surface area.

ILs have demonstrated unique advantages in the field of diesel extractive desulfurization. Their highly tunable molecular structures allow precise recognition and efficient extraction of various sulfur compounds by adjusting the combinations of cations and anions, exhibiting particularly high selectivity toward benzothiophene derivatives, which are difficult to remove via conventional methods. Owing to their extremely low vapor pressure and excellent thermal stability, ILs effectively avoid volatilization losses common in traditional organic solvents. Moreover, the incorporation of metal coordination centers or acidic functional groups can further enhance their affinity for sulfur-containing compounds.

However, several challenges hinder their practical industrial application. These include high material costs, limited mass transfer due to relatively high viscosity, potential miscibility with aromatic components in diesel, and issues related to large-scale synthesis and economic feasibility. Future research is expected to focus on the development of novel ILs with reduced viscosity, exploration of integrated processes coupling ILs with other separation technologies, and advancement of recyclable extraction systems. In addition, utilizing low-cost raw materials and implementing continuous production strategies will be key to reducing the overall process cost and promoting the industrialization of IL-based desulfurization technologies.

5.2.3. Deep eutectic solvent extractants. Deep eutectic solvents (DESs), often referred to as “ionic liquid analogues”, represent a rapidly emerging class of green solvents formed by hydrogen bond acceptors, such as choline, quaternary ammonium salts, etc., and hydrogen bond donors, such as glycerol, urea, ethylene glycol, carboxylic acids etc., mixed in specific molar ratios to form stable liquids.²⁴⁹ DESs feature flexible molecular design, low toxicity, biodegradability, and significantly lower cost than ILs, making them promising alternatives.²⁵⁰ Studies indicate that DESs interact with aromatic sulfur compounds mainly via hydrogen-bonding networks and π–π stacking, yielding high selectivity toward DBT, 4-MDBT, 4,6-DMDBT, and similar species. Systems such as choline chloride/ethylene glycol and choline chloride/lactic acid generally exhibit distribution coefficients for DBT above 1.5, with good recyclability. Moreover, metal-based DESs incorporate metal–sulfur coordination mechanisms to further enhance recognition of refractory sulfur species, demonstrating excellent desulfurization performance in diesel EDS applications. Currently, there are several different types of DESs that have been employed in EDS of diesel, including glycol-based DES, sulfone-based DES, aromatic-based DES, metal-based DES, organic acid-based DES, amide-based DES, and so on.

Li et al.²⁵¹ designed and synthesized a series of deep eutectic solvents (DESs) using typical hydrogen bond acceptors (HBAs), such as choline chloride (ChCl), tetramethylammonium chloride (TMAC), and tetrabutylammonium chloride (TBAC), in combination with various hydrogen bond donors, including malonic acid (MA), glycerol (Gl), tetraethylene glycol (TEG), ethylene glycol (EG), polyethylene glycol (PEG), and propionic acid (Pr). By employing different HBA–HBD combinations, a range of DES systems was constructed and evaluated for extractive desulfurization of model fuel, and all the synthesized DESs were effective in removing sulfur compounds from fuel. Under optimized conditions, the TBAC/PEG system achieved a removal rate of up to 82.83% for thiophene-based sulfur compounds in a single extraction cycle, significantly outperforming both conventional and functionalized ionic liquid systems (Fig. 14a). After five consecutive extraction cycles, the cumulative desulfurization efficiency reached 99.48%, reducing the sulfur content in the fuel to below 8.5 ppm.

Fig. 14 EDS of diesel by different DESs. (a) Effect of DESs types on extraction efficiency, reproduced from ref. 251 with permission from Royal Society of Chemistry,²⁵¹ copyright 2013. (b) Extraction efficiencies of Th, DBT, Py, and Carb from n-heptane using four different DES, PEG400, and sulfolane, reproduced from ref. 252 with permission from American Chemical Society,²⁵² copyright 2019. (c) Formation mechanism of the DES-T, reproduced from ref. 253 with permission from American Chemical Society,²⁵³ copyright 2015. (d) Extraction of the model oils with different sulfur compounds by [C₁₂DMEA]Cl/FeCl₃ DES, reproduced from ref. 254 with permission from Elsevier,²⁵⁴ copyright 2017. (e) Optimized structures of DBT interacting with MIM/PA (left) or n-octane (right), reproduced from ref. 255 with permission from American Chemical Society,²⁵⁵ copyright 2016.

Marrucho et al.²⁵² found that DESs formulated from tetrabutylphosphonium hydrogen sulfate and sulfonic acids exhibited excellent extraction capabilities not only for sulfur-containing compounds such as Th and DBT, but also showed superior performance in removing nitrogen-containing compounds like pyridine (Py) and carbazole (Carb). In multicomponent systems, the desulfurization efficiency was slightly reduced due to competitive interactions; however, the extraction of nitrogen compounds was actually enhanced under these conditions. Moreover, the DESs demonstrated stronger selectivity for nitrogen compounds in model gasoline systems, while in model diesel systems, the presence of complex hydrocarbon components interfered to some extent with sulfur removal (Fig. 14b).

Li et al.²⁵³ prepared a series of DESs composed of AlCl₃, chlorinated paraffin-52, and various aromatic hydrocarbons, which exhibited excellent extraction performance toward a range of thiophenic sulfur compounds. The type of aromatic solvent had a significant influence on desulfurization efficiency, following the order: toluene > p-xylene > o-xylene > ethylbenzene > benzene > chlorobenzene. Among them, the toluene-based DES (DES-T) achieved highly efficient removal of 3-methylthiophene, benzothiophene, and dibenzothiophene at room temperature (293 K) within just 10 minutes. The desulfurization mechanism was primarily attributed to the synergistic effects of π–π interactions between aromatic rings and Lewis acid–base coordination between the sulfur atom and the Lewis acidic Al³⁺ centers (Fig. 14c). Moreover, these DESs also demonstrated impressive performance in real fuel oils, significantly improving fuel quality and highlighting their potential for industrial-scale desulfurization applications.

Li et al.²⁵⁴ synthesized and characterized a series of metal-based choline-derived deep eutectic solvents (DESs), featuring quaternary ammonium salts modified with various linear alkyl substituents. Among the studied systems, the [C₁₂DMEA]Cl/FeCl₃ DES, incorporating a dodecyl-substituted ammonium salt, exhibited the highest desulfurization efficiency, achieving up to 52.9% removal of thiophenic sulfur compounds. To further elucidate the extraction mechanism, a comparative study was conducted using a phenyl-substituted DES ([BzMDEA]Cl/FeCl₃). The results suggested that CH–π interactions may play a more prominent role than traditional π–π interactions in the desulfurization process of these DESs (Fig. 14d). Moreover, this class of DESs demonstrated excellent recyclability, maintaining stable desulfurization performance after simple water-washing regeneration, indicating strong potential for practical applications.

Li et al.²⁵⁵ further synthesized and characterized a novel class of ionic liquid-based deep eutectic solvents, in which the ionic liquid is generated in situ within the system. Three DESs were constructed using 1-methylimidazole (MIM) and diethanolamine as hydrogen bond acceptors, and propionic acid or nitric acid as hydrogen bond donors. Among them, the MIM/propionic acid system exhibited the highest desulfurization performance, with a sulfur distribution coefficient (K_N) reaching 2.31. This system possessed significantly lower viscosity compared with other reported DESs and low-viscosity ionic liquids, thereby promoting mass transfer during the extraction process. Furthermore, the presence of aromatic hydrocarbons such as p-xylene or cyclic olefins like cyclohexene in model fuel had negligible impact on the extraction of thiophenic compound (Fig. 14e).

DESs, as a new class of green solvents for extractive desulfurization of diesel fuels, have demonstrated unique advantages alongside several limitations that remain to be addressed. The notable benefits of DESs include their low-cost and readily available raw materials, such as choline chloride combined with urea or other hydrogen bond donors, simple preparation procedures (without the need for complex purification), and selective solubility toward thiophenic sulfur compounds via hydrogen-bonding networks. Particularly, DESs exhibit superior extraction efficiency for DBT compared with conventional solvents. In addition, DESs are biodegradable, virtually non-volatile, and environmentally benign.

However, several technical bottlenecks still hinder their practical application: (1) incomplete phase separation between DESs and aromatic-rich diesel fractions can lead to product losses; (2) chemical stability of DESs under prolonged high-temperature operation is inadequate; and (3) current systems still struggle to achieve ultra-deep desulfurization for diesel fuels with sulfur content below 10 ppm, and multi-stages are required.

Future advancements are anticipated to focus on three main aspects. At the molecular design level, the development of novel metal-based DESs, such as systems incorporating ZnCl₂ or FeCl₃ as hydrogen bond acceptors, may enhance sulfur coordination capabilities, while designing hydrophobic DESs could reduce mutual solubility with fuel oils. From a process intensification perspective, the application of microwave-assisted extraction or ultrasound-enhanced mass transfer technologies, as well as the integration of DES-based systems with membrane separation processes for in situ regeneration, offer promising directions. In terms of system integration, coupling DES extraction with subsequent catalytic oxidation or adsorption techniques can enable multi-stage deep desulfurization. Moreover, establishing a comprehensive environmental risk assessment framework for the degradation products of DESs is essential to ensure sustainable development.

5.2.4. Porous ILs and DESs. Porous liquids are a class of functional materials that combine the fluidity of conventional liquids with permanent intrinsic porosity. Traditionally, porosity has been considered an inherent property of solid materials—such as porous carbons, metal–organic frameworks (MOFs), covalent organic frameworks (COFs), and zeolites—while liquids are typically viewed as densely packed systems lacking internal cavities. Porous liquids, however, transcend this conventional boundary by achieving the coexistence of “fluidity” and “permanent porosity” through molecular design, mesoscale structural control, or composite assembly strategies.^256,257 Owing to their unique combination of tunable porosity, flowability, and selective pore structures, PLs have attracted significant attention in a wide range of fields, including gas capture, separation and purification, energy storage and conversion, and as reaction media.²⁵⁸

The concept of porous liquids was first proposed by James and co-workers in 2007.²⁵⁹ Later, in 2015, the same group, in collaboration with Dr Sheng Dai from Oak Ridge National Laboratory, successfully synthesized the first porous liquids and conducted a systematic classification and in-depth investigation of their structural characteristics and application potential.^260,261 These efforts have since propelled the rapid development of porous liquid research. In recent years, our group and researchers from other groups have employed porous ionic liquid and porous DESs in EDS.

Our group reported a porous ionic liquid (PIL) system that was constructed by incorporating n-butylpyridinium tetrachloroferrate ([BPy][FeCl₄]) as the organic guest within a microporous boron nitride (m-BN) framework.²⁶² This design effectively promoted electron transfer between [BPy][FeCl₄] and m-BN, enhanced the electron density of the pyridinium cation, and improved the dispersion of m-BN, thereby facilitating the full exposure of extraction sites. Molecular size simulation and gas adsorption experiments confirmed the presence of permanent porosity within the material. Compared with pure [BPy][FeCl₄], the m-BN-PIL composite exhibited significantly improved mass transfer capacity, ultimately leading to superior desulfurization performance. Under optimized conditions, the removal efficiencies of the m-BN-PIL for DBT and 4-methyldibenzothiophene (4-MDBT) reached 59.2% and 61.8%, respectively. After four-stage stepwise extraction, a cumulative desulfurization efficiency of up to 98.2% for DBT was achieved, enabling ultra-deep desulfurization of model fuel. Moreover, the PIL exhibited excellent structural stability during the extraction process.

Zhang et al.²⁶³ developed a novel quaternary ammonium-functionalized organosilicon-based porous ionic liquid (OS-PIL) for the efficient extractive desulfurization of thiophenic compounds (represented by DBT) from model fuels. This material exhibited excellent extraction performance, with a Nernst partition coefficient as high as 4.22, significantly outperforming conventional ionic liquid systems. The OS-PIL structure integrates the synergistic advantages of a porous framework and ionic liquids, while the multiple active sites in the quaternary ammonium groups further enhance the interactions with sulfur compounds. DFT calculations and spectroscopic analyses revealed that the desulfurization mechanism is primarily governed by strong interactions between the ionic pairs and the porous structure with DBT, especially dominated by C–H⋯π interactions. Such porous ionic liquids, featuring high surface area and multiple active sites, show great promise in liquid–liquid extractive desulfurization applications.

Building upon this, Zhang et al.²⁶⁴ further developed a novel proton-functionalized porous ionic liquid (PPIL) via an innovative one-step coupling neutralization reaction strategy for extractive desulfurization. Experimental results demonstrated that the PPIL achieved a single-cycle extraction efficiency of 75.0% for DBT. Upon addition of aromatic hydrocarbon interferents, the extraction efficiency of PPIL showed only slight decreases (from 45.2% to 37.3%, 37.9%, and 33.5%), highlighting its excellent selectivity. Combined experimental measurements and density functional theory (DFT) calculations revealed that the surface channels of the porous structure possess strong electrophilicity, enabling selective capture of DBT. Meanwhile, multiple active sites within the ionic pairs synergistically facilitate DBT enrichment and transport from the oil phase to the PPIL via π⋯π, C–H⋯π, and hydrogen bonding interactions.

5.3. Challenges of EDS and further perspective

EDS of diesel has attracted considerable attention due to its operational simplicity and mild process conditions. However, several critical challenges remain to be addressed. Firstly, existing solvent systems, including conventional organic solvents, ionic liquids, and DESs, exhibit notable limitations in selectivity. They generally show low removal efficiency for structurally complex and high molecular weight aromatic sulfur compounds such as DBT. Moreover, these solvents tend to co-extract aromatic hydrocarbons present in diesel, resulting in reduced fuel yield and compromised economic viability.

Besides, certain solvents suffer from high volatility or emulsification with the oil phase, causing solvent degradation or residual contamination during recycling and regeneration steps, thereby increasing operational costs and undermining process stability. To achieve ultra-deep desulfurization targets (sulfur content below 10 ppm), multi-stage extraction or coupling with other desulfurization technologies is often required, leading to significant capital investment and energy consumption, which restrict large-scale industrial application.

From an environmental perspective, solvent recovery and disposal processes may generate secondary pollution, including challenges in safe treatment of spent organic solvents. Additionally, the long-term ecological toxicity of some DESs remains insufficiently evaluated, posing potential environmental risks that warrant attention.

Addressing these challenges, future research should focus on the design of novel functionalized solvents that combine high selectivity for sulfur compounds with low solubility in the oil phase. Promising directions include solvent systems derived from hydrophobic MOFs, where machine learning and molecular simulation techniques can accelerate solvent screening and performance optimization. Process intensification strategies, such as coupling with ultrasonic or microwave-assisted mass transfer enhancement, or development of microchannel continuous extraction, are expected to improve desulfurization rate and energy efficiency.

Innovative integrated approaches combining extraction with catalytic oxidation (e.g., photocatalysis, electrochemical oxidation) and adsorption technologies are likely to form “extraction–conversion” unified systems, enabling efficient sulfur compound transformation and recovery. Simultaneously, green chemistry principles will drive the development of biodegradable solvent systems, closed-loop recovery, and recycling processes to minimize waste discharge and environmental impact, steering extractive desulfurization technology towards higher efficiency, economic feasibility, and sustainability.

6. Oxidative desulfurization

6.1. Mechanism of oxidative desulfurization

ODS is a non-hydrodesulfurization technology that achieves efficient removal of organic sulfur compounds in diesel under mild conditions, typically at ambient temperature and pressure.²⁶⁵ The core principle of ODS lies in the use of oxidants to convert sulfur-containing compounds into more polar and easily separable oxidation products such as sulfoxides and sulfones. This approach is particularly effective for refractory aromatic sulfur compounds that are difficult to remove by conventional HDS, including DBT and 4,6-DMDBT.²⁶⁶ Compared with traditional HDS processes, ODS does not require high temperature, high pressure, or large amounts of hydrogen, resulting in lower energy consumption and reduced equipment costs.²⁶⁷ Therefore, ODS is well suited for upgrading ultra-low sulfur fuels such as aviation kerosene and clean diesel, which demand stringent sulfur specifications.

The ODS process typically involves two main steps: the first step is the catalytic oxidation of organic sulfur compounds in diesel under the action of a catalyst, where the oxidant selectively oxidizes sulfur species into more polar sulfoxides; the second step is the separation of the oxidation products (Fig. 15). The generated sulfoxides or sulfones are separated from the fuel through methods such as extraction, adsorption, or membrane separation, thereby achieving the removal of aromatic sulfur compounds (Fig. 15).²⁶⁸ The overall efficiency of the ODS process hinges on the selectivity of the oxidation reaction, the catalytic activity and stability of the catalyst, and the effectiveness of product separation. In the catalytic oxidation stage, the type of oxidant directly determines the oxidation conversion efficiency of sulfur compounds. Common oxidants include H₂O₂, peracetic acid (CH₃COOOH), organic peroxides, ozone (O₃), and molecular oxygen (O₂) in air.

Fig. 15 Schematic diagram of the basic mechanism of diesel oxidative desulfurization.

6.2. Commonly used oxidants

H₂O₂ is currently one of the most widely used oxidants due to its low cost, safety, and environmentally friendly byproduct (water), making it an ideal green oxidant.^269–273 However, the O–O bond in H₂O₂ possesses a relatively high bond energy, and its reactivity is low without catalysts, making it difficult to efficiently oxidize aromatic sulfur compounds. Catalysts play a key role by promoting the cleavage of the O–O bond and generating more reactive oxygen species with higher oxidation potential. Metal oxide catalysts, such as MoO₃, V₂O₅, TiO₂, etc., can form peroxo-metal complexes with H₂O₂ (such as MoO(O₂)₂ and VO(O₂)₂) via coordination with metal centers. These complexes exhibit strong electrophilicity and can form transition state complexes with sulfur atoms in aromatic sulfides like DBT, facilitating oxygen transfer.^274,275 Heteropoly acids, such as H₃PW₁₂O₄₀, and their metal derivatives can react with H₂O₂ to form peroxo-heteropoly acid intermediates (such as [PO₄WO(O₂)₄]³⁻), which possess excellent oxygen transfer abilities, efficiently oxidizing sulfur to sulfoxides or sulfones.²⁷⁶ Furthermore, under certain conditions (e.g., the presence of Fe²⁺), H₂O₂ can undergo Fenton-like reactions to generate hydroxyl radicals (˙OH), which further oxidize aromatic sulfides to highly polar sulfones.²⁷⁷ However, the lifetime and diffusion of free radicals in organic oil phases are limited; thus, in extractive–oxidative desulfurization systems, radical generation typically occurs in the extractant phase to prolong radical lifetime. In direct oxidative desulfurization of oil, non-radical pathways dominate, where metal-centered oxygen transfer mechanisms form peroxo intermediates that catalytically oxidize sulfur compounds.

Organic peroxides, such as tert-butyl hydroperoxide (t-BuOOH), are often used in oil-phase systems due to their stronger hydrophobicity and higher oxidative stability. These oxidants coordinate with metal centers, such as Ti⁴⁺, Mo⁶⁺, W⁶⁺, etc., on catalysts or within pores to form metal–peroxo complexes with enhanced catalytic oxidation activity, thereby oxidizing aromatic sulfur compounds to sulfones.²⁷⁸ For example, in titanium silicalite-1 (TS-1) catalysts, t-BuOOH forms Ti–O–O-tBu complexes with four-coordinate Ti⁴⁺ centers. The Ti–O–O-tBu complex exhibits excellent spatial orientation and selectivity, allowing a single oxygen atom to be specifically transferred to sulfur atoms, resulting in efficient oxidation of DBT and related compounds. Additionally, in the presence of certain transition metal catalysts, t-BuOOH can generate alkoxy (t-BuO˙) or peroxy radicals (t-BuOO˙). The t-BuOO˙ radical, a highly electrophilic intermediate, attacks sulfur-containing aromatic rings, selectively oxidizing them to sulfones with high efficiency.¹³⁶ Compared with other oxidants, tert-butyl hydroperoxide generates fewer byproducts and better preserves fuel quality. However, tert-butyl hydroperoxide is relatively expensive and is sensitive to temperature and light, which may impact its stability and storage.

O₃ is a strong oxidant that decomposes to generate highly reactive species such as atomic oxygen (˙O), ˙OH, and ˙O₂⁻.²⁷⁹ Under non-catalytic conditions, ozone can electrophilically add to the aromatic π electrons or lone pairs of DBT, forming ozonide intermediates that rapidly cleave and oxidize sulfur atoms. However, in practical ODS applications, ozone exhibits poor catalytic oxidation selectivity and tends to oxidize unsaturated hydrocarbons in the fuel matrix. Therefore, catalysts are often used to enhance selective recognition of aromatic sulfur compounds. In the presence of catalysts, O₃ can adsorb and activate on catalyst surfaces, producing highly reactive oxygen atoms or surface oxide species that transiently coordinate with sulfur atoms to facilitate oxygen transfer reactions and improve desulfurization selectivity. Nevertheless, the requirement for specialized ozone generators and the associated high equipment cost, coupled with difficulties in maintaining stable long-term operation, limit its widespread industrial adoption. In addition, the low selectivity of ozone may induce side reactions involving non-target components in diesel, potentially altering fuel properties or causing over-oxidation of hydrocarbon compounds.

O₂ is theoretically an inexpensive and green oxidant. However, due to its triplet ground state, O₂ is highly stable and exhibits low reactivity under ambient conditions, requiring highly active catalysts for effective activation.²⁸⁰ Under the catalysis of transition metal oxides, nitrides, carbides, or noble metals (e.g., Pt, Pd), O₂ can be activated via single-electron transfer processes to generate reactive oxygen species such as ˙O₂⁻.^281,282 In certain bifunctional catalysts (e.g., Fe–Mo composites), metal centers provide electron transfer pathways that convert O₂ into surface-active oxygen species, which selectively attack sulfur atoms in organic sulfides to oxidize them to sulfones, thus accomplishing desulfurization. Recent studies also indicate that some MOFs, covalent organic frameworks (COFs), boron nitride, carbon materials, and heteropoly acids can efficiently activate O₂ for effective oxidative desulfurization.^283–285

Besides the commonly seen oxidants, some other oxidants such as NaClO, KMnO₄, etc., have also been employed in ODS. NaClO, as a strong oxidizing agent, exhibits high desulfurization efficiency and can effectively oxidize aromatic sulfur compounds in diesel into sulfones, which are more easily removed through subsequent extraction. Additionally, NaClO has relatively low environmental impact due to its ease of degradation and operational simplicity. However, its limited stability and corrosive nature may pose challenges to reaction equipment and long-term operation. Similarly, KMnO₄ is an excellent oxidant capable of facilitating efficient oxidation reactions at relatively low temperatures, making it effective for the removal of organic sulfur compounds from diesel. Its widespread application and straightforward handling are advantageous, yet its high cost restricts its scalability for industrial use. Moreover, the oxidation process involving KMnO₄ may produce manganese oxides or other byproducts that could affect the quality of diesel or result in over-oxidation, leading to the formation of undesired products.

6.3. Separation of oxidation products in oxidative desulfurization

In ODS systems, the separation of oxidized sulfur compounds is a critical step for achieving efficient desulfurization. Effective separation of oxidation products can restore catalytic active sites in time, thus maintaining continuous ODS reactions. Traditional ODS technologies typically adopt a two-step approach: firstly, oxidizing sulfur compounds in diesel into more polar sulfoxides or sulfones; secondly, separating these oxidation products from the oil phase through solvent extraction, adsorption, or membrane separation. Although this method provides relatively good separation performance for sulfur compounds, it suffers from complex operational procedures, limited separation efficiency, high interfacial mass transfer resistance, and high energy consumption for solvent recovery.

Traditional separation mainly relies on polarity differences: oxidation products such as DBTO and DBTO₂ possess stronger polarity and lower solubility in nonpolar diesel, enabling their extraction by polar solvents, such as acetonitrile, DMSO, ethanol, etc. However, multiple extraction cycles require large solvent volumes, and problems such as emulsion instability and solvent recovery restrict industrial applicability. Alternatively, solid adsorbents, such as functionalized silica, activated carbon, MOFs, and hybrid materials, have been employed for solid–liquid separation. While this avoids solvent use, mass transfer limitations and rapid saturation of adsorption sites in the oil phase limit performance.

To enhance overall ODS efficiency and reduce energy consumption, recent research has focused on integrating oxidation and separation steps, especially oxidation–adsorption coupling systems and extraction–oxidation coupling systems, enabling simultaneous reaction and separation and thus process intensification.

Adsorption coupled with oxidative desulfurization (ACODS) achieves concurrent oxidation and efficient capture/separation of oxidation products within one system, overcoming drawbacks of the conventional two-step approach such as low adsorbent selectivity and complex post-reaction handling (Fig. 16).²⁸⁷ In ACODS, adsorbents must exhibit strong physical adsorption and selective polarity recognition, often functioning as catalytic supports or active centers, forming bifunctional catalytic–adsorptive materials.²⁸⁸ Current high-performance ACODS catalysts are typically prepared by loading catalysts with strong oxidative activity (e.g., heteropoly acids, metal oxides, transition metal complexes) onto high-surface-area adsorbents such as mesoporous silica, activated carbon, MOFs, and covalent organic frameworks (COFs). These supported catalysts not only highly disperse active sites, enhancing catalytic oxidation performance, but also adsorb oxidation products in situ, achieving effective oxidative desulfurization.^289,290 During reaction, aromatic sulfur compounds (e.g., DBT and derivatives) diffuse to catalyst surfaces and oxidize into polar sulfoxides or sulfones, which are subsequently adsorbed via hydrogen bonding, polarity interactions, or π–π stacking with the support, thus being removed from the reaction medium. This in situ separation prevents catalyst fouling or deactivation and facilitates catalyst recycling, simplifying post-treatment and improving system stability.

Fig. 16 Schematic diagram of the mechanism of ACODS, reproduced from ref. 286 with permission from American Chemical Society,²⁸⁶ copyright 2014.

Extraction Coupled with Oxidation Desulfurization Systems (ECODS) integrate extraction and oxidation in one solvent environment based on liquid–liquid phase behavior, enhancing sulfur transfer and transformation at the reaction interface via functionalized extractants (Fig. 17).^292,293 In conventional biphasic ODS, oxidants usually reside in the polar phase, such as water, acetonitrile, etc., whereas organic sulfur compounds exist in the nonpolar oil phase, limiting interfacial contact area and reaction rate. To overcome this, ECODS systems have developed solvents capable of dissolving both oxidants and selectively separating sulfur compounds, such as ionic liquids and DESs.^294,295 These extractants offer high polarity, selectivity, and tunability, preferentially forming stable complexes or hydrogen bonds with aromatic sulfur compounds and their oxidation products, and provide a microenvironment that stabilizes and enhances oxidant activity (e.g., H₂O₂, t-BuOOH). For example, imidazolium-based ionic liquids in ECODS can dissolve DBT-type sulfides and stabilize H₂O₂ to form peroxo intermediates, achieving coupled “reaction–extraction”. Choline/acetate-based DESs strongly interact with sulfone products to enable efficient extraction, reducing their solubility in diesel. Substrates migrate into the polar extractant phase for oxidation, where oxidation products remain stable without back-diffusion into the oil phase, enabling ultra-deep desulfurization.

Fig. 17 Schematic diagram of the mechanism of ECODS, reproduced from ref. 291 with permission from American Chemical Society,²⁹¹ copyright 2015.

Recently, catalyst design and performance tuning have become central to ODS research, with main catalyst categories including transition metal catalysts, heteropoly acid catalysts, metal–organic frameworks (MOFs), functional ionic liquid catalysts, carbon-based catalysts, and other emerging functional materials. The following sections detail the advances in each catalyst type.

6.4. Commonly used catalysts

6.4.1. Transition metal oxide catalysts. Transition metal oxides have attracted extensive attention in ODS systems due to their ability to efficiently activate oxidants and selectively oxidize aromatic sulfur compounds under mild conditions, offering advantages of green, efficient, and economical catalysis. These catalysts utilize the variable valence metal centers on their surface to activate oxidants such as H₂O₂, TBHP, and O₂, generating highly oxidative active species (e.g., peroxo complexes, hydroxyl radicals, singlet oxygen) that oxidize refractory aromatic sulfides into more polar sulfoxides and sulfones, facilitating their removal from fuels. Common transition metal oxides include single oxides, mixed oxides, or supported composites composed of Mo, W, V, Fe, Cu, Co, Mn, etc.

Molybdenum-based catalysts are widely studied due to their superior ability to activate H₂O₂, forming highly active Mo–peroxo complexes such as [MoO(O₂)₂]²⁻, which nucleophilically attack the sulfur atoms in DBT and other aromatic sulfides, converting them into sulfoxides or sulfones.²⁹⁶ The catalytic cycle involves the redox transition of Mo between Mo⁶⁺ and Mo⁵⁺/Mo⁴⁺ states. Vanadium oxides function similarly, with the V⁵⁺/V⁴⁺ redox pair facilitating efficient electron transfer and activation of O₂ or H₂O₂ to generate ˙OH, exhibiting excellent ODS performance.²⁹⁷ Iron-based catalysts are valued for their abundance, low cost, and low toxicity;²⁹⁸ Fe³⁺ species can induce Fenton-like reactions with H₂O₂ to produce ˙OH radicals that oxidize aromatic sulfides to sulfones, enabling ultra-deep desulfurization.²⁹⁹ Tuning the coordination environment and dispersion of Fe³⁺ significantly enhances catalytic oxidation efficiency and suppresses side reactions. Copper, cobalt, and manganese oxides typically participate in electron transfer via multivalent states, with catalytic activity closely linked to surface oxygen vacancies and crystal lattice distortions.³⁰⁰ Synergistic effects can be achieved by doping or combining multiple metal oxides; e.g., Co–Mo bimetallic oxides exhibit enhanced electronic structure modulation and oxidant activation efficiency.³⁰¹

Chen et al.³⁰² prepared CoFeMo mixed metal oxides (MMOs) by calcining cobalt–iron layered double hydroxides intercalated with molybdate ions, yielding catalysts with optimized electronic structures and active site distributions through precise MoO₄²⁻ loading (Fig. 18a). The catalyst achieved up to 98.7% DBT oxidation at 100 °C under mild conditions, selectively forming sulfones. It exhibited excellent stability, retaining over 95% activity after multiple cycles. Mechanistic studies highlighted Co–Mo synergistic active centers on the catalyst surface as key to performance. This work offers new strategies for designing high-efficiency oxidation catalysts and elucidates structure–performance relationships.

Fig. 18 Metal-based catalysts for ODS. (a) Synthesis route to CoFeMo–MMO catalysts, reproduced from ref. 302 with permission from Elsevier,³⁰² copyright 2021. (b) Evaluation of catalytic performances the control samples in the ACODS of DBT, reproduced from ref. 304 with permission from Elsevier,³⁰⁴ copyright 2023. (c) Magnetic mesoporous silica-supported tungsten oxide catalysts and corresponding TEM image, reproduced from ref. 305 with permission from Springer Nature,³⁰⁵ copyright 2020. (d) High-entropy metal oxide catalyst for ACODS and magnetic separation, reproduced from ref. 306 with permission from Elsevier,³⁰⁶ copyright 2023. Reaction conditions of 18b: m_catal = 5 mg, V_oil = 30 mL, T = 110 °C, and t = 4 h.

To enhance activity and stability, metal oxides are often supported on high surface area supports such as mesoporous silicas (MCM-41, SBA-15), metal oxides, boron nitride, and carbon materials.^274,303 Supported structures improve metal dispersion, optimize pore architectures, enhance oil–catalyst interfacial mass transfer, and prevent aggregation or leaching in liquid phase. Surface acidity and hydrophobicity/hydrophilicity critically affect aromatic sulfur compound enrichment and oxidation. Novel hybrid catalysts combining transition metal oxides with carbon materials show promising performance due to excellent electronic conductivity and structural stability. Magnetic oxides, such as Fe₃O₄@MoO_x, Fe₃O₄@V₂O₅, not only offer high catalytic activity but also facilitate magnetic recovery, aligning with green chemistry principles and industrial application prospects.

Chen et al.³⁰⁴ developed a novel spinel porous metal oxide nanosheet catalyst loaded with MoO_x nanoclusters for petroleum fraction ACODS. The metal oxide support effectively modulated the electronic structure of MoO_x clusters, identifying Mo(V) species as active centers. Under atmospheric air at 90 °C, MoO_x/CuCo₂O₄ catalyzed DBT oxidation with a turnover frequency of 16.2 h⁻¹, maintaining stable performance and chemical structure over six reuse cycles, providing guidance for designing highly active and stable nanocluster catalysts (Fig. 18b).

Our group³⁰⁵ reported a magnetic mesoporous silica-supported tungsten oxide catalyst achieving deep diesel desulfurization with molecular oxygen (Fig. 18c). Calcination temperature tuning showed that catalysts calcined at 500 °C exhibited optimal activity. Under 120 °C and 8-hour reaction, the catalyst reached 99.9% DBT removal and 98.2% and 92.3% removal of 4-MDBT and 4,6-DMDBT, respectively. Quenching experiments and FT-IR analysis elucidated the oxidation mechanism. The catalyst's magnetic properties enabled facile recovery by magnetic separation, offering new avenues for efficient, recyclable oxidative desulfurization technologies.

More recently, our group developed high entropy metal oxides, which contain five or more metal elements, for ODS using oxygen as the oxidant. Compared with traditional metal oxide catalysts, high entropy metal oxides possess lattice distortion, high entropy effect, cocktail effect, etc., which favors the improvement of catalytic performance. Zhu et al.³⁰⁶ designed and developed a magnetically separable and recyclable high-entropy metal oxide catalyst (HEMO-900), and systematically evaluated its performance in oxidative desulfurization of fuels using molecular oxygen as the oxidant (Fig. 18d). The catalyst was synthesized via a mechanochemically assisted calcination method and characterized using various analytical techniques. The results revealed that the high-entropy structure effectively modulates the charge distribution of active components, thereby enhancing O₂ activation capability. Under optimized conditions, HEMO-900 achieved a sulfur removal efficiency of up to 96.9%, and was capable of efficiently oxidizing and removing a broad range of aromatic sulfur compounds, enabling deep desulfurization of various fuel samples. More importantly, the catalyst exhibited excellent magnetic recoverability, allowing for rapid separation from the fuel phase under an external magnetic field. Even after five consecutive cycles, HEMO-900 maintained a high desulfurization efficiency of 91.1%.

To briefly summarize, metal oxide catalysts have demonstrated high efficiency in activating oxidants like H₂O₂ or O₂ under mild conditions to convert refractory sulfur compounds in diesel into more polar and easily separable sulfones or sulfoxides or sulfones. Compared with conventional HDS, the ODS process does not require high-pressure hydrogen, offers lower energy consumption, and is particularly effective for removing aromatic sulfur compounds such as 4,6-dimethyldibenzothiophene, which are resistant to HDS. Moreover, some mixed metal oxides exhibit high stability and can be regenerated via calcination or washing, thus extending catalyst lifetime. However, catalyst deactivation remains a major challenge due to blockage of active sites by sulfur oxidation products. The relatively high cost of oxidants like H₂O₂ and the need for polar solvents to extract the resulting sulfones further increase process complexity.

Future improvements should focus on catalyst structure optimization and the development of greener processes. Strategies include the construction of low-cost and highly active multi-metal composite oxides and the use of porous supports to enhance catalyst dispersion and mass transfer. Nanostructure engineering, such as tailoring exposed high-energy crystal facets of TiO₂, and bio-inspired catalysis (e.g., metal porphyrin/O₂ systems) can further improve catalytic activity and selectivity. Incorporating photocatalysis or electrocatalysis may enable the use of solar or electric energy to drive the reaction, reducing reliance on chemical oxidants. From a process integration perspective, catalytic membrane reactors offer a promising route to couple oxidation and separation in a single step. Machine learning-assisted catalyst design is also emerging as a powerful tool to accelerate the screening and optimization of catalyst materials. Furthermore, the development of low-temperature regeneration techniques and self-cleaning catalyst surfaces could mitigate deactivation issues, thereby improving long-term stability and sustainability of the ODS process.

6.4.2. Heteropoly acid catalysts (HPAs). Heteropoly acids (HPAs) are well-defined multinuclear metal–oxo clusters exhibiting excellent redox activity and strong acidity, making them promising catalysts for ODS reactions.^307,308 Compared with traditional transition metal oxides, HPAs can efficiently activate oxidants such as H₂O₂, TBHP, and O₃ under mild conditions, possess stable molecular structures, multifunctional catalytic centers, and are suitable for deep removal of sterically hindered aromatic sulfides like DBT and 4,6-DMDBT from diesel and jet fuels. Common HPAs include Keggin-, Dawson-, and Anderson-type structures, with Keggin-type HPAs such as phosphotungstic acid (H₃PW₁₂O₄₀), phosphomolybdic acid (H₃PMo₁₂O₄₀), and silicotungstic acid (H₄SiW₁₂O₄₀) widely applied due to facile synthesis, controllable oxidation states, and high selectivity.^309,310 Their cage-like structure consists of central heteroatoms (P⁵⁺, Si⁴⁺) linked via oxygen bridges to twelve MO₆ (M = Mo⁶⁺ or W⁶⁺) octahedra, providing stable electronic frameworks and superior redox properties. In ODS, HPAs react with H₂O₂ to form peroxo intermediates (e.g., [PMo(O₂)₄]³⁻, [PW(O₂)₄]³⁻), which nucleophilically attack sulfur atoms in aromatic sulfides at oil/water interfaces or organic phases, inducing two-electron oxidation to sulfoxides or sulfones at ambient conditions with minimal side reactions, favorable for green and low-energy desulfurization.

Li et al.³¹¹ synthesized seven surface-active heteropoly acid (HPA) catalysts and systematically evaluated their catalytic performance for the ODS under various conditions. Among them, hexadecylmethylimidazolium phosphotungstate ([C₁₆MIM]₃PWO) was identified as one of the most effective catalysts, achieving efficient oxidation of DBT using only H₂O₂ as the oxidant. The study revealed that an ideal amphiphilic catalyst cation should contain at least one long alkyl chain (C₁₂–C₁₈) to enhance hydrophobicity and oil-phase dispersibility (Fig. 19a). This structural feature enables the catalyst to form a stable water-in-oil (W/O) emulsion, even in the presence of excess water, thereby promoting the interfacial catalytic efficiency. Under optimized conditions, DBT was completely oxidized within 40 minutes.

Fig. 19 ODS of diesel using different heteropoly acid catalysts. (a) Lipophilicity of amphiphilic phosphotungstates in catalytic oxidative desulfurization of oil by H₂O₂, reproduced from ref. 311 with permission from Elsevier,³¹¹ copyright 2019. (b) Proposed reaction mechanism involving V-peroxo species, reproduced from ref. 312 with permission from Springer Nature,³¹² copyright 2013. (c) Oxidative desulfurization mechanism of oxygen-deficient phosphomolybdate-vanadium catalyst, reproduced from ref. 313 with permission from Elsevier,³¹³ copyright 2025. (d) The proposed mechanism for dioxygen activation and aerobic ODS by [(C₁₈H₃₇)₂N(CH₃)₂]₅[IMo₆O₂₄], reproduced from ref. 314 with permission from Royal Society of Chemistry,³¹⁴ copyright 2010. (e) Two possibilities for C–S bond activation in benzothiophene, reproduced from ref. 315 with permission from American Chemical Society,³¹⁵ copyright 2017. (f) IL-modified heteropolyacids for ODS, reproduced from ref. 310 with permission from American Chemical Society,³¹⁰ copyright 2018.

Hu et al.³¹² designed and synthesized three Keggin-type hydrophobic heteropoly acid catalysts by reacting phosphomolybdovanadic acid with imidazolium bromide salts bearing alkyl chains of varying lengths. These catalysts exhibited excellent catalytic activity in an ECODS system employing H₂O₂ as the oxidant and acetonitrile as the phase transfer agent. The desulfurization efficiency increased with the length of the alkyl chain in the imidazolium cation. Under optimized conditions, nearly complete sulfur removal was achieved. Mechanistically, the Keggin-type anion abstracts active oxygen from H₂O₂ to form a peroxo intermediate, which subsequently oxidizes DBT into the corresponding sulfone products (Fig. 19b). This study highlights that tuning the alkyl chain length of the imidazolium cation can significantly enhance the catalyst's hydrophobicity and dispersibility in the oil phase, thereby promoting efficient interfacial oxygen transfer and accelerating the oxidative conversion of sulfur compounds.

Zhu et al.³¹³ successfully synthesized an oxygen-vacancy-rich phosphomolybdovanadic heteropoly acid catalyst (P₆MoV) using phytic acid, a phosphorus-rich organic molecule, as the structural building block (Fig. 19c). This catalyst exhibited exceptional ODS performance toward aromatic sulfur compounds in diesel. Owing to the engineered oxygen vacancies, the catalyst achieved a remarkable sulfur removal efficiency of up to 99.55% even under mild conditions as low as 0 °C, demonstrating its outstanding oxidation capability against refractory sulfur species. More importantly, the P₆MoV catalyst maintained high desulfurization efficiency after 20 cycles. The study revealed that the introduction of oxygen vacancies not only modulated the catalyst's electronic structure and enhanced the generation of reactive oxygen species, but also improved the adsorption and activation of sulfur-containing molecules on the catalyst surface.

Lü et al.³¹⁴ reported that the Anderson-type heteropoly acid catalyst [(C₁₈H₃₇)₂N(CH₃)₂]₅IMo₆O₂₄ could efficiently catalyze the oxidation of benzothiophene (BT), DBT, and 4,6-dimethyldibenzothiophene (4,6-DMDBT) into the corresponding sulfone compounds under mild conditions using molecular oxygen as the sole oxidant, without requiring any sacrificial agent (Fig. 19d). This demonstrates its remarkable capacity for deep oxidation of refractory aromatic sulfur compounds.

However, HPAs’ water solubility leads to leaching in polar extractants, hampering recovery and reuse. To address this, supported catalysts and structural modifications have been developed to enhance stability, hydrophobicity, and interfacial interactions. Supports include mesoporous silica, metal oxides, and carbon materials, with immobilization via electrostatic adsorption, covalent bonding, or in situ encapsulation improving structural stability and active site exposure. For instance, PW₁₂/SBA-15 catalysts enhance H₂O₂ activation and suppress HPA dissolution, improving recyclability.³¹⁶ Functional ionic liquids are also explored to regulate HPA interfacial properties via hydrogen bonding or ionic interactions, enhancing dispersion and catalytic efficiency in organic phases.

Song et al.³¹⁷ synthesized a series of efficient HPW(x)/MIL-100(Fe) catalysts by hydrothermal methods, maintaining MIL-100(Fe) structural features with high surface area promoting phosphotungstic acid dispersion. Optimizing HPW loading, catalyst dosage, temperature, and O/S molar ratio yielded complete removal (100%) of BT, DBT, and 4,6-DMDBT within 90 minutes at 50 °C. The high activity results from highly dispersed HPW active sites on the high surface area support.

Besides engineering supported HPAs or dispersing HPAs in ILs to form ECODS systems, heteroatom substitution of HPAs is another efficient strategy to promote the catalytic performance of HPAs. For instance, Albert et al.³¹⁵ proposed a green oxidative desulfurization strategy employing molecular oxygen as the oxidant and H₈PV₅Mo₇O₄₀ (HPA-5) as a water-phase catalyst system, which significantly enhanced the conversion of sulfur compounds. In this process, approximately 60–70% of the organic sulfur species in fuel were oxidized to sulfates, accompanied by the formation of water-soluble byproducts such as sulfoacetic acid (SAA, 10–20%), 2-sulfobenzoic acid (2-SBA), and 2-(sulfinyloxy)benzoic acid (2-SOBA) (Fig. 19e). The method exhibited excellent performance for the removal of BT from isooctane, achieving up to 99% desulfurization efficiency under optimized conditions. Furthermore, when applied to commercial fuel oil with a sulfur content of 973 ppmw, the desulfurization efficiency reached 28%. This study offers a novel approach for converting sulfur compounds in oxidative desulfurization systems into water-soluble products.

Zhao et al.³¹⁰ synthesized three ionic liquid-modified heteropoly acid catalysts by employing quaternary ammonium cations with different nitrogen configurations: single N atom, double N atoms, and cyclic N structure, designated as S-POM, D-POM, and C-POM, respectively (Fig. 19f). Based on this, a hybrid catalyst was constructed by self-assembly using MOF-199 as a template, followed by a one-step hydrothermal process to co-immobilize the modified heteropoly acids within the mesopores of MCM-41. In ODS reactions, the cyclic quaternary ammonium-modified catalyst exhibited 100% removal efficiency of DBT within 90 minutes. DFT calculations further revealed the competitive adsorption–oxidation behaviors among sulfur compounds such as BT and 4,6-DMDBT, indicating a significant structure-dependent catalytic activity. This work demonstrates a highly efficient and stable catalytic strategy through ionic liquid modification, MOF mediation, and heteropoly acid immobilization, providing both a new pathway and theoretical foundation for the development of practical oxidative desulfurization technologies for diesel fuels.

To briefly summarize, HPAs are known for their strong Brønsted acidity, high redox potential, and tunable structures, enabling them to efficiently catalyze the oxidation of sulfur-containing compounds in diesel under mild conditions, converting them into sulfones or sulfoxides that can be more readily removed by subsequent separation processes. Supported HPAs, in which active components are dispersed onto solid carriers, not only exhibit increased specific surface area and thermal stability but also reduce leaching of active species and enhance catalyst recyclability. Moreover, these catalysts are generally environmentally benign, operate under mild reaction conditions, and require lower energy input. The catalytic performance can be further optimized by tuning the HPA composition and selecting appropriate supports.

However, several limitations hinder their broader application. HPAs tend to dissolve in polar media, making catalyst recovery challenging in liquid-phase systems. Additionally, some HPAs exhibit poor thermal or chemical stability under high temperatures or in strongly polar solvents, leading to decomposition or deactivation. The synthesis of supported HPA catalysts is often complex, and the choice of support material and loading method significantly impacts performance. Uneven dispersion may lead to reduced availability of active sites. Furthermore, the oxidation products, sulfones, typically require an additional separation step, such as extraction or adsorption, increasing process complexity and cost. Industrial-scale application remains challenging due to concerns over catalyst lifetime, regeneration efficiency, and overall economic feasibility.

Future research should aim to develop novel heteropoly acid derivatives or composite materials—such as transition metal-substituted or ionic liquid-modified HPAs—to enhance catalytic activity and stability. At the same time, optimization of support materials, particularly porous carriers and mesoporous molecular sieves, is expected to improve mass transfer efficiency and dispersion of active species. In-depth mechanistic studies, aided by in situ characterization techniques, will be essential to clarify the roles played by active sites and establish structure–activity relationships. Moreover, integrating HPAs with other desulfurization technologies may enable low-temperature and high-efficiency sulfur removal through synergistic effects. To facilitate industrial application, future work must also address practical challenges including scalable catalyst synthesis, regeneration processes, and environmentally responsible handling of spent catalysts.

6.4.3. Metal–organic frameworks (MOFs). MOFs are porous crystalline materials constructed from metal ions/clusters and organic ligands via coordination bonds, characterized by high specific surface areas, tunable pore sizes, designable functionalities, and ordered structures.³¹⁸ Since 2000, MOFs have been extensively investigated in heterogeneous catalysis, gas adsorption/separation, and drug delivery. Recently, the urgent demand for green, efficient, and mild diesel deep desulfurization has spurred interest in MOFs for ODS due to their structural tunability and catalytic performance. MOF-based ODS research mainly focuses on the role they play as catalysts or catalyst supports. MOFs’ metal nodes often exhibit Lewis acidity, synergistically activating oxidants to generate reactive oxygen species for selective oxidation of aromatic sulfides. For example, Zr-based MOFs such as UiO-66, UiO-67, and MOF-808 feature robust frameworks and abundant Zr–OH Lewis acid sites, efficiently activating H₂O₂ for oxidative desulfurization.³¹⁹

To enhance catalytic performance, functional ligands or post-synthetic modifications have been introduced to improve oxidation ability and hydrophobicity. Incorporating nitrogen-containing heterocycles (pyridine, imidazole), aldehyde, or hydroxyl groups strengthens π–π or hydrogen bonding with sulfur compounds, enriching reactants and facilitating oxidation. Functionalization also adjusts MOF affinity toward organic solvents, optimizing oil–catalyst interfacial reaction kinetics.^324,325

Ti-based MOFs are commonly seen in ODS of fuel oils, because of the center metal Ti can effectively activate oxygen to form active intermediates. Li et al.³²⁶ synthesized a titanium-based MOF material, MIL-125, and its amino-functionalized derivative, NH₂-MIL-125, via a solvothermal method for application in an ECODS system using H₂O₂ as the oxidant. The study found that the –NH₂ functional groups partially hindered the interaction between sulfur compounds and the active Ti sites, thereby reducing the catalytic activity. Furthermore, it was confirmed that both the quantity and accessibility of Ti active centers are critical factors determining the ODS performance of the catalyst. Feng et al.³²⁰ addressed the limited redox catalytic activity of Ti-based metal–organic frameworks (Ti-MOFs) caused by coordination saturation by proposing a plasma post-treatment strategy (Fig. 20a). By tuning the treatment duration and intensity of argon plasma, they achieved controllable defect engineering. The resulting defective MUV-10 material exhibited a ligand deficiency rate of nearly 50% and demonstrated significantly enhanced catalytic performance in oxidative desulfurization. Song et al.³²¹ developed a novel amorphous titanium-based terephthalate catalyst (Ti-IPA) featuring a hierarchical pore structure and highly dispersed Ti–OH defect sites (Fig. 20b). The Ti-IPA catalyst is compatible with various oxidant systems and exhibits broad adaptability for oxidative desulfurization. Under optimized conditions, it achieved rapid and efficient oxidation of 1000 ppm thiophenic sulfur compounds to their corresponding sulfones within 12 minutes at 40 °C, with a sulfur conversion rate of up to 99.5% at an oxidant-to-sulfur molar ratio of only 2. The terminal hydroxyl groups on Ti defect sites can interact with oxidants to form peroxo-titanium intermediates, which are identified as the key species driving efficient sulfur oxidation. To address the limited diversity of Ti⁴⁺-based MOFs, the synthesis of COK-47 has filled a critical gap in this field. COK-47 represents the first titanium carboxylate MOF constructed from TiIVO₆ octahedral layer structures, whose framework can be modulated and synthesized using various organic linkers.³²² As a nanoparticulate material inherently containing structural defects, COK-47 exhibits excellent catalytic activity for the oxidation of thiophenic sulfur compounds (Fig. 20c).

Fig. 20 MOFs for ODS of diesel. (a) Synthesis procedure of MUV-10 and the defect generation using plasma treatment, reproduced from ref. 320 with permission from Elsevier,³²⁰ copyright 2024. (b) The preparation process of Ti-IPA with a hierarchical pore structure and highly dispersed Ti–OH defect sites, reproduced from ref. 321 with permission from Elsevier,³²¹ copyright 2025. (c) COK-47, a pristine or inherently defective nanoparticulate material as a highly efficient catalyst for the oxidation of thiophenes, reproduced from ref. 322 with permission from John Wiley and Sons,³²² copyright 2019. (d) Proposed mechanism of photocatalytic oxidation of DBT on Ce/MIL-125-NH₂, reproduced from ref. 323 with permission from Elsevier,³²³ copyright 2023.

Besides thermal catalytic ODS, MOFs also have been employed in photocatalytic ODS. However, MOFs have shown only relatively low photocatalytic ODS performance, and modifications are required to improve their photocatalytic ODS activities. Wang et al.³²³ synthesized a series of Ce-doped MIL-125-NH₂ photocatalysts via a simple in situ doping method, exhibiting excellent DBT desulfurization performance under mild conditions. Among them, 1.0 mol% Ce/MIL-125-NH₂ achieved 100% sulfur removal within 22 minutes under visible light irradiation at 30 °C (Fig. 20d). The outstanding performance is mainly attributed to the formation of Ce–Ti–oxo clusters, which enhanced electronic coupling and thus promoted the separation and transfer of photogenerated charge carriers. Meanwhile, the reversible redox transitions between Ce⁴⁺/Ce³⁺ and Ti⁴⁺/Ti³⁺ not only exposed abundant Lewis acid sites but also improved the generation efficiency of reactive species during the reaction.

In addition, Zr-based MOFs are another important catalyst for catalytic ODS of diesel oil, since H₂O₂ can be readily activated by Zr clusters to form active species, while MOFs can adsorb aromatic sulfides to form a high-concentration microenvironment, favoring the promotion of ODS performance. Zhang et al.³²⁷ employed the metal–organic framework NU-1000 as a catalyst for ODS of sulfur compounds in model diesel. Under optimized conditions, DBT was completely converted within 180 minutes. NU-1000 exhibited excellent stability during the reaction, with its catalytic activity remaining nearly unchanged after four consecutive cycles.

In addition, the morphology also plays an important role in tuning the catalytic ODS performance of MOFs. Zhang et al.³²⁸ further synthesized a hierarchical porous zirconium-based metal–organic framework, HP-UiO-66, for ODS of fuel oils. Under ambient conditions, the material achieved complete removal of DBT, meeting the requirements for deep desulfurization. Remarkably, even for the refractory sulfur compound 4,6-DMDBT, complete oxidation and efficient extraction were accomplished within 80 minutes. Comparative performance evaluations showed that HP-UiO-66 outperformed various other MOF-based catalysts, which was primarily attributed to its larger pore size and higher porosity. Mechanistic investigations revealed that the oxidation reaction catalyzed by HP-UiO-66 followed a radical pathway dominated by hydroxyl radicals. Lin et al.³²⁹ systematically evaluated the catalytic performance of four representative Zr-based MOFs, namely UiO-66, UiO-67, NU-1000, and MOF-808, in an ODS system. Among them, MOF-808 exhibited the best performance, achieving complete removal of DBT from a 1000 ppm sulfur model oil within just 5 minutes at 40 °C. This outstanding activity was primarily attributed to the high concentration of accessible Zr–OH(H₂O) active sites in MOF-808, which effectively promoted the decomposition of H₂O₂ to generate both ˙O₂⁻ and ˙OH radicals involved in the oxidation of DBT. Additionally, the synergistic effect of Brønsted and Lewis acidity was also confirmed to positively contribute to the catalytic performance.

Introducing more abundant defects in MOFs favors the exposures of metal centers, which benefits the activation of oxidants for an enhanced ODS performance. De Vos et al.³³⁰ employed a post-synthetic modification strategy on MOF-808 and UMCM-309 to remove coordinated formate ligands, thereby generating more accessible Zr⁴⁺ open metal sites and significantly enhancing catalytic activity. The resulting modified materials, MOF-808-M, UMCM-309-M1, and UMCM-309-M2, possessed a greater number of exposed active centers. Even at low catalyst loading, MOF-808-M exhibited high oxidative activity toward sulfur compounds, along with excellent selectivity and good reusability. Sun et al.³³¹ treated MOF-808 with ammonium bicarbonate (NH₄HCO₃) solution to selectively break partial Zr–O bonds, thereby introducing abundant structural defects into the framework. The study systematically investigated the effects of NH₄HCO₃ concentration and treatment duration on the structural characteristics of MOF-808. The resulting defect-engineered MOF-808 exhibited significantly enhanced catalytic activity in the oxidative desulfurization of DBT compared with the untreated sample. This performance improvement was mainly attributed to the formation of numerous defect sites induced by NH₄HCO₃ treatment, which provided more active centers for H₂O₂ activation and reactant adsorption.

Moreover, engineering bimetallic MOFs is also an efficient strategy to improve the ODS performance. Wang et al.³³² first reported the synthesis of bimetallic Ce/Zr MOF-801 under ambient aqueous conditions, along with an acid post-treatment strategy. Through defect engineering, Ce species were preferentially exposed within the stable Zr-MOF-801 framework, significantly enhancing the ODS performance of this host–guest catalyst. The Ce/Zr-MOF-801-H catalyst achieved complete removal of DBT with an initial sulfur content of 1000 ppm. The exposed Ce sites played a dual role: (1) activating the substrate by decreasing the electron density of the sulfur atom; (2) facilitating the generation of hydroxyl and superoxide radicals from H₂O₂, thereby driving the ODS reaction.

MOFs also serve as supports for heteropoly acids, transition metal oxides, or organic oxidants, constructing supported catalytic systems that combine high oxidative activity with efficient reactant enrichment and active site distribution.^309,333 For instance, H₃PMo₁₂O₄₀@MIL-101 composites confine HPAs within MOF pores, enhancing stability and preventing leaching for repeated use without activity loss.³³⁴ Single-atom metal or nanoparticle-doped MOFs demonstrate excellent electronic modulation, facilitating oxidant charge polarization and O–O bond activation, increasing active oxygen species generation.³³⁵

As mentioned above, HPAs are generally active species in ODS, and they are often dispersed or confined in MOFs to construct highly active catalysts. Wang et al.³³⁶ developed a novel hybrid material by integrating phosphomolybdic acid (PMA) with a Zr-based metal–organic framework (Zr-MOF), forming the PMA/UiO-66 catalyst. In this system, photogenerated electrons from the conduction band of the Zr-MOF are transferred to PMA, inducing the formation of Mo⁵⁺ species with high electron density. These newly generated Mo⁵⁺ sites exhibit excellent intrinsic catalytic activity in ODS reactions. Radical scavenging experiments and DFT calculations further confirmed that Mo⁵⁺ facilitates the formation of reactive oxygen species such as ˙OH, thereby enhancing the ODS performance. Haruna et al.³³⁷ prepared PW₁₂@MOF-808 and PW₁₁@MOF-808 composite catalysts via a facile encapsulation strategy, in which POMs were successfully and uniformly dispersed within the MOF-808 framework. This homogeneous incorporation of POMs significantly enhanced the catalytic performance of the resulting materials. Zhao et al.³³⁸ designed and synthesized a novel heterogeneous catalyst via a one-pot method—composite catalysts composed of POMs supported on carbon nanotube-modified MOF materials (CNTs@MOF-199-POM). Under optimized conditions, the CNTs@MOF-199-Mo₁₆V₂ catalyst achieved a DBT removal efficiency as high as 98.30%, demonstrating excellent catalytic activity and reusability. Cao et al.³³⁹ employed the structurally tunable and water-stable MOF-808X as a support to immobilize phosphotungstic acid (PTA) via a simple encapsulation strategy, achieving a synergistic regulation of window aperture and PTA loading. As a result, complete removal of DBT with an initial sulfur concentration of 1000 ppm in model fuel was accomplished within just 30 minutes, significantly surpassing standard fuel sulfur limits. The remarkable catalytic activity is primarily attributed to the synergistic effect between the metal clusters in the MOF framework and the encapsulated PTA species. This study not only offers a green, stable, and efficient catalyst for ODS but also provides a new approach and reference paradigm for the sustainable development of ultra-deep desulfurization technologies.

Moreover, the HPA-based ionic liquids also have been employed as catalytically active sites, which can combine the advantages of ionic liquids and MOFs. Qiu et al.³⁴⁰ reported a novel approach for constructing efficient heterogeneous catalysts by integrating MOFs with POM-based ionic liquids containing carboxyl-functionalized anions, specifically [mim(CH₂)₃COOH]₃PW, for application in fuel oxidative desulfurization. Using the carboxyl-functionalized ionic liquid as a molecular bridge, a POM-based MOF composite material, [mim(CH₂)₃COO]₃PW@UiO-66, was synthesized in situ. The resulting catalyst exhibited outstanding ODS performance, achieving 100% removal of DBT within 60 minutes. This superior catalytic activity is primarily attributed to the synergistic effect between W=O groups and Lewis acid sites, where the latter promotes the decomposition of H₂O₂ to form peroxotungstate species (W(O₂)_n), thereby significantly enhancing catalytic efficiency. Ye et al.³⁴¹ reported function of MOFs by ILs, significantly improve the stability of ionic liquids. They proposed a novel strategy based on post-synthetic ligand exchange (PSLE) to achieve the stable anchoring of monocarboxylic acid-functionalized ionic liquids within the MOF framework, successfully constructing a new type of Fenton-like catalyst, [mim(CH₂)₃COO]FeCl₄@UiO-66. This catalyst exhibited excellent oxidative desulfurization (ODS) performance toward DBT in a biphasic system using H₂O₂ as the oxidant, achieving a DBT removal efficiency of up to 99.1% under optimized conditions.

Besides HPAs, transition metals or transition metal oxides are also commonly employed active species. Similar to HPAs, the specific surface areas of metals or metal oxides are limited, making the exposure of catalytically active sites limited. Thus, dispersing or confining them on high specific area MOFs favors the dispersing and exposure of catalytically active sites. Li et al. utilized tungsten-modified MoO₃ as the active component to enhance desulfurization activity, while employing a MOF material to provide a stable support structure. Additionally, magnetic Fe₃O₄ was successfully incorporated to enable rapid magnetic separation and catalyst recyclability. The resulting composite catalyst, Fe₃O₄@W-MoO₃@MOF, achieved complete oxidation and removal of DBT within 60 minutes, maintaining stable catalytic performance after at least 18 consecutive reuse cycles. Jhung et al.³⁴² prepared a uniformly dispersed tungsten trioxide (WO₃) catalyst immobilized on the NU-1000, wherein the defect sites of NU-1000 facilitate the firm anchoring of WO₃ species within its pores. This material exhibited excellent performance in the oxidation of DBT using H₂O₂ as the oxidant, achieving near-complete conversion of 1000 ppm DBT within 120 minutes at room temperature, with an activation energy as low as 25.9 kJ mol⁻¹. The outstanding catalytic activity is attributed to the well-dispersed WO₃ species that efficiently activate H₂O₂ and the material's high porosity. Cunha-Silva et al.³⁴³ reported for the first time the application of aluminum-based 2-aminoterephthalate MOFs in the preparation of efficient heterogeneous catalysts for desulfurization processes. Sandwich-type polyoxometalates [Eu(PW₁₁O₃₉)₂]¹¹⁻ (POM) were supported on Al(III)- and Cr(III)-based MIL-type MOF carriers, NH₂-MIL-53(Al) and MIL-101(Cr), exhibiting excellent catalytic activity and good stability in ODS reactions.

Moreover, because of the high carbon contents in MOFs as well as their high specific surface areas, MOFs also have been employed as precursors to engineer carbon-supported catalysts, inducing remarkable ODS performances. Our group²⁸⁷ developed a MoO₂@CNT composite catalyst by employing a urea-assisted strategy to in situ anchor nanoscale MoO₂ particles onto the surface of carbon nanotubes (CNTs), using a molybdenum-based metal–organic framework as the precursor. Powder X-ray diffraction analysis confirmed that the particle size of MoO₂ and the nitrogen content in the catalyst could be effectively tuned by adjusting the amount of urea and the calcination temperature. The resulting catalyst exhibited excellent oxidative desulfurization performance, achieving over 99% removal efficiency for various refractory sulfur compounds, including DBT, 4-MDBT, and 4,6-DMDBT. Jhung et al.³⁴⁴ developed a novel ODS catalytic system based on a Ti/Zn bimetallic MOF composite, denoted as ZIF(30)@H₂N-MIL-125 (abbreviated as ZIF(30)@MOF). Through pyrolysis, they synthesized a mesoporous carbon material, MDC-C, which was compared with MDC-P, being derived from pyrolyzed H₂N-MIL-125 alone. MDC-C exhibited significantly enhanced physicochemical properties, including higher specific surface area, larger pore size, and a richer mesoporous architecture. Moreover, it featured more uniform and finer TiO₂ nanoparticle dispersion. These structural advantages led to superior catalytic performance in the oxidative desulfurization of DBT. Notably, despite having a lower TiO₂ content than MDC-P, MDC-C achieved a catalytic rate approximately three times higher and demonstrated the lowest activation energy reported to date, enabling efficient DBT conversion even at very low catalyst dosage.

MOF materials exhibit great potential in ODS of diesel fuel due to their high specific surface area, tunable pore structure, and abundant active sites, offering both adsorption and catalytic functionalities. However, their relatively poor hydrothermal stability and insufficient exposure of active sites limit their practical applicability. To address these issues, MOF-supported catalysts have been developed by incorporating active species, such as metal oxides or polyoxometalates, into MOF pores, effectively combining the adsorption capacity of the MOF host with the catalytic activity of the guest components, thereby enhancing desulfurization efficiency. Nonetheless, such loading processes may lead to pore blockage or agglomeration of active species, compromising long-term stability. Alternatively, MOF-derived catalysts, including carbon-based materials and metal oxides obtained via pyrolysis or chemical transformation, typically exhibit improved structural stability and conductivity. Yet, some derivatives suffer from the loss of the parent MOF's ordered porosity, which negatively affects the diffusion of sulfur compounds.

Future research should focus on the design of MOF-based materials that simultaneously exhibit high structural stability and abundant active sites. This may include strategies such as defect engineering or the construction of bimetallic MOFs to modulate electronic structures. Optimizing loading techniques to improve the dispersion of catalytic species—particularly the development of atomically dispersed catalysts—could significantly enhance catalytic efficiency. Additionally, developing low-temperature synthetic routes for MOF derivatives that preserve hierarchical porosity is critical. The integration of machine learning approaches for predicting optimal catalyst compositions, as well as the exploration of photo- or electro-catalytic synergies, may offer promising routes to reduce energy consumption. Moreover, understanding the effects of complex components in real diesel fuels—such as olefins and nitrogen-containing compounds—on ODS performance will be essential for advancing MOF-based catalysts toward industrial applications.

6.4.4. Functional ionic liquids/porous ionic liquids. Functional ILs have shown great promise in ODS and emerged as key research directions recently.^345,346 ILs feature low melting points, negligible vapor pressure, good thermal stability, strong polarity, and structural designability, serving as both reaction solvents and extraction media in ODS for sulfur compound enrichment and removal.^347,348 Functional ILs introduce catalytic active sites into traditional ILs, enabling oxidant activation and simultaneously acting as catalysts and reaction media to achieve synergistic oxidation and extraction.

The catalytic activity of functional ILs primarily arises from tailored anion structures.^349,350 Commonly, ILs incorporate oxidation-active polyoxoanion systems such as phosphotungstates, phosphomolybdates, peroxotungstates, peroxomolybdates, or metal-coordinated complex anions.³⁵¹ Heteropolyacid-type anions activate diverse oxidants for ultra-deep oxidation desulfurization due to abundant lone pairs and variable valence metal centers forming peroxo heteropolyacid intermediates, which selectively attack sulfur atoms to produce polar sulfones or sulfoxides. Compared with conventional HPAs, IL-based heteropolyacid anions exhibit better dispersion and enhanced interfacial reactivity between catalytic and organic phases, significantly reducing mass transfer resistance.

Functional ILs also possess molecular recognition capabilities toward target sulfur compounds. By tuning the hydrophobicity and steric hindrance of cation side chains, they form π–π stacking and hydrophobic aggregation with DBT and other aromatic sulfides, increasing enrichment efficiency and local reaction concentration, thereby improving catalytic kinetics. This structure–function relationship renders functional ILs highly effective for diverse organic sulfides oxidation.

Fe-based ILs have been widely employed in ODS using H₂O₂ as the oxidant, since the Fe³⁺ can react with H₂O₂ to form a Fenton-like reaction, inducing the formation of ˙OH, which are regarded as catalytically active species in ODS processes. In addition, as mentioned in the EDS section, because of the Lewis acid–base interaction between aromatic sulfides and Fe-based ILs, the IL can well concentrate the aromatic sulfides, favoring the ODS process. Generally, ODS processes using Fe-based ILs can be finished in a very short time. Zhang et al.³⁵² synthesized a series of Lewis acidic ionic liquids (ILs), including dialkylpyridinium tetrachloroferrate salts such as [C₄₃MPy] and [C₈₃MPy], and employed them as both extractants and catalysts for the oxidative removal of DBT from model oil using 30 wt% H₂O₂ as the oxidant. The results demonstrated that these ILs possess excellent dual functionality in extraction and catalysis, enabling complete removal of DBT within just 1 minute under mild conditions, even with low IL dosage and at relatively low temperatures. Our group also has reported a series of Fe-based ILs for ODS. We prepared a thermosensitive magnetic ionic liquid catalyst, N-butylpyridinium tetrachloroferrate ([BPy][FeCl₄]),³⁵³ which exhibits reversible melting and solidification behavior. During the oxidative desulfurization process, [BPy][FeCl₄] forms a liquid–liquid extraction and catalytic oxidation coupled system (ECODS) above approximately 40 °C, significantly enhancing the removal efficiency of sulfur compounds. Upon completion of the reaction and subsequent cooling, the system reverts to a liquid–solid state, enabling controllable operation and efficient catalyst recovery. This catalyst achieved over 95% desulfurization of DBT in model oil within 10 minutes. In addition, we also reported a series of novel dialkylpiperidinium tetrachloroferrate catalysts, including [C₂OHmpip]FeCl₄, [C₄mpip]FeCl₄, [C₈mpip]FeCl₄, and [C₁₂mpip]FeCl₄.³⁴⁹ Under conditions of 30 °C and 60 minutes reaction time, using [C₄mpip]FeCl₄ as the catalyst and [C₈mim]BF₄ as the ionic liquid, a DBT removal efficiency of up to 97.1% was achieved. The optimal molar ratio of H₂O₂ to sulfur was as low as 3.5 : 1, indicating that this catalyst is among the most efficient reported to date. However, an inevitable disadvantage using Fe-based ILs in ODS systems is that the Fe-based ILs can catalyze self-decomposition of H₂O₂, making the utilization efficiency of oxidant relatively low.³⁵⁴ Thus, we synthesized two hexacyanoferrate-based ionic liquid catalysts, [C₄Py]₃Fe(CN)₆ and [C₁₆Py]₃Fe(CN)₆.³⁵⁵ Under optimized conditions, using [C₄Py]₃Fe(CN)₆ as the catalyst and 1-octyl-3-methylimidazolium hexafluorophosphate ([C8mim]PF₆) as the extractant, a DBT removal efficiency of up to 97.1% was achieved. The desulfurization reactivity for different sulfur compounds followed the order: DBT > 3-MBT > BT > 4-MDBT > 4,6-DMDBT. In addition, the influence of water content on the desulfurization performance was investigated by introducing various concentrations of H₂O₂ into the system. The results showed that excess water favored sulfur removal, while the catalyst exhibited lower sensitivity to moisture compared with [FeCl₄]⁻-based catalysts.

In addition, polyoxometalate-based ionic liquids (POM-ILs) represent another class of high-performance catalysts for oxidative desulfurization (ODS). Most of the reported POM-ILs are based on molybdenum (Mo) and tungsten (W), with some examples incorporating heteroatom-substituted polyoxometalates. In these systems, the catalytic active sites are primarily located on the polyoxometalate anions, while the ionic liquid cations mainly serve to enhance mass transfer. In typical ODS processes employing POM-ILs, H₂O₂ is used as the oxidant, although a few studies have explored molecular oxygen as an alternative. Due to the variable oxidation states of polyoxometalates, the oxidant first reacts with the POM to generate peroxo species, which can subsequently oxidize aromatic sulfur compounds, thereby achieving efficient desulfurization. Our group has developed a series of hybrid catalytic materials based on Keggin-type polyoxometalate (POM) ionic liquids, including [MIMPS]₃PW₁₂O₄₀·2H₂O, [Bmim]₃PW₁₂O₄₀, [Bmim]₃PMo₁₂O₄₀, and [Bmim]₄SiW₁₂O₄₀.³⁵⁸ In oxidative desulfurization reactions, these POM-ILs were employed as catalysts, with H₂O₂ as the oxidant and an ionic liquid as the reaction medium. Among them, [MIMPS]₃PW₁₂O₄₀·2H₂O exhibited the highest catalytic activity, achieving 100% removal of DBT at 30 °C within 1 hour. Furthermore, a series of surface-active polyoxometalate ionic liquids (SPILs) were successfully synthesized and applied in the ODS of model diesel using H₂O₂ as the oxidant.³⁵⁶ Experimental results demonstrated that these SPILs exhibited excellent catalytic performance for the removal of DBT under mild conditions. In the SPIL-catalyzed ODS system, the reaction follows a metal-peroxo-centered catalytic cycle. Taking molybdenum-based SPILs as an example, the catalyst initially reacts with H₂O₂ to generate highly active Mo–peroxo species (Fig. 21a and b). These species progressively transfer oxygen atoms to organic sulfur compounds such as DBT, oxidizing them to sulfoxides and subsequently sulfones, while releasing reduced Mo=O species. The Mo=O species can then be reoxidized by H₂O₂ to regenerate the active Mo–peroxo species, thus completing the catalytic cycle. Wu et al.³⁵⁷ designed and synthesized a novel polyoxometalate-based ionic liquid salt catalyst, 3-(1-methylimidazolium-3-yl)propane-1-sulfonic acid phosphomolybdotungstate ([MIMPs]₃PMo₆W₆O₄₀), and applied it in ODS reactions. In a model diesel system using DBT as the target sulfur compound, [MIMPs]₃PMo₆W₆O₄₀ exhibited outstanding catalytic performance, achieving complete DBT removal even at a relatively low oxidant-to-sulfur molar ratio (Fig. 21c). Remarkably, the catalyst demonstrated excellent reusability, maintaining its catalytic activity over 33 consecutive cycles with negligible loss in efficiency. Furthermore, it effectively reduced the sulfur content of real diesel to as low as 9 μg g⁻¹, highlighting its strong potential for practical applications.

Fig. 21 ODS of diesel using POM-ILs. (a) Mechanism proposed for oxidation of sulfur compound to sulfone over the active species, reproduced from ref. 356 with permission from Elsevier,³⁵⁶ copyright 2013. (b) The proposed sequence involved in the oxidation reaction, reproduced from ref. 356 with permission from Elsevier,³⁵⁶ copyright 2013. (c) Supposed mechanism of the [MIMPs]₃PMo₆W₆O₄₀ in ODS process, reproduced from ref. 357 with permission from Elsevier,³⁵⁷ copyright 2021.

It is worth noting that, based on years of research experience in our group, Mo-based polyoxometalate ionic liquids typically require relatively lower reaction temperatures but exhibit lower hydrogen peroxide utilization efficiency. In contrast, W-based polyoxometalate ionic liquids generally require higher reaction temperatures, yet demonstrate significantly higher efficiency in H₂O₂ utilization.

To overcome high viscosity, difficult separation/recycling, and limited catalytic site exposure, supported IL catalysts have been developed by immobilizing catalytically active ILs onto porous solid supports via physical adsorption, electrostatic interaction, or chemical anchoring, creating heterogeneous catalysts.^359,360 Typical supports include high surface area inorganic oxides, ordered mesoporous materials, MOFs, porous carbons, and magnetic nanoparticles. These supports enhance IL dispersion and uniformity in reaction media, while pore confinement and interfacial effects improve mass transfer and oxidant activation. Supported ILs maintain efficient oxidant activation, structural stability, easy operation, and recyclability through centrifugation, magnetic separation, or filtration, avoiding homogeneous catalyst loss and secondary pollution, promoting sustainable industrial processes. Optimizing support pore size, surface polarity, and IL spatial arrangement constructs “enrichment-reaction” microenvironments at oil/water interfaces, accelerating oxidation rates with excellent kinetic advantages.

Currently, different types of supports, such as MOFs, carbons, boron nitride, molecular sieves, and some other materials with high specific surface areas, have been employed to disperse ILs, making the exposure of catalytically active sites significantly improved. After many years of research, most designed supported ILs can possess ODS performance comparable to that of pure ILs, combining both the advantages of homogeneous ILs and heterogeneous catalysts. Our group first employed few-layered hexagonal boron nitride as supports to disperse ILs. Graphene-like hexagonal boron nitride (G-h-BN), as a novel few-layer two-dimensional material, has been synthesized and employed as a support for tungsten-based ILs in ODS systems.³⁶⁰ The few-layer structure endows G-h-BN with a high specific surface area and excellent dispersibility, which significantly reduces the required amount of IL while maintaining catalytic activity. Consequently, the desulfurization performance surpasses that of the corresponding homogeneous ionic liquid systems. Mokhtarani et al.³⁶¹ reported the preparation of a novel heterogeneous catalyst by immobilizing the ionic liquid 1-octyl-3-methylimidazolium hydrogen sulfate ([Omim][HSO₄]) onto a silica matrix via a sol–gel method. This catalyst was applied to ODS of model oil containing DBT and real diesel fuel, achieving a DBT removal efficiency as high as 99.1%. Compared with desulfurization systems using bulk ionic liquids, this catalyst significantly reduced the amount of ionic liquid required. Akbari et al.³⁶² reported the effective immobilization of the ionic liquid NMP·FeCl₃ (NMP = N-methyl-2-pyrrolidone, C₅H₉NO) onto a γ-Al₂O₃ support, which was applied in ODS using 30 wt% H₂O₂ as the oxidant. The catalyst achieved a maximum DBT desulfurization efficiency of 99%, and the use of the support significantly reduced the amount of ionic liquid required during the ODS process. Xia et al.³⁶³ designed and synthesized a novel graphene-based “rewritable” functional catalyst, composed of poly(1-vinyl-3-ethylimidazolium bromide)-modified reduced graphene oxide (denoted as poly[ViEtIm]Br-rGO). The polymeric ionic liquid poly[ViEtIm]Br serves as a bridge between polar catalytic anions and the nonpolar graphene support, imparting excellent dispersibility of the composite in ionic liquids and thereby ensuring full exposure of catalytic active sites during desulfurization. More importantly, poly[ViEtIm]Br exhibits reversible anion exchange properties, enabling sequential “writing” and “erasing” of different Brønsted acids or heteropolyacid anions on the reduced graphene oxide nanosheets, establishing a green recyclable platform for screening suitable catalysts based on the same support. To improve the stability of ILs, our group synthesized a Fenton-like ionic liquid catalyst supported on MCM-41 mesoporous material via a grafting method and applied it to the removal of sulfur compounds in model oil. The supported catalyst exhibited high catalytic activity.³⁶⁴

ILs in diesel ODS offer advantages such as high desulfurization efficiency, low volatility, and tunable structures. However, they face challenges including high cost, high viscosity, and difficulties in recovery. Supported ionic liquids immobilized on carriers can enhance specific surface area and mass transfer efficiency, reduce IL dosage, and facilitate recovery, but uneven loading or insufficient carrier stability may impair catalytic performance. Future research should focus on designing low-cost, highly stable functionalized ionic liquids; optimizing carrier structures and loading techniques to maximize active site utilization; developing integrated catalytic–adsorptive materials; and exploring green synthesis and recycling methods to advance industrial applications.

Porous ionic liquids (PILs) combine the intrinsic advantages of ionic liquids—such as high thermal stability, structural tunability, and negligible volatility—with the desirable features of porous materials, including large specific surface area and well-developed pore networks. These unique properties endow PILs with excellent catalytic performance. In oxidative desulfurization of diesel, the porous architecture of PILs can efficiently adsorb sulfur compounds and concentrate oxidants, while their active sites can significantly catalyze the oxidation of sulfur species into sulfone products that are readily separable. Moreover, the low volatility, recyclability, and tunable structure of PILs further enhance desulfurization efficiency under mild, energy-efficient conditions. In recent years, our research group, among others, has successfully applied PILs in the catalytic ODS of diesel fuels.

Our group³⁶⁵ has focused on the construction of type III PILs and their application in ECODS. To address the limitations of conventional PILs, such as the scarcity of catalytic active sites and the difficulty in tuning hydrophilicity during liquid–liquid biphasic reactions, we designed and synthesized a novel material, PILs-M. This material employs ZIF-8-confined phosphomolybdic acid (HPMo) as the microporous framework and utilizes N-butylpyridinium bis(trifluoromethylsulfonyl)imide ([Bpy][NTf₂]) as the ionic liquid medium. The resulting system not only integrates the advantages of traditional ionic liquids and microporous materials, such as excellent extraction capability, high dispersion of active species, and robust structural stability, but also enhances the hydrophilicity of the internal surface of ZIF-8. This modification improves the contact efficiency between the oxidant (H₂O₂) and the active catalytic sites, thereby significantly boosting the oxidative desulfurization performance. This study not only enriches the design strategies of porous ionic liquid systems and broadens the selection of organic guests and porous framework materials, but also offers new insights for developing efficient and renewable green catalytic systems. It holds great potential for applications in energy purification and environmental remediation.

Meanwhile, our group developed a highly efficient and stable solid–liquid interfacial strategy by constructing a porous ionic liquid (PIL) material based on the UiO-66(Zr) for the selective catalytic removal of sulfur compounds from diesel.³⁶⁶ The synthesized PIL material integrates the permanent porous structure of the MOF with the superior solvent properties of ionic liquids, exhibiting excellent activity and selectivity in catalytic ODS, particularly in the removal of DBT. Experimental results, in combination with DFT calculations, revealed that the ionic liquid component not only acts as a solvent and extractant, enriching DBT molecules via π–π stacking and C–H⋯π interactions, but also stabilizes the solid–liquid interface through electrostatic solvation effects. This interfacial stabilization facilitates the activation of H₂O₂ by the UiO-66 framework, promoting the generation of ˙OH radicals and enabling the selective oxidation of DBT to its sulfoxide/sulfone derivatives. This study demonstrates the great potential of porous ionic liquids in solid–liquid interface modulation and green catalytic oxidative desulfurization, providing a new research pathway and theoretical foundation for the design of high-performance desulfurization materials and solid–liquid interface catalysis.

On the other hand, focusing on the structural modulation of ionic liquid organic guests, we constructed a novel bifunctional porous ionic liquid catalyst (UiO-66-BAPILs) by dispersing the metal–organic framework UiO-66 into a Brønsted acidic pyridinium-based ionic liquid, [BSPy][CF₃SO₃], for ODS of diesel.³⁶⁷ Under optimized reaction conditions, using H₂O₂ as the oxidant, this catalytic system achieved a sulfur removal efficiency of up to 99.5%. Mechanistic investigations revealed that UiO-66-BAPILs served both as an extractant and a catalyst during the reaction: on one hand, aromatic sulfur compounds were enriched through π–π stacking and C–H⋯π interactions; on the other hand, they were efficiently oxidized into sulfone compounds via the generation of peroxysulfonic acid and ˙OH. This study proposed a new strategy for designing Brønsted acidic bifunctional porous ionic liquids to enhance catalytic oxidative desulfurization performance.

Moreover, the currently reported type III PILs are predominantly constructed based on size-exclusion effects, featuring microporous frameworks whose pore sizes limit substrate mass transfer and diffusion. To address this challenge, we propose a novel construction strategy for porous ionic liquids based on differences in oleophilicity and oleophobicity. We innovatively designed a class of type-III-B porous ionic liquids (PILS-B), effectively overcoming the conventional “size-effect” limitation imposed by pore size matching in traditional PIL designs. This PILS-B material employs a hydrophobic internal surface mesoporous SBA-15 framework loaded with high-entropy single-atom catalysts (HESAC@SBA-15) as the porous scaffold, combined with an oleophilic ionic liquid 1-butyl-3-methylimidazolium tetrafluoroborate as the organic guest to form the composite.³⁶⁸ The resulting material not only retains the excellent gas storage properties of porous ionic liquids but also achieves synergistic enhancement of structural stability and catalytic efficiency through the ultra-high catalytic activity of HESAC@SBA-15 in oxygen activation. The ordered mesoporous structure of SBA-15 significantly facilitates mass transfer during the reaction process, thereby substantially increasing the catalytic reaction rate. In aerobic oxidative desulfurization of aromatic sulfur compounds in fuel oils as model substrates, PILS-B exhibits outstanding catalytic performance, achieving sulfur removal efficiency exceeding 99%. This study not only enriches the construction strategies of porous ionic liquids by broadening the selection of organic guests and porous framework materials but also provides a novel approach for developing efficient and recyclable green catalytic systems. It holds broad application prospects in the fields of energy purification and environmental remediation.

PILs, as a novel class of functional materials, integrate the structural advantages of porous materials with the unique functional properties of ionic liquids, demonstrating distinctive value in diesel oxidative desulfurization. Their porous architecture not only provides abundant active sites and efficient mass transfer channels but also enables precise catalytic performance tuning through the designability of ionic liquids, achieving efficient adsorption and catalytic oxidation of organic sulfur compounds. This material can effectively realize deep removal of aromatic sulfur compounds in diesel under mild conditions. However, the technology currently faces challenges including mass transfer limitations caused by viscosity effects, cost pressures from complex preparation processes, and stability concerns during long-term operation. Future research should focus on multidimensional breakthroughs in molecular design, pore engineering, and process optimization, emphasizing the development of novel composite systems with high stability and low cost, as well as exploring their scale-up applications in practical industrial settings.

6.4.5. Deep eutectic solvents. Deep eutectic solvents (DESs), also known as ionic liquid analogues, are a class of special solvent systems formed through intermolecular hydrogen bonding between hydrogen bond donors (such as carboxylic acids, alcohols, and amides) and hydrogen bond acceptors (such as quaternary ammonium salts and metal salts). These solvents exhibit eutectic temperatures significantly lower than those of their individual components and offer numerous advantages, including ease of preparation, low cost of raw materials, extremely low volatility, relatively low toxicity, and biodegradability.³⁶⁹ The physicochemical properties of DESs can be precisely tuned by varying the types and ratios of hydrogen bond donors and acceptors, granting them exceptional design flexibility. As a result, DESs have attracted widespread attention in green chemistry, especially in fields such as catalysis, separation and purification, electrochemistry, and materials synthesis.³⁷⁰

DESs have shown promising progress in deep desulfurization of fuels in recent years. The role they play in ODS processes mainly manifests in three ways: first, certain DESs exhibit selective solubility toward sulfur compounds and can act directly as extractants; second, acidic DESs can activate oxidants such as hydrogen peroxide, promoting the generation of reactive oxygen species; and third, metal-containing DESs provide catalytic activity through their metal components. Studies have demonstrated that by rationally designing the composition of DESs, it is possible to achieve synergistic enhancement of oxidation and extraction processes, thereby significantly improving desulfurization efficiency. In recent years, research on DESs for diesel ODS has continued to advance, with the development of various novel DES systems, particularly those based on combinations of quaternary ammonium salts and organic acids. These new solvents not only exhibit higher sulfur capacity but also maintain stability across a wide temperature range. In terms of process optimization, incorporating catalysts or applying external field technologies has been shown to markedly accelerate oxidation kinetics and enhance desulfurization performance.

Similar to IL-based ODS systems, traditional DESs lack catalytically active sites, and can only function as reaction media. Thus, because of the tunable structure in both hydrogen donors and hydrogen acceptors, catalytically active species can be introduced to DESs, making the DESs work both as the extractant and the catalyst. Wang et al.³⁷¹ designed and synthesized a series of novel deep eutectic solvents (DESs) using choline chloride (ChCl) as a representative hydrogen bond acceptor (HBA) and various hydrogen bond donors (HBDs) such as polyethylene glycol (PEG), 1,3-butanediol (BG), ethylene glycol (EG), glycerol (Gl), propionic acid (Pr), malonic acid (MA), and urea (U), thereby constructing a diverse range of neutral, acidic, and basic DESs. Using phosphotungstic acid (HPW) as the catalyst and 30 wt% H₂O₂ as the oxidant, their performance was evaluated in an ECODS system. The optimized system achieved a DBT removal efficiency of up to 99.1%, demonstrating the high potential of DESs in deep desulfurization applications. To address the limited intrinsic catalytic activity of DESs, a key bottleneck restricting their practical application, Wang et al.³⁷² developed and screened a series of DES systems with multiple catalytic active sites. Among them, the DES composed of 5-sulfosalicylic acid (5-SSA) and formic acid (FA) in a 1 : 3 molar ratio exhibited exceptional performance, achieving a DBT removal efficiency of up to 99.8% under mild conditions (25 °C, 2.5 h, O/S = 2.5). In this system, H₂O₂ interacts with the DES to form highly oxidative peroxide intermediates. Furthermore, density functional theory (DFT) calculations revealed that the functional groups –SO₃H, –COOH, and –OH present in the DES molecules effectively activate H₂O₂, significantly lowering the reaction energy barrier and thereby enhancing the overall oxidative desulfurization efficiency. Ren et al.³⁷³ synthesized four types of dual-acid deep eutectic solvents (DESs) and applied them in ODS systems using hydrogen peroxide as the oxidant. Among these, the L-pyroglutamic acid/trifluoroacetic acid (L-Pyro/TFA) DES demonstrated outstanding performance in the removal of aromatic sulfur compounds. This DES system could be easily regenerated by simple washing with ultrapure water, maintaining a desulfurization efficiency above 98% upon reuse. Lü further developed a DES that can activate O₂ as the oxidant for ODS.³⁷⁴ They designed a biomimetic catalytic system by coupling ammonium molybdate polyoxometalates (POMs) with four different DESs to achieve ODS under aerobic conditions. DFT calculations revealed that the differences in desulfurization performance were mainly attributed to the inherent oxidative capabilities of the organic acids present in the DESs. Our group³⁷⁵ constructed a series of Brønsted acidic DESs and systematically investigated the structure–activity relationships (SARs) at both molecular and atomic scales. Experimental and theoretical results revealed that shorter distances between the oxygen atoms in the active sites of DESs and the sulfur atoms in sulfur compounds correlated with higher catalytic activity. Notably, the effective O⋯S distance showed a strong correlation with EODS efficiency, with an R² value exceeding 0.9. This geometric configuration could be optimized by tailoring the DES composition, significantly accelerating the interaction between intermediates and substrates. Among the studied systems, TEAC/2BSA exhibited the optimal O⋯S distance and demonstrated excellent deep desulfurization performance toward various sulfur compounds. This study not only deepened the mechanistic understanding of the EODS process but also provided theoretical and methodological guidance for the efficient screening of functional DESs. In addition, our group³⁷⁶ further developed an inorganic DES system composed of choline chloride (ChCl), polyethylene glycol (PEG), and boric acid (BA) for ECODS of diesel fuel. Compared with conventional organic acid-based DESs, this ternary system exhibits several advantages, including low volatility, low toxicity, and enhanced catalytic activity. Under mild conditions, it achieved a high sulfur removal efficiency of up to 99.2%. Albayati et al.³⁷⁷ synthesized a series of DESs based on choline chloride (ChCl) and polyethylene glycol (PEG), and applied them to the desulfurization of heavy crude oil with an initial sulfur content as high as 37 900 ppm (3.79 wt%). Using 30 wt% H₂O₂ as the oxidant and formic acid as the co-catalyst, an ultrasound-assisted extractive–oxidative desulfurization (EUAODS) system was constructed, achieving a sulfur removal efficiency of 62%. The study demonstrated that the synergistic effect between DESs and ultrasound significantly enhances the desulfurization efficiency of heavy oils, providing a novel and environmentally friendly strategy for high-sulfur crude oil treatment.

DESs have demonstrated significant advantages in ODS of diesel. Their green, environmentally friendly nature—characterized by low toxicity and biodegradability—aligns well with the goals of sustainable development. DESs exhibit high desulfurization efficiency, particularly showing excellent selective oxidation capability toward thiophenic sulfur compounds, which are difficult to remove by conventional methods. Moreover, DESs are low cost and easy to synthesize, with readily available raw materials. Their physicochemical properties can be finely tuned by adjusting the combination of HBDs and HBAs, enabling tailored performance for different sulfur-containing systems.

However, some challenges remain. The relatively high viscosity of DESs may hinder mass transfer efficiency, while limited thermal stability under elevated temperatures can affect their performance. Additionally, their desulfurization efficiency may decrease after multiple regeneration cycles, and issues related to process compatibility and economic feasibility must be addressed for large-scale industrial applications.

Future research should focus on designing novel DESs with lower viscosity and enhanced thermal stability, and on integrating catalytic oxidation strategies to accelerate reaction rates. The development of efficient recovery and regeneration methods is also crucial for reducing operational costs. Furthermore, molecular simulations and artificial intelligence (AI) tools could be employed to predict correlations between DES composition and desulfurization performance. Finally, exploring synergistic coupling with other desulfurization technologies may further promote the industrial-scale application of DES-based systems.

6.4.6. Carbon-based catalysts. Carbon-based catalysts have gained increasing attention in ODS due to their excellent physicochemical stability, abundant surface functional groups, tunable pore structures, good electrical conductivity, and loading capacity.³⁷⁸ Carbon materials include activated carbon, porous carbons, carbon nanotubes (CNTs), graphene and its oxides (GO), carbon quantum dots (CQDs), and nitrogen-doped carbons (N–C), which can be structurally and functionally modified to efficiently activate oxidants and catalyze the oxidation of aromatic sulfur compounds into polar sulfones for deep desulfurization.^379–381

Carbon materials inherently adsorb and enrich oxidants and facilitate electron transfer, especially after surface oxidation, nitrogen doping, or heterogeneous loading of transition metals, which significantly enhance catalytic activity. For example, GO contains abundant oxygen-containing groups (carboxyl, hydroxyl, epoxy), providing hydrophilic environments and inducing directional cleavage of oxidants to generate selective active oxygen species. In an alkaline or metal ion presence, this process is further accelerated, forming mild and efficient oxidation systems. GO's large surface area also promotes sulfur compound adsorption, enabling enrichment-reaction coupling.³⁸²

Carbon-based catalysts have been widely employed in ODS systems using H₂O₂ as the oxidant. Zhu et al.³⁸³ investigated the hydrogen peroxide-mediated ODS of diesel fuel using activated carbon as the catalyst, focusing on the adsorption capacity of activated carbon for DBT and its correlation with catalytic performance. The results demonstrated that stronger adsorption capacity of activated carbon corresponded to higher catalytic oxidation activity toward DBT. Once DBT was adsorbed on the activated carbon surface, reactive oxygen species effectively promoted its oxidation. The catalytic performance was significantly influenced by the aqueous phase pH, with the oxidation rate of DBT markedly increasing when the pH dropped below 2. Additionally, the addition of formic acid enhanced DBT oxidation in systems containing lignin-based activated carbon. Timko et al.³⁸⁴ employed hydrogen peroxide and formic acid as oxidants, with activated carbon serving as a reaction promoter, combined with power ultrasound to enhance interfacial dispersion, to study the ODS of jet fuel and diesel. Following ODS treatment and subsequent adsorption purification using activated alumina, sulfur removal efficiencies reached 98% for JP-8 fuel, 94% for diesel, and over 88% for ultra-low sulfur diesel. Fischer et al.³⁸⁵ systematically characterized and evaluated the performance of activated carbons derived from various sources and preparation methods, focusing on their catalytic promotion role in the ODS process. The study revealed that acid-treated activated carbon from lignocellulosic materials exhibited significantly higher catalytic activity than other types of activated carbon. In JP-8 fuel desulfurization experiments using hydrogen peroxide and acetic acid as oxidants, the optimized activated carbon achieved 69% oxidation conversion of 2,3-dimethyl dibenzothiophene (2,3-DMBT). A comprehensive analysis suggested that activated carbon promotes the formation of carboxylic acid-like active species through its surface defects, thereby enhancing ODS reaction efficiency. Generally, carbon-based materials are generally employed as co-catalyst in ODS systems with H₂O₂ as the oxidant, and they do not function as catalytically active sites.

The very first report using a carbon-based catalyst for ODS with O₂ as the oxidant can be dated back to 2014; Xiao et al. reported application of carbon nanotubes for ODS using O₂ as the oxidant.³⁸⁰ They found that using three different types of carbon nanotubes (CNT-SZ, CNT-TS, and CNT-CD) as catalysts, the oxidation conversion rate of DBT reached 100% at 150 °C. The deactivated CNT catalysts could be effectively regenerated by heat treatment under an argon atmosphere at 900 °C. Raman spectroscopy analysis indicated that the degree of graphitization of the CNTs played a decisive role in their catalytic activity. Higher graphitization resulted in better electrical conductivity and enhanced electron transfer capability, thereby promoting efficient electron transfer in redox reactions and significantly improving oxidative desulfurization performance. In 2017, Su, an expert in carbon catalysis, and co-workers employed reduced graphene oxide (rGO) in ODS with O₂ as the oxidant, and remarkable desulfurization performance was gained.³⁸² By employing XPS analysis, chemical titration, and a series of comparative experiments, they demonstrated that carbonyl (C=O) clusters play a crucial role in the oxidation process. Under reaction conditions, chemically active defect sites contribute to the enhancement of catalytic performance, as these defect sites are capable of in situ generating carbonyl groups. Afterward, the same group combined oxidized carbon nanotubes (oCNTs) with ionic liquids (ILs) for the oxidation of aromatic sulfur compounds, using H₂O₂ as the oxidant.³⁸⁶ In this system, oCNTs exhibited significant catalytic activity and stability, outperforming some reported metal catalysts. The study demonstrated that ionic liquids play an indispensable role in enhancing the catalytic performance of oCNTs. It was found that carbonyl groups act as active sites in the oxidation process, and this finding was further supported by deactivation and model catalyst experiments. In 2021, our group proposed a heteroatom-bridging strategy to enhance the catalytic activity of carbon-based catalysts.³⁸⁷ As a validation of this strategy, a series of boron (B)-doped graphite catalysts were synthesized. Through detailed characterization, it was found that the doped boron atoms were uniformly distributed within the graphite. More importantly, the study revealed that the doped boron atoms, acting as bridges for electron transfer, played a crucial role in oxygen activation, significantly enhancing the catalytic activity of graphite in oxidation processes, thereby achieving ultra-deep oxidative desulfurization performance.

Carbon-based supported catalysts have demonstrated outstanding performance and great potential for application in diesel ODS. Carbon materials such as graphene and carbon nanotubes, with their unique physicochemical properties, have become ideal catalyst supports. Their advantages are mainly reflected in the following aspects: first, these materials possess an extremely high specific surface area and abundant surface functional groups, which provide ample anchoring sites for active components such as transition metal oxides and polyoxometalates, allowing for the highly dispersed and stable loading of these active components, thereby significantly enhancing catalytic efficiency.³⁸⁸ Second, the excellent electronic conductivity of carbon materials promotes electron transfer during the catalytic process, effectively activating oxidants like hydrogen peroxide and accelerating the oxidation of stubborn sulfur-containing compounds such as dibenzothiophene. Furthermore, by doping with heteroatoms such as nitrogen and boron, or using defect engineering modifications, the electronic structure and surface properties of carbon supports can be precisely tuned to further enhance their selective adsorption and catalytic oxidation capabilities for sulfur compounds. Additionally, the good chemical stability and mechanical strength of carbon materials provide catalysts with a longer lifespan. Carbon-based supported catalysts are capable of achieving deep desulfurization under relatively mild reaction conditions.

Kozhevnikov et al.³⁰⁸ loaded Keggin-type HPA onto activated carbon (HPA/C), which effectively catalyzed the ODS of diesel. H₃PMo₁₂O₄₀/C exhibited the best catalytic performance, achieving 100% removal of dibenzothiophene in model diesel at 60 °C. Additionally, this catalyst is recyclable and does not undergo deactivation. Compared with other recently reported similar heterogeneous catalysts, this catalyst demonstrated superior performance in comparable systems. Besides the effect of HPA type on desulfurization performance, they also investigated the effect of carbon support on the desulfurization performance.³⁸⁹ By testing commercial activated carbons with different acidic and basic properties as HPA carriers, the study revealed a strong influence of the basicity of the carbon support on the retention of the HPA structure on the carbon surface and its catalytic activity. Zhao et al. employed carbon nanotubes as the support to disperse HPAs.³⁹⁰ They studied two types of catalysts based on carbon nanotubes (CNTs), one by impregnating POM into the channels of CNTs using an impregnation method, and the other by encapsulating modified CNTs within polyoxometalates through an ion-exchange method. The catalysts achieved a DBT conversion rate of up to 99.4%, and the desulfurization system maintained stable catalytic activity after 8 cycles without significant degradation. Our group employed carbon as the support to disperse polyoxometalate-based ionic liquids.³⁹¹ We used [N-(3-sulfonic acid propyl)-pyridine]₃PMo₁₂O₄₀ ([PSPy]PMo) as the active component and layered graphite carbon as the support to successfully immobilize the ionic liquid onto graphite carbon. After hydrothermal impregnation, the structure of the catalyst was preserved. The prepared catalyst exhibited excellent catalytic performance in a solvent-free oxidative desulfurization system, achieving 100% removal of DBT. We further proposed a novel single-atom catalyst (SAC) consisting of uniformly distributed single chromium atoms anchored on multi-walled carbon nanotubes, applied to ODS reaction with O₂ as the oxidant.²⁸⁹ Unlike traditional noble metal-based nanocatalysts, this catalyst is based on abundant earth metals, with a stable nanoporous support, effectively overcoming the deactivation issues caused by carbon deposition and sintering. Using aromatic sulfur compounds as model reactants and molecular oxygen as the oxidant, the study shows that this catalytic system converts oxygen molecules into active oxygen species ˙O₂⁻via chromium atomic active sites, which then effectively transform aromatic sulfur compounds into sulfone. This transformation is not only attributed to the catalytic activity of the chromium atoms but also to the robust structure of the carbon nanotube support material, its nanoporous characteristics, and the π-conjugated structure. Besides, graphene, a 2D material with high specific surface areas has been employed as support as well. Mokhtarani et al.³⁹² employed phosphomolybdic acid (H₃PMo₁₂O₄₀) supported on graphene oxide (GO). By evaluating the catalyst's characteristics and its performance in the extraction–oxidative desulfurization process, they achieved complete oxidative desulfurization within 30 minutes, with desulfurization efficiencies of 100% for both DBT and 4,6-DMDBT. Mortaheb et al.³⁸⁸ proposed a novel graphene oxide (GO)-based heterogeneous catalyst, synthesized by loading varying amounts of phosphotungstic acid (H₃PW₁₂O₄₀) onto GO. The catalyst achieved 100% desulfurization yield within 30 minutes. Compared with the homogeneous HPW catalyst, the heterogeneous HPW-GO catalyst demonstrated a higher desulfurization yield, exhibiting a better synergistic effect. As summarized above, MOF-derived catalysts are another important type of carbon-supported catalysts, which can in situ disperse catalytically active sites. Jhung et al.³⁹³ investigated the ODS reaction of model fuels using tungsten-based catalysts under microwave irradiation. The study primarily employed a high-porosity W₂N@C catalyst, obtained from a ternary MOF composite with a pure W₂N phase. Microwave heating significantly promoted the ODS reaction, particularly for the most difficult-to-remove thiophene (Th) compounds. At 60 °C, the W₂N@C catalyst, under microwave radiation, achieved a 97% removal of thiophene in just 30 minutes.

Carbon materials possess unique advantages and potential in the field of diesel ODS, yet they also face some critical challenges. In terms of material properties, carbon-based materials such as activated carbon, carbon nanotubes, and graphene, with their well-developed structures, tunable surface chemistry, and excellent electronic conductivity, can efficiently adsorb organic sulfur compounds in diesel and facilitate their catalytic oxidation conversion. Especially, carbon materials modified with heteroatoms can significantly enhance surface acidic sites and redox activity, thereby improving the removal efficiency of thiophene-based sulfur compounds. Moreover, carbon materials boast the advantages of abundant raw materials, good environmental compatibility, and recyclability. However, existing carbon-based catalysts still face issues such as insufficient active site density, poor long-term stability, and low selectivity in complex systems, particularly when dealing with high-sulfur diesel, where pore blockage and loss of active components are common.

Future research should focus on developing novel multifunctional carbon-based catalytic systems, such as enhancing catalytic activity by precisely constructing atomically dispersed metal species on carbon frameworks or using interface engineering strategies to create strong interactions between carbon materials and metal oxides to improve stability. Additionally, exploring the role played by carbon materials in photochemical/electrocatalytic coupled desulfurization processes and developing efficient, green regeneration technologies are crucial directions for overcoming the current technological bottlenecks. Furthermore, combining advanced characterization techniques and theoretical calculations to deeply understand the structure–activity relationships of surface-active sites on carbon materials will provide a scientific basis for designing high-performance desulfurization catalysts. From an engineering perspective, developing scalable production processes for carbon materials, optimizing reactor design to improve mass transfer efficiency, and establishing comprehensive lifespan evaluation methods are key steps to bring this technology to industrial applications.

6.4.7. Boron nitride catalysts. Hexagonal boron nitride (h-BN) is a material with a graphene-like layered structure, composed of alternating boron and nitrogen atoms arranged in a hexagonal honeycomb plane. The layers are tightly bound by strong covalent bonds, while the interlayer interactions are governed by weaker van der Waals forces, which make it easy to exfoliate into ultra-thin nanosheets.³⁹⁴ Unlike graphite, h-BN is an insulator with a wide band gap (approximately 5–6 eV). However, its high thermal stability and chemical inertness allow it to retain structural stability even in high-temperature and strongly oxidative environments. The surface of h-BN is rich in boron sites and nitrogen lone pair electrons, with boron atoms exhibiting Lewis acidity and nitrogen atoms showing Lewis basicity. This bifunctional property enables h-BN to simultaneously activate oxidants (such as H₂O₂ or O₂) and adsorb sulfur-containing compounds, playing a crucial role in ODS reactions. Furthermore, exfoliated nanosheets of h-BN possess a high specific surface area, and the edge or defect sites (such as boron vacancies) can further enhance its catalytic activity.

Our research group was the first to introduce h-BN into diesel ODS systems, achieving activation of H₂O₂, O₂, and other oxidants. In ODS applications, h-BN demonstrates significant advantages. Its boron atoms can effectively adsorb and activate hydrogen peroxide, generating highly reactive hydroxyl radicals, which oxidize sulfur compounds in fuel (e.g., DBT) to sulfone or sulfoxide, making them easier to remove through subsequent extraction or adsorption steps. At the same time, the nitrogen atoms on the surface of h-BN can adsorb sulfur-containing compounds through their Lewis basicity, increasing the local concentration of reactants on the catalyst surface and promoting the oxidation reaction. The high thermal stability of h-BN ensures that it maintains catalytic activity under high-temperature reaction conditions, while its chemical inertness prevents catalyst degradation in strong oxidative environments. Compared with traditional metal-based catalysts, h-BN, as a non-metal material, does not suffer from metal leaching, thus reducing secondary pollution and aligning better with green chemistry principles. Additionally, h-BN exhibits high selectivity towards sulfur compounds, avoiding the over-oxidation of hydrocarbon components in fuel, thus preserving the fuel's calorific value and quality.

Many h-BN catalysts for ODS have been reported, and Li et al. have systemically summarized all our works.³⁹⁵ The very first report using h-BN as the catalyst for ODS using O₂ as the oxidant can be dated back to 2016; our group successfully synthesized a h-BN with high specific surface area of 1900 m² g⁻¹ using a solvent-mediated, template-free nanostructuring strategy, and proposed a new lattice plane control mechanism to explain the formation process.³⁹⁶ Furthermore, h-BN exhibited excellent catalytic activity in air oxidation for deep desulfurization. This study not only provides new insights into controlling the structure of h-BNNs through rational precursor selection but also offers a novel strategy for the large-scale preparation of high-surface-area nanomaterials. The obtained h-BN possessed the highest specific surface areas at that time. Moreover, this work is the very first work using h-BN for catalytic oxidation reactions. In 2017,³⁹⁷ we further clarified the catalytically active sites in h-BN for O₂ activation. We employed a DFT calculation as well as experimental verification methods in determining the catalytically active sites, and found that N–N terminated edges in h-BN are the main catalytically active sites for O₂ activation. In addition, based on such conclusion, we proposed a Zn-salt-induced strategy to engineer defects with more N–N terminated defective edges, inducing a significantly improved ODS performance. In addition, to expose more catalytically active edges, our group first developed a gas-driven exfoliation strategy to produce h-BN with a fewer-layer structure from commercially available bulk h-BN.³⁹⁸ When the few-layered h-BN was employed in ODS with O₂ as the oxidant, a significantly improved desulfurization performance was gained.³⁹⁹ Afterward, electronic structure tuning strategies were further proposed to fine modulate the structure of h-BN for an enhanced ODS performance. For example, we introduced a carbon-doped strategy for tuning the electron dispersion in h-BN (BCN), thus making the catalyst more easily activate O₂, reducing the ODS temperature from 120 °C to 100 °C.⁴⁰⁰ We also investigated the influence of different oxygen doping modes on the catalytic performance of h-BN in aerobic oxidative desulfurization.⁴⁰¹ Oxygen-doped BN (BNO) was strategically modified, and the effects of various oxygen doping modes were systematically studied. Detailed results showed that lattice oxygen doping significantly enhances catalytic activity, while the impact of edge hydroxyl oxygen is relatively minor. Catalytic experiments demonstrated that, under mild conditions, BNO exhibits excellent catalytic performance, achieving a sulfur removal rate of 98.4% for aromatic sulfides. In 2020, we further employed h-BN as a metal-free catalyst for ODS using H₂O₂ as the oxidant, and we found that the boron radical in h-BN can activate H₂O₂, generating ˙OH species.⁴⁰² These ˙OH species can rapidly oxidize the sulfur compounds to their corresponding sulfone and facilitate their separation. Furthermore, we proposed a strategy to modify a metal-free catalyst through organic modification using DES to tune the electronic structure of h-BN.⁴⁰³ Detailed experimental results showed that the electronic properties of h-BN can be effectively adjusted by the introduced DES, with the boron atoms in h-BN carrying a positive charge. Furthermore, the tuned electronic properties lower the activation energy for H₂O₂ to generate hydroxyl radicals, thus enhancing the catalytic performance of the oxidative desulfurization reaction.

h-BN can also serve as a high-surface-area support, and by combining it with other active components, we can construct supported catalysts to further optimize its catalytic performance. For instance, loading ionic liquids, metal oxides, and other compounds can significantly enhance ODS activity. Furthermore, introducing defects, such as boron or nitrogen vacancies, into h-BN through defect engineering can significantly enhance its activation capacity for oxidants, further improving desulfurization efficiency. Experimental studies have shown that catalytic systems composed of h-BN nanosheets and H₂O₂ or O₂ can achieve highly efficient removal of dibenzothiophene from diesel under mild conditions, with desulfurization rates exceeding 95%.

Similarly, our group was also the very first group to employ h-BN as a support to engineer a supported catalyst for ODS. As mentioned above in the ILs for ODS section, in 2015, we first employed h-BN as a support to disperse ILs for promoted ODS performance.³⁶⁰ Afterward, h-BN has been widely employed as a support to synthesize supported catalysts. We introduced a one-step high-temperature in situ synthesis method to confine tungsten oxide nanoparticles (WO_xNPs) within graphite-like boron nitride (g-BN), which enhances the interaction between WO_xNPs and the support, and controls the particle size distribution of WO_xNPs to approximately 4–5 nanometers.⁴⁰⁴ The resulting catalyst was applied to the catalytic oxidation of aromatic sulfides and demonstrated high catalytic activity. The study revealed that the balance between the tungsten loading and the particle size distribution of WO_xNPs plays a decisive role in catalytic activity. Moreover, h-BN has been employed as support to disperse noble metals, such as platinum nanoparticles (Pt NPs), which have been long regarded as active species for activation of O₂.^405–407 We found that electron transfer between Pt NPs and h-BN can be formed, and the Pt NPs are generally located at edges of h-BN, which are often electron-rich structures. Such an electron transfer can significantly improve the catalytic performance of Pt NPs for a boosted ODS performance.

h-BN, with its unique graphene-like layered structure and highly ordered honeycomb lattice, demonstrates excellent performance advantages in the field of diesel ODS. Not only does h-BN exhibit some intrinsic catalytic activity, but it can also serve as a metal-free catalyst to activate O₂ or H₂O₂ for ODS. Additionally, h-BN can act as a catalyst support, providing a platform for the uniform dispersion of active metal species, while maintaining structural stability in highly oxidative environments. This effectively addresses the technical challenge of corrosion and deactivation commonly faced by traditional supports. However, the intrinsic chemical inertness of h-BN leads to insufficient surface reactivity, often requiring precise defect engineering and atomic-level doping to create sufficient active sites. This process involves complex preparation techniques and strict control of reaction conditions, posing significant challenges during industrial-scale implementation.

Future development will focus on the exploration of new green synthetic routes, as well as the development of scalable green methods to achieve large-scale production of high-quality materials. In addition, a deeper understanding of the bond breakage and recombination mechanisms of boron–nitrogen chemical bonds at the material surface is necessary to reveal the electronic transfer patterns between active metals and supports, providing theoretical guidance for the design of efficient and stable catalytic systems. Furthermore, combining h-BN with other functional materials to construct hierarchical porous structures and developing adaptive catalysts with high efficiency will be critical breakthroughs in this field. From an engineering application perspective, systematic studies on the catalyst's cycling regeneration performance and long-term operational stability, along with the establishment of a comprehensive lifetime evaluation system, will provide reliable guidance for the design of industrial devices.

Research on diesel ODS catalysts spans a wide range of material systems and catalytic mechanisms, encompassing not only traditional systems such as heteropoly acids (HPAs), metal–organic frameworks (MOFs), metal oxides, noble metals, ionic liquids, carbon materials, and boron nitride but also many emerging or non-traditional catalytic systems. Some non-noble metal complexes, such as metal phthalocyanines (FePc, CoPc) and metal porphyrins (MnTPP, FeTPP), can mimic the active sites of biological enzymes, efficiently catalyzing oxidative desulfurization under mild conditions. Their macrocyclic structures stabilize high-valent metal active species, promoting the oxidation of sulfides and thiophene compounds. However, issues such as poor stability and high cost have led to optimization efforts primarily through immobilization or biomimetic catalytic strategies.

Rare earth metal oxides exhibit unique oxygen storage and release capabilities, facilitating the generation of reactive oxygen species and demonstrating excellent desulfurization performance. Furthermore, some transition metal nitrides, with tunable electronic structures, show good catalytic activity in oxidative desulfurization, especially when present in monolayer or few-layer forms, which significantly enhance their catalytic performance. Photocatalytic desulfurization systems, such as TiO₂, BiVO₄, and g-C₃N₄, produce reactive oxygen species under UV or visible light irradiation, which subsequently oxidize sulfur compounds. Electrochemical desulfurization, on the other hand, uses conductive electrodes to directly or indirectly oxidize sulfides under an applied voltage or generates hydroxyl radicals through the electro-Fenton process, a method particularly suited for deep desulfurization of low-sulfur diesel.

There are also some unconventional strategies, such as plasma-assisted oxidative desulfurization, which utilizes low-temperature plasma to generate active species to oxidize sulfur compounds. Ultrasound-assisted catalysis enhances mass transfer and free radical generation through cavitation effects. Additionally, some natural minerals, such as montmorillonite and zeolite, can serve as inexpensive catalysts after acid treatment or metal modification. Their layered structures or acidic sites aid in the adsorption and activation of sulfur compounds.

ODS faces several techno-economic challenges that hinder its widespread adoption in industrial applications. The most significant barrier is the high energy consumption required for the oxidation reactions, which often necessitate high temperatures or additional chemicals to activate the process. Overall, research on diesel oxidative desulfurization catalysts is moving towards diversification, high efficiency, and greenness. However, balancing catalytic activity, stability, and economic feasibility remains a core challenge for future research.

7. Other desulfurization methods

The research field of diesel desulfurization technology is vast and diverse, encompassing not only conventional methods such as HDS, biodesulfurization (BDS), extraction desulfurization, adsorption desulfurization, and oxidative desulfurization (ODS), but also a variety of innovative or non-traditional desulfurization techniques that show unique advantages under specific conditions.

One emerging technology is ultrasound-assisted desulfurization, which uses the cavitation effect of ultrasound to generate localized high temperature and pressure, while producing free radicals and excited-state oxygen atoms. This promotes the oxidation of sulfur compounds into sulfone or sulfate, which can then be separated by solvent extraction. Similarly, photochemical oxidative desulfurization uses ultraviolet or visible light to excite the reaction system, such as the diesel–hydrogen peroxide mixture undergoing sulfur compound oxidation under light exposure. Research in Japan has shown that this method can reduce sulfur content from 500 ppm to 50 ppm. Plasma-assisted desulfurization is another cutting-edge technology that uses low-temperature plasma to generate reactive oxygen species (such as ozone and singlet oxygen) for the direct oxidation of sulfur compounds. This method offers advantages such as low energy consumption and no need for catalysts but is still in the experimental research phase.

Additionally, electrochemical oxidative desulfurization utilizes an external voltage to drive the electrocatalytic conversion of sulfur compounds, such as generating hydroxyl radicals (˙OH) through the electro-Fenton process to oxidize refractory sulfur compounds. This is particularly suited for the production of ultra-low-sulfur diesel. Membrane separation desulfurization works by selectively permeating membranes based on differences in molecular size or polarity to separate sulfur compounds, making it particularly suitable for deep desulfurization. However, it faces challenges such as membrane fouling, low flux, and long-term stability issues.

8. Summary and outlook

8.1. Summary

Diesel desulfurization technology is a crucial method for reducing sulfur content in diesel, reducing air pollution, and complying with increasingly stringent environmental regulations. As global attention on environmental protection and sustainable development continues to grow, the removal of sulfur compounds from diesel has become one of the significant challenges faced by the petroleum refining industry. This review systematically summarizes the current diesel desulfurization methods, mainly covering technologies such as HDS, biodesulfurization, extraction desulfurization, adsorption desulfurization, and oxidative desulfurization (ODS), while analyzing their respective advantages and disadvantages, as well as future research directions.

HDS, as the most mature and widely applied desulfurization technology, effectively removes sulfur compounds, particularly aromatic sulfur compounds. However, HDS requires high temperature and pressure reaction conditions, as well as expensive noble metal catalysts, increasing energy consumption and costs. Future research should focus on developing low-temperature hydrodesulfurization techniques, designing novel catalysts to reduce reaction conditions and dependence on precious metals, thus improving the process's economic and environmental sustainability. Furthermore, catalyst stability and durability are key research areas, especially addressing catalyst deactivation to extend catalyst lifetime and enhance reaction efficiency.

Biodesulfurization, as an environmentally friendly desulfurization technology, utilizes microorganisms to degrade sulfur compounds, offering low energy consumption and mild operating conditions. Although this method has significant theoretical advantages, it faces challenges in practical applications, including low catalytic activity, slow reaction rates, complex microorganism cultivation, and reaction control. Scaling up to industrial applications remains a major hurdle. Therefore, future research should focus on screening and optimizing microbial catalysts to improve reaction rates and stability, while addressing by-product issues during the process. Additionally, integrating biodesulfurization with other desulfurization technologies, such as oxidative desulfurization, to enhance desulfurization efficiency could be an important future direction.

Extractive desulfurization relies on the liquid–liquid distribution principle, where solvents react with sulfur compounds in diesel to remove them. This technology has advantages such as simple operation and low energy consumption, but its effectiveness is limited by the selectivity, stability, and regeneration capacity of the extractant. Currently, the selectivity of extractants is a bottleneck, particularly when dealing with complex diesel components, as traditional solvents have low selectivity, leading to suboptimal desulfurization performance. Future research needs to focus on developing highly selective and stable green solvents, while improving solvent recovery and regeneration technologies to ensure economic viability in industrial applications. Furthermore, the combination of extraction desulfurization with other technologies, especially ODS and adsorption desulfurization, may further improve desulfurization efficiency.

Adsorptive desulfurization removes sulfur compounds using adsorbent materials, offering advantages such as simplicity and low energy consumption. However, challenges remain regarding the selectivity, stability, and regenerability of adsorbents. Research has shown that functionalized adsorbent materials, such as metal–organic frameworks (MOFs), nanocarbon materials, and nitrides, have significant potential in enhancing adsorption performance. Future research should focus on developing new adsorbents, especially multifunctional composite materials (such as metal–carbon composites and single-atom catalysts), to address issues of adsorbent lifetime and stability by improving their adsorption capacity, selectivity, and regenerability.

Oxidative desulfurization, as an emerging desulfurization technology, offers unique advantages in oxidizing aromatic sulfur compounds. ODS can be performed under relatively mild conditions and efficiently remove sulfur compounds. However, the industrial application of ODS still faces challenges such as poor catalyst selectivity and harsh reaction conditions. Future research will need to focus on the multifunctionalization of catalysts, designing catalysts with optimized structures and functional groups to enhance their oxidation activity and selectivity. Specifically, increasing oxidation efficiency and minimizing by-product formation for hard-to-oxidize aromatic sulfur compounds (such as dibenzothiophene) will be a research hotspot. Additionally, the development of new oxidants, such as supercritical carbon dioxide-based or other green solvent-based oxidants, is expected to offer more options for oxidative desulfurization technology.

8.2. Outlook

Future research in diesel desulfurization technology will not only focus on desulfurization efficiency and cost-effectiveness but will also promote the development of multi-process coupling and sulfur compound resource utilization to achieve more efficient, green, and sustainable molecular refining. With increasingly stringent environmental regulations, traditional desulfurization technologies are facing technical and economic challenges, making the integration and innovation of multiple technologies even more important.

8.2.1 Multi-process coupling. Traditional desulfurization technologies often work independently, but in practical industrial applications, multi-process coupling strategies can significantly enhance the overall effectiveness of desulfurization. For instance, combining ODS with adsorption desulfurization or hydrogen desulfurization (HDS) with biodesulfurization can leverage the advantages of each technology and overcome the limitations of single methods. ODS can efficiently convert sulfur compounds, while adsorption desulfurization can further remove by-products from the oxidation reaction, improving both the desulfurization efficiency and selectivity. Additionally, coupling extraction desulfurization with catalytic desulfurization not only accelerates the desulfurization rate but also allows the effective recovery of organic chemicals generated during the desulfurization process, reducing resource waste. Future research should focus on designing multi-technique combined reactors, optimizing the process flow by strategically integrating different desulfurization techniques, and reducing energy consumption while improving desulfurization efficiency. For example, combining membrane separation technology with catalyst-directed regulation within reactors during the process can further enhance the selectivity and efficiency of desulfurization.

8.2.2 Sulfur compound resource utilization. As sulfur compounds continue to impact the environment and catalysts, traditional desulfurization technologies primarily focus on removing sulfur compounds. However, sulfur compound resource utilization (i.e., converting sulfur compounds into valuable by-products) is gradually becoming a research hotspot. Sulfur compounds themselves hold potential economic value in the chemical, pharmaceutical, and agricultural industries. By catalytic or biological conversion technologies, sulfur compounds can be transformed into useful chemicals such as sulfuric acid, hydrogen sulfide, and sulfur-containing alcohols, further enhancing the economic efficiency of the desulfurization process. For example, hydrogen sulfide conversion technologies (such as efficient catalysts or microbial catalysts) can convert hydrogen sulfide into sulfur compounds or other organic sulfur compounds, which can be used for the synthesis of important chemicals like sulfuric acid. Additionally, the reuse of sulfur-containing polymers or the regeneration of catalysts may also become important future research directions to improve resource utilization efficiency and reduce environmental pollution.

8.2.3 Achieving molecular refining. The combination of multi-process coupling and sulfur compound resource utilization ultimately contributes to the goal of molecular refining. Traditional refining processes rely on large-scale, low-precision chemical transformations, whereas molecular refining focuses on optimizing the conversion of every molecular component of crude oil through fine-tuned processes. By precisely identifying, separating, and converting different types of sulfur compounds, the value of each molecular component in crude oil can be maximized, particularly by converting sulfur compounds into valuable by-products, achieving more efficient and greener resource utilization. Molecular refining can further reduce sulfur content and improve the quality of petrochemical products through efficient catalysts and process design. For example, by using catalytic cracking, hydrogen treatment, and molecular sieve separation techniques, each component can be selectively converted to produce high-quality chemicals, fuels, and feedstocks that meet various demands. In the future, molecular refining will focus more on precise process control, optimizing each reaction and conversion step to achieve the best outcome, thereby improving overall production efficiency and reducing raw material consumption.

8.2.4 Intelligent and efficient refining process control. With the development of modern information technology and artificial intelligence (AI), the future diesel desulfurization process will gradually shift toward intelligence and automation. By continuously monitoring reaction conditions, catalyst status, reactant and product quality, and other parameters, combined with big data analysis and machine learning algorithms, it will be possible to precisely control every step of the desulfurization process, optimizing process conditions, and reducing energy consumption and waste generation. Particularly in multi-process coupling and molecular refining, intelligent control will play a crucial role in enhancing production flexibility and sustainability.

The future innovations in diesel desulfurization technology will not be limited to single desulfurization methods but will increasingly integrate multi-process coupling and resource utilization technologies into the overall refining process. This approach aims to achieve more efficient, economical, and environmentally friendly goals. Such advancements will not only drive the upgrading of traditional petroleum refining technologies but also pave the way for greener and more sustainable development in the petrochemical industry. This evolution will enhance both the efficiency of desulfurization processes and the utilization of by-products, contributing to a more sustainable refining ecosystem. By combining different technologies, such as oxidation desulfurization, adsorption, bio-desulfurization, and advanced catalytic methods, the industry can move toward minimizing waste, reducing energy consumption, and improving overall environmental performance.

Author contributions

Peiwen Wu: conceptualization; resources; data curation; funding acquisition; visualization; writing – original draft; writing – review & editing. Shaojie Ma: resources; data curation; writing – review & editing. Wenshuai Zhu: conceptualization; resources; funding acquisition; project administration; writing – review & editing.

Conflicts of interest

The authors declare no conflict of interest.

Data availability

No primary research results, software or code have been included and no new data were generated or analysed as part of this review.

Acknowledgements

All authors appreciate the financial support from the National Key R&D Program of China (no. 2022YFA1504403 and 2022YFA1504404), National Science Foundation for Distinguished Young Scholars (no. 22425808), National Natural Science Foundation of China (no. 22178154), and Natural Science Foundation of Jiangsu Province (no. BK20230068).

References

1. N. Abas, A. Kalair and N. Khan, Futures, 2015, 69, 31–49.
2. İ. Yıldız, in Comprehensive Energy Systems, ed. I. Dincer, Elsevier, Oxford, 2018, pp. 521–567,  DOI:10.1016/B978-0-12-809597-3.00111-5.
3. A. N. Tutuncu, Fossil fuels, Oxford University Press, 2020.
4. Y. Zhang, E. Xing, W. Han, P. Yang, S. Zhang, S. Liu, D. Cao and M. Li, Engineering, 2024, 43, 99–114.
5. R. Diercks, J. D. Arndt, S. Freyer, R. Geier, O. Machhammer, J. Schwartze and M. Volland, Chem. Eng. Technol., 2008, 31, 631–637.
6. G. Lopez, D. Keiner, M. Fasihi, T. Koiranen and C. Breyer, Energy Environ. Sci., 2023, 16, 2879–2909.
7. P. J. Megía, A. J. Vizcaíno, J. A. Calles and A. Carrero, Energy Fuels, 2021, 35, 16403–16415.
8. J. Tickner, K. Geiser and S. Baima, Environment, 2021, 63, 4–15.
9. Global Energy & CO2 Status Report 2017, International Energy Agency, 2018.
10. M. Farghali, A. I. Osman, I. M. A. Mohamed, Z. Chen, L. Chen, I. Ihara, P. S. Yap and D. W. Rooney, Environ. Chem. Lett., 2023, 21, 1–37.
11. M. K. G. Deshmukh, M. Sameeroddin, D. Abdul and M. Abdul Sattar, Mater. Today: Proc., 2023, 80, 1756–1759.
12. H. Yan, S. J. Acquah, J. Zhang, G. Wang, C. Zhang and R. O. Darko, Int. J. Agric. Biol. Eng., 2021, 14, 1–8.
13. H. Ritchie, P. Rosado and M. Roser, Energy Production and Consumption, 2020, https://ourworldindata.org/energy-production-consumption.
14. A. Ferrari, J. Pet. Technol., 2024, 76, 10–11.
15. Y. Wang, Z. Zhang and M. Xu, Energy, 2023, 284, 128656.
16. I. A. Lakhiar, H. Yan, J. Zhang, G. Wang, S. Deng, R. Bao, C. Zhang, T. N. Syed, B. Wang, R. Zhou and X. Wang, Agronomy, 2024, 14, 548.
17. Y. Yuan, L. Zhou, H. Robatjazi, J. L. Bao, J. Zhou, A. Bayles, L. Yuan, M. Lou, M. Lou and S. Khatiwada, Science, 2022, 378, 889–893.
18. Z. Shi, X. Zhang, X. Lin, G. Liu, C. Ling, S. Xi, B. Chen, Y. Ge, C. Tan and Z. Lai, Nature, 2023, 621, 300–305.
19. Z. Zhuo, E. Du, N. Zhang, C. P. Nielsen, X. Lu, J. Xiao, J. Wu and C. Kang, Nat. Commun., 2022, 13, 3172.
20. J. Hong, C. Xu, B. Deng, Y. Gao, X. Zhu, X. Zhang and Y. Zhang, Adv. Sci., 2022, 9, e2103926.
21. N. Zwane, H. Tazvinga, C. Botai, M. Murambadoro, J. Botai, J. De Wit, B. Mabasa, S. Daniel and T. Mabhaudhi, Energies, 2022, 15, 5520.
22. R. Verduci, V. Romano, G. Brunetti, N. Yaghoobi Nia, A. Di Carlo, G. D'Angelo and C. Ciminelli, Adv. Energy Mater., 2022, 12, 2200125.
23. B. E. Layton, Int. J. Green Energy, 2008, 5, 438–455.
24. T.-Z. Ang, M. Salem, M. Kamarol, H. S. Das, M. A. Nazari and N. Prabaharan, Energy Strategy Rev., 2022, 43, 100939.
25. D. Maradin, Int. J. Energy Econ. Policy, 2021, 11, 176–183.
26. M. J. Hossain, S. I. Ahmed, M. U. Monir, H. Salman, K. Techato and S. Chowdhury, in Crises in Oil, Gas and Petrochemical Industries, Elsevier, 2023, pp. 393–418.
27. M. Larsen and J. P. Tilsted, Energy Res. Soc. Sci., 2024, 118, 103750.
28. X. Zhou, Z. Cheng, K. Ren, Y. Zhai, T. Zhang, X. Shen, Y. Bai, Y. Jia and J. Hong, Sustainable Prod. Consumption, 2022, 34, 105–113.
29. Q. Lyu and Z. Chai, Bull. Chin. Acad. Sci., 2022, 37, 541–548.
30. N. Weiland, R. Dennis, R. Ames, S. Lawson and P. Strakey, in Fundamentals and Applications of Supercritical Carbon Dioxide (sCO₂) Based Power Cycles, Elsevier, 2017, pp. 293–338.
31. M. J. Nazir, G. Li, M. M. Nazir, F. Zulfiqar, K. H. M. Siddique, B. Iqbal and D. Du, Soil Tillage Res., 2024, 237, 105959.
32. B. Muhammad, M. K. Khan, M. I. Khan and S. Khan, Environ. Sci. Pollut. Res., 2021, 28, 21789–21798.
33. M. Z. Jacobson, A.-K. von Krauland, S. J. Coughlin, E. Dukas, A. J. Nelson, F. C. Palmer and K. R. Rasmussen, Energy Environ. Sci., 2022, 15, 3343–3359.
34. W. Qu, Y. Gao, S. He, J. Zhang, K. Peng, H. Wu, R. Wang and H. Hong, Energy Convers. Manage., 2023, 276, 116574.
35. P. Gao, L. Zhong, B. Han, M. He and Y. Sun, Angew. Chem., Int. Ed., 2022, 61, e202210095.
36. J. Artz, T. E. Muller, K. Thenert, J. Kleinekorte, R. Meys, A. Sternberg, A. Bardow and W. Leitner, Chem. Rev., 2018, 118, 434–504.
37. C. F. Shih, T. Zhang, J. Li and C. Bai, Joule, 2018, 2, 1925–1949.
38. K. Ukoba, K. O. Olatunji, E. Adeoye, T.-C. Jen and D. M. Madyira, Energy Environ., 2024, 35, 3833–3879.
39. B. Li and W. Tan, Sol. Energy, 2023, 262, 111869.
40. C. J. Ejiyi, D. Cai, D. Thomas, S. Obiora, E. Osei-Mensah, C. Acen, F. O. Eze, F. Sam, Q. Zhang and O. O. Bamisile, J. Big Data, 2025, 12, 169.
41. S. Dawn, A. Ramakrishna, M. Ramesh, S. S. Das, K. D. Rao, M. M. Islam and T. Selim Ustun, Adv. Energy Sustainability Res., 2024, 5, 2400088.
42. A. Haruna, Z. M. A. Merican and S. G. Musa, J. Ind. Eng. Chem., 2022, 112, 20–36.
43. Q. Shi and J. Wu, Energy Fuels, 2021, 35, 14445–14461.
44. A. Vetere, D. Pröfrock and W. Schrader, Angew. Chem., Int. Ed., 2017, 56, 10933–10937.
45. Y. A. Abd Al-Khodor and T. M. Albayati, Process Saf. Environ. Prot., 2020, 136, 334–342.
46. M. I. S. de Mello, E. V. Sobrinho, V. Teixeira da Silva and S. B. C. Pergher, Ind. Eng. Chem. Res., 2018, 57, 15663–15669.
47. H. Zheng, Y. Li, W. He, Y. Han and D. Xu, Fuel, 2021, 289, 119846.
48. V. Perraud, J. R. Horne, A. S. Martinez, J. Kalinowski, S. Meinardi, M. L. Dawson, L. M. Wingen, D. Dabdub, D. R. Blake and R. B. Gerber, Proc. Natl. Acad. Sci. U. S. A., 2015, 112, 13514–13519.
49. M. Sun, R. Abazari, J. Chen, C. M. Hussain, Y. Zhou and A. M. Kirillov, ACS Appl. Mater. Interfaces, 2023, 15, 52581–52592.
50. E. D. Salonikidou, D. A. Giannakoudakis, J. Prekodravac, S. Kiartzis, E. Nanaki, M. Barczak, F. Kyriakos, P. Baltzopoulou, G. Karagiannakis and T. J. Bandosz, Carbon, 2025, 120347.
51. E.-L. S. Hinckley, J. T. Crawford, H. Fakhraei and C. T. Driscoll, Nat. Geosci., 2020, 13, 597–604.
52. G. Ren, M. Cui, H. Yu, X. Fan, Z. Zhu, H. Zhang, Z. Dai, J. Sun, B. Yang and D. Du, J. Soil Sci. Plant Nutr., 2024, 24, 2402–2412.
53. A. Vetere, D. Profrock and W. Schrader, Energy Fuels, 2021, 35, 8723–8732.
54. J. Zhai, G. Yu, J. Zhang, S. Shi, Y. Yuan, S. Jiang, C. Xing, B. Cai, Y. Zeng, Y. Wang, A. Zhang, Y. Zhang, T.-M. Fu, L. Zhu, H. Shen, J. Ye, C. Wang, S. Tao, M. Li, Y. Zhang and X. Yang, Environ. Sci. Technol., 2023, 57, 12341–12350.
55. T. Yuan, H. Song, R. Wood, C. Wang, L. Oreopoulos, S. E. Platnick, S. von Hippel, K. Meyer, S. Light and E. Wilcox, Sci. Adv., 2022, 8, eabn7988.
56. I. Shafiq, S. Shafique, P. Akhter, W. Yang and M. Hussain, Catal. Rev., 2022, 64, 1–86.
57. R. Cordtz, J. Schramm, A. Andreasen, S. S. Eskildsen and S. Mayer, Energy Fuels, 2013, 27, 1652–1660.
58. A. Gómez, P. Fernández-Yáñez, J. A. Soriano, L. Sánchez-Rodríguez, C. Mata, R. García-Contreras, O. Armas and M. D. Cárdenas, Fuel, 2021, 289, 119836.
59. D. V. Nelyubov, M. I. Fakhrutdinov, A. A. Sarkisyan, E. A. Sharin, M. A. Ershov, U. A. Makhova, A. E. Makhmudova, N. A. Klimov, M. Y. Rogova and V. D. Savelenko, J. Mar. Sci. Eng., 2023, 11, 1460.
60. A. V. Akopyan, E. A. Eseva, R. A. Mnatsakanyan, D. A. Davtyan, M. O. Lukashov, I. S. Levin, A. A. Sadovnikov, A. V. Anisimov, A. M. Terzyan, A. M. Agoyan and E. Karakhanov, Chem. Eng. J., 2023, 464, 142641.
61. L. Sun, Z. Zhu, T. Su, W. Liao, D. Hao, Y. Chen, Y. Zhao, W. Ren, H. Ge and H. Lü, Appl. Catal., B, 2019, 255, 117747.
62. X. Liao, Y. Huang, Y. Zhou, H. Liu, Y. Cai, S. Lu and Y. Yao, Appl. Surf. Sci., 2019, 484, 917–924.
63. Z. Zhu, H. Lin, M. Chi, X. Gao, Y. Feng, K. Yang and H. Lü, Fuel, 2022, 308, 122070.
64. N. S. Nesterov, V. P. Pakharukova, A. A. Philippov, I. P. Prosvirin, A. S. Shalygin and O. N. Martyanov, Molecules, 2023, 28, 7041.
65. A. A. Turkin, F. Rammal, S. Eyley, W. Thielemans, E. V. Makshina and B. F. Sels, ACS Sustainable Chem. Eng., 2023, 11, 8968–8977.
66. M. Mohamad Ali, V. Dutta, H. Guy, S. Durécu, E. Abi-Aad, C. Ciotonea and C. Gennequin, Sustainable Energy Fuels, 2025, 9, 3424–3457.
67. X. Zhang, Y. Zhou, G. Li, L. Zhang, C. Yin, Y. Yang, H. Wang, F. Feng, L. Wei and Q. Zhang, Appl. Catal., B, 2022, 315, 121566.
68. H. Chen, Z. Ling, Y. Wang, H. Dai, Z. Zhao and Q. Yang, ChemCatChem, 2025, 17, e202401506.
69. S. S. Bello, C. Wang, M. Zhang, H. Gao, Z. Han, L. Shi, F. Su and G. Xu, Energy Fuels, 2021, 35, 10998–11016.
70. M. Vakili, P. Koutnik and J. Kohout, Sustainability, 2024, 16, 1661.
71. M. C. Bruzzoniti, R. M. De Carlo, C. Sarzanini, R. Maina and V. Tumiatti, Ind. Eng. Chem. Res., 2014, 53, 8675–8684.
72. S. Narayanan, J. Kasture and M. S. Dangate, Biomass and Solar–Powered Sustainable Digital Cities, 2024, pp. 197–207.
73. X. Hou, Y. Wang, L. Dai, Y. Yang, J. Du, Y. Wang and H. Wan, Eng. Failure Anal., 2023, 147, 107151.
74. A. S. Morshedy, S. I. El-Hout, A. M. Salem, A. Helal, C. Chen, A. Alsalme and A. Barhoum, J. Cleaner Prod., 2024, 449, 141713.
75. D. Yu, Z. Li, J. Li, J. He, B. Li and Y. Wang, J. Hazard. Mater., 2024, 462, 132618.
76. M. Flores-Granobles and M. Saeys, Energy Environ. Sci., 2020, 13, 1923–1932.
77. D. D. Pham, T. M. Nguyen, T. H. Ho, Q. V. Le and D. L. Nguyen, Fuel, 2024, 372, 132082.
78. D. D. Pham, T. M. Nguyen, T. H. Ho, Q. V. Le and D. L. T. Nguyen, Fuel, 2024, 372, 132082.
79. E. Wang, F. Yang, M. Song, G. Chen, Q. Zhang, F. Wang, L. Bing, G. Wang and D. Han, Fuel Process. Technol., 2022, 235, 107386.
80. T.-Y. Cui, A. Rajendran, H.-X. Fan, J. Feng and W.-Y. Li, Ind. Eng. Chem. Res., 2021, 60, 3295–3323.
81. T. Chen, Y. Wu, D. Xing and R. Duan, Horticulturae, 2022, 8, 185.
82. P. S. Kulkarni and C. A. M. Afonso, Green Chem., 2010, 12, 1139–1149.
83. J. P. M. Sousa, P. Ferreira, R. P. P. Neves, M. J. Ramos and P. A. Fernandes, Green Chem., 2020, 22, 7604–7621.
84. R. Abro, A. A. Abdeltawab, S. S. Al-Deyab, G. R. Yu, A. B. Qazi, S. R. Gao and X. C. Chen, RSC Adv., 2014, 4, 35302–35317.
85. D. T. Tran, J. M. Palomino and S. R. J. Oliver, RSC Adv., 2018, 8, 7301–7314.
86. S. N. Wei, H. J. He, Y. Cheng, C. P. Yang, G. M. Zeng and L. Qiu, RSC Adv., 2016, 6, 103253–103269.
87. X. Qin, L. Ye, A. Murad, J. Liu, Q. Ying, J. Long, W. Yu, J. Xie, L. Hou and X. Pu, Fuel, 2022, 319, 123567.
88. P. Aleksandrov, S. Reshetnikov, G. Bukhtiyarova and A. Noskov, Chem. Eng. J., 2022, 446, 137059.
89. M. T. Timko, E. Schmois, P. Patwardhan, Y. Kida, C. A. Class, W. H. Green, R. K. Nelson and C. M. Reddy, Energy Fuels, 2014, 28, 2977–2983.
90. A. R. Othman, S. M. Zain and K. H. Low, J. Near Infrared Spectrosc., 2022, 30, 237–245.
91. H. Li, X. Chen, J. Wu, Y. Zhang, X. Liu, Q. Shi, S. Zhao, C. Xu and C. S. Hsu, Energy Fuels, 2019, 33, 1797–1802.
92. O. Konur, Petrodiesel Fuels, 2021, 1035–1053.
93. A. A. Grin'ko, R. S. Min, T. A. Sagachenko and A. K. Golovko, Pet. Chem., 2012, 52, 221–227.
94. M. Prabhu, J. Louwen, E. Vogt and I. Groot, Catal. Lett., 2025, 155, 1–10.
95. L. Gao, W. He, W. Zhai, M. Xin, W. Wang, Y. Xiang, H. Yuan, S. Wang, F. Liu and G. Xu, Fuel, 2025, 393, 135052.
96. R. Hua, Y. Li, W. Liu, J. Zheng, H. Wei, J. Wang, X. Lu, H. Kong and G. Xu, J. Chromatogr. A, 2003, 1019, 101–109.
97. X. Qin, L. Ye, A. Murad, J. Liu, Q. Ying, J. Long, W. Yu, J. Xie, L. Hou, X. Pu, X. Han, J. Zhao, H. Sun and H. Ling, Fuel, 2022, 319, 123567.
98. M. A. Ershov, V. D. Savelenko, A. E. Makhmudova, E. S. Rekhletskaya, U. A. Makhova, V. M. Kapustin, D. Y. Mukhina and T. M. Abdellatief, J. Mar. Sci. Eng., 2022, 10, 1828.
99. T. Yarmola, P. Topilnytskyy and V. Romanchuk, Chem. Chem. Technol., 2023, 17, 195–202.
100. A. Tirado, G. Félix, F. Trejo, M. A. Varfolomeev, C. Yuan, D. K. Nurgaliev, V. Sámano and J. Ancheyta, Catalytic In–Situ Upgrading of Heavy and Extra–Heavy Crude Oils, 2023, pp. 1–38.
101. J. Zhu, J. Han, G. Tao and X. Zhang, Pet. Sci. Technol., 2023, 41, 2161–2172.
102. X. Qin, L. Ye, L. Hou, T. Wang, M. Ma, X. Pu, X. Han, J. Liu, B. Luan and P. Liu, Chem. Eng. J., 2023, 466, 143078.
103. J. Zhao, L. Dai, W. Wang, T. Liu, L. Ren, L. Zhang, W. Han and D. Li, Fuel Process. Technol., 2021, 224, 107025.
104. P. Wu, C. Deng, P. Liu, F. Liu, L. Chen, B. Wang, W. Zhu and C. Xu, Chem. Eng. J., 2023, 474, 145850.
105. S. Wang, Y. Li, Y. Fu, X. Li, C. Tao, W. Zhang and H. Ma, Adv. Funct. Mater., 2024, 34, 2312341.
106. A. A. Toutov, M. Salata, A. Fedorov, Y.-F. Yang, Y. Liang, R. Cariou, K. N. Betz, E. P. A. Couzijn, J. W. Shabaker, K. N. Houk and R. H. Grubbs, Nat. Energy, 2017, 2, 17008.
107. C. Lamonier, Nat. Energy, 2017, 2, 17019.
108. I. Shafiq, S. Shafique, P. Akhter, W. Yang and M. Hussain, Catal. Rev. Sci. Eng., 2022, 64, 1–86.
109. E. Wang, F. Yang, M. Song, G. Chen, Q. Zhang, F. Wang, L. Bing, G. Wang and D. Han, Fuel Process. Technol., 2022, 235, 107356.
110. Y. S. Al-Degs, A. H. El-Sheikh, R. Z. Al Bakain, A. P. Newman and M. A. Al-Ghouti, Energy Technol., 2016, 4, 679–699.
111. V. C. Srivastava, RSC Adv., 2012, 2, 759–783.
112. A. Tanimu and K. Alhooshani, Energy Fuels, 2019, 33, 2810–2838.
113. B. Wang, Z. Chen, T. Jiang, J. Yu, H. Yang, A. Duan and C. Xu, AIChE J., 2022, 68, e17577.
114. Y. Yin, W. Chen, G. Wu, F. Xin, K. Qin, Y. Lu, L. Zhang and M. Li, AIChE J., 2021, 67, e17188.
115. A. Vutolkina, I. Baygildin, A. Glotov, K. Cherednichenko, A. Maksimov and E. Karakhanov, Appl. Catal., B, 2021, 282, 119616.
116. Y. Zou, C. Xiao, X. Yang, L. Qi, H. Wu, J. Liu, Z. Liu, A. Wang, A. Duan and C. Xu, Chem. Eng. J., 2025, 507, 160479.
117. L. Qi, P. Zheng, Z. Zhao, A. Duan, C. Xu and X. Wang, J. Catal., 2022, 408, 279–293.
118. A. V. Vutolkina, I. G. Baygildin, A. P. Glotov, K. A. Cherednichenko, A. L. Maksimov and E. A. Karakhanov, Appl. Catal., B, 2021, 282, 119616.
119. J. Díaz de León, C. Ramesh Kumar, J. Antúnez-García and S. Fuentes-Moyado, Catalysts, 2019, 9, 87.
120. X. Fang, R. Guo and C. Yang, Chin. J. Catal., 2013, 34, 130–139.
121. X. Weng, L. Cao, G. Zhang, F. Chen, L. Zhao, Y. Zhang, J. Gao and C. Xu, Ind. Eng. Chem. Res., 2020, 59, 21261–21274.
122. S. Zhou, Y. Pan, Y. Wang, H. Cheng, P. Wu, H. Li, Y. Huang, M. Hua, J. Liu and W. Zhu, Inorg. Chem., 2024, 63, 16928–16939.
123. B. Wang, Z. Chen, T. Jiang, J. Yu, H. Yang, A. Duan and C. Xu, AIChE J., 2022, 68, e17577.
124. L. Oliviero, A. Travert, E. D. Garcia, J. Chen and F. Maugé, J. Catal., 2021, 403, 87–97.
125. A. Gaur, M. Stehle, M.-A. Serrer, M. Z. Stummann, C. La Fontaine, V. Briois, J.-D. Grunwaldt and M. Høj, ACS Catal., 2021, 12, 633–647.
126. Z. Fanani, W. R. Asri, B. Dwiyanti, N. Rinaldi, R. Maryana, M. Al Muttaqii, F. Riyanti, A. Mara, R. Zainul and H. Hasanudin, Period. Polytech., Chem. Eng., 2024, 68, 552–560.
127. J. Liang, M. Wu, J. Wang, P. Wei, B. Sun, Y. Lu, D. Sun, Y. Liu and C. Liu, Catal. Sci. Technol., 2018, 8, 6330–6345.
128. X. Wang, C. Xiao, M. H. Alabsi, P. Zheng, Z. Cao, J. Mei, Y. Shi, A. Duan, D. Gao, K.-W. Huang and C. Xu, Green Energy Environ., 2022, 7, 324–333.
129. Y. Sun and R. Prins, Angew. Chem., Int. Ed., 2008, 47, 8478–8481.
130. J. N. Díaz de León, J. Antunes-García, G. Alonso-Nuñez, T. A. Zepeda, D. H. Galvan, J. A. de los Reyes and S. Fuentes, Appl. Catal., B, 2018, 238, 480–490.
131. M. Umar, I. Abdulazeez, A. Tanimu, S. A. Ganiyu and K. Alhooshani, J. Environ. Chem. Eng., 2021, 9, 106738.
132. Z. Rahimi-Ahar, A. Sohrabi and A. Ghareghashi, Gas Process. J., 2022, 10, 25–36.
133. K. Chen, Y. Cao, H. Zhu, H. Ge, T. Yang, Z. Zhou and Z. Cheng, Processes, 2022, 10, 1822.
134. S. Zhang, Q. Zhang and Z. C. Zhang, Ind. Eng. Chem. Res., 2004, 43, 614–622.
135. X. Meng, Y. Zhao, Q. Wang and L. Duan, Energy Environ. Focus, 2014, 3, 53–59.
136. A. A. Botin, R. E. Boldushevskii, A. V. Mozhaev, M. Ghambarian, M. Balar, M. Ghashghaee and P. A. Nikulshin, Appl. Catal., B, 2023, 337, 122946.
137. M. Zheng, L. Zhao, L. Cao, Y. Zhang, J. Gao and C. Xu, Fuel, 2021, 300, 120941.
138. T. Jiang, F. Yu, Q.-M. Jiang, Y.-X. Song, Q.-F. Tan, W. Su, X. Yang, Z.-T. Chen and C.-M. Xu, Pet. Sci., 2025, 22, 1360–1369.
139. M. S. Nazir, S. Ahmad, Z. Tahir, Z. Ali, M. N. Akhtar, K. Azam and M. Abdullah, Curr. Anal. Chem., 2021, 17, 815–830.
140. R. Abro, N. Kiran, S. Ahmed, A. Muhammad, A. S. Jatoi, S. A. Mazari, U. Salma and N. V. Plechkova, J. Environ. Chem. Eng., 2022, 10, 107369.
141. L. Dong, X. Dai, C. Peng, C. Yang, G. Li, S. Chen, G. Miao and J. Xiao, AIChE J., 2022, 68, e17493.
142. F. Liu, J. Yu, A. B. Qazi, L. Zhang and X. Liu, Environ. Sci. Technol., 2021, 55, 1419–1435.
143. S. Chen, M. Zang, L. Li, J. Chen, Q. Liu, X. Feng, S. Sun, C. Zang and C. Zhao, Biochem. Eng. J., 2021, 174, 108094.
144. S. Guchhait, D. Biswas, P. Bhattacharya and R. Chowdhury, Chem. Eng. J., 2005, 112, 145–151.
145. C. Song and X. Ma, Appl. Catal., B, 2003, 41, 207–238.
146. H. N. Nassar, H. R. Ali and N. S. El-Gendy, Fuel, 2021, 294, 120534.
147. P. Ferreira, R. P. P. Neves, F. P. Miranda, A. V. Cunha, R. W. A. Havenith, M. J. Ramos and P. A. Fernandes, J. Chem. Inf. Model., 2024, 64, 4218–4230.
148. M. Stylianou, I. Vyrides and A. Agapiou, J. Chromatogr. B, 2021, 1171, 122602.
149. M. Stylianou, C. G. Samanides, I. Vyrides and A. Agapiou, Energy, 2023, 282, 128950.
150. Z. Xiong, T. Shi, W. Zhang, Y. Kong, L. Yuan and R. Gao, LWT, 2021, 145, 111303.
151. M. Kalita, P. A. Sangannavar, M. Chutia, D. K. Jha, K. Sathyanarayana, J. S. Kumar and G. Subrahmanyam, in Microbial Resource Technologies for Sustainable Development, Elsevier, 2022, pp. 333–351.
152. A. Ahmad, M. A. Zamzami, V. Ahmad, S. Al-Thawadi, M. S. Akhtar and M. J. Khan, Fermentation, 2023, 9, 211.
153. A. Bagchi and P. Srivastava, Front. Bioeng. Biotechnol., 2024, 12, 1482270.
154. Z. Gou, Sci. China, Ser. B: Chem., 2002, 45, 521–531.
155. T. P. Silva, S. M. Paixão and L. Alves, Green Chem., 2023, 25, 6416–6431.
156. P. Das, Umesh, L. Barbora and V. S. Moholkar, Prep. Biochem. Biotechnol., 2025, 55, 719–735.
157. M. Yang, J. Yun, H. Zhang, T. A. Magocha, H. Zabed, Y. Xue, E. Fokum, W. Sun and X. Qi, Food Technol. Biotechnol., 2018, 56, 3–15.
158. P. Wang, Dissertation/Thesis, University Microfilms International, 1993.
159. M. S. Borhani and Z. Etemadifar, Microbiol., Metab. Biotechnol., 2019, 2, 89–99.
160. J. Wang, R. R. Butler III, F. Wu, J.-F. Pombert, J. J. Kilbane and B. C. Stark, PLoS One, 2017, 12, e0168833.
161. G. Martín-Cabello, L. Terrón-González, M. Ferrer and E. Santero, Environ. Microbiol., 2020, 22, 91–106.
162. S. Parveen, N. Akhtar, M. A. Ghauri and K. Akhtar, Crit. Rev. Microbiol., 2020, 46, 300–320.
163. A. M. M. Mawad, M. Hassanein, E. S. Aldaby and N. Yousef, J. Appl. Microbiol., 2021, 130, 1181–1191.
164. G. Mohebali and A. S. Ball, Int. Biodeterior. Biodegrad., 2016, 110, 163–180.
165. C. R. Cattaneo, R. Munoz, G. V. Korshin, V. Naddeo, V. Belgiorno and T. Zarra, Sci. Total Environ., 2023, 893, 164689.
166. Z. Chen, Y. Chen, X. Hao, Q. Wen, X. Wang, A. Liu, D. Gao, K. Hu, N. A. Samak and M. Yang, Chem. Eng. J., 2025, 160632.
167. Z. Chen, X. Hao, Q. Wen, Y. Jia, N. A. Samak, M. Yang and J. Xing, Chem. Eng. J., 2024, 496, 153683.
168. R. S. Malani, A. H. Batghare, J. B. Bhasarkar and V. S. Moholkar, Bioresour. Technol. Rep., 2021, 14, 100668.
169. D. Kumar Bal and J. B. Bhasarkar, in Ultrasound Technology for Fuel Processing, Bentham Science Publishers, 2023.
170. L. Sun, Y. Wang, M. Yue, X. Ding, X. Yu, J. Ge, W. Sun and L. Song, Foods, 2022, 11, 4078.
171. B. Saha, S. Vedachalam and A. K. Dalai, Fuel Process. Technol., 2021, 214, 106685.
172. S. A. Ganiyu and S. A. Lateef, Fuel, 2021, 294, 120273.
173. N. R. Yeole, Environ. Prog. Sustainable Energy, 2023, 42, e13960.
174. R. A. Omar and N. Verma, Ind. Eng. Chem. Res., 2022, 61, 8595–8606.
175. J. D. C. Hossi, D. Nkazi, J. van der Merwe, A. S. Joel and K. G. Harding, Chem. Eng. Technol., 2024, 47, 905–913.
176. B. Yoosuk, B. Chareonteraboon, N. Hinchiranan and P. Prasassarakich, Ind. Eng. Chem. Res., 2023, 62, 7080–7092.
177. J. Xu, B. Zhang, Y. Lu, L. Wang, W. Tao, X. Teng, W. Ning and Z. Zhang, J. Hazard. Mater., 2022, 421, 126680.
178. N. M. Khalse and M. De, Sep. Purif. Technol., 2024, 331, 125693.
179. P. Zhang, X. Mao and B. Yang, Ind. Eng. Chem. Res., 2019, 58, 10589–10598.
180. J. Wu, J. Wang, T. Guan, G. Zhang, N. Wang and K. Li, Fuel, 2021, 292, 120251.
181. Y. Chai, W. Dai, G. Wu, N. Guan and L. Li, Acc. Chem. Res., 2021, 54, 2894–2904.
182. X. Wang, Y. Ma, Q. Wu, Y. Wen and F.-S. Xiao, Chem. Soc. Rev., 2022, 51, 2431–2443.
183. I. D. Boateng, X. M. Yang, H. Yin and W. Liu, Food Chem., 2024, 450, 139284.
184. S. Ma, M. Wang, T. You and K. Wang, J. Agric. Food Chem., 2019, 67, 8035–8044.
185. W. Zhou, C. Li, C. Sun and X. Yang, Food Chem., 2016, 192, 351–357.
186. M. Sarker, H. J. An, D. K. Yoo and S. H. Jhung, Chem. Eng. J., 2018, 338, 107–116.
187. R. Menzel, D. Iruretagoyena, Y. Wang, S. M. Bawaked, M. Mokhtar, S. A. Al-Thabaiti, S. N. Basahel and M. S. Shaffer, Fuel, 2016, 181, 531–536.
188. E. Svinterikos, I. Zuburtikudis and M. Al-Marzouqi, J. Nanotechnol., 2019, 2019, 2809867.
189. J. Rong, Z. Zhou, Y. Wang, J. Han, C. Li, W. Zhang and L. Ni, Food Technol. Biotechnol., 2019, 57, 260–271.
190. Y. Shi, X. Zhang, L. Wang and G. Liu, AIChE J., 2014, 60, 2747–2751.
191. T. Zhang, B. Zhao, Q. Chen, X. Peng, D. Yang and F. Qiu, Appl. Biol. Chem., 2019, 62, 12.
192. X. Li, Z. Liu, S. Shao, Y. Yu, W. Wang, T. Sun and S. Wu, Ind. Crops Prod., 2025, 228, 120891.
193. J. Yuan, Y. Zhu, J. Wang, L. Gan, M. He, T. Zhang, P. Li and F. Qiu, Food Bioprod. Process., 2021, 126, 293–304.
194. Y. Zhang, Y. Yang, H. Han, M. Yang, L. Wang, Y. Zhang, Z. Jiang and C. Li, Appl. Catal., B, 2012, 119, 13–19.
195. L. Yin, J. Xu, B. Zhang, L. Wang, W. Tao, X. Teng, W. Ning and Z. Zhang, Chem. Eng. J., 2022, 428, 132581.
196. Y. Yin, Z.-H. Wen, X.-Q. Liu, A.-H. Yuan and L. Shi, Adsorpt. Sci. Technol., 2018, 36, 953–966.
197. X. Yang, C. Cao, K. J. Klabunde, K. L. Hohn and L. E. Erickson, Ind. Eng. Chem. Res., 2007, 46, 4819–4823.
198. T. Nkomzwayo, L. Mguni, D. Hildebrandt and Y. Yao, Chem. Eng. J., 2022, 444, 136484.
199. J. Xiao, S. Sitamraju, Y. Chen, S. Watanabe, M. Fujii, M. Janik and C. Song, AIChE J., 2015, 61, 631–639.
200. N. Liang, B. Shi, X. Hu, Y. Shi, T. Wang, X. Huang, Z. Li, X. Zhang, X. Zou and J. Shi, Food Chem., 2025, 472, 142891.
201. Z. Zhang, Y. Zhang, H. Jayan, S. Gao, R. Zhou, N. Yosri, X. Zou and Z. Guo, Food Chem., 2024, 448, 139051.
202. L. Kang, Q. Liang, Q. Abdul, A. Rashid, X. Ren and H. Ma, Food Chem., 2023, 420, 136142.
203. I. Ahmed and S. H. Jhung, J. Hazard. Mater., 2016, 301, 259–276.
204. Y.-X. Li, W.-J. Jiang, P. Tan, X.-Q. Liu, D.-Y. Zhang and L.-B. Sun, J. Phys. Chem. C, 2015, 119, 21969–21977.
205. S. Kitagawa, Chem. Soc. Rev., 2014, 43, 5415–5418.
206. I. Ahmed and S. H. Jhung, J. Hazard. Mater., 2016, 301, 259–276.
207. T. Zhou, S. Wang, C. Zhang, Y. Yao, Y. Chen, S. Lu and X. Liao, Fuel, 2023, 339, 127396.
208. L. L. Mguni, Y. Yao, J. Ren, X. Liu and D. Hildebrandt, J. Cleaner Prod., 2021, 323, 129196.
209. H. Song, X. Li, B. Jiang, M. Gong and T. Hao, Ind. Eng. Chem. Res., 2019, 58, 19586–19598.
210. Y. Guo, W. Xie, H. Li, J. Li, J. Hu and H. Liu, Sep. Purif. Technol., 2022, 285, 120287.
211. M. Ghassa, F. Khorashe, Z. Hajjar and S. Soltanali, Energy Fuels, 2023, 37, 6490–6502.
212. M. Saeed, A. Firdous, M. S. Zaman, F. Izhar, M. Riaz, S. Haider, M. Majeed and S. Tariq, J. Chin. Chem. Soc., 2023, 70, 789–824.
213. N. A. Khan and S. H. Jhung, J. Hazard. Mater., 2012, 237, 180–185.
214. H. Yang, L. Chen, C. Zhou, X. Yu, A. E. A. Yagoub and H. Ma, Food Chem., 2018, 245, 346–352.
215. X. Li, K. Chen, R. Guo and Z. Wei, Chem. Rev., 2023, 123, 10432–10467.
216. N. A. Khan, Z. Hasan and S. H. Jhung, Chemistry, 2014, 20, 376–380.
217. W. S. He, H. H. Wang, Z. M. Jing, D. D. Cui, J. Q. Zhu, Z. J. Li and H. L. Ma, J. Am. Oil Chem. Soc., 2018, 95, 89–100.
218. A. Khodadadi Dizaji, H. R. Mortaheb and B. Mokhtarani, J. Mater. Sci., 2016, 51, 10092–10103.
219. G. Yu, D. Jin, F. Zhang, S. Tian, Z. Zhou and Z. Ren, Chem. Eng. J., 2023, 453, 139935.
220. M. Shirani, A. Semnani, S. Habibollahi, H. Haddadi and M. Narimani, J. Porous Mater., 2016, 23, 701–712.
221. B.-L. Jiang, N. Jiang and Y.-X. Chang, Pet. Sci., 2020, 18, 295–306.
222. F. Subhan, S. Aslam, Z. Yan, L. Zhen, A. Ahmad, M. Naeem, J. Zeng, R. Ullah and U. J. Etim, Chem. Eng. J., 2017, 330, 372–382.
223. F. Subhan, S. Aslam, Z. Yan, L. Zhen, M. Ikram, R. Ullah, U. J. Etim and A. Ahmad, Chem. Eng. J., 2018, 339, 557–565.
224. L. Ma, Y. Zhong, L. Qu, X. Feng, Y. Zhang, H. Hou and G. Yang, Sep. Purif. Technol., 2025, 358, 130324.
225. S. Kumar, V. C. Srivastava and S. M. Nanoti, Sep. Purif. Rev., 2017, 46, 319–347.
226. N. H. Ko, J. S. Lee, E. S. Huh, H. Lee, K. D. Jung, H. S. Kim and M. Cheong, Energy Fuels, 2008, 22, 1687–1690.
227. Z. Li, D. Liu, Z. Men, L. Song, Y. Lv, P. Wu, B. Lou, Y. Zhang, N. Shi and Q. Chen, Green Chem., 2018, 20, 3112–3120.
228. J. Li, X.-J. Lei, X.-D. Tang, X.-P. Zhang, Z.-Y. Wang and S. Jiao, Energy Fuels, 2019, 33, 4079–4088.
229. J.-j. Li, H. Xiao, X.-d. Tang and M. Zhou, Energy Fuels, 2016, 30, 5411–5418.
230. C. Li, J. Zhang, Z. Li, J. Yin, Y. Cui, Y. Liu and G. Yang, Green Chem., 2016, 18, 3789–3795.
231. P. Makoś and G. Boczkaj, J. Mol. Liq., 2019, 296, 111916.
232. S. Gao, X. Chen, X. Xi, M. Abro, W. Afzal, R. Abro and G. Yu, ACS Sustainable Chem. Eng., 2019, 7, 5660–5668.
233. J. Yan, F. Luo, Y. Du, Q. Yi, X. Hao, H. Dong and L. Sun, Fluid Phase Equilib., 2023, 565, 113655.
234. E. Kianpour and S. Azizian, Fuel, 2014, 137, 36–40.
235. H. Cheng, Y. Cui, Z. Ge, R. Wang, Z. Qin, L. Chen and Z. Qi, Sep. Purif. Technol., 2021, 262, 118342.
236. D. Jha, P. Maheshwari, Y. Singh, M. B. Haider, R. Kumar and M. S. Balathanigaimani, J. Energy Inst., 2023, 110, 101313.
237. W. S. He, Q. Liu, H. Yu, X. J. Si and J. K. Zhang, J. Am. Oil Chem. Soc., 2016, 93, 509–517.
238. W.-S. He, L.-L. Li, Q.-J. Huang, J. Yin and X.-C. Cao, Food Chem., 2018, 263, 1–7.
239. M. Li, H. Jiang, L. Zhang, X. Yu, H. Liu, A. E. A. Yagoub and C. Zhou, Ind. Crops Prod., 2020, 149, 112361.
240. T. Welton, Green Chem., 2011, 13, 225–225.
241. P. C. Marr and A. C. Marr, Green Chem., 2016, 18, 105–128.
242. T. Shi, Z. Xiong, W. Jin, L. Yuan, Q. Sun, Y. Zhang, X. Li and R. Gao, Food Hydrocolloids, 2020, 102, 105596.
243. J. Lin, R. Lü, C. Wu, Y. Xiao, F. Liang and T. Famakinwa, J. Mol. Model., 2017, 23, 145.
244. H. S. Butt, K. C. Lethesh and A. Fiksdahl, Sep. Purif. Technol., 2020, 248, 116959.
245. H. Li, W. Zhu, Y. Chang, W. Jiang, M. Zhang, S. Yin, J. Xia and H. Li, J. Mol. Graphics Modell., 2015, 59, 40–49.
246. L. C. Player, B. Chan, M. Y. Lui, A. F. Masters and T. Maschmeyer, ACS Sustainable Chem. Eng., 2019, 7, 4087–4093.
247. Z. Ding, T. Fang, G. Zhou, X. Tang, Y. Wang and X. Liu, J. Mol. Liq., 2022, 368, 120719.
248. N. Jin, J. Yue, Y. Zhao, H. Lü and C. Wang, Chem. Eng. J., 2021, 413, 127419.
249. T. Lemaoui, Y. Benguerba, A. S. Darwish, F. A. Hatab, S. E. Warrag, M. C. Kroon and I. M. Alnashef, Sep. Purif. Technol., 2021, 256, 117861.
250. Z. Naseem, J. Iqbal, M. Zahid, A. Shaheen, S. Hussain and W. Yaseen, J. Food Meas. Charact., 2021, 15, 1487–1498.
251. C. Li, D. Li, S. Zou, Z. Li, J. Yin, A. Wang, Y. Cui, Z. Yao and Q. Zhao, Green Chem., 2013, 15, 2793–2799.
252. F. Lima, M. Dave, A. J. D. Silvestre, L. C. Branco and I. M. Marrucho, ACS Sustainable Chem. Eng., 2019, 7, 11341–11349.
253. X.-d. Tang, Y.-f. Zhang, J.-j. Li, Y.-q. Zhu, D.-y. Qing and Y.-x. Deng, Ind. Eng. Chem. Res., 2015, 54, 4625–4632.
254. W. Jiang, L. Dong, W. Liu, T. Guo, H. Li, S. Yin, W. Zhu and H. Li, Chem. Eng. Process., 2017, 115, 34–38.
255. W. Jiang, H. Li, C. Wang, W. Liu, T. Guo, H. Liu, W. Zhu and H. Li, Energy Fuels, 2016, 30, 8164–8170.
256. M. Wysokowski, P. Makoś-Chełstowska, A. Brzęczek-Szafran, A. Sikora, A. Gorczyński and T. Jesionowski, Adv. Sci., 2025, 12, 2412622.
257. T. D. Bennett, F. X. Coudert, S. L. James and A. I. Cooper, Nat. Mater., 2021, 20, 1179–1187.
258. P. F. Fulvio and S. Dai, Chem, 2020, 6, 3263–3287.
259. N. O'Reilly, N. Giri and S. L. James, Chem. – Eur. J., 2007, 13, 3020–3025.
260. J. Zhang, S. H. Chai, Z. A. Qiao, S. M. Mahurin, J. Chen, Y. Fang, S. Wan, K. Nelson, P. Zhang and S. Dai, Angew. Chem., 2015, 127, 946–950.
261. N. Giri, M. G. Del Popolo, G. Melaugh, R. L. Greenaway, K. Ratzke, T. Koschine, L. Pison, M. F. Gomes, A. I. Cooper and S. L. James, Nature, 2015, 527, 216–220.
262. L. He, J. He, P. Cui, Y. Feng, M. Hua, J. Zhang, P. Wu, W. Zhu, H. Li, Z. Liu and C. Xu, Sep. Purif. Technol., 2023, 307, 122781.
263. J. Zhang, J. Yin, W. Fu, H. Ran, W. Jiang, H. Li, W. Zhu, H. Li and M. Zhang, Fuel Process. Technol., 2023, 243, 107687.
264. J.-R. Zhang, J. Yin, J. He, H.-S. Ran, W. Jiang, H.-P. Li, W.-S. Zhu, H.-M. Li and M. Zhang, Pet. Sci., 2024, 21, 2817–2829.
265. F. Boshagh, M. Rahmani and W. Zhu, Energy Fuels, 2022, 36, 12961–12985.
266. X. B. Lim and W. J. Ong, Nanoscale Horiz., 2021, 6, 588–633.
267. Z. Jiang, L. Hongying, Y. Zhang and L. Can, Chin. J. Catal., 2011, 32, 707–715.
268. F. Boshagh, M. Rahmani, K. Rostami and M. Yousefifar, Energy Fuels, 2021, 36, 98–132.
269. L. Chen, J.-T. Ren, H.-Y. Wang, M.-L. Sun and Z.-Y. Yuan, ACS Catal., 2023, 13, 12125–12133.
270. Y. Wang, X. Wang, Y. H. Dong, M. L. Peng, L. N. Guo, M. Y. Cui, Y. He, J. B. Yi, H. J. Ma, H. Zhang and H. M. Fan, Green Chem., 2024, 26, 6131–6138.
271. Z. Wang, Y. M. Chen, Z. Y. Long, Y. F. Wang, M. M. Fan, P. B. Zhang and Y. Leng, Green Chem., 2025, 27, 3091–3098.
272. Z. B. Zhou, H. Xu, J. M. Ma, X. H. Zeng and Y. G. Wei, Green Chem., 2024, 26, 4161–4167.
273. X. F. Zhu, X. Y. Liu, Y. Ding, S. N. Li, Y. C. Jiang and Y. Chen, Green Chem., 2024, 26, 6039–6045.
274. M. A. Shadmehri, M. R. Housaindokht and A. Nakhaei Pour, Appl. Organomet. Chem., 2023, 37, e7272.
275. Y. Chen, Q. Tian, Y. Tian, J. Cui and G. Wang, Appl. Sci., 2021, 11, 2018.
276. R. Ghubayra, Oxidative desulfurization of model diesel fuel catalyzed by polyoxometalates, The University of Liverpool, United Kingdom, 2022.
277. G. Zhao, Y. Wang, C. Wang, H. Lei, B. Yi and R. Tong, Green Chem., 2022, 24, 4041–4049.
278. M. A. Safa, X. Ma, R. Bouresli and H. Albazzaz, Catal. Today, 2021, 371, 258–264.
279. I. Nigmatullin, V. Zuber and R. Nigmatullin, Pet. Chem., 2023, 63, 531–538.
280. X. Lu, J. Li, L. Liang and G. Li, Fuel, 2024, 362, 130849.
281. H. Ye, W. Xing, F. Zhao, J. Wang, C. Yang, Y. Hou, J. Zhang, J. C. Yu and X. Wang, Angew. Chem., Int. Ed., 2025, 64, e202419630.
282. S. Bo, H. Lin, J. Zhang, W. Liao, K. Yang, T. Su, H. Lü and Z. Zhu, Green Chem., 2024, 26, 2661–2672.
283. J. Zhong, Z. Qi, Y. Liu, Y. Zhu, J. Chen, C. Ye and T. Qiu, Ind. Eng. Chem. Res., 2021, 61, 977–986.
284. S. Lyu, J. Wang, Y. Zhou, C. Wei, X. Liang, Z. Yu, W. Lin, Y. Hou and C. Yang, Adv. Funct. Mater., 2024, 34, 2310286.
285. Y. Gao, C. M. Granadeiro, L. Cunha-Silva, J. Zhao and S. S. Balula, Catal. Sci. Technol., 2023, 13, 4785–4801.
286. J. Dou and H. C. Zeng, ACS Catal., 2014, 4, 566–576.
287. X. Gao, W. Jiang, X. An, L. Xu, J. He, H. Li, M. Zhang, W. Zhu and H. Li, Fuel Process. Technol., 2023, 239, 107526.
288. X. Zeng, A. Adesina, P. Li, H. Wang and R. Zhou, Korean J. Chem. Eng., 2022, 39, 96–108.
289. W. Jiang, X. An, J. Xiao, Z. Yang, J. Liu, H. Chen, H. Li, W. Zhu, H. Li and S. Dai, ACS Catal., 2022, 12, 8623–8631.
290. S. Zhou, J. He, P. Wu, L. He, D. Tao, L. Lu, Z. Yu, L. Zhu, Y. Chao and W. Zhu, Fuel, 2022, 309, 122143.
291. Y. Tian, Y. Yao, Y. Zhi, L. Yan and S. Lu, Energy Fuels, 2015, 29, 618–625.
292. M. A. Rezvani, H. H. Ardeshiri and Z. Aghasadeghi, Energy Fuels, 2023, 37, 2245–2254.
293. S. Guan, Z. Li, B. Xu, J. Wu, N. Wang, J. Zhang, J. Han, T. Guan, J. Wang and K. Li, Chem. Eng. J., 2022, 441, 136022.
294. D. Jin, G. Yu, X. Li, T. Li, F. Zhang, S. Tian, Z. Zhou and Z. Ren, Fuel, 2022, 329, 125513.
295. P. S. Sinhmar, A. Tiple and P. R. Gogate, Environ. Technol. Innovation, 2021, 22, 101487.
296. X. An, X. Gao, J. Yin, L. Xu, B. Zhang, J. He, H. Li, H. Li and W. Jiang, Chem. Eng. J., 2024, 480, 147879.
297. H. Niu, S. Xie, X. Zhao, H. Yang, L. Yang, L. Bai, D. Wei, W. Wang and H. Chen, Energy Fuels, 2022, 36, 15080–15087.
298. Y. Tian, X. Zhou, C. Wang, P. Zhou, W. Wang, Z. Song and X. Zhao, Energy, 2022, 258, 124865.
299. M. Naghavi, G. Mazloom, A. Akbari and F. Banisharif, Chem. Eng. Res. Des., 2021, 174, 454–462.
300. E. Eseva, A. Dunko, S. Latypova, O. Grafov, K. Cherednichenko, M. Motyakin, A. Anisimov and A. Akopyan, Fuel, 2024, 357, 129689.
301. X. Cao, R. Tong, J. Wang, L. Zhang, Y. Wang, Y. Lou and X. Wang, Molecules, 2023, 28, 5073.
302. Y. Song, J. Bai, S. Jiang, H. Yang, L. Yang, D. Wei, L. Bai, W. Wang, Y. Liang and H. Chen, Fuel, 2021, 306, 121751.
303. L. Chen and Z.-Y. Yuan, J. Ind. Eng. Chem., 2022, 108, 1–14.
304. Z. Liu, Y. Zhang, J. Bai, H. Yang, L. Yang, L. Bai, D. Wei, W. Wang, Y. Liang and H. Chen, Fuel, 2023, 334, 126753.
305. W. Jiang, X. Gao, L. Dong, J. Xiao, L.-H. Zhu, G.-Y. Chen, S.-H. Xun, C. Peng, W.-S. Zhu and H.-M. Li, Pet. Sci., 2020, 17, 1422–1431.
306. P. Wu, C. Deng, F. Liu, H. Zhu, L. Chen, R. Liu, W. Zhu and C. Xu, Chin. J. Catal., 2023, 54, 238–249.
307. J. Li, Z. Yang, S. Li, Q. Jin and J. Zhao, J. Ind. Eng. Chem., 2020, 82, 1–16.
308. R. Ghubayra, C. Nuttall, S. Hodgkiss, M. Craven, E. F. Kozhevnikova and I. V. Kozhevnikov, Appl. Catal., B, 2019, 253, 309–316.
309. J. Li, Z. Yang, G. Hu and J. Zhao, Chem. Eng. J., 2020, 388, 124325.
310. S.-W. Li, R.-M. Gao and J.-s. Zhao, ACS Sustainable Chem. Eng., 2018, 6, 15858–15866.
311. L. Li, Y. Lu, H. Meng and C. Li, Fuel, 2019, 253, 802–810.
312. J. Zhuang, B. Hu, J. Tan and X. Jin, Transition Met. Chem., 2014, 39, 213–220.
313. Q. Yuan, B. B. Xu, Y. Su, F. Xiao, S. Zhi, Y. Xiao, L. Zhu and C. Dai, Fuel, 2025, 390, 134662.
314. H. Lü, Y. Zhang, Z. Jiang and C. Li, Green Chem., 2010, 12, 1954–1958.
315. B. Bertleff, J. Claußnitzer, W. Korth, P. Wasserscheid, A. Jess and J. Albert, ACS Sustainable Chem. Eng., 2017, 5, 4110–4118.
316. E. K. Nazloo, N. R. Moheimani and H. Ennaceri, Sci. Total Environ., 2023, 859, 160000.
317. D. Zhang, H. Song and D. Yuan, Chem. Eng. Commun., 2019, 206, 1706–1714.
318. H. Li, A. Murugesan, M. Shoaib, W. Sheng and Q. Chen, Crit. Rev. Food Sci. Nutr., 2025, 65, 5218–5250.
319. L. Chu, J. Guo, Z. Huang, H. Yang, M. Yang and G. Wang, Fuel, 2023, 332, 126012.
320. G. Hou, Y. Liu, Y. Gai, G. Han, Y. Ji, G. Hou, A. Wang, Y. Liu, P. Van Der Voort and X. Feng, Chem. Eng. J., 2024, 492, 152139.
321. G. Ye, Z. Yang, G. Shi, R. Huang, X. Liu, Q. Zhang and R. Song, Sep. Purif. Technol., 2025, 354, 128669.
322. S. Smolders, T. Willhammar, A. Krajnc, K. Sentosun, M. T. Wharmby, K. A. Lomachenko, S. Bals, G. Mali, M. B. J. Roeffaers, D. E. De Vos and B. Bueken, Angew. Chem., Int. Ed., 2019, 58, 9160–9165.
323. K. Zhang, F. Chu, Y. Hu, X. Huang, G. Zhao and G. Wang, Chin. Chem. Lett., 2023, 34, 107766.
324. C. Piscopo, C. Granadeiro, S. S. Balula and D. Bošković, ChemCatChem, 2020, 12, 4721–4731.
325. J. Li, X. Jiang, M. Zhu and B. Dai, Fuel, 2023, 341, 127674.
326. Y. Zhang, G. Li, L. Kong and H. Lu, Fuel, 2018, 219, 103–110.
327. L. Hao, M. J. Hurlock, X. Li, G. Ding, K. W. Kriegsman, X. Guo and Q. Zhang, Catal. Today, 2020, 350, 64–70.
328. L. Hao, S. A. Stoian, L. R. Weddle and Q. Zhang, Green Chem., 2020, 22, 6351–6356.
329. H.-Q. Zheng, Y.-N. Zeng, J. Chen, R.-G. Lin, W.-E. Zhuang, R. Cao and Z.-J. Lin, Inorg. Chem., 2019, 58, 6983–6992.
330. G. Fu, B. Bueken and D. De Vos, Small Methods, 2018, 2, 1800203.
331. Y. Gu, H. Luo, W. Xu, W. Zhou and Y. Sun, J. Ind. Eng. Chem., 2022, 105, 435–445.
332. L. Chu, Z. Wang, Z. Liu, T. Zhao, L. Wang, M. Yang and G. Wang, Fuel, 2025, 401, 135846.
333. S.-W. Li, W. Wang and J.-S. Zhao, Appl. Catal., B, 2020, 277, 119224.
334. R. Canioni, C. Roch-Marchal, M. Haouas, A. Vimont, P. Horcajada, F. Sécheresse and C. Serre, Curr. Inorg. Chem., 2017, 7, 145–156.
335. S. Wang, X. Zhang, X. Chang, M.-Y. Zong, C.-Z. Fan, D.-X. Guo, J. Xu, D.-H. Wang and X.-H. Bu, Appl. Catal., B, 2021, 298, 120594.
336. X.-m. Zhang, Z. Zhang, B. Zhang, X. Yang, X. Chang, Z. Zhou, D.-H. Wang, M.-H. Zhang and X.-H. Bu, Appl. Catal., B, 2019, 256, 117804.
337. A. Haruna, Z. M. A. Merican, S. G. Musa and M. B. A. Rahman, J. Taiwan Inst. Chem. Eng., 2023, 147, 104919.
338. Y. Gao, Z. Lv, R. Gao, G. Zhang, Y. Zheng and J. Zhao, J. Hazard. Mater., 2018, 359, 258–265.
339. Z.-J. Lin, H.-Q. Zheng, J. Chen, W.-E. Zhuang, Y.-X. Lin, J.-W. Su, Y.-B. Huang and R. Cao, Inorg. Chem., 2018, 57, 13009–13019.
340. Z. Qi, Z. Huang, H. Wang, L. Li, C. Ye and T. Qiu, Chem. Eng. Sci., 2020, 225, 115818.
341. Z. Qi, T. Qiu, H. Wang and C. Ye, Fuel, 2020, 268, 117336.
342. I. Ahmed and S. H. Jhung, Fuel, 2025, 393, 135043.
343. C. M. Granadeiro, L. S. Nogueira, D. Julião, F. Mirante, D. Ananias, S. S. Balula and L. Cunha-Silva, Catal. Sci. Technol., 2016, 6, 1515–1522.
344. B. N. Bhadra, J. Y. Song, N. A. Khan and S. H. Jhung, ACS Appl. Mater. Interfaces, 2017, 9, 31192–31202.
345. A. W. Bhutto, R. Abro, S. Gao, T. Abbas, X. Chen and G. Yu, J. Taiwan Inst. Chem. Eng., 2016, 62, 84–97.
346. B. Jiang, H. Yang, L. Zhang, R. Zhang, Y. Sun and Y. Huang, Chem. Eng. J., 2016, 283, 89–96.
347. J. Gui, D. Liu, Z. Sun, D. Liu, D. Min, B. Song and X. Peng, J. Mol. Catal. A: Chem., 2010, 331, 64–70.
348. G. Yu, J. Zhao, D. Song, C. Asumana, X. Zhang and X. Chen, Ind. Eng. Chem. Res., 2011, 50, 11690–11697.
349. W. Jiang, W. Zhu, Y. Chang, Y. Chao, S. Yin, H. Liu, F. Zhu and H. Li, Chem. Eng. J., 2014, 250, 48–54.
350. H. A. Elwan, M. T. Zaky, A. S. Farag, F. S. Soliman and M. E. D. Hassan, J. Mol. Liq., 2020, 310, 113146.
351. J. Han, Y. Cai, L. Wang, Y. Mao, L. Ni and Y. Wang, Food Chem., 2020, 309, 125749.
352. Y. Nie, Y. Dong, L. Bai, H. Dong and X. Zhang, Fuel, 2013, 103, 997–1002.
353. W. Zhu, P. Wu, L. Yang, Y. Chang, Y. Chao, H. Li, Y. Jiang, W. Jiang and S. Xun, Chem. Eng. J., 2013, 229, 250–256.
354. W. Jiang, W. Zhu, H. Li, Y. Chao, S. Xun, Y. Chang, H. Liu and Z. Zhao, J. Mol. Catal. A: Chem., 2014, 382, 8–14.
355. W. Jiang, H. Li, S. Yin, X. Wang, W. Liu, H. Liu, W. Zhu and H. Li, Appl. Organomet. Chem., 2016, 30, 753–758.
356. W. Zhu, G. Zhu, H. Li, Y. Chao, M. Zhang, D. Du, Q. Wang and Z. Zhao, Fuel Process. Technol., 2013, 106, 70–76.
357. J. Wang, B. Yang, X. Peng, Y. Ding, S. Yu, F. Zhang, L. Zhang, H. Wu and J. Guo, Chem. Eng. J., 2022, 429, 132446.
358. W. Zhu, W. Huang, H. Li, M. Zhang, W. Jiang, G. Chen and C. Han, Fuel Process. Technol., 2011, 92, 1842–1848.
359. S. Xun, W. Zhu, D. Zheng, L. Zhang, H. Liu, S. Yin, M. Zhang and H. Li, Fuel, 2014, 136, 358–365.
360. W. Zhu, B. Dai, P. Wu, Y. Chao, J. Xiong, S. Xun, H. Li and H. Li, ACS Sustainable Chem. Eng., 2015, 3, 186–194.
361. M. Safa, B. Mokhtarani, H. R. Mortaheb, K. Tabar Heidar, A. Sharifi and M. Mirzaei, Energy Fuels, 2017, 31, 10196–10205.
362. H. Azimzadeh, A. Akbari and M. R. Omidkhah, Chem. Eng. J., 2017, 320, 189–200.
363. H. Zhang, Q. Zhang, L. Zhang, T. Pei, L. Dong, P. Zhou, C. Li and L. Xia, Chem. Eng. J., 2018, 334, 285–295.
364. J. Xiong, W. Zhu, H. Li, Y. Xu, W. Jiang, S. Xun, H. Liu and Z. Zhao, AIChE J., 2013, 59, 4696–4704.
365. P. Wu, B. Wang, L. Chen, J. Zhu, N. Yang, L. Zhu, C. Deng, M. Hua, W. Zhu and C. Xu, Adv. Sci., 2024, 11, 2401996.
366. H. Li, W. Fu, J. Yin, J. Zhang, H. Ran, M. Zhang, W. Jiang, W. Zhu, H. Li and S. Dai, J. Colloid Interface Sci., 2024, 662, 160–170.
367. J. Yin, W. Fu, J. Zhang, X. Zhang, W. Qiu, W. Jiang, L. Zhu, H. Li and H. Li, Chem. Eng. J., 2024, 492, 152349.
368. R. Liu, C. Deng, L. Chen, Y. Chao, P. Wu, W. Zhu and C. Xu, AIChE J., 2025, 71, e18751.
369. W. Ashraf, A. Rehman, A. Hussain, A. Karim, H. R. Sharif, M. Siddiquy and Z. Lianfu, Food Bioprocess Technol., 2023, 17, 887–903.
370. W. Huang, X. He, J. Wu, X. Ma, J. Han, L. Wang and Y. Wang, Ind. Crops Prod., 2023, 197, 116555.
371. S. Guan, Z. Li, B. Xu, J. Wu, J. Han, T. Guan, J. Wang and K. Li, ACS Sustainable Chem. Eng., 2023, 11, 6292–6301.
372. S. Guan, Z. Li, B. Xu, J. Wu, J. Han, T. Guan, J. Wang and K. Li, Fuel, 2023, 333, 126329.
373. W. Liu, T. Li, G. Yu, J. Wang, Z. Zhou and Z. Ren, Fuel, 2020, 265, 116967.
374. Z. Zhu, H. Lin, M. Chi, X. Gao, Y. Feng, K. Yang and H. Lü, Fuel, 2022, 308, 122070.
375. D. Zhu, J. Yin, L. Xu, B. Zhang, L. Zhu, J. He, H. Li, W. Zhu, H. Li and W. Jiang, Appl. Catal., B, 2025, 363, 124774.
376. W. Jiang, H. Jia, H. Li, L. Zhu, R. Tao, W. Zhu, H. Li and S. Dai, Green Chem., 2019, 21, 3074–3080.
377. M. Y. Mohammed, A. M. Ali and T. M. Albayati, Chem. Eng. Res. Des., 2022, 182, 659–666.
378. Z. Shao, C. Xing, M. Xue, Y. Fang and P. Li, J. Sci. Food Agric., 2023, 103, 6929–6939.
379. F. F. Roman, J. L. Diaz de Tuesta, A. M. Silva, J. L. Faria and H. T. Gomes, Catalysts, 2021, 11, 1239.
380. W. Zhang, H. Zhang, J. Xiao, Z. Zhao, M. Yu and Z. Li, Green Chem., 2014, 16, 211–220.
381. K. Zeng, W. Wei, L. Jiang, F. Zhu and D. Du, J. Agric. Food Chem., 2016, 64, 7874–7881.
382. Q. Gu, G. Wen, Y. Ding, K.-H. Wu, C. Chen and D. Su, Green Chem., 2017, 19, 1175–1181.
383. G. Yu, S. Lu, H. Chen and Z. Zhu, Energy Fuels, 2005, 19, 447–452.
384. L. A. Gonzalez, P. Kracke, W. H. Green, J. W. Tester, L. M. Shafer and M. T. Timko, Energy Fuels, 2012, 26, 5164–5176.
385. M. T. Timko, J. A. Wang, J. Burgess, P. Kracke, L. Gonzalez, C. Jaye and D. A. Fischer, Fuel, 2016, 163, 223–231.
386. Q. Gu, Y. Ding, Z. Liu, Y. Lin, R. Schlögl, S. Heumann and D. Su, J. Energy Chem., 2019, 32, 131–137.
387. P. Wu, Y. Sun, L. Chen, Q. Jia, J. He, W. Ma, L. Lu, Y. Chao, L. Fan and W. Zhu, Inorg. Chem., 2022, 61, 633–642.
388. A. K. Dizaji, B. Mokhtarani and H. R. Mortaheb, Fuel, 2019, 236, 717–729.
389. R. Ghubayra, R. Yahya, E. F. Kozhevnikova and I. V. Kozhevnikov, Fuel, 2021, 301, 121083.
390. Y. Gao, R. Gao, G. Zhang, Y. Zheng and J. Zhao, Fuel, 2018, 224, 261–270.
391. W. Jiang, D. Zheng, S. Xun, Y. Qin, Q. Lu, W. Zhu and H. Li, Fuel, 2017, 190, 1–9.
392. A. Khodadadi Dizaji, H. R. Mortaheb and B. Mokhtarani, Chem. Eng. J., 2018, 335, 362–372.
393. B. N. Bhadra, M. M. H. Mondol and S. H. Jhung, Chem. Eng. J., 2022, 440, 135841.
394. Y. Chao, J. Pang, Y. Bai, P. Wu, J. Luo, J. He, Y. Jin, X. Li, J. Xiong, H. Li and W. Zhu, Food Chem., 2020, 320, 126666.
395. A. Rajendran, H. X. Fan, J. Feng and W. Y. Li, Chem. – Asian J., 2020, 15, 2038–2059.
396. P. Wu, W. Zhu, Y. Chao, J. Zhang, P. Zhang, H. Zhu, C. Li, Z. Chen, H. Li and S. Dai, Chem. Commun., 2016, 52, 144–147.
397. P. Wu, S. Yang, W. Zhu, H. Li, Y. Chao, H. Zhu, H. Li and S. Dai, Small, 2017, 13, 1701857.
398. W. Zhu, X. Gao, Q. Li, H. Li, Y. Chao, M. Li, S. M. Mahurin, H. Li, H. Zhu and S. Dai, Angew. Chem., Int. Ed., 2016, 55, 10766–10770.
399. Y. Wu, P. Wu, Y. Chao, J. He, H. Li, L. Lu, W. Jiang, B. Zhang, H. Li and W. Zhu, Nanotechnology, 2018, 29, 025604.
400. L. Lu, J. He, P. Wu, Y. Wu, Y. Chao, H. Li, D. Tao, L. Fan, H. Li and W. Zhu, Green Chem., 2018, 20, 4453–4460.
401. P. Wu, X. Song, L. Lu, C. Deng, H. Lü, L. Chen, H. Ji, D. Tao and W. Zhu, Appl. Catal., B, 2024, 347, 123784.
402. P. Wu, L. Lu, J. He, L. Chen, Y. Chao, M. He, F. Zhu, X. Chu, H. Li and W. Zhu, Green Energy Environ., 2020, 5, 166–172.
403. L. Lu, B. Zhang, H. Li, Y. Chao, Y. Li, L. Chen, H. Li, H. Ji, P. Wu and W. Zhu, Mol. Catal., 2021, 512, 111757.
404. P. Wu, W. Zhu, A. Wei, B. Dai, Y. Chao, C. Li, H. Li and S. Dai, Chem. – Eur. J., 2015, 21, 15421–15427.
405. P. Wu, X. Song, L. Chen, L. He, Y. Wu, D. Tao, J. He, C. Deng, L. Lu, Y. Chao, M. Hua and W. Zhu, Green Energy Environ., 2024, 9, 495–506.
406. P. Wu, Y. Wu, L. Chen, J. He, M. Hua, F. Zhu, X. Chu, J. Xiong, M. He, W. Zhu and H. Li, Chem. Eng. J., 2020, 380, 122526.
407. W. Zhu, Z. Wu, G. S. Foo, X. Gao, M. Zhou, B. Liu, G. M. Veith, P. Wu, K. L. Browning, H. N. Lee, H. Li, S. Dai and H. Zhu, Nat. Commun., 2017, 8, 15291.

---