Modified porous ceramic catalysts derived from titanium-bearing blast furnace slag for efficient toluene oxidation Authors: He Tao, Fanshu Kong, Jiahua Li, Shuaibo Huang, Zehua Wang, Shucheng Tian, Xueqing Li, Haiping Liu Published: 2025 ABSTRACT Titanium-bearing blast furnace slag (TBBFS) is a solid waste generated by the iron and steel industry, rich in valuable components such as TiO2, SiO2, Al2O3, and CaO. Conventional utilization methods mainly involve low-value applications or direct landfill disposal, leading to resource waste and potential environmental risks. In this study, TBBFS was innovatively employed to construct a porous ceramic catalyst support (TC). By regulating sintering temperature, holding time, and pore-forming agent content, a synergistic optimization between porosity and mechanical strength was achieved. After acid leaching, Cu, Ce, and Mn oxides were sequentially loaded onto the support through a stepwise hydrothermal method to prepare a series of hierarchical porous ceramic catalysts. The samples were systematically characterized by XRD, SEM, EDS, BET, and H2-TPR analyses. The results show that the raw TBBFS catalyst exhibited negligible activity, with toluene conversion remaining below 10% within the tested temperature range. The acid-leached ceramic support (HNO3@TC) achieved a T50 of 411.6 °C but failed to reach T90. For the single-metal catalyst Cu@TC, the T50 and T90 were 403.9 °C and 499.1 °C, respectively. The bimetallic Cu–Ce@TC achieved T50 and T90 at 345.4 °C and 469.3 °C, respectively. Compared with the raw TBBFS and the single- or bimetallic catalysts, the ternary Cu–Ce–Mn@TC catalyst exhibited outstanding performance, with T50 and T90 reaching 285.6 °C and 396.4 °C, respectively, fully demonstrating the remarkable enhancement of catalytic activity due to the synergistic effect among multiple metals. Furthermore, even under high space velocity (150,000 mL g⁻1 h⁻1) and high toluene concentration (3000 ppm), the Cu–Ce–Mn@TC catalyst maintained high activity and stability. During a continuous 36 h reaction test, toluene conversion was consistently maintained between 97 and 100% with minimal deactivation, highlighting its excellent efficiency, durability, and adaptability under complex operating conditions. In addition, kinetic fitting based on the Mars–van Krevelen (MVK) model revealed that the catalytic oxidation followed a lattice oxygen participation mechanism, with the oxidation step identified as the rate-determining step. Therefore, this study demonstrates the efficient application of TBBFS-derived ceramic supports in VOCs abatement, not only expanding the resource utilization pathway of metallurgical solid waste but also providing a new technical route and theoretical foundation for the development of low-cost, efficient, and environmentally friendly catalyst systems. Introduction With the rapid development of the iron and steel metallurgy industry, global blast furnace slag (BFS) production has exceeded 400 million tons annually. Among these, titanium-bearing blast furnace slag (TBBFS) is typically a byproduct generated during the blast furnace smelting of ilmenite1,2. In regions rich in ilmenite resources—such as Panzhihua, China—ilmenite concentrate is used as a raw material for smelting. During this process, iron is extracted in metallic form, while titanium reacts with other minerals to form TBBFS3,4. Additionally, in the production of titanium alloys, titanium is introduced into the smelting system as an alloying element, which also leads to the formation of TBBFS as a byproduct5. At present, China’s annual production of titanium slag has exceeded 3 million tons, with an accumulated reserve of approximately 80 million tons. However, these titanium-containing slag resources have not yet been efficiently or fully utilized, resulting in substantial resource waste and increasing pressure on the future development of titanium ore resources6,7. The large-scale stockpiling of TBBFS not only consumes land resources but also poses environmental pollution risks. Some TBBFS contains trace amounts of potentially harmful metals such as chromium and vanadium, which, when left in long-term open storage, may leach into groundwater systems or be absorbed by plants, thus posing potential threats to ecosystems and human health8,9. For a long time, many studies have focused on the use of TBBFS as a raw material for the production of cement and concrete10. Due to its high content of calcium, silicon, and iron, TBBFS exhibits excellent wear resistance and has been applied in the construction industry in several countries. For instance, Mao et al. utilized the porous structure of expanded titanium-bearing blast furnace slag (ETS) to prepare phase change aggregates (PCA) by vacuum-impregnating paraffin, which was further used to produce phase change energy storage mortar (PCEM) for building materials11. In another study, Jiang et al. employed lightweight porous Ti-BFS as support for phase change materials to fabricate paraffin/Ti-BFS phase change aggregates (PTA) for construction applications12. In recent years, however, research on the resource utilization of TBBFS has gradually shifted from traditional building materials to functional materials, showing promising potential in fields such as antibiotic removal from water13,14, photocatalytic dye degradation15, and gaseous pollutant control16–20. For example, Cheng et al. developed a TBBFS-based porous geopolymer that exhibited excellent synergistic adsorption–photodegradation performance (97.18 mg/g) for methylene blue under visible light13. Kong et al. constructed a low-temperature NH₃-SCR catalyst by acid modification and loading MnOx–CeO2 active components onto TBBFS, significantly enhancing NO removal efficiency and SO2 resistance20. In another case, a Ce-doped perovskite catalyst achieved a toluene degradation efficiency as high as 92.08% under visible light, verifying the role of rare-earth dopants in improving the dispersion of active components and the concentration of surface oxygen vacancies18. These studies not only highlight the diversified pathways for using TBBFS as a precursor for functional materials but also underscore its potential in catalytic oxidation applications. Therefore, developing green and efficient resource utilization technologies to achieve high-value applications of TBBFS has become a key issue in the implementation of the “ carbon peaking and carbon neutrality” strategy and the construction of a circular economy system. In the context of addressing volatile organic compounds (VOCs)—a pressing environmental concern—the further development of TBBFS-derived materials shows great promise for application. VOCs, as major atmospheric pollutants, are widely present in industrial emissions and domestic sources. They readily react with nitrogen oxides under sunlight to form photochemical smog and ground-level ozone, posing serious threats to both the environment and human health21,22. Among various VOCs, toluene is regarded as a typical representative. It is commonly emitted from industries such as solvent-based paints, inks, adhesives, rubber, pesticides, furniture manufacturing, and petrochemical refining, and can also be released from vehicle exhaust and residential combustion activities23. Toluene is highly volatile and toxic, capable of participating in ozone and particulate matter (PM) formation, and poses potential risks to the central nervous system, liver and kidney function, and may even induce mutagenic and carcinogenic effects24. Therefore, to meet the urgent demand for the effective removal of refractory VOCs, thermal catalytic oxidation (TCO) technology has emerged as a mainstream solution in industrial VOC abatement. Compared with other treatment methods, TCO offers the advantages of high reaction efficiency, excellent selectivity, low energy consumption, operational stability, and broad adaptability to various working conditions, while achieving complete oxidation of VOCs into CO2 and H2O25–34. TCO is particularly effective in treating high-concentration and difficult-to-degrade organic exhaust gases, demonstrating strong industrial compatibility and application potential35–37. Consequently, the development of low-cost, efficient, and stable catalytic systems for the deep oxidation of toluene has become a key research focus in the field of air pollution control38. Meanwhile, in recent years, extensive research has systematically explored the mechanisms of toluene removal via thermal catalytic oxidation, as well as the performance of metal oxide catalysts. Among various catalytic materials, metal oxide catalysts have gradually become a research focus due to their low cost and high thermal stability. Common transition metals used include V, Cr, Mn, Fe, Co, Ni, Cu, and Zn39–45. Zou et al. reported that CuO/TiO2 catalysts demonstrated high toluene conversion rates even at relatively low temperatures, along with excellent resistance to carbon deposition46. Wang et al. revealed that the MnOx catalyst relies on a reversible redox cycle between Mn3⁺ and Mn4⁺, which facilitates the generation and migration of active oxygen species, thereby improving toluene oxidation activity47. Sui et al. proposed that Cu–Mn composite oxides, through a solid solution structure, enhance the formation and mobility of oxygen vacancies, leading to a synergistic improvement in catalytic performance48. Sun et al. synthesized a highly dispersed Cu–CeOₓ catalyst by modifying a Cu–Ce MOF precursor with EDTA, finding that the introduction of CeO2 significantly increased the oxygen storage capacity and improved resistance to steam and sulfur poisoning49. Therefore, transition metal oxide catalysts, with advantages such as low raw material cost and simple preparation processes, exhibit excellent application potential and broad prospects in catalytic oxidation technologies50–52. Overall, although some studies have used TBBFS to prepare catalysts for NOX17,20 or dye degradation13, there are few studies on the construction of porous ceramic carriers based on TBBFS and their application in the thermal catalytic oxidation of VOCs. Therefore, this study selected a pore-forming agent-assisted sintering method that combines structural controllability with process simplicity, using titanium-containing blast furnace slag (TBBFS) as raw material. Through high-temperature calcination, a porous ceramic carrier (TC) was prepared, which was then acid-leached and modified. Subsequently, a stepwise hydrothermal method was employed to load transition metal oxides such as Cu, Cu–Ce, and Cu–Ce–Mn, resulting in the preparation of multi-level structured composite catalysts. The physicochemical properties of the catalyst were systematically characterized using XRD, SEM, XPS, and H2-TPR techniques, followed by toluene catalytic oxidation performance tests, aiming to effectively improve the catalyst’s treatment efficiency while expanding the high-value application pathways of TBBFS. This also provides new insights for low-cost, green, and sustainable VOCs treatment technologies. Experimental section Materials Cu(NO3)2·3H2O, Ce(NO3)3·6H2O, Mn(NO3)2·4H2O, and urea were purchased from Shanghai McLean Biochemical Technology Co., Ltd. All reagents were of analytical grade and used directly without further purification. TBBFS was provided by a steel mill in Shijiazhuang, Hebei Province. Catalyst preparation Pretreatment of titanium-bearing blast furnace slag The detailed composition is shown in Table 1. As observed, the main components of titanium-bearing blast furnace slag (TBBFS) include SiO2 (25.5%), CaO (22.33%), TiO2 (21.7%), Al2O3 (16.7%), along with alkaline earth metal oxides and trace amounts of Fe. Notably, the abundant presence of Ti, Al, Si, and Ca in TBBFS closely matches the primary constituents of conventional ceramic materials, endowing it with excellent ceramization potential. The TBBFS was placed into a ball mill for 30 min of grinding until the particle size was reduced to below 0.5 mm, after which the treated slag was collected using a mesh sieve.Table 1Chemical composition of titanium-bearing blast furnace slag (wt.%).IngredientsCaOSiO2Fe2O3Al2O3TiO2MgOK2ONa2OContent22.33%25.5%1.5%16.7%21.7%7.05%0.76%0.46% Table 1Chemical composition of titanium-bearing blast furnace slag (wt.%).IngredientsCaOSiO2Fe2O3Al2O3TiO2MgOK2ONa2OContent22.33%25.5%1.5%16.7%21.7%7.05%0.76%0.46% Chemical composition of titanium-bearing blast furnace slag (wt.%). Sintering and optimization of porous ceramic supports The pretreated TBBFS was used as the main raw material and thoroughly mixed with pore-forming agents (limestone) and fluxing agents (boric acid and phosphates) in appropriate proportions until a mixing uniformity exceeding 98.5% was achieved. The mixture was then placed in a muffle furnace for high-temperature sintering under various conditions (sintering temperature ranging from 940 to 1060 °C, holding time from 0 to 15 min, and pore-forming agent content from 0 to 10 wt%). The sintered samples were subjected to apparent porosity and compressive strength tests to evaluate their structural compactness and pore structure regulation. The schematic diagram illustrating the ceramic sintering and structure regulation mechanism is shown in Fig. 1.Fig. 1Schematic diagram of ceramic sintering structure regulation mechanism. Schematic diagram of ceramic sintering structure regulation mechanism. Acid leaching treatment of ceramic supports According to common pretreatment methods reported in related literature, to further remove surface impurities and enhance hydroxyl group exposure for improved metal loading, the optimized ceramic samples (TC) were soaked in a 1 mol/L nitric acid (HNO3) solution and subjected to ultrasonic treatment at 60 °C for 2 hours53–55. Afterward, the samples were repeatedly rinsed with deionized water until the solution pH stabilized between 6.5 and 7.0. Finally, the acid-leached ceramic supports were dried at 80 °C for 12 h to obtain the acid-washed ceramic support samples, denoted as HNO3@TC. Preparation of Cu@TC A total of 2.41 g of Cu(NO3)2·3H2O and 0.5 g of urea were dissolved in 50 mL of deionized water and magnetically stirred for 30 min. Subsequently, 10 g of the acid-leached ceramic support (HNO3@TC) was dispersed into the solution and subjected to hydrothermal treatment at 120 °C for 6 h to generate the Cu(OH)2 precursor. The theoretical loading of the active component (w) was 8.0 wt%. The obtained sample was then calcined at 400 °C for 2 h with a heating rate of 3 °C min⁻1 to form a CuO inner layer, yielding the CuO-loaded ceramic catalyst (Cu@TC). Post-calcination weighing results indicated that the deviation between the actual and theoretical loadings of the active component in the monolithic catalyst was within 10%. Therefore, the theoretical loading value was adopted as the basis for discussion in this study. Preparation of Cu–Ce@TC Using Cu@TC as the support, 4.34 g of cerium nitrate hexahydrate (Ce(NO3)3·6H2O) was dissolved in 80 mL of deionized water to obtain a Ce3+/Cu2+ molar ratio of approximately 1:1. After magnetic stirring to form a homogeneous precursor solution, the pH was adjusted to 9.0 using ammonia water, followed by hydrothermal treatment at 80 °C for 4 h to promote uniform deposition of Ce(OH)3 on the support surface. The theoretical loading of the active component was 7.9 wt%. The obtained sample was then calcined at 500 °C for 3 h with a heating rate of 2 °C min⁻1 to form a CeO2 interlayer, yielding the bimetallic composite catalyst Cu–Ce@TC. Post-calcination weighing results indicated that the deviation between the actual and theoretical loading amounts was less than 10%. Preparation of Cu–Ce–Mn@TC Based on the Cu–Ce@TC catalyst, 2.87 g of manganese nitrate tetrahydrate (Mn(NO3)2·4H2O) and 0.3 g of PVP-K30 were dissolved in deionized water, followed by ultrasonic dispersion for 30 min. The mixture was then subjected to hydrothermal deposition at 80 °C for 2 h. After drying, the obtained sample was calcined at 350 °C for 2 h with a heating rate of 2 °C min⁻1 to form an outer layer enriched with MnO2. The resulting material was the ternary composite catalyst with a MnO2-rich surface (Cu–Ce–Mn@TC). The theoretical loading of the active component was 8.4 wt%, and post-calcination weighing revealed that the deviation between the actual and theoretical loading was less than 10%. The complete process flow for preparing the catalyst is shown in Fig. 2.Fig. 2Schematic diagram of the catalyst preparation process. Schematic diagram of the catalyst preparation process. Catalyst activity evaluation Catalytic activity tests were conducted in a fixed-bed reaction system consisting of a gas supply unit, VOCs generator, tubular reactor, and gas chromatographic analysis system (Fig. 3). The prepared catalyst (1 g) was mixed with quartz sand at a mass ratio of 1:3, packed into the middle section of a quartz tube, and fixed with quartz wool to ensure uniform gas flow through the catalyst bed. The reaction conditions were as follows: toluene was selected as the target pollutant, and the reaction was carried out under atmospheric pressure. The temperature was controlled by a programmable tubular furnace within the range of 50–600 °C, with a heating rate of 2–5 °C/min. The total inlet gas flow rate was maintained at 0.5 L/min, with an initial toluene concentration of approximately 1000 ppm, and air was used as the carrier gas. During the experiments, the weight hourly space velocity (WHSV) was maintained at 30,000 mL g-1 h-1. A gas chromatograph (Shimadzu 2010 Plus) was employed to monitor the toluene concentrations at the inlet and outlet in real time. Each experiment was repeated, and the average value was used to calculate toluene conversion and evaluate the catalytic activity.Fig. 3Schematic diagram of the fixed-bed reactor. Schematic diagram of the fixed-bed reactor. The catalytic activity of different catalysts was evaluated by measuring the temperature at which toluene conversion reached 50% (T50) and 90% (T90). Toluene conversion was calculated using Eq. (1):1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathrm{X}} = \left( {1 - \frac{{{\mathrm{C}}_{{{\mathrm{out}}}} }}{{{\mathrm{C}}_{{{\mathrm{in}}}} }}} \right) \times 100\%$$\end{document} In the equation, X represents the toluene conversion rate; Cout denotes the toluene concentration at the outlet; Cin indicates the toluene concentration at the inlet. Material characterization The crystal structure and phase composition of the samples were analyzed using X-ray powder diffraction (XRD, Bruker D8 advance) with a scanning range of 10–80° and a scanning speed of 2°/min. Scanning electron microscopy (SEM, Hitachi HITSCHI SU8010) was employed to observe the surface microstructure of the porous ceramics and catalysts loaded with different metal oxides. Additionally, energy-dispersive spectroscopy (EDS, Oxford ultimax 40) was used for elemental analysis and two-dimensional elemental mapping to identify the actual distribution of metals such as Cu, Ce, and Mn on the catalyst surface. X-ray photoelectron spectroscopy (XPS, Thermo ESCALAB 250XI) was performed to analyze the surface chemical states and elemental compositions of the catalysts. Hydrogen temperature-programmed reduction (H2-TPR, Bayer BETCAT) was carried out to investigate the redox properties of the catalysts. N₂ adsorption–desorption measurements (BET, Mike ASAP2460) were employed to determine the specific surface area, pore volume, and pore size distribution of the porous ceramics and catalysts. Results and discussion Structural and performance optimization of porous ceramic supports In this study, TBBFS was used as the raw material to prepare porous ceramic supports. To optimize the pore structure and mechanical properties of the porous ceramics, the effects of sintering temperature, holding time, and pore-forming agent content were systematically investigated. As shown in Fig. 4a, the sintering temperature has a significant influence on the structural density and mechanical strength of the porous ceramics. When the temperature increased from 940 °C to 1060 °C, the compressive strength markedly increased from 6.19 MPa to 11.22 MPa, while the apparent porosity decreased from 39.07% to 17.38%. As shown in Fig. 5, XRD analysis revealed that fewer crystalline phases with lower crystallinity were formed at 940 °C and 970 °C. When the temperature increased to 1000 °C and 1030 °C, the formation of major crystalline phases—including calcium aluminosilicate, aluminum titanate (Al2TiO5), and silica (SiO2)—increased significantly, and the crystallinity of the material was markedly enhanced. At 1060 °C, although the crystalline phase content reached its maximum, excessive grain growth and pore shrinkage likely exerted adverse effects on the porosity. The evolution of crystallization behavior is mainly attributed to the reaction between TiO2 and Al2O3/SiO2 components in the slag at elevated temperatures, forming the stable aluminum titanate phase. Meanwhile, low-melting-point oxides such as Na2O and K2O form a liquid phase under the fluxing action of borates and phosphates, which further reacts with SiO2 and Al2O3 to generate aluminosilicate compounds. Among these, calcium aluminosilicate not only contributes to constructing the ceramic matrix structure but also plays a crucial role in enhancing the mechanical properties, pore structure, and chemical stability of the ceramics, thus imparting high mechanical strength and chemical durability. Therefore, considering the balance required between porosity and compressive strength for catalyst supports, a sintering temperature of 1000 °C was selected for subsequent studies to ensure structural stability while achieving a desirable porosity.Fig. 4process optimization results of titanium-bearing blast furnace slag porous ceramics (a) effect of sintering temperature on porous ceramics (b) effect of holding time on porous ceramics (c) effect of pore-forming agent content on porous ceramics.Fig. 5XRD patterns of TBBFS porous ceramics at different calcination temperatures. process optimization results of titanium-bearing blast furnace slag porous ceramics (a) effect of sintering temperature on porous ceramics (b) effect of holding time on porous ceramics (c) effect of pore-forming agent content on porous ceramics. XRD patterns of TBBFS porous ceramics at different calcination temperatures. As shown in Fig. 4b, extending the holding time from 0 to 15 min resulted in an overall decrease in apparent porosity (from 23.74% to 18.78%) and a significant increase in compressive strength (from 5.51 MPa to 10.64 MPa), with the most notable improvement of 3.53 MPa occurring between 5 and 15 min. This trend is mainly attributed to the gradual formation of the liquid phase during the holding period, which promotes the densification of the ceramic framework and further enhances the stability of the ceramic lattice. However, the continuous increase of the liquid phase also leads to the closure of some pores, causing porosity to decline. Although extending the holding time to 15 min yields higher compressive strength, considering energy consumption and process efficiency, a holding time of 5 min has achieved satisfactory densification and mechanical performance, making it more practical for engineering applications. The addition of pore-forming agents has a significant effect on regulating the pore structure of the ceramics. As shown in Fig. 4c, even without the addition of pore-forming agents, the material exhibits an apparent porosity of 12.18% and a compressive strength of 10.1 MPa, which results from pore formation caused by gas release at grain boundaries and organic decomposition in the mixed raw materials. With increasing pore-forming agent content, the apparent porosity continuously rises, reaching a maximum of 43.34% at 10 wt% addition, while the compressive strength decreases to 6.01 MPa. This is because the pore-forming agent completely decomposes during sintering, generating numerous pores that weaken the bonding and packing density between ceramic particles, leading to a looser structure and reduced mechanical strength. In summary, considering both apparent porosity and compressive strength, although an 8% pore-forming agent slightly reduces mechanical strength, its enhancement of the pore structure benefits subsequent loading of active components and gas mass transfer. Therefore, the sintering parameters of 1000 °C sintering temperature, 5 min holding time, and 8% pore-forming agent content were determined as the optimal conditions for this study. To further verify the microstructure and pore distribution of the ceramics, scanning electron microscopy (SEM) analysis was conducted on the TBBFS-derived porous ceramics prepared under the optimized sintering parameters, as shown in Fig. 6. The images reveal a typical porous structure, with hierarchically distributed internal pores composed primarily of macropores and smaller pores of varying sizes. In some regions, interconnected pore channels can be observed, while the pore walls exhibit rough edges accompanied by microcracks, which contribute to an increased specific surface area. Overall, the pore structure is highly open with good connectivity, forming a three-dimensional skeletal network. This architecture is favorable for the infiltration and uniform deposition of metal precursors, while also promoting efficient diffusion of reactants and removal of products. Thus, it provides an ideal structural foundation for anchoring metal oxides and enhancing their catalytic performance in subsequent applications.Fig. 6SEM morphology images of titanium-bearing blast furnace slag porous ceramics. SEM morphology images of titanium-bearing blast furnace slag porous ceramics. Catalyst characterization and evaluation Based on the porous ceramic supports with excellent mechanical properties and pore structure, stepwise loading of Cu, Ce, and Mn was carried out via acid leaching and sequential hydrothermal methods to construct hierarchically structured catalysts. The specific characterizations are as follows. Structural characteristics of the catalyst As shown in Fig. 7, the XRD patterns of different catalysts indicate that when Cu is hydrothermally loaded alone, diffraction peaks of CuO appear at 2θ≈35.5°and 38.7°, corresponding to the (− 111) and (111) crystal planes, respectively, which are consistent with JCPDS No. 48–1548. No metallic Cu peaks are observed, suggesting that copper mainly exists in the Cu2+ state and is present on the support surface as CuO nanocrystals56,57. After Ce introduction, weak and broadened peaks appear at 2θ≈28.6°(111), 33.1°(200), and 47.5°(220), fully matching the fluorite structure of CeO2 (JCPDS No. 34–0394), indicating the formation of nanoscale CeO2 crystallites58. With further addition of Mn, diffraction peaks corresponding to Mn3O4 at 2θ≈37.0°(211) and 65.1°(440) are observed (JCPDS No. 24–0734)59. Throughout the loading process, the main diffraction peaks of the porous ceramic matrix phases calcium silicoaluminate, aluminum titanate, and silica show some intensity attenuation and slight broadening but no phase transformation, indicating that the framework structure remains intact and is covered by the metal oxide layers60. The stability of the main phases suggests that the support maintains its structural integrity during metal loading, which helps the catalyst retain good mechanical properties and chemical stability under high temperatures and repeated redox conditions, thereby enhancing its long-term operational stability. The increased peak broadening and raised baseline of the metal oxide peaks also support the presence of nanocrystalline size effects and a small amount of amorphous coverage61,62. Similar systems with synergistically dispersed multimetal oxides on ceramic or carbon supports have been applied in VOC catalytic oxidation, selective reduction, and bioabsorption, all confirming the high dispersion and synergistic enhancement characteristics of Cu-Ce-Mn ternary oxides at the nanoscale63,64.Fig. 7XRD patterns of catalysts with different compositions. XRD patterns of catalysts with different compositions. Morphological characteristics of the catalysts Systematic analysis of the SEM images for the four catalyst samples (Fig. 8) shows the following features: As illustrated in Fig. 8a, the HNO3@TC support exhibits a distinct porous structure, providing a favorable platform for metal loading and gas transport. In the Cu@TC sample (Fig. 8b), noticeable aggregation of CuO particles is observed, with local agglomeration and partial blockage of the pore channels, which may limit the exposure of active sites. For the Cu-Ce@TC sample (Fig. 8c), the metal particles are more densely distributed with increasingly diverse morphologies. This microstructural evolution corresponds well with the appearance of characteristic CeO2 diffraction peaks in the XRD patterns, confirming that Ce is stably present in the form of nanocrystals and is successfully incorporated onto the catalyst surface.Fig. 8SEM morphology images of catalysts with different compositions. SEM morphology images of catalysts with different compositions. In contrast, the Cu-Ce-Mn@TC sample (Fig. 8d) exhibits the most favorable microstructural characteristics: the surface is uniformly covered with smaller particles, forming a dense and continuous catalytic active layer. Although metal oxides form localized coverage in certain regions, the overall pore structure remains highly permeable. As a result, the catalyst achieves an optimal balance between a highly active catalytic layer and efficient gas transport. Previous studies indicate that Cu and Mn can be incorporated into the CeO2 lattice to form solid solutions, enhancing oxygen mobility and reactivity; an appropriate amount of Cu promotes the formation of CuOX/MnCeOX interfacial regions enriched with active oxygen species65. This structural synergy improves redox performance and increases oxygen vacancy concentration, enabling the Cu-Ce-Mn@TC catalyst to exhibit superior activity in toluene catalytic combustion. Elemental distribution of the catalysts Based on the characteristic X-ray wavelengths of specific elements, EDS analysis was performed to identify the elemental compositions of the catalysts with different loaded components. The corresponding results are presented in Fig. 9, including the energy-dispersive spectra of catalysts with various compositions (a, c, e, g) and the semi-quantitative elemental statistics (b, d, f, h). The elemental mapping images of the catalysts are shown in Fig. 10.Fig. 9Energy spectrum analysis of catalysts with different compositions (a,c,e,g); semi-quantitative statistical plots of catalysts with different compositions (b,d,f,h).Fig. 10Overall elemental mapping of different catalysts (a–c); Cu elemental mapping of different catalysts (d–f); Ce elemental mapping of different catalysts (g,h); Mn elemental mapping of Cu-Ce-Mn@TC (i). Energy spectrum analysis of catalysts with different compositions (a,c,e,g); semi-quantitative statistical plots of catalysts with different compositions (b,d,f,h). Overall elemental mapping of different catalysts (a–c); Cu elemental mapping of different catalysts (d–f); Ce elemental mapping of different catalysts (g,h); Mn elemental mapping of Cu-Ce-Mn@TC (i). As shown in Fig. 9a, the acid-leached ceramic substrate mainly contains Ca, Si, Al, Ti, Mg, and Fe, which are intrinsic components of titanium-bearing blast furnace slag. No signals from loaded metals are detected, indicating that the sample is free from external metal contamination and can serve as a reliable reference. In Fig. 9c, after Cu loading, distinct Cu characteristic peaks appear at approximately 1.0, 8.0, and 8.9 keV with moderate intensities, confirming that Cu is successfully deposited on the ceramic surface in the form of copper oxides. Meanwhile, the original Ca, Si, Ti, and Fe peaks remain visible, implying good structural stability of the support. In Fig. 9e, with the introduction of Ce, several additional Ce characteristic peaks (4.8 and 5.3 keV) emerge alongside the Cu peaks, with significantly increased intensities. This indicates the formation of CeO2 on the surface and more pronounced loading signals. The further enhancement of the O signal suggests that CeO2 contributes abundant surface active oxygen species. Figure 9g presents the EDS spectrum of the Cu-Ce-Mn ternary composite catalyst, where Mn signals (5.8–6.0 keV) appear in addition to those of Cu and Ce, confirming the successful dispersion of MnOₓ on the surface. All metal elements exhibit clear and distinguishable signals without noticeable peak overlap. Combined with the elemental mapping images (Fig. 10), the results confirm that the various metal oxides are uniformly distributed across the catalyst surface, providing a structural basis for the potential synergistic effects among the components. This observation is consistent with the morphological features revealed by SEM images and the stable crystalline phases detected by XRD. Such a synergistically dispersed structure is conducive to forming a catalytic layer enriched with surface active oxygen species, thereby enabling higher catalytic efficiency in the practical oxidation of toluene. Analysis of BET characterization results Based on the BET nitrogen adsorption–desorption isotherms (Fig. 11a), all four samples exhibit typical Type IV isotherms, indicating that the materials are primarily composed of mesoporous structures66. BET test results (Table 2) show that acid-treated HNO3@TC possesses the highest specific surface area and total pore volume, providing an excellent loading platform and gas pathways. Following metal loading, both Cu@TC and Cu-Ce@TC exhibited decreases in specific surface area and pore volume. Combined with the pore size distribution (Fig. 11b), a significant reduction in the 2–5 nm pore volume was observed, indicating that metal oxide particles partially blocked the original pores during deposition. This phenomenon is common in supported catalytic materials and is consistent with SEM observations66. Notably, despite the decrease in specific surface area, the average pore size of the metal-loaded material significantly increased. This is primarily due to the partial blockage of small pores, shifting the pore size contribution toward medium-to-large mesopores. This increase in overall average pore size enhances the transport of reaction gases within the pore network. Furthermore, after Mn introduction, both specific surface area and total pore volume markedly recovered. Combined with the pore size distribution analysis, the mesoporous volume in the 10–30 nm range significantly increased, indicating that Mn addition promoted more uniform dispersion of the metal oxide. This partially alleviated excessive pore blockage by the metal oxide, improved pore connectivity, and exposed more accessible mesopores67. This more open mesoporous structure may facilitate oxygen species migration and reactant diffusion within the pores, offering potential advantages for catalytic reactions.Fig. 11N2 adsorption–desorption isotherms and pore size distributions of the catalysts with different compositions.Table 2Specific surface area of different catalysts.Specific surface area (m2/g)Total pore volume (cm3/g)Average pore diameter (nm)HNO3@TC21.9730.0314732.865Cu@TC5.91890.023177.829Cu-Ce@TC3.72830.0182868.886Cu-Ce-Mn@TC7.54810.0297619.809 N2 adsorption–desorption isotherms and pore size distributions of the catalysts with different compositions. Table 2Specific surface area of different catalysts.Specific surface area (m2/g)Total pore volume (cm3/g)Average pore diameter (nm)HNO3@TC21.9730.0314732.865Cu@TC5.91890.023177.829Cu-Ce@TC3.72830.0182868.886Cu-Ce-Mn@TC7.54810.0297619.809 Specific surface area of different catalysts. Analysis of H2-TPR characterization results The H2-TPR measurements were conducted to investigate the reducibility and metal–metal interactions of the catalysts. As shown in the Fig. 12, the main reduction peaks of Cu@TC, Cu-Ce@TC, and Cu-Ce-Mn@TC appear at 264.7℃, 252.9℃, and 242.6℃, respectively. The gradual shift of the reduction peaks toward lower temperatures indicates an enhancement in the reducibility of the catalysts upon the introduction of Ce and Mn.Fig. 12H2-TPR images of catalysts with different components. H2-TPR images of catalysts with different components. Based on relevant research68,69, it is speculated that for Cu@TC, the reduction peak corresponds to the Cu2+ → Cu+ transition. After Ce incorporation, the reduction peak shifts to a lower temperature, suggesting that Ce facilitates the reduction of Cu species and enhances oxygen migration through the Ce4+/Ce3+ redox cycle70. The further incorporation of Mn leads to an additional shift, implying that Mn promotes stronger electronic interactions and redox synergy among the metals, thereby improving the activation and migration of surface oxygen species71. Such multimetal synergy enhances oxygen circulation and surface reactivity during the catalytic oxidation process. Analysis of XPS characterization results As shown in Fig. 13, XPS analysis reveals that multiple valence states of Cu, Ce, and Mn coexist on the catalyst surface, exhibiting synergistic interactions. In the O 1 s spectrum, peaks at 529.5–530.9 eV correspond to lattice oxygen (O2-), while peaks at 531.8–532.9 eV are attributed to adsorbed oxygen or surface hydroxyl groups. With the introduction of Ce and Mn, the intensity of the lattice oxygen peak (OⅠ) increases significantly, and its proportion rises from 26.61% to 35.56%, indicating a substantial enhancement in oxygen vacancy concentration and lattice oxygen mobility72.Fig. 13XPS images of catalysts with different components. XPS images of catalysts with different components. In the Cu 2p spectrum, peaks at 931.0–942.0 eV are assigned to Cu+/CuO, and peaks at 943.8–954.6 eV correspond to Cu2+, accompanied by satellite peaks at 962.4–962.9 eV, indicating that Cu2+ remains the dominant species. Doping with Ce and Mn increases the Cu+/Cu2+ ratio from 0.409 to 0.461, suggesting electron transfer from Ce3+ and Mn3+ to Cu2+, which promotes partial reduction of Cu2+ to Cu+ 73. In the Ce 3d spectrum, characteristic peaks of Ce4+ appear at 885.9–908.1 eV, accompanied by satellite peaks at 907.4 eV and 916.5 eV, while Ce3+ peaks are observed at 882.5 eV and 882.6 eV. The introduction of Mn increases the Ce3+ proportion from 18.23% to 18.55%, indicating partial reduction of the CeO2 lattice and an increased number of oxygen vacancies74,75. The Mn 2p3/2 spectrum shows two peaks at 641.5 eV and 653.1 eV, corresponding to Mn3+ and Mn4+, respectively, indicating the coexistence of mixed-valence Mn. The presence of Mn3+ can induce additional oxygen vacancies and enhance oxygen migration rates76. Taken together, the synergistic changes in the O 1 s, Cu 2p, Ce 3d, and Mn 2p spectra suggest the formation of hierarchical electron coupling and redox cycles within the Cu-Ce-Mn system (Cu2+ + Ce3+ ↔ Cu+ + Ce4+; Mn4+ + Ce3+ ↔ Mn3+ + Ce4+), which promotes the generation of oxygen vacancies and dynamic replenishment of lattice oxygen. This substantially enhances the catalyst’s oxygen activation capability and structural stability77. These findings are consistent with trends observed in H2-TPR measurements, further confirming the excellent electronic structure regulation achieved in the catalyst. Catalyst activity testing and performance comparison To further elucidate the intrinsic relationship between structure and performance, this section presents a systematic evaluation of the catalytic activity of catalysts with different loaded components in the toluene oxidation reaction. The results are shown in Fig. 14.Fig. 14Toluene degradation efficiency of catalysts with different components. Toluene degradation efficiency of catalysts with different components. The catalytic activity was evaluated based on the temperatures at which toluene conversion reached 50% and 90% (T50 and T90). The activity results indicate that the catalysts follow the order from high to low activity: Cu-Ce-Mn@TC > Cu-Ce@TC > Cu@TC > HNO3@TC > TBBFS. Modification of the catalyst with different metal oxides led to varying degrees of activity enhancement, with Cu-Ce-Mn@TC exhibiting markedly superior activity for toluene oxidation compared to the other catalysts. Combining previous experimental results, it is evident that loading different metal oxide components onto titanium-bearing blast furnace slag-derived porous ceramics enhances their catalytic activity for toluene oxidation. Among them, the ternary Cu-Ce-Mn@TC catalyst shows significantly higher activity than the corresponding single- or dual-metal catalysts. SEM and EDS analyses reveal that the catalyst surface possesses fine, uniformly distributed particles forming a continuous and dense active layer, which facilitates full exposure of active sites. BET measurements show that the specific surface area and pore structure of Cu-Ce-Mn@TC surpass those of the single- and dual-metal catalysts, promoting reactant diffusion and contact within the pores. XPS analysis further indicates an increased proportion of lattice oxygen on the catalyst surface, while the synergistic changes in Cu2+/Cu+, Ce3+/Ce4+, and Mn3+/Mn4+ enhance the redox cycling ability and generation of active oxygen species. H2-TPR results show that, with the introduction of Ce and Mn, the reduction peaks shift to lower temperatures, indicating significantly improved reduction activity and greater participation of oxygen species in toluene oxidation78,79. XRD results confirm that CeO2 and MnOX are stably loaded on the support surface, providing a structural basis for multi-metal synergy. Furthermore, prior studies provide mechanistic evidence supporting the rationality and effectiveness of such multi-metal oxide synergy. Jin et al., based on a spinel-type CuM2O4 system, revealed significant electronic coupling between surface Cu2+ and oxygen vacancies, which promotes lattice oxygen activation and migration, thereby enhancing the supply of active oxygen species during oxidation reactions80. Similarly, Xia et al. demonstrated in Cu-Mn composite oxide catalysts that multiphase structures such as Cu1.4Mn1.6O4 can, under synergistic regulation, strengthen redox cycling, optimize oxygen migration pathways, and improve reaction stability and intermediate conversion efficiency81. In summary, the Cu, Ce, and Mn oxides exhibit excellent synergistic effects: Ce provides oxygen vacancies and oxygen storage capacity, Mn enhances redox cycling and active oxygen generation, and Cu supplies the primary active sites. Together, they establish an efficient oxygen migration and reaction network, significantly improving the catalyst performance in VOCs oxidation, such as toluene82. Effect of reaction conditions In practical industrial applications, catalysts are often required to maintain stability and high efficiency under various operating conditions. Therefore, it is necessary to systematically investigate the effects of space velocity (WHSV), toluene concentration, and oxygen concentration on the catalytic performance. As shown in Figs. 15 and 16, both the space velocity (WHSV) and the inlet toluene concentration have significant effects on catalytic oxidation efficiency. As the WHSV increased from 30,000 to 150,000 mL g-1 h-1, T50 and T90 rose from 285.6℃ to 369.2℃ and from 396.4 ℃ to 485.1℃, respectively. This is attributed to the shortened residence time of reactant gases on the catalyst surface, which limits the contact between toluene molecules and active sites, thereby requiring higher temperatures to achieve the same conversion. Similarly, when the inlet toluene concentration increased from 500 to 3000 ppm, the catalytic activity slightly decreased (T50: 273.5 → 321.1℃ ; T90: 382.4 → 434.5℃ ), primarily due to insufficient supply of surface active oxygen under high-concentration conditions. Nevertheless, even under high WHSV and high toluene concentration, Cu-Ce-Mn@TC still achieved approximately 90% toluene conversion, indicating strong adaptability and resistance to operational fluctuations.Fig. 15Effect of space velocity on the toluene oxidation activity of the Cu-Ce-Mn@TC catalyst.Fig. 16Effect of toluene concentration on the toluene oxidation activity of Cu–Ce–Mn@TC catalyst. Effect of space velocity on the toluene oxidation activity of the Cu-Ce-Mn@TC catalyst. Effect of toluene concentration on the toluene oxidation activity of Cu–Ce–Mn@TC catalyst. To further investigate the effects of oxygen content and catalyst stability on the reaction, dynamic oxygen supply and long-term stability experiments were conducted (Fig. 17). When oxygen in the reaction atmosphere was cut off, toluene conversion gradually decreased from 35.1% to 0.9%, indicating that the surface lattice oxygen was being depleted and the catalytic reaction nearly ceased. Upon reintroduction of oxygen, toluene conversion rapidly recovered to 88.4% within 10 min and eventually stabilized above 98%, demonstrating the irreplaceable role of oxygen in maintaining reaction activity. This observation confirms the excellent oxygen storage and release capability of CeO2 (Ce4+/Ce3+ cycle) and the oxygen migration promoted by Cu and Mn components83. Moreover, the Cu-Ce-Mn@TC catalyst maintained nearly 100% conversion during continuous operation for 36 h at 400℃ with a WHSV of 30,000 mL g-1 h-1, demonstrating outstanding thermal stability. In summary, the synergistic interaction among Cu, Ce, and Mn oxides provides abundant and regenerable active oxygen species, ensuring efficient and stable toluene oxidation and laying a solid foundation for practical applications in VOCs treatment.Fig. 17(a) Effect of O2 concentration on toluene oxidation activity. (b) Stability test of the Cu-Ce-Mn@TC catalyst during continuous operation. (a) Effect of O2 concentration on toluene oxidation activity. (b) Stability test of the Cu-Ce-Mn@TC catalyst during continuous operation. Kinetic studies To investigate the activity differences among catalysts with different compositions and the mechanism of toluene oxidation, kinetic experiments were conducted using the best-performing Cu-Ce-Mn@TC catalyst in a fixed-bed reactor under varying inlet toluene concentrations (600–1200 ppm) and reaction temperatures (170–200℃). The experimental data were subsequently fitted and analyzed based on the Power-rate Law and Mars–van Krevelen models84,85. Dynamic model Power-rate law Power-rate law Power-rate law To characterize the reaction rate of toluene catalytic oxidation, this study employed two kinetic models, the Power-rate Law and Mars-van Krevelen (MVK) models, for data fitting. The Power-rate Law model, based on an empirical rate equation, can be simplified as:2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$- {\mathrm{r}}_{{\mathrm{i}}} = {\mathrm{k}}^{\prime } {\mathrm{C}}_{{\mathrm{i}}}^{{\mathrm{n}}}$$\end{document}where (Ci) is the reactant concentration, (k') is the simplified rate constant, and (n) is the reaction order. The rate constant (k') and reaction order (n) can be determined by performing a linear fit of ln(-ri) versus ln Ci. 2.Mars-van-Krevelen 2.Mars-van-Krevelen Mars-van-Krevelen The MVK model describes the catalytic oxidation process based on redox reactions. The reaction occurs in two stages86–90:VOCs are adsorbed onto the catalyst’s surface active sites and react with the metal oxides in the catalyst, being oxidized to CO2 and H2O, while the metal oxides are reduced.The reduced metal oxides in the catalyst adsorb oxygen from the gas phase and are re-oxidized, replenishing the oxygen vacancies and restoring the catalyst to its oxidized state. VOCs are adsorbed onto the catalyst’s surface active sites and react with the metal oxides in the catalyst, being oxidized to CO2 and H2O, while the metal oxides are reduced. VOCs are adsorbed onto the catalyst’s surface active sites and react with the metal oxides in the catalyst, being oxidized to CO2 and H2O, while the metal oxides are reduced. The reduced metal oxides in the catalyst adsorb oxygen from the gas phase and are re-oxidized, replenishing the oxygen vacancies and restoring the catalyst to its oxidized state. The reduced metal oxides in the catalyst adsorb oxygen from the gas phase and are re-oxidized, replenishing the oxygen vacancies and restoring the catalyst to its oxidized state. This reaction can be represented as:3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathrm{VOCs}} + {\mathrm{Cat}} - {\mathrm{O}} \to {\mathrm{Cat}} + {\mathrm{CO}}_{2} + {\mathrm{H}}_{2} {\mathrm{O}}$$\end{document}4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathrm{Cat}} + {\mathrm{O}}_{2} \to {\mathrm{Cat}} - {\mathrm{O}}$$\end{document} The corresponding rate equation is:5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$- {\mathrm{r}}_{{\mathrm{i}}} = {\mathrm{k}}_{{\mathrm{i}}} {\mathrm{C}}_{{\mathrm{i}}} \theta$$\end{document}6\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$- {\mathrm{r}}_{{{\mathrm{o}}2}} = {\mathrm{k}}_{{{\mathrm{o}}2}} {\mathrm{C}}_{{{\mathrm{o}}2}} (1 - \theta )$$\end{document}where (ri) and (rO2) are the reaction rates of VOCs and the oxygen consumption rate (mol m-3 s-1), (ki) and (kO2) are the rate constants for the reduction and oxidation steps, (Ci) is the VOC concentration, (CO2) is the oxygen concentration, and θ is the surface coverage of active oxygen on the catalyst. The MVK model can be simplified as91–93:7\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\frac{1}{{ - {\mathrm{r}}_{{\mathrm{i}}} }} = \frac{1}{{{\mathrm{k}}_{{\mathrm{i}}} {\mathrm{C}}_{{\mathrm{i}}} }} + {\mathrm{A}}$$\end{document} By plotting 1/Ci on the x-axis and 1/(-ri) on the y-axis for linear fitting, the reaction rate constants (ki) and (kO2) can be determined. 3.The Arrhenius equation 3.The Arrhenius equation The Arrhenius equation The relationship between the reaction rate constant and temperature follows the Arrhenius equation94,95.8\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\mathrm{k}=\mathrm{Aexp}\left(\frac{{-\mathrm{E}}_{\mathrm{a}}}{\mathrm{RT}}\right)$$\end{document}where A is the pre-exponential factor, Ea is the activation energy (kJ mol-1), R is the ideal gas constant (8.314 × 10–3 kJ mol-1 K-1), and T is the reaction temperature (K). Taking the natural logarithm gives:: 9\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\mathrm{lnk}=\frac{{-\mathrm{E}}_{\mathrm{a}}}{\mathrm{RT}}+\mathrm{lnA}$$\end{document} By plotting 1/T on the x-axis and ln k on the y-axis for linear fitting, the activation energy (Ea) can be determined from the slope of the resulting straight line. Kinetic data analysis Based on the Power-rate Law model, a linear fit of ln C(VOCs) versus ln (-r(VOCs)) was performed, as shown in Fig. 18a, to obtain the rate constant (k) and reaction order (n) (Table 3). The results show that (k) increases with temperature, while (n) varies with temperature, indicating that this model cannot accurately describe the kinetics of the present system96. Alternatively, the Mars-van-Krevelen (MVK) model was employed, and a linear fit of 1/Ci versus 1/(-ri) was used to determine the rate constants for the reduction and oxidation steps, (ki) and (kO2) (Table 3), both of which increase with temperature following the Arrhenius law. By fitting lnk versus 1/T (Fig. 18b,c), the activation energies for surface oxidation and reduction were calculated to be 135.65 and 48.02 kJ mol-1, respectively, indicating that the reaction is controlled by the surface oxidation step of the catalyst. The experimentally measured toluene conversion was then used in the MVK model to calculate the reaction rate, and the results were compared with the experimental values (Fig. 18d), yielding (R2 = 0.959). This demonstrates that the MVK model provides a reasonable and reliable description of toluene catalytic combustion over the Cu-Ce-Mn@TC catalyst.Fig. 18Arrhenius fitting plots (a–c) and comparison of theoretical versus calculated reaction rates (d).Table 3Calculated data based on power-rate-law and MVK kinetic model fitting.ModelTemperature170℃180℃190℃200℃Power-rate lawln \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathrm{k}}{\prime}$$\end{document} − 4.5135 − 2.6733 − 1.20690.6055\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathrm{k}}{\prime}$$\end{document}0.01100.06900.29911.8322n0.74680.842910.906520.9871R20.99710.98420.99910.9964Mars–van–Krevelen1/ki0.87300.82890.60330.4343\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\frac{{\mathrm{a}}_{\mathrm{i}}}{{\mathrm{k}}_{\mathrm{o2}}{{\mathrm{C}}}_{\mathrm{o2}}}$$\end{document}0.85890.40980.18970.0083ki1.14541.20551.65822.3027kO20.10830.22610.48911.6857R20.99720.98620.99890.9961 Arrhenius fitting plots (a–c) and comparison of theoretical versus calculated reaction rates (d). Table 3Calculated data based on power-rate-law and MVK kinetic model fitting.ModelTemperature170℃180℃190℃200℃Power-rate lawln \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathrm{k}}{\prime}$$\end{document} − 4.5135 − 2.6733 − 1.20690.6055\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathrm{k}}{\prime}$$\end{document}0.01100.06900.29911.8322n0.74680.842910.906520.9871R20.99710.98420.99910.9964Mars–van–Krevelen1/ki0.87300.82890.60330.4343\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\frac{{\mathrm{a}}_{\mathrm{i}}}{{\mathrm{k}}_{\mathrm{o2}}{{\mathrm{C}}}_{\mathrm{o2}}}$$\end{document}0.85890.40980.18970.0083ki1.14541.20551.65822.3027kO20.10830.22610.48911.6857R20.99720.98620.99890.9961 Calculated data based on power-rate-law and MVK kinetic model fitting. Reaction mechanism Based on the results of the structural-performance-kinetic analysis, combined with the lattice oxygen migration mechanism, the oxidation reaction mechanism involving the synergistic participation of Cu-Ce-Mn is inferred as follows. In this study, porous ceramic catalysts containing titanium blast furnace slag were prepared as catalyst supports, and composite metal oxide catalysts loaded with transition metal elements such as Cu, Ce, and Mn were prepared. The catalyst was comprehensively studied using characterization techniques such as XRD, SEM, BET, XPS, EDS, and H2-TPR. Under a stepwise hydrothermal method, the composite metal catalyst Cu-Ce-Mn@TC was prepared. The catalyst’s crystal structure was improved, with more uniformly distributed active sites and more thorough synergistic interaction between elements, successfully achieving efficient degradation of toluene during the catalytic combustion stage, with a degradation efficiency of 98.36%. In the reaction system, transition metal oxide catalysts exhibit high catalytic activity for toluene degradation due to their flexible crystal structure and variable cation valence states. The catalytic combustion reaction of toluene follows the Mars-vanKrevelen (MVK) mechanism97. Since oxygen vacancies have the ability to regulate the number of catalytic active sites, chemical valence, and oxide distribution, the oxidation of toluene primarily involves catalytic oxidation reactions between toluene and lattice oxygen on the catalyst surface, reducing the active centers of surface metal ions. While gaseous oxygen is activated through oxygen vacancies, filling the vacancies to form new lattice oxygen98, thereby more efficiently degrading toluene molecules. The catalytic oxidation process is illustrated in Fig. 19. Therefore, it can be inferred that the primary process in this reaction involves the reduction followed by oxidation of surface materials on the catalyst. The higher the number of oxygen-rich vacancies in the transition metal oxide catalyst, the higher the activation efficiency of the oxidation products.Fig. 19Reaction mechanism diagram. Reaction mechanism diagram. Conclusion In this study, industrial by-product titanium-bearing blast furnace slag (TBBFS) was used as the raw material to successfully construct a porous ceramic support (HNO3@TC) with high porosity and good mechanical strength through optimized sintering (1000℃, 5 min, 8wt% pore-forming agent) combined with acid-leaching treatment. Based on this support, a series of supported catalysts, including Cu@TC, Cu-Ce@TC, and Cu-Ce-Mn@TC, were prepared using a stepwise hydrothermal and multi-stage calcination strategy. Systematic performance evaluation demonstrated that the ternary composite Cu-Ce-Mn@TC exhibited the best catalytic activity for toluene oxidation, with T50 and T90 of 285.6℃ and 396.4 ℃, respectively, significantly outperforming the single- and dual-metal systems. This enhanced performance primarily arises from the synergistic effects among Cu, Ce, and Mn: Cu provides the primary active sites, Ce promotes the formation and migration of oxygen vacancies, and Mn enhances redox cycling and the generation of surface active oxygen species. Moreover, the Cu-Ce-Mn@TC catalyst maintained excellent activity even under high space velocity (150,000 mL g-1 h-1) and high toluene concentration (3000 ppm), while exhibiting remarkable oxygen responsiveness and regenerability. During a continuous 36-h reaction, toluene conversion remained stable at 97–100% with negligible deactivation, fully demonstrating the catalyst’s high efficiency, stability, and adaptability under demanding conditions. Further Mars-van Krevelen kinetic fitting (R2≈1) indicated that the reaction follows a lattice-oxygen-participated gas-phase oxidation-regeneration mechanism, with the oxidation step being rate-determining. Mechanistically, this confirms the high efficiency and reliability of the catalyst in toluene catalytic oxidation, providing a feasible material and process solution for volatile organic compound (VOC) treatment.