Enhanced Plasma-Catalytic Oxidation of Toluene over Spinel Oxide-Mesoporous SiO₂ Composites

Abstract

Plasma-catalytic oxidation is a promising approach for the abatement of volatile organic compounds (VOCs), yet its efficiency is often limited by the ineffective utilization of plasma-generated reactive oxygen species and incomplete oxidation pathways. In this work, a composite catalyst was constructed by integrating spinel-type NiCo₂O₄ with three-dimensional cubic mesoporous KIT-6 to couple efficient mass transfer with redox-active surface functionality for plasma-catalytic degradation of toluene. The performance of NiCo/KIT-6 was systematically evaluated in a dielectric barrier discharge (DBD) reactor and compared with Ni/KIT-6, Co/KIT-6, and NTP-only systems. XPS, O₂-TPD, H₂-TPR, and apparent dielectric measurements were employed to elucidate catalyst properties relevant to plasma–surface interactions. NiCo/KIT-6 exhibits superior overall performance in terms of toluene conversion, CO_x selectivity, and CO₂ selectivity over a wide range of specific input energies. This enhancement is closely associated with the integrated regulation of surface redox properties, oxygen activation capability, and apparent dielectric response by the NiCo₂O₄/KIT-6 composite structure, which may promote reactive oxygen utilization and facilitates effective plasma–surface redox processes. These results provide insights into the rational design of composite catalysts for plasma-assisted oxidation of aromatic VOCs.

1. Introduction

Volatile organic compounds (VOCs) are major contributors to atmospheric pollution, posing direct threats to human health and playing an important role in the formation of tropospheric ozone and secondary organic aerosols, thereby exacerbating photochemical smog and fine particulate matter pollution [1,2]. In recent years, non-thermal plasma (NTP) technology has attracted increasing interest as an advanced VOCs abatement technique because it can generate energetic electrons and a variety of reactive species under ambient conditions, enabling rapid activation and decomposition of VOCs molecules [3]. Nevertheless, NTP-only systems often suffer from limited utilization efficiency of plasma-generated reactive species and incomplete oxidation pathways, which results in high energy consumption and insufficient mineralization. To overcome these intrinsic limitations, the integration of catalysts with plasma to construct plasma-catalytic systems has been widely recognized as an effective strategy for improving VOCs removal efficiency and oxidation performance [4,5].

Even in plasma-catalytic systems, the capture and effective utilization of reactive oxygen species such as •O, •OH, and O₃ remain challenging. Although these species can be generated efficiently under non-thermal plasma conditions, their extremely short lifetimes in the gas phase mean that, without suitable surface enrichment and interfacial coupling with active sites, a considerable fraction may undergo self-decomposition or participate in non-selective side reactions, thereby limiting overall reaction efficiency and mineralization [3,6]. Consequently, the development of composite catalytic materials that integrate high surface area, efficient mass transport, and surface sites capable of coupling plasma-generated species with redox-active centers is essential for improving plasma-assisted VOCs degradation. Ordered mesoporous materials have therefore attracted considerable attention owing to their regular pore structures, high surface areas, and thermal stability, which are advantageous for adsorption and diffusion in VOCs catalytic oxidation [7,8]. Among them, three-dimensional cubic KIT-6 (Ia3d) possesses a bicontinuous mesoporous network with interconnected channels, offering enhanced molecular diffusion and mass transfer compared with one-dimensional mesoporous materials such as SBA-15 [9]. This unique architecture provides an effective platform for dispersing active components and facilitating the transport and utilization of plasma-generated reactive oxygen species. Previous studies have demonstrated that loading noble metals or transition-metal oxides onto KIT-6 can significantly improve catalytic activity and stability in VOCs oxidation, highlighting its role as both an adsorption-diffusion medium and an interface for heterogeneous reactions [10,11].

Among transition-metal oxides, Co₃O₄-based catalysts have been widely investigated for the catalytic oxidation of aromatic VOCs due to the reversible Co²⁺/Co³⁺ redox couple, the presence of oxygen vacancies, and favorable π–d interactions with aromatic rings, which collectively facilitate adsorption and initial activation of aromatic molecules [12,13,14,15]. However, the catalytic performance of monometallic oxides is often constrained by limited redox flexibility and oxygen mobility. In contrast, bimetallic oxides can provide enhanced electronic tunability and interfacial reactivity through synergistic interactions between different metal centers. As a typical spinel-type bimetallic oxide, NiCo₂O₄ combines the aromatic affinity associated with Co-based oxides with multiple reversible Ni²⁺/Ni³⁺ and Co²⁺/Co³⁺ redox couples, offering improved redox adaptability and surface oxygen activation capability [16,17]. These structural and electronic features are expected to favor the enrichment and transformation of plasma-generated reactive oxygen species at the catalyst surface and to facilitate their coupling with variable-valence metal centers, thereby providing a promising active phase for plasma-assisted oxidation of aromatic VOCs.

Based on the above analysis, this study aims to integrate three-dimensional cubic KIT-6 with spinel-type NiCo₂O₄ to construct a composite catalyst that combines high surface area, efficient mass transport, and redox-active metal-oxygen functionality, and to systematically evaluate its performance in plasma-catalytic degradation of toluene. To clarify the role of the bimetallic spinel phase and the contribution of the mesoporous framework, the catalytic behavior of the NiCo₂O₄/KIT-6 composite is further compared with that of NiO- and Co₃O₄-based counterparts. Through this comparative approach, the synergistic effects arising from the coupling of the spinel oxide with the ordered mesoporous structure are examined, with particular emphasis on reactive oxygen utilization, plasma–surface interaction, and the oxidation efficiency of toluene.

2. Results and Discussion

2.1. Structural and Surface Properties

The XRD patterns of all samples are shown in Figure 1. In the low-angle region (Figure 1a), two distinct reflections at 0.94° and 1.36° can be indexed to the (211) and (200) planes of the Ia3d cubic symmetry, respectively [18,19]. These features confirm the typical three-dimensional bicontinuous Ia3d mesostructure characteristic of KIT-6. After loading the active phases, the low-angle diffraction peaks are noticeably weakened, which can be attributed to partial filling of the mesoporous channels by the deposited species. This filling reduces the electron-density contrast between the silica framework and the pore voids, thereby diminishing the scattering intensity associated with the ordered mesostructure. Nevertheless, the characteristic Ia3d symmetry remains discernible, indicating that the long-range mesoscopic order of KIT-6 is largely preserved. The wide-angle XRD patterns of the samples are shown in Figure 1b. The broad diffraction feature centered at approximately 27.4° is assigned to the amorphous silica walls of the KIT-6 support. For Ni/KIT-6, diffraction peaks located at 37.3°, 43.3°, 62.9°, and 75.5° can be indexed to the (111), (200), (220), and (311) planes of cubic NiO (PDF#47-1049). In the case of Co/KIT-6, the reflections at 19.1°, 31.3°, 37.0°, 44.8°, 59.6°, and 65.5° correspond to the (111), (220), (311), (400), (511), and (440) planes of the spinel Co₃O₄ phase (PDF#43-1003). For NiCo/KIT-6, the diffraction peaks exhibit a slight shift toward lower angles relative to those of Co/KIT-6, which can be ascribed to partial substitution of Co by the larger Ni²⁺ ions. This substitution leads to a modest expansion of the spinel lattice and reflects the incorporation of Ni and Co cations within the octahedral and tetrahedral sites of the spinel framework [20].

Figure 1. (a) Low-angle and (b) wide-angle XRD patterns of samples.

The nitrogen adsorption–desorption isotherms and pore size distributions of the samples are shown in Figure 2. All four samples display typical type IV isotherms with step-like hysteresis loops in the P/P₀ range of 0.6–0.9, corresponding to H1-type hysteresis [21], which indicates well-ordered mesoporous structures. Upon metal loading, the mesoporous framework of KIT-6 is largely preserved, although a slight decrease in adsorption capacity and a modest broadening of the hysteresis loops are observed. These changes indicate that a fraction of the mesopores is partially occupied by the deposited metal oxides or that minor structural adjustments occur upon impregnation and calcination. The textural properties of the samples, including specific surface area, pore volume, and average pore diameter, are summarized in Table 1. Brunauer-Emmett-Teller (BET) analysis shows that pristine KIT-6 exhibits a high specific surface area of 910.9 m²/g and a pore volume of 1.31 cm³/g. After metal oxide loading, both parameters decrease, which can be primarily attributed to partial pore filling by the active phases. Notably, NiCo/KIT-6 retains a comparatively higher surface area and pore volume among the modified samples, suggesting that its mesoporous framework remains more accessible. Furthermore, the average pore diameter derived from the desorption branch using the Barrett–Joyner–Halenda (BJH) method shows only a slight reduction after metal loading, indicating that the ordered mesoporous structure of KIT-6 is largely maintained.

Figure 2. (a) Nitrogen adsorption–desorption isotherms and (b) pore size distributions of the samples.

Table 1. Physicochemical properties of the samples.

Sample	SBET (m2/g)	Vp (cm3/g)	Dp (nm)	Co2+/Co3+	Chemisorbed Oxygen (10−3 mmol/g)
KIT-6	910.9	1.31	6.13	-	0.3
Ni/KIT-6	464.3	0.84	6.07	-	4.2
Co/KIT-6	386.9	0.67	6.02	0.79	4.0
NiCo/KIT-6	542.1	0.93	6.07	0.92	5.9

TEM was employed to examine the mesoporous structure of the samples, as illustrated in Figure 3. KIT-6 displays a highly ordered cubic mesoporous framework, with an interfringe distance of approximately 10 nm. For the metal-oxide-loaded samples, well-dispersed oxide nanoparticles are clearly observed both on the external surfaces and within the mesoporous channels. In detail, the lattice spacing of about 0.21 nm in Figure 3b can be indexed to the (200) plane of NiO, whereas the lattice fringes with a spacing of ~0.24 nm in Figure 3c,d are attributed to the (311) planes of Co₃O₄ and NiCo₂O₄, respectively. Although slight diffraction peak shift are detected in the XRD patterns, the interplanar spacings measured by HRTEM remain essentially unchanged, which can be ascribed to the localized nature of HRTEM observation and its limited precision relative to bulk-averaged XRD analysis.

Figure 3. TEM and HRTEM images of (a) KIT-6, (b) Ni/KIT-6, (c) Co/KIT-6, and (d) NiCo/KIT-6.

XPS was performed to elucidate the surface chemical states of the catalysts, and the corresponding high-resolution Ni 2p, Co 2p, and O 1s spectra are presented in Figure 4. In the Ni 2p region (Figure 4a), spectrum is split into the Ni 2p_3/2 and Ni 2p_1/2 components as expected from spin–orbit (j-j) coupling, and deconvolution of the Ni 2p_3/2 envelope reveals that Ni/KIT-6 is characterized by a Ni²⁺ contribution at 854.7 eV together with a shake-up satellite at 861.0 Ev [22], whereas NiCo/KIT-6 requires three components corresponding to Ni²⁺ (853.6 eV), Ni³⁺ (855.6 eV), and a satellite peak at 861.2 eV [23,24]. Compared with Ni/KIT-6, NiCo/KIT-6 exhibits an overall negative shift of the Ni²⁺ binding energy accompanied by the emergence of Ni³⁺ species, indicating that the incorporation of Co modifies the local electronic environment around Ni and leads to a redistribution of its oxidation-state population reflected in the Ni 2p spectra. As shown in Figure 4b, the Co 2p spectra of Co/KIT-6 and NiCo/KIT-6 can both be deconvoluted into three components corresponding to Co³⁺, Co²⁺, and their associated satellite peaks, with the Co³⁺ species in both samples exhibiting Co 2p_3/2 and Co 2p_1/2 binding energies centered at 779.5 eV and 794.5 eV, respectively [25,26]. Compared with Co/KIT-6, the Co²⁺ species in NiCo/KIT-6 display an overall shift toward higher binding energy, and quantitative analysis further reveals that the Co²⁺/Co³⁺ ratio increases from 0.79 for Co/KIT-6 to 0.92 for NiCo/KIT-6, indicating that the incorporation of Ni modifies the valence-state distribution of Co and leads to a higher relative fraction of Co²⁺ species, which reflects an altered electronic environment and a strengthened redox flexibility of Co centers in the Ni-Co spinel system. The high-resolution O 1s spectra, shown in Figure 4c, exhibit very similar peak shapes and binding-energy positions for all samples, indicating that no pronounced differences can be resolved based solely on O 1s XPS analysis. This behavior can be primarily attributed to the use of KIT-6 as a SiO₂ support, whose surface is dominated by abundant Si-O-Si species that contribute overwhelmingly to the O 1s signal and thereby mask subtle variations associated with metal-related oxygen species [27].

Figure 4. (a) Ni 2p, (b) Co 2p, and (c) O 1s XPS spectra of samples.

2.2. Redox and Dielectric Behaviors

Accordingly, O₂-TPD was employed to further characterize the surface oxygen species, with particular emphasis on their availability and activity, and the corresponding profiles are shown in Figure 5a. Oxygen species on catalyst surfaces are commonly classified as physisorbed oxygen (O₂), chemisorbed oxygen species (O₂⁻, O₂²⁻, and O⁻), and lattice oxygen (O²⁻) [28]. Under the present experimental conditions, no low temperature desorption signals associated with physisorbed oxygen were detected. In contrast, the desorption peaks appearing between 100 °C and 300 °C are assigned to chemisorbed oxygen species, which represent readily removable and highly reactive surface oxygen. Because these species are often anchored on oxygen vacancy related defect sites, their abundance can be taken as an indicator of the surface capability to capture and utilize reactive oxygen species under plasma-dominated conditions, thereby promoting subsequent oxidation reactions during VOCs degradation [29]. As shown in the O₂-TPD profiles, all metal loaded catalysts exhibit markedly enhanced chemisorbed oxygen desorption in this temperature region compared with pristine KIT-6, indicating a substantially higher abundance of surface active oxygen species. Quantitative results summarized in Table 1 demonstrate that NiCo/KIT-6 exhibits the highest proportion of chemisorbed oxygen among all investigated samples. This enrichment of surface chemisorbed oxygen indicates a greater density of readily accessible active sites that can effectively participate in the adsorption, activation, and regeneration of reactive oxygen species, thereby providing a favorable surface environment for the deep oxidation of toluene, particularly under plasma-dominated conditions. Desorption features observed at temperatures above 300 °C are attributed to lattice oxygen species [30]. Given the relatively low operating temperature of the NTP-catalysis system, lattice oxygen species that are typically activated through high-temperature desorption processes are expected to play a limited role via conventional thermally driven Mars–van Krevelen pathways. Instead, the observed catalytic activity under plasma-dominated conditions is more plausibly associated with surface chemisorbed oxygen species and defect-related active oxygen, which can be readily generated and continuously replenished through plasma-induced reactions, including ozone decomposition and the formation of reactive oxygen species [31,32].

Figure 5. (a) O₂-TPD and (b) H₂-TPR profiles of samples.

Since the generation and stabilization of surface active oxygen species are closely related to the reducibility of metal-oxygen entities, H₂-TPR was further conducted to elucidate the redox characteristics underlying the O₂-TPD observations, as illustrated by the profiles in Figure 5b. Pristine KIT-6 shows no detectable reduction features because framework Si⁴⁺ in mesoporous SiO₂ is essentially unreducible over the tested temperature window, consistent with previous studies [33]. Ni/KIT-6 exhibits two reduction events. The peak at ~351 °C is attributed to weakly bound or highly dispersed NiO species, whereas the broad feature centered at ~509 °C is assigned to Ni²⁺ species strongly coupled to the SiO₂ surface, for which reduction is delayed by pronounced metal-support interactions [34]. Co/KIT-6 displays three peaks at 319 °C, 362 °C, and 384 °C. The low-temperature signal (319 °C) is associated with the initial reduction of surface or highly dispersed Co³⁺ to Co²⁺, the intermediate peak (362 °C) corresponds to the predominant bulk conversion of Co₃O₄ to CoO, and the high-temperature peak (384 °C) is assigned to the subsequent reduction of CoO to metallic Co [35]. A similar reduction pattern is observed for NiCo/KIT-6, with peaks at ~289 °C, 327 °C, and 386 °C. The first peak (289 °C) is ascribed to the preferential reduction of highly dispersed trivalent cations (Ni³⁺/Co³⁺), followed by the bulk reduction of the spinel lattice to mixed (Ni,Co)O at 327 °C, and finally the reduction of divalent oxides to metallic Ni and Co at 386 °C [17,36]. Notably, relative to Ni/KIT-6 and Co/KIT-6, the reduction peaks of NiCo/KIT-6 shift consistently toward lower temperatures, indicating an enhanced low-temperature reducibility arising from the synergistic interaction between Ni and Co. In plasma-catalytic systems, where catalyst surfaces are frequently exposed to alternating reducing and oxidizing environments, such facilitated reducibility is expected to lower the kinetic barrier for redox transitions. This redox behavior provides a favorable basis for the enhanced surface oxygen activity observed in O₂-TPD and supports the improved utilization and transformation of reactive oxygen species under plasma excitation.

To characterize the dielectric behavior of the samples at the plasma discharge frequency, their dielectric properties were measured using a parallel-plate capacitor configuration at 10 kHz and room temperature. As shown in Figure 6, pristine KIT-6 exhibits the highest apparent relative permittivity (ε = 65) with a dielectric loss factor (tan δ) of about 0.10. Comparable high apparent dielectric responses have also been reported in SiO2-containing heterogeneous systems, where the enhanced dielectric response was associated with interfacial effects [37]. Upon loading single metal oxides, the apparent relative permittivity decreases to 34 for Ni/KIT-6 and 26 for Co/KIT-6, accompanied by an increase in dielectric loss to 0.24 and 0.12, respectively. In comparison, NiCo/KIT-6 displays an intermediate apparent relative permittivity (ε = 42) together with the lowest dielectric loss among all samples (tan δ = 0.09). Notably, the combination of a moderate apparent relative permittivity and suppressed dielectric loss distinguishes NiCo/KIT-6 from the single-metal counterparts, suggesting that the formation of the NiCo₂O₄ spinel may contribute to a more balanced dielectric response at the discharge frequency. It is widely recognized that the dielectric properties of packing materials constitute an important physical factor in packed-bed DBD systems. Previous studies on plasma-catalysis have pointed out that material properties, including relative permittivity and dielectric loss, can influence the electrical environment of the discharge, such as electric-field distribution, charge accumulation, and energy dissipation under alternating electric fields [38,39]. In particular, excessively high permittivity may induce pronounced local field distortion and charge buildup, whereas high dielectric loss is generally associated with increased energy dissipation through polarization or conductive pathways rather than effective discharge processes. By contrast, materials with moderate permittivity can help mitigate excessive charge accumulation and field localization, while a low dielectric loss minimizes dielectric energy dissipation, together helping to sustain a stable discharge environment under plasma excitation [40,41].

Figure 6. Dielectric properties of samples.

2.3. Plasma-Catalytic Performance

Figure 7 summarizes the plasma-catalytic performance of the investigated systems for toluene degradation. Regardless of catalyst loading, the toluene conversion increases monotonically with increasing SIE, indicating that a higher energy input enhances the availability of effective reaction energy and facilitates the activation and oxidation of toluene molecules under plasma excitation [42]. In the NTP-only system, increasing the SIE from 153 J/L to 676 J/L raises the toluene conversion from 34.8% to 77.8%, accompanied by CO_x selectivities of 54.1–89.4%. Nevertheless, the oxidation of converted carbon toward CO₂ remains limited based on the detected carbon oxides, as evidenced by the relatively low CO₂ selectivity of only 23.1–33.7%, indicating that part of the converted carbon may remain in incompletely oxidized intermediates rather than being fully oxidized to CO₂. Among all catalysts, NiCo/KIT-6 exhibits the most favorable plasma-catalytic performance across the investigated energy range. When the SIE is increased from 60 J/L to 663 J/L, the toluene conversion, CO_x selectivity, and CO₂ selectivity reach 25.5–96.0%, 50.3–99.9%, and 29.2–74.3%, respectively. Since the CO_x selectivity was calculated on a carbon basis, the fraction of converted carbon not detected as CO or CO₂ can be estimated as 100% − CO_x selectivity. For NiCo/KIT-6, this fraction decreases from 49.7% to approximately 0.1% with increasing SIE, suggesting that most of the converted toluene carbon was transformed into carbon oxides and that the accumulation of incompletely oxidized carbon-containing intermediates was limited. In addition, NiCo/KIT-6 also outperformed the mechanically mixed Ni/KIT-6 + Co/KIT-6 (Figure S1), suggesting the enhanced performance is not simply derived from the additive contribution of separated Ni and Co oxide species. Moreover, compared with previously reported plasma catalysts under similar operating conditions [43,44,45,46,47], NiCo/KIT-6 delivers higher energy efficiency while maintaining high toluene conversion and CO₂ selectivity (Table S1), highlighting its favorable energy utilization for plasma-catalytic toluene oxidation. By comparison, although Ni/KIT-6 also displays relatively high toluene conversion, its CO_x and CO₂ selectivities are consistently lower than those observed over NiCo/KIT-6. A distinct performance trend is observed for Co/KIT-6. Under low SIE conditions, its toluene conversion is even lower than that obtained over KIT-6 and in the NTP-only system, while comparatively higher CO_x and CO₂ selectivities are achieved. With further increases in SIE, the toluene conversion over Co/KIT-6 improves markedly but remains inferior to that of NiCo/KIT-6 throughout the investigated energy window. For KIT-6, the high specific surface area combined with the three-dimensionally interconnected mesoporous structure favors the initial conversion of toluene under low SIE conditions. However, further increases in input energy density result in only marginal enhancement of toluene conversion. Moreover, owing to the absence of metal active sites, KIT-6 exhibits a limited ability to promote the mineralization of intermediates, leading to low CO₂ selectivity. Its CO_x selectivity is even lower than that of the NTP-only system, which may be attributed to the influence of the packing material on the gas-phase plasma reaction environment. This may alter certain gas-phase oxidation pathways and reduces the overall efficiency of CO_x formation.

Figure 7. Effect of SIE on the plasma-catalytic performance of samples: (a) toluene conversion, (b) CO_x selectivity, (c) CO₂ selectivity, and (d) O₃ concentration.

Figure 7d illustrates the O₃ concentration measured at the outlet of the DBD reactor. For all investigated systems, the ozone concentration exhibits a gradual decrease with increasing SIE, which can be attributed to the enhanced decomposition and consumption of O₃ under conditions where high-energy electrons and reactive species become increasingly abundant. Compared with the catalyst-packed configurations, the NTP-only system yields a higher O₃ concentration in the effluent, indicating that the introduction of packing materials markedly influences O₃ generation and stability within the discharge environment. Further comparison among the catalyst-packed systems reveals that the O₃ concentrations produced over Ni/KIT, Co/KIT, and NiCo/KIT show an overall positive correlation with their dielectric constants, whereas KIT-6, despite possessing the highest dielectric constant, generates a relatively low O₃ concentration at the reactor outlet. This observation suggests that O₃ formation is not governed solely by the dielectric constant, and that the support structure, together with its combined modulation of discharge behavior and ozone stability, plays an important role in determining the observed ozone levels [41,48].

To evaluate the stability of the catalyst, a 12 h stability test was conducted over the NiCo/KIT-6 catalyst at SIE of 600 J/L, as shown in Figure 8a. After prolonged plasma-catalytic reaction, the catalyst maintained relatively stable catalytic performance, with no obvious decrease in toluene conversion, CO_x selectivity, and CO₂ selectivity, indicating the good stability of NiCo/KIT-6. To further evaluate the post-reaction structural stability of the catalyst from the perspectives of crystalline phase structure and surface chemical state, XRD and XPS characterizations were performed on the fresh and spent NiCo/KIT-6 catalysts, as shown in Figure 8b–e. The XRD results (Figure 8b) show that the spent catalyst still retained the main diffraction peaks of the NiCo₂O₄ spinel phase, indicating that its main crystalline structure was basically preserved during the long-term plasma-catalytic reaction. Notably, weak new diffraction peaks appeared at around 42.8° and 62.3° after the reaction, which may be associated with the formation of a small amount of rock-salt NiO/CoO-like species or Ni-Co mixed monoxide, suggesting that partial Ni-Co oxide species underwent reaction-induced reconstruction during the reaction [49].

Figure 8. (a) Stability test of NiCo/KIT-6 at 600 J/L; (b) XRD patterns, (c) C 1s XPS spectra, (d) Ni 2p XPS spectra, and (e) Co 2p XPS spectra of fresh and spent NiCo/KIT-6.

The surface chemical states of the fresh and spent NiCo/KIT-6 catalysts were further analyzed by XPS. The C 1s spectra (Figure 8c) were deconvoluted into three components at approximately 284.8, 286.5, and 288.8 eV, corresponding to C-C/C=C, C-O, and O-C=O species, respectively [32,50]. The O-C=O component is generally associated with oxygen-containing carbonaceous intermediates formed during toluene oxidation, such as carboxyl or carboxylate species. After the stability test, no obvious π–π* shake-up satellite peak was observed in the high binding energy region, and the relative proportion of O-C=O changed only marginally, indicating that no significant accumulation of carbon-containing byproducts or coke-like deposits occurred on the catalyst surface. The Ni 2p spectra (Figure 8d) showed only minor changes after the stability test, suggesting that the chemical environment of surface Ni species was slightly affected. In contrast, the Co 2p spectra (Figure 8e) exhibited more pronounced changes. After the stability test, the Co³⁺ component shifted toward higher binding energy, whereas the Co²⁺ component shifted toward lower binding energy, indicating a redistribution of the local electronic environment of surface Co species. In particular, the negative shift of Co²⁺ suggests increased electron density around Co²⁺ species, which is consistent with the formation of CoO-like or Ni-Co mixed monoxide species. Combined with the newly emerged diffraction peaks in the XRD pattern, these XPS results further support the partial structural reconstruction of Ni-Co oxide species during the plasma-catalytic reaction. Although a small amount of new crystalline species appeared after the stability test, the main spinel structure and catalytic performance were well maintained, demonstrating the good stability of NiCo/KIT-6 during long-term plasma-catalytic oxidation of toluene.

2.4. Discussion

In plasma-catalytic toluene degradation, the SIE primarily determines the discharge intensity and the generation of gas-phase reactive species, thereby governing the initial activation of toluene molecules. With increasing energy input, the accompanying temperature rise promotes O₃ decomposition [51]. Meanwhile, the population of energetic electrons and reactive species is enhanced, promoting molecular excitation and bond cleavage while further accelerating O₃ decomposition and consumption [52]. This behavior accounts for the universal decrease in ozone concentration observed with increasing SIE across all systems. Under NTP-only conditions, the lack of packing materials results in a reaction regime dominated by gas-phase plasma chemistry, in which catalyst interfaces are absent for O₃ activation into more reactive oxygen species, leading to relatively high toluene conversion but limited further oxidation of intermediates toward CO₂ [53].

Upon introducing catalyst packing, the reaction pathway progressively shifts from a purely gas-phase process toward a gas-surface coupled oxidation process, in which the overall oxidation efficiency becomes increasingly dependent on the surface redox capability of the catalyst [38,54]. Among the investigated materials, NiCo/KIT-6 exhibits a clear advantage in this respect. The electronic interaction between Ni and Co in the NiCo₂O₄ spinel may contribute to more flexible metal valence states, a higher availability of low-temperature reactive oxygen species, and more readily accessible redox behavior, as inferred from XPS, O₂-TPD, and H₂-TPR analyses. These complementary characterizations provide insight into the intrinsic redox properties and oxygen activation capability of NiCo/KIT-6, which may facilitate the surface conversion of plasma-activated intermediates and are consistent with the observed enhancement in toluene conversion and CO₂ selectivity over a wide range of energy densities [55,56].

At low energy densities, the outlet O₃ concentration reflects the net balance between O₃ formation governed by the discharge environment and its consumption via surface-mediated pathways. The dielectric properties of the packing materials modulate the local electric field and microdischarge behavior, thereby influencing ozone generation, while the catalyst surface determines the efficiency of O₃ utilization and subsequent oxidation reactions [57,58]. In this regime, Co/KIT-6 exhibits a lower O₃ concentration, which can be attributed to the combined effect of a relatively less favorable discharge environment and effective surface consumption of available oxidizing equivalents. However, its relatively limited redox flexibility and lower availability of readily activated surface oxygen species may constrain the initial activation of toluene, resulting in low overall conversion despite relatively high CO₂ selectivity. By contrast, NiCo/KIT-6 promotes more efficient initial activation of toluene, which is closely associated with the Ni-Co interactions that provide more accessible redox states and enhanced surface oxygen activation. Meanwhile, its dielectric characteristics sustain a more intense and stable discharge, leading to increased ozone generation. Under low SIE conditions, residual ozone may persist when surface oxidation becomes kinetically limited by competitive adsorption of toluene and partially oxidized intermediates [58,59,60]. These results suggest that ozone concentration reflects the balance between plasma-driven oxidant generation over NiCo/KIT-6 and surface-mediated consumption, rather than the extent of oxidation toward CO₂. The elevated ozone level observed at low energy input thus may originate from strong oxidant formation coupled with limited surface utilization, a mismatch that is progressively alleviated at higher energy input, where markedly enhanced CO_x and CO₂ selectivities are consistent with the improved ability of NiCo/KIT-6 to effectively utilize reactive oxidants.

3. Conclusions

This work systematically investigates the plasma-catalytic degradation of toluene over KIT-6-supported NiO, Co₃O₄, and NiCo₂O₄ catalysts. NiCo/KIT-6 exhibits the most favorable overall performance in terms of toluene conversion, CO_x and CO₂ selectivity. Based on the combined results of catalytic performance and physicochemical characterizations, the superior activity of NiCo/KIT-6 is proposed to be closely associated with the integrated regulation of surface redox properties, oxygen activation capability, and apparent dielectric response by the NiCo₂O₄ spinel structure. These features may facilitate plasma–surface interaction and promote the further oxidation of toluene-derived intermediates toward CO₂. Overall, the study highlights the importance of jointly optimizing catalytic redox properties and the discharge environment for efficient plasma-assisted VOCs abatement.

4. Experimental

4.1. Chemical Reagents

The reagents used in this study included P123 (poly(ethyleneglycol)-block-poly(propyleneglycol)-block-poly(ethylene glycol), MW = 5800), hydrochloric acid (HCl, 35 wt.%), n-butanol (CH₃(CH₂)₃OH), tetraethyl orthosilicate (TEOS), nickel nitrate hexahydrate (Ni(NO₃)₂·6H₂O), cobalt nitrate hexahydrate (Co(NO₃)₂·6H₂O), anhydrous ethanol (CH₃CH₂OH), and sodium hydroxide (NaOH). P123 was obtained from Sigma-Aldrich (St. Louis, MO, USA), HCl and CH₃CH₂OH were supplied by Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China), and all other chemicals were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. (Shanghai, China). All chemicals were of analytical grade and used without further purification.

4.2. Preparation of KIT-6

KIT-6 was synthesized following the previously reported method with some modification [9]. Briefly, 3.6 g of P123 was dissolved in a mixture of 7.08 g of HCl and 130.2 g of deionized water. Subsequently, 3.6 g of n-butanol was added dropwise, and the mixture was stirred for an additional 1 h. Then, 7.74 g of TEOS was introduced into the solution, followed by continuous stirring at 35 °C for 24 h. The resulting mixture was transferred to a Teflon-lined stainless-steel autoclave and subjected to hydrothermal treatment at 100 °C for 24 h. The product was collected by filtration, washed thoroughly with deionized water, and vacuum-dried at 90 °C overnight. Finally, the white powder was calcined in air at 550 °C for 5 h with a heating rate of 2 °C/min to obtain the KIT-6.

4.3. Preparation of Catalysts

A series of metal-modified KIT-6 mesoporous catalysts were synthesized via an incipient wetness impregnation method. In a typical procedure, a certain amount of metal precursors (Ni(NO₃)₂·6H₂O and Co(NO₃)₂·6H₂O) was dissolved in ethanol to form a homogeneous solution. The pre-prepared KIT-6 powder was then introduced into the solution and magnetically stirred at room temperature for 12 h. The resulting mixture was dried at 90 °C overnight, followed by calcination at 400 °C for 3 h with a heating rate of 2 °C/min to yield the final catalysts. The mass fraction of NiCo₂O₄, NiO, or Co₃O₄ in each corresponding catalyst was controlled at approximately 12%. The resulting catalysts were hereafter referred to as NiCo/KIT-6, Ni/KIT-6, and Co/KIT-6, respectively. For comparison, a mechanically mixed control sample was prepared by physically mixing the as-prepared Ni/KIT-6 and Co/KIT-6 catalysts. The mixing ratio was calculated to maintain the Ni/Co molar ratio of NiCo₂O₄ and a comparable total metal oxide loading of approximately 12 wt.%. The obtained sample was denoted as mechanically mixed Ni/KIT-6 + Co/KIT-6.

4.4. Catalysts Characterization

The phase structures of the catalysts were examined by X-ray diffraction (XRD) on an Anton Paar XRDynamic500 diffractometer (Anton Paar GmbH, Graz, Austria) using Cu Kα radiation. Nitrogen adsorption–desorption isotherms were recorded on a Micromeritics ASAP 2460 analyzer (Micromeritics Instrument Corporation, Norcross, GA, USA), from which the specific surface areas were calculated using the Brunauer-Emmett-Teller (BET) method and the pore volumes and pore size distributions were derived using the Barrett–Joyner–Halenda (BJH) method. Transmission electron microscopy (TEM) was performed on a FEI Talos F200x microscope (FEI Company, Hillsboro, OR, USA), with the samples dispersed in ethanol, and deposited onto Cu grids. Surface chemical states were analyzed by X-ray photoelectron spectroscopy (XPS, Thermo Scientific K-Alpha, Thermo Fisher Scientific, East Grinstead, UK). H₂ temperature-programmed reduction (H₂-TPR) and O₂ temperature-programmed desorption (O₂-TPD) measurements were conducted over the temperature range of 50–800 °C. Prior to testing, all samples were pretreated under an inert gas flow. H₂-TPR experiments were carried out at a heating rate of 10 °C/min in a flow of 10% H₂/Ar (30 mL/min), while O₂-TPD measurements were performed after oxygen adsorption under 10% O₂/He (30 mL/min), followed by temperature-programmed desorption under He atmosphere. The signals were monitored using a thermal conductivity detector (TCD). The dielectric properties of the catalysts were measured using a parallel-plate capacitor method at 10 kHz and 25 °C with a LCR meter (E4980AL, Keysight Technologies, Santa Rosa, CA, USA). The samples were pressed into disk-shaped pellets and coated with gold on both sides as electrodes, allowing direct measurement of capacitance and dielectric loss. The relative permittivity values obtained from these pressed powder pellets are therefore referred to as apparent relative permittivity, which were used for qualitative comparison of the apparent dielectric responses among different catalysts.

4.5. Plasma-Catalytic Performance Evaluation

The experimental setup for evaluating the catalytic performance is illustrated in Figure 9. It mainly consists of a toluene generation unit, a dielectric barrier discharge (DBD) reactor, a power supply, and an exhaust gas analysis system. Air was supplied by an air compressor and passed through 3A molecular sieve to remove moisture. The dried air then entered the evaporator, where it swept over the toluene liquid continuously injected by a micro-syringe pump, thereby generating a toluene gas stream with a stable concentration (550 ppm).

Figure 9. Schematic diagram of experimental setup.

The DBD reactor was constructed based on a quartz tube with an outer diameter of 10 mm and an inner diameter of 8 mm. A stainless steel rod with an outer diameter of 2 mm was positioned along the central axis of the quartz tube and served as the high-voltage electrode. A 40 mm-long Cu foil was wrapped around the outer surface of the quartz tube, functioning as the grounded electrode. A total of 500 mg of catalyst (40–60 mesh) was placed in the discharge region and fixed with quartz wool, and the gas hourly space velocity (GHSV) was controlled at 30,000 mL/(g·h). A high-voltage AC power supply (CTP-2000, Suman, Nanjing, China) was employed to generate non-thermal plasma, with the operating frequency maintained at approximately 10 kHz during all experiments. The discharge voltage and current waveforms in the DBD reactor were monitored using an oscilloscope (TBS 1102C, Tektronix, Inc., Beaverton, OR, USA). The discharge power of the DBD reactor was determined by the V-Q Lissajous method, and the specific input energy (SIE) was calculated as follows:

$$SIE\left(\mathrm{J}/\mathrm{L}\right)={\displaystyle \frac{P\left(\mathrm{W}\right)}{Q(\mathrm{L}/\min )}}\times 60$$

where P represents the discharge power, and Q denotes the total gas flow rate.

The concentrations of toluene, CO, and CO₂ were analyzed using an online gas chromatograph (F70, Fuli, Taizhou, China) equipped with a flame ionization detector (FID) and a methane conversion furnace. The O₃ concentration was measured by an ozone monitor (Model 106-M, 2B Technologies, Boulder, CO, USA). All measurements were conducted after the reaction reached a steady state and were repeated five times. The toluene conversion (η_toluene), mineralization efficiency (S_COx), and CO₂ (S_CO2) selectivity were calculated as follows:

$${\mathsf{\eta }}_{\mathrm{t}\mathrm{o}\mathrm{l}\mathrm{u}\mathrm{e}\mathrm{n}\mathrm{e}}\left(\%\right)={\displaystyle \frac{C_{inlet}-C_{outlet}}{C_{inlet}}}\times 100\%$$

$$S_{{CO}_x}\left(\%\right)={\displaystyle \frac{C_{CO}+C_{{CO}_2}}{7\times (C_{inlet}-C_{outlet})}}\times 100\%$$

$$S_{{CO}_2}\left(\%\right)={\displaystyle \frac{C_{{CO}_2}}{7\times (C_{inlet}-C_{outlet})}}\times 100\%$$

where $C_{inlet}$ and $C_{outlet}$ represent the inlet and outlet concentrations of toluene, respectively, and $C_{\mathrm{CO}}$ and $C_{{CO}_2}$ denote the outlet concentrations of CO and CO₂.