Enhanced plasma-catalytic decomposition of toluene over Co–Ce binary metal oxide catalysts with high energy efficiency

Abstract

In-plasma catalysis has been considered as a promising technology to degrade volatile organic compounds. Heterogeneous catalysts, especially binary metal oxide catalysts, play an important role in further advancing the catalytic performance of in-plasma catalysis. This work investigates the toluene decomposition performance over Co–Ce binary metal oxide catalysts within the in-plasma catalysis. Co–Ce catalysts with different Co/Ce molar ratios are synthesized by a citric acid method. Results show that the catalytic activity of Co–Ce catalysts is obviously superior to those of monometallic counterparts. Especially, Co_0.75Ce_0.25O_x catalyst simultaneously realizes highly efficient toluene conversion (with a decomposition efficiency of 98.5% and a carbon balance of 97.8%) and a large energy efficiency of 7.12 g kW h⁻¹, among the best performance in the state-of-art literature (0.42 to 6.11 g kW h⁻¹). The superior catalytic performance is further interpreted by the synergistic effect between Co and Ce species and the significant plasma–catalyst interaction. Specifically, the synergistic effect can decrease the catalyst crystallite size, enlarge the specific surface area and improve the amount of oxygen vacancies/mobility, providing more active sites for the adsorption of surface active oxygen species. Meanwhile, the plasma–catalyst interaction is able to generate the surface discharge and reinforce the electric field strength, thereby accelerating the plasma-catalytic reactions. In the end, the plasma-catalytic reaction mechanism and pathways of toluene conversion are demonstrated.

---

Introduction

Volatile organic compounds (VOCs) are atmospheric contaminations derived from industrial, commercial and agricultural processes, which can be detrimental to human health and the environment.^1,2 Non-thermal plasma (NTP) consisting of energetic electrons and active species has exhibited unique advantages for VOC decomposition, such as quick response, moderate capital cost and easy operation under ambient conditions.^3,4 However, the formation of undesired by-products and poor energy efficiency significantly restrict the industrial application of NTP technology.

Combining NTP with catalysts (i.e., plasma catalysis) has been considered as a prospective approach to overcome these drawbacks.^2,5 Based on the location of catalysts, the plasma catalysis can be classified into two configurations, i.e., in-plasma catalysis (IPC, placing the catalysts in the discharge zone) and post-plasma catalysis (PPC, placing the catalysts downstream the plasma reactor). In IPC, the short-lived active species including O₂⁺, O₂⁻ and produced by NTP can reach the catalyst surface for efficient utilization, meanwhile the electric field strength of plasma discharge is reinforced by the integration of catalysts, thus yielding more highly energetic species for accelerating the catalytic reactions.^6,7 Catalysts are identified to play an important role in further advancing the plasma-catalytic performance and energy efficiency of IPC system.⁸ Heterogeneous catalysts have been employed for VOC decomposition, e.g., noble metals,^9,10 metal oxides,^11–14 and their combinations.^15–21 Among these catalysts, metal oxide catalysts (e.g., CoO_x,¹¹ MnO_x,¹² CuO¹³ and CeO₂ ¹⁴) attract tremendous interest in consideration of the high catalytic activity, good economics and superior poisoning resistance. Especially, binary metal oxide catalysts (e.g., MnCoO_x,¹⁶ CuCeO_x,¹⁷ CeMnO_x ¹⁸), synthesized by two types of metal salt precursors, are being regarded as preeminent candidates for IPC because of the potential complementary advantages from different components for improving the physicochemical property and catalytic activity.

Cobalt oxide (CoO_x) with excellent reduction ability, abundant oxygen vacancy and electrophilic oxide species has been proved as one of the most efficient metal oxide catalysts for VOC decomposition.^22,23 Meanwhile, cerium oxide (CeO_x) is being considered as a promising catalyst for VOC oxidation due to its high oxygen mobility/storage capacity associated with the redox shift between Ce³⁺ and Ce⁴⁺.^17,24 As a consequence, developing Co–Ce binary metal oxide catalysts exhibits great potential to further promote the catalytic activity in consideration of the possible synergistic effect between Co and Ce species. In fact, the collaborative interaction has been demonstrated in the thermal catalytic oxidation of VOCs.^25,26 For example, Akram et al.²⁶ reported that the decomposition temperature of ethyl acetate using CoCeO_x catalyst (T = 195 °C) was exceptionally lower than those of CeO₂ (T = 215 °C) and Co₃O₄ (T = 244 °C). However, the catalytic performance of VOC decomposition over Co–Ce binary metal oxide catalysts in plasma catalysis has not yet been reported. Moreover, an in-depth insight into the reaction mechanism in terms of Co–Ce synergistic effect and plasma–catalyst interaction is also crucial to advance our understanding towards VOC conversion process.

In this work, the toluene decomposition performance over Co–Ce binary metal oxide catalysts in IPC was investigated. Co–Ce catalysts with different Co/Ce molar ratios were synthesized by the citric acid method. The catalytic activity in terms of toluene decomposition efficiency, CO₂ selectivity, carbon balance and the formation of by-products was comprehensively evaluated. Subsequently, a detailed characterization on the structure and redox properties of Co–Ce catalysts was conducted to unveil the synergistic effect between Co and Ce species. Moreover, the influence of catalysts on the plasma discharge behavior (including current/voltage magnitude, discharge mode and electric field strength) was also explored. In the end, the underlying reaction mechanism and pathways of toluene decomposition over Co–Ce binary metal oxide catalysts within IPC were demonstrated.

Experimental

Preparation of Co–Ce binary metal oxide catalysts

Co–Ce binary metal oxide catalysts were prepared by the citric acid method. Cobalt nitrate and ceria nitrate were mixed in deionized water under stirring. Then the citric acid (1.2 : 1 molar ratio of the acid to metal nitrate) was dissolved in the solution as the complexing agent, followed by the addition of polyglycol (10% weight of citric acid). The as-obtained solution was continuously stirred in the magnetic stirrer at 80 °C until a gel formed. The samples were dried at 110 °C overnight and subsequently calcined at 500 °C during the next 4 h. Co/Ce molar ratios (i.e., Co_yCe_1−yO_x, y = 0, 0.25, 0.5, 0.75 and 1) could be regulated with adjusting the molar ratio of cobalt nitrate to ceria nitrate. Conceptually, y = 0 and y = 1 denoted the CeO_x and CoO_x catalyst, respectively. All the samples were sieved to 40–60 meshes before the catalytic performance measurement.

Characterization

The X-ray diffraction (XRD) results were recorded by an XRD-6000 Diffractometer with Cu Kα radiation. A DXR 532 Raman spectrometer (532 nm excitation wavelength, Thermo Fisher Scientific) was used to gain the Raman spectra of catalysts. The specific surface area, total pore volume and average pore diameter of the Co–Ce binary metal oxide catalysts were determined by N₂ adsorption–desorption measurement. The oxidation states of catalyst surface species were obtained by X-ray photoelectron spectroscopy (XPS) analysis using a VG Escalab Mark II equipment. The reducibility of Co–Ce binary metal oxides was measured through hydrogen temperature-programmed reduction (H₂-TPR) experiments with a chemisorption analyser (Micrometrics, Autochem II 2920).

Experimental setup

The experimental setup for catalytic decomposition of toluene in IPC is presented in Fig. 1. The plasma-catalytic system consisted of the gas supply system, the dielectric barrier discharge (DBD) reactor, the high voltage power supply as well as the gas analysis system. The gaseous toluene was obtained with a bubbler, which was diluted with pure air (99.999%, Jingong, Hangzhou). The mass flow controller (MFC, Sevenstars D08-3F) was used to maintain the gas flow rate constant during the test (i.e., 0.5 L min⁻¹). The initial concentration of toluene was 210 ppm. The DBD reactor is illustrated detailedly in Fig. 1. An aluminium foil (40 mm long) wrapped around a cylindrical quartz tube (10 mm outer diameter and 0.8 mm wall thickness, respectively), which was used as the ground electrode. A coaxial stainless steel rod (3.5 mm diameter) served as the high voltage electrode. The catalyst (0.2 g) was packed into the gap between the stainless steel rod and quartz wall (approx. 80% volume fraction of the whole discharge zone) and a pair of polytetrafluoroethylene (PTFE) heads were used to seal the plasma reactor.

Fig. 1 Schematic diagram of the experimental setup.

The high voltage power supply (CTP-2000K) connected with the DBD reactor was used to generate the plasma discharge. The electrical signals were recorded using the oscilloscope (MDO 3034). The discharge powers of DBD reactor were calculated by a Q-U Lissajous method, and the specific input energy (SIE) could be calculated as follows:

SIE (J L⁻¹) = P/Q × 60, (1)

where P (W) is the discharge power in IPC and Q (L min⁻¹) is 0.5 L min⁻¹ in this work.

The analysis of outlet gas samples was continuously performed by gas chromatography (GC, GC9790 Plus, Fuli) equipped with two flame ionization detectors. One was connected to a nickel conversion furnace for measuring the methane, which was converted from carbon dioxide and carbon monoxide, and the other one was used for toluene measurement. The concentrations of O₃ in the effluent were determined by an ozone monitor (Models 106-MH). The concentrations of NO_x were measured by a NO_x analyzer (Testo 350XL). The analysis of organic by-products in the effluent was performed with gas chromatography with mass spectrometry (GC-MS, Agilent 7890A GC/5957 MS). To ensure the reliability of tests, the plasma-catalytic system was run for 30 min before measurements until a steady state of effluent compositions was achieved.

The toluene decomposition efficiency (η_toluene), selectivity of CO₂ (S_CO2), carbon balance (S_COx) and energy efficiency (η_energy) could be calculated as:

η_toluene (%) = (c_inlet − c_outlet)/c_inlet × 100% (2)

S_CO2 (%) = c_CO2/[7 × (c_inlet − c_outlet)] × 100% (3)

S_COx (%) = (c_CO2 + c_CO)/[7 × (c_inlet − c_outlet)] × 100% (4)

η_energy (g kW h⁻¹) = 3.6 × M × c_inlet × η_toluene/(SIE × 24.04), (5)

where c_inlet (ppm) and c_outlet (ppm) are toluene concentrations of inlet and outlet, respectively; c_CO2 (ppm) and c_CO (ppm) are the concentrations of carbon dioxide and monoxide in the effluent, respectively; M is the molar mass of toluene.

Results and discussion

Plasma-catalytic performance of Co–Ce catalysts

The catalytic activity of Co–Ce binary metal oxide catalysts towards toluene oxidation in IPC is evaluated. As presented in Fig. 2a, the toluene removal efficiency increases monotonically with enlarging the SIE. The higher SIE can reinforce the electric field strength in the discharge region, thus generating more highly energetic electrons and active particles to accelerate the reactions.²⁷ Compared with NTP system, the introduction of catalysts remarkably improves the toluene decomposition efficiency (by 35.0%), indicating the strong plasma–catalyst interaction. More importantly, the removal efficiency of Co–Ce binary metal oxide catalysts is much higher than those of monometallic counterparts. To be specific, the decomposition efficiencies of Co_0.75Ce_0.25O_x, Co_0.5Ce_0.5O_x and Co_0.25Ce_0.75O_x are 98.5% (at 401 J L⁻¹), 91.9% (at 394 J L⁻¹) and 86.5% (at 381 J L⁻¹), respectively, superior to those of CeO_x (75.5% at 391 J L⁻¹) and CoO_x (82.5% at 400 J L⁻¹), demonstrating an optimum catalytic performance when using the Co_0.75Ce_0.25O_x catalyst. The excellent toluene oxidation can be further validated by the lowest O₃ concentration (81 ppm at 401 J L⁻¹) with the Co_0.75Ce_0.25O_x catalyst (Fig. S1a†).

Fig. 2 (a) Decomposition efficiency, (b) CO₂ selectivity, (c) carbon balance of Co–Ce catalysts as a function of SIE. (d) Comparison of toluene decomposition efficiency and energy efficiency between this work (using Co_0.75Ce_0.25O_x catalyst) and previous studies.

The reaction rate of toluene oxidizing with atomic oxygen (7.6 × 10⁻¹⁴ cm³ per molecule per s) is much higher than that with O₃ counterpart (3.9 × 10⁻²² cm³ per molecule per s).²⁸ Thus, the high conversion efficiency of O₃ to atomic oxygen is beneficial for superior toluene oxidation performance. Besides, the concentration of NO_x is also measured (Fig. S1b†), in which the lowest NO₂ content (19 ppm at 401 J L⁻¹) is recognized when using the Co_0.75Ce_0.25O_x catalyst.

In addition to the efficient decomposition, the complete conversion of toluene to CO_x, especially CO₂, is also highly desired. As presented in Fig. 2b and c, coupling the NTP with Co–Ce catalysts significantly improves the CO₂ selectivity (by 44.8%) and carbon balance (by 37.7%). Among the catalysts, Co_0.75Ce_0.25O_x catalyst exhibits the optimum CO₂ selectivity of 87.3% and carbon balance of 97.8% at 401 J L⁻¹, demonstrating a high degree of toluene conversion to CO₂ and CO_x. The durability and reusability of Co_0.75Ce_0.25O_x catalyst are further measured at 298 J L⁻¹. The catalytic performance of Co_0.75Ce_0.25O_x catalyst decreases slightly after 12 h test (Fig. S2a†), demonstrating the excellent durability to resist deactivation. Moreover, the Co_0.75Ce_0.25O_x catalyst could be reused for 5 times with slight activity loss (less than 4%, Fig. S2b†), exhibiting a superior reusability.

A comparison of plasma-catalytic performance with respect to toluene decomposition efficiency and energy efficiency (i.e., the electrical energy input for per unit mass of VOCs) between this work and previous studies is conducted (Fig. 2d). Detailed boundary conditions of different catalytic systems for comparison are available in Table S1.† Using binary metal oxide catalyst of Co_0.75Ce_0.25O_x, this work simultaneously realizes enhanced toluene conversion (decomposition efficiency of 98.5%) and high energy efficiency of 7.12 g kW h⁻¹, superior to those in the state-of-art literatures (0.42 to 6.11 g kW h⁻¹).^18,29–37

Structure and redox properties of catalysts

The outstanding plasma-catalytic performance can be interpreted by the synergistic effect between Co and Ce species and obvious plasma–catalyst interaction. A comprehensive characterization on the structure and redox property of Co–Ce catalysts is thus carried out. According to the XRD patterns in Fig. 3a, Co–Ce binary metal oxide catalysts are composed of spinel phase of Co₃O₄ (JCPDS 43-1003) and fluorite structure of CeO₂ (JCPDS 34-0394). In particular, the characteristic peak of CeO₂ (111) crystal face in the Co–Ce binary metal oxide catalysts slightly shifts to higher Bragg angle region, i.e., from 28.7° of CeO₂ to 28.95° of Co–Ce catalysts (Fig. 3b). This phenomenon is primarily stimulated by the incorporation of Co cations into CeO₂ fluorite lattice, leading to the formation of Co–Ce solid solution. When Ce⁴⁺ (0.097 nm in ionic radius) is replaced by Co³⁺ (0.065 nm) and Co²⁺ (0.082 nm), a contraction of CeO₂ lattice occurs, thus developing plenty of defects in the lattice (Fig. 3c). These defects can increase the oxygen vacancies and contribute to the generation of surface oxygen species for improved catalytic activity.

Fig. 3 (a) Angle patterns and (b) enlarged zone patterns from XRD analysis. (c) Schematic diagram of Co–Ce solid solution. (d) Raman spectra and (e) N₂ adsorption–desorption isotherms.

The structural deformation in the crystallite lattice of Co–Ce catalysts can also be demonstrated by the Raman spectra. As shown in Fig. 3d, CeO_x shows a characteristic F_2g band of CeO₂ cubic fluorite structure at around 471.6 cm^−1 38 and CoO_x presents five characteristic peaks of Co₃O₄ spinel phase distinguished as F¹_2g, E_g, F²_2g, F³_2g and A_1g.³⁹ In comparison with CeO_x, the peak of F_2g bands in the binary metal oxide catalyst shifts to lower frequency (452.3 cm⁻¹), confirming the incorporation of Co cations into the CeO₂ fluorite lattice and the deformation of lattice framework.⁴⁰ Similar phenomenon is recognized on the A_1g bands of Co–Ce catalysts, highlighting the obvious Co–Ce interaction.

Moreover, the characteristic diffraction peaks of Co–Ce oxides are much broader and weaker than those of monometallic counterparts (Fig. 3a), implying the decrease of crystallite size in the catalyst. To quantitatively characterize the catalyst structure, the CeO₂ crystallite size in Co–Ce catalysts is calculated by the Scherrer's equation:⁴¹

d = 0.89λ/(β cos θ), (6)

where d (nm) is the calculated crystallite size; λ (nm) is the wavelength of incident X-ray from Cu source; β is the peak width at half peak height; θ is the X-ray incidence angle. The CeO₂ crystallite sizes in Co_0.75Ce_0.25O_x, Co_0.5Ce_0.5O_x, Co_0.25Ce_0.75O_x and CeO_x are calculated to be 9.6, 11.9, 12.6 and 14.2 nm, respectively, exhibiting a minimum value at Co_0.75Ce_0.25O_x (decreased by 32.4% in comparison with the monometallic counterpart) (Table S2†). This observation is presumably affiliated with the suppressed growth of CeO₂ crystallite by Co doping.

As presented in Fig. 3e, N₂ adsorption–desorption isotherms are further carried out to quantify the catalyst structure. Based on this, the specific surface area, total pore volume and average pore size of samples can be calculated (Table S2†). The specific surface area of Co–Ce binary metal oxide catalysts (46.0–56.0 m² g⁻¹) is much larger than those of CoO_x and CeO_x (12.1 and 25.6 m² g⁻¹, respectively). Among Co–Ce catalysts, Co_0.75Ce_0.25O_x sample exhibits the largest specific surface area (56.0 m² g⁻¹), which is 3.6 folds greater than that of CoO_x. The higher surface area can provide abundant active sites for efficient oxidation reaction, consistent with the as-obtained superior plasma-catalytic performance using Co_0.75Ce_0.25O_x catalyst.

XPS analysis is performed to unveil the oxidation state of catalyst surface species, which plays a crucial role in determining the oxygen vacancies and oxygen mobility for reaction.^15,42 Two peaks located at 779.8 and 795.0 eV are recognized (Fig. 4a), indexed to Co 2p_3/2 and Co 2p_1/2, respectively. The spin–orbit splitting within Co 2p peaks (15.2 eV) reveals that Co₃O₄ is the primary phase of cobalt oxides in the catalyst,²⁶ in good accordance with XRD results. As shown in Fig. 4b, the Ce 3d spectra are deconvolved into eight peaks, in which the spin–orbit components of Ce 3d_5/2 and Ce 3d_3/2 are denoted by U and V, respectively. Ratios of Ce³⁺/(Ce³⁺ + Ce⁴⁺) are calculated by integrating the peak area (Table S3†). The content of Ce³⁺ increases with incorporating the Co species, which are estimated to be 12.7%, 16.2% and 19.2% for Co_0.25Ce_0.75O_x, Co_0.5Ce_0.5O_x and Co_0.75Ce_0.25O_x samples, respectively. The higher Ce³⁺ content is able to generate abundant defects in the CeO₂ lattice,^43,44 which facilitates more oxygen vacancies over catalyst surface and adsorption centers for active oxygen species, thus boosting the toluene conversion reaction.

Fig. 4 (a) Co 2p, (b) Ce 3d and (c) O 1s spectra of XPS analysis. (d) H₂-TPR profiles of Co–Ce binary metal oxide catalysts.

As shown in Fig. 4c, O 1s spectra are fitted into two peaks. The peaks at around 532.0 and 529.8 eV can be identified as the surface absorbed oxygen species (labeled by O_s) and lattice oxygen (labeled by O_l), respectively.^44,45 With increasing the amount of Co dopant, the ratio of O_s/(O_s + O_l) increases remarkably from 20.6% to 31.0% (Table S3†). In consideration of the high mobility of O_s, the larger ratio of O_s/(O_s + O_l) is beneficial for yielding improved catalytic activity. As expected, Co_0.75Ce_0.25O_x catalyst exhibits the highest Ce³⁺ and O_s concentration, agreeing well with the catalytic performance.

H₂-TPR tests are carried out to evaluate the reducibility of catalysts. CeO_x sample exhibits two broad peaks at 483.6 and 773.9 °C,⁴⁶ and CoO_x presents a distinct peak at 395.9 °C with a small shoulder of 341.4 °C ⁴⁷ (Fig. 4d). As for the Co–Ce composite, three peaks labeled as α (285.6 °C), β (344.3 °C) and γ (484.7 °C) are recognized. The α peak stems from the reduction of Co³⁺ to Co²⁺ cations while β and γ peaks are predominantly correlated with the reduction of Co²⁺ to Co in the Co₃O₄ phase and solid solution of Co–Ce, respectively.⁴² The temperature of α peak in the Co–Ce catalyst (285.6 °C) is obviously lower than those of CoO_x (341.4 °C) and CeO_x (483.6 °C), indicating the improved reducibility. This phenomenon is associated with the smaller crystallite size of Co–Ce samples, which promotes the formation of easily reducible sites and improved oxygen mobility for efficient toluene conversation. Detailed characterizations on the structure and redox property of used Co_0.75Ce_0.25O_x catalyst after plasma catalysis are available in Fig. S3, Tables S2 and S3.† It is found that the plasma-catalytic reaction exhibits a negative impact on the structure and redox property of the Co_0.75Ce_0.25O_x catalyst, which is mainly ascribed to the accumulation of intermediates on the catalyst surface,¹⁷ the partly oxidation from Ce³⁺ to Ce^4+ 48 and the consumption of O_s during the toluene decomposition.⁴⁹

Plasma–catalyst interaction

In IPC, catalysts are integrated in the discharge region, which will affect the discharge behavior and thus determine the catalytic activity.^50–52 Fig. 5a and b show the electrical signals within the NTP and IPC (using the Co_0.75Ce_0.25O_x catalyst at 250 J L⁻¹). A typical filamentary discharge is observed for the NTP system, exhibiting obvious discharge current pulses in the waveform. Importantly, introducing catalyst into the discharge zone of the NTP noticeably affects the discharge behavior. On one hand, the presence of catalyst suppresses the current amplitude. The filamentary discharge can only be generated in the small void gap between catalyst particles and quartz wall. As a consequence, the packed catalysts in the NTP will reduce the available discharge volume, thus decreasing the current amplitude of filamentary discharge. On the other hand, the integration of catalyst markedly increases the amount of current pulses, which is predominantly attributed to the surface discharge generated over the catalyst surface. As schematically shown in Fig. 5c, the discharge mode is changed from the filamentary discharge (NTP) to the combined filamentary discharge and surface discharge (IPC), implying the enhanced ionization travelling along the catalyst surface for advanced catalytic reaction.⁵³

Fig. 5 Electrical signals in (a) NTP and (b) IPC. (c) Schematic diagram of discharge behavior. (d) Discharge voltages within the NTP and IPC.

The discharge voltage in the NTP and IPC system is also measured and compared. As presented in Fig. 5d, the magnitude of discharge voltage increases with enlarging the SIE. Importantly, the integration of catalysts can obviously extend the voltage amplitude (by > 0.5 kV), e.g., from 9.45 kV at 419 J L⁻¹ to 9.97 kV at 401 J L⁻¹. Enhanced electric field strength in the DBD reactor will produce more highly energetic electrons for accelerating the plasma-catalytic oxidation, demonstrating the significance of plasma–catalyst interaction in boosting the catalytic performance.

Mechanism of plasma-catalytic reaction

In order to reveal the reaction mechanism and pathways, the by-products in the outlet of the NTP and IPC system over the Co_0.75Ce_0.25O_x catalyst are measured by GC-MS (Fig. 6). Large amounts of toluene and organic intermediates, e.g., aldehydes, hydrocarbons, acids and benzene series, are observed after NTP system treated. While types and amounts of by-products significantly decrease in IPC, in good accordance with the efficient toluene conversion.

Fig. 6 Outlet gases of the NTP and IPC system at 250 J L⁻¹ measured by GC-MS method.

The toluene decomposition process in IPC system can be identified into two parts:^12,18 the gas-phase plasma reactions and the catalytic reactions on catalyst surface. According to GC-MS analysis, the reaction pathways of toluene degradation in gas phase are presented (Fig. 7).

Fig. 7 Mechanism of plasma-catalytic reaction over Co–Ce binary metal oxide catalysts.

The dissociation energy of C–H (methyl group), C–H (aromatic ring), C–C (methyl group-aromatic ring), C–C (aromatic ring) and C=C (aromatic ring) are 3.7 eV, 4.3 eV, 4.4 eV, 5.0–5.3 eV and 5.5 eV, respectively.^54,55 In NTP system, the energy of highly energetic electrons reaches up to 10 eV. Therefore, these chemical bonds can be destroyed by electrons. The destruction of C–H within methyl group might be primary process of plasma reactions due to the lowest dissociation energy, resulting in the generation of benzyl radicals. After that, the formed benzyl radicals further react with oxygen species (·O and ·OH) from plasma discharge to produce benzaldehyde, as demonstrated in GC-MS results.

Moreover, methyl and phenyl radicals can also be generated from the breaking of C–C connecting the methyl group with the aromatic ring. The reaction between methyl radicals and ·NO₂ contributes to the formation of nitromethane. Phenyl radicals could react with ·CH₃, ·H and ·NO₂ to produce o-xylene/p-xylene, benzene and nitrobenzene. Subsequently, the aromatic intermediates are attacked by highly energetic electrons (>5.5 eV) and active species in the plasma to produce ring-opening intermediates, which are finally decomposed to CO₂ and H₂O. Noteworthy, the introduction of Co–Ce binary metal oxide catalysts into the NTP system can yield stronger electric field and more highly energetic electrons to improve the catalytic activity.

Additionally, toluene and organic intermediates are also adsorbed on the surface of Co–Ce catalysts placed in the discharge region, which are further decomposed by surface active oxygen species (Fig. 7). On one hand, the surface active oxygen species are generated from adsorbed oxygen molecules attacked by highly energetic electrons (eqn (7)) and O₃ conversion reaction on catalyst surface (eqn (8) and (9)):^56,57

O₂ + e → O + O (7)

O₃ + * → O* + O₂ (8)

where * represents active centers on catalyst surface. On the other hand, surface active oxygen species could also be supplied from the redox cycle between Co and Ce cations (eqn (10)) according to the Mars-Van Krevelen mechanism:^18,25,58

Ce³⁺/Ce⁴⁺ ⇌ Co²⁺/Co³⁺ (10)

During the catalytic reactions on the surface of catalysts, active oxygen species are consumed by decomposition of absorbed organic molecules, meanwhile the Co³⁺ sites are reduced to Co²⁺. The redox pair of Ce³⁺/Ce⁴⁺ in CeO₂ facilitates the storage and release of lattice oxygen species, re-oxidizing Co²⁺ to Co³⁺. This redox cycle of Co²⁺/Co³⁺ is able to transmit the oxygen species to the surface of catalysts, thereby supplying the consumed surface active oxygen species. Additionally, the oxygen vacancies generated from lattice oxygen consumption via Co–Ce redox cycle can be replenished by gas phase oxygen species. Therefore, the synergistic effect between Co and Ce species promotes the generation and mobility of active surface oxygen species for superior catalytic performance.

Conclusion

This work investigated the catalytic performance of toluene decomposition over Co–Ce binary metal oxide catalysts in IPC. In comparison with the NTP system, the integration of Co–Ce catalysts significantly advances the toluene removal efficiency (by 35.0%), CO₂ selectivity (by 44.8%) and carbon balance (by 37.7%). Among the catalysts, the Co–Ce binary metal oxide catalysts present better catalytic activity than those of monometallic counterparts. Especially, Co_0.75Ce_0.25O_x simultaneously realizes high-performance toluene conversion (decomposition efficiency of 98.5%, CO₂ selectivity of 87.3% and carbon balance of 97.8% at 401 J L⁻¹) and large energy efficiency of 7.12 g kW h⁻¹, which is obviously higher than or comparable to those in the state-of-art literatures (0.42 to 6.11 g kW h⁻¹). The excellent catalytic performance could be attributed to the synergistic effect between Co and Ce species and crucial plasma–catalyst interaction. On one hand, structural characterizations including XRD, Raman, N₂ adsorption–desorption, XPS and H₂-TPR reveal that the collaborative interactions can decrease the catalyst crystallite size (by 32.4%), enlarge the specific surface area (by 3.6 times), enhance the amount of oxygen vacancies/mobility (6.5% and 10.4% increase in the Ce³⁺ and O_s concentration, respectively). These features are able to provide more active adsorption centers for the surface oxygen species and thus accelerate the toluene decomposition reactions. On the other hand, the plasma–catalyst interaction can reinforce the discharge voltage in the DBD reactor (by > 0.50 kV) and transform the discharge mode from filamentary discharge to the combined filamentary discharge and surface discharge, yielding improved catalytic reactions. In the end, based on the GC-MS observations, the plasma-catalytic reaction mechanism and pathways of toluene decomposition are demonstrated. The as-obtained insights open an alternative way of employing Co–Ce binary metal oxide catalysts to realize enhanced toluene conversion with high energy efficiency in the plasma catalysis.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors acknowledge the financial support from the National Natural Science Foundation of China (No. 51576175), the Zhejiang Provincial Natural Science Foundation of China (No. LR17E060002) and the Fundamental Research Funds for the Central Universities (No. 2018XZZX002-17).

References

1. H. H. Kim, Y. Teramoto, N. Negishi and A. Ogata, Catal. Today, 2015, 256, 13–22.
2. S. Veerapandian, C. Leys, N. De Geyter and R. Morent, Catalysts, 2017, 7, 113.
3. E. C. Neyts, K. K. Ostrikov, M. K. Sunkara and A. Bogaerts, Chem. Rev., 2015, 115, 13408–13446.
4. X. X. Feng, H. X. Liu, C. He, Z. X. Shen and T. B. Wang, Catal. Sci. Technol., 2018, 8, 936–954.
5. J. C. Whitehead, J. Phys. D: Appl. Phys., 2016, 49, 243001.
6. J. Van Durme, J. Dewulf, C. Leys and H. Van Langenhove, Appl. Catal., B, 2008, 78, 324–333.
7. H. L. Chen, H. M. Lee, S. H. Chen, M. B. Chang, S. J. Yu and S. N. Li, Environ. Sci. Technol., 2009, 43, 2216–2227.
8. H. H. Kim, Y. Teramoto, A. Ogata, H. Takagi and T. Nanba, Plasma Chem. Plasma Process., 2015, 36, 45–72.
9. H. H. Kim, S. M. Oh, A. Ogata and S. Futamura, Appl. Catal., B, 2005, 56, 213–220.
10. C. Ayrault, J. Barrault, N. Blin-Simiand, F. Jorand, S. Pasquiers, A. Rousseau and J. M. Tatibouët, Catal. Today, 2004, 89, 75–81.
11. X. B. Zhu, X. Gao, X. N. Yu, C. H. Zheng and X. Tu, Catal. Today, 2015, 256, 108–114.
12. M. J. Lu, R. Huang, P. T. Wang, L. M. Chen, J. L. Wu, M. L. Fu, W. Wen, B. C. Huang and D. Q. Ye, Plasma Chem. Plasma Process., 2014, 34, 1141–1156.
13. C. He, L. Cao, X. Liu, W. Fu and J. Zhao, Int. J. Environ. Sci. Technol., 2015, 12, 3531–3540.
14. J. L. Wu, Y. X. Huang, Q. B. Xia and Z. Li, Plasma Chem. Plasma Process., 2013, 33, 1073–1082.
15. N. Jiang, J. Hu, J. Li, K. F. Shang, N. Lu and Y. Wu, Appl. Catal., B, 2016, 184, 355–363.
16. H. Song, F. Hu, Y. Peng, K. Li, S. Bai and J. Li, Chem. Eng. J., 2018, 347, 447–454.
17. X. B. Zhu, X. Gao, R. Qin, Y. X. Zeng, R. Y. Qu, C. H. Zheng and X. Tu, Appl. Catal., B, 2015, 170–171, 293–300.
18. B. W. Wang, C. M. Chi, M. Xu, C. Wang and D. Meng, Chem. Eng. J., 2017, 322, 679–692.
19. Z. Bo, M. Yuan, S. Mao, X. Chen, J. H. Yan and K. F. Cen, Sens. Actuators, B, 2018, 256, 1011–1020.
20. Z. Bo, W. G. Zhu, W. Ma, Z. H. Wen, X. R. Shuai, J. H. Chen, J. H. Yan, Z. H. Wang, K. F. Cen and X. L. Feng, Adv. Mater., 2013, 25, 5799–5806.
21. Z. Bo, S. Mao, Z. J. Han, K. Cen, J. Chen and K. K. Ostrikov, Chem. Soc. Rev., 2015, 44, 2108–2121.
22. A. Rokicińska, P. Natkański, B. Dudek, M. Drozdek, L. Lityńska-Dobrzyńska and P. Kuśtrowski, Appl. Catal., B, 2016, 195, 59–68.
23. A. Białas, M. Mazur, P. Natkański, B. Dudek, M. Kozak, A. Wach and P. Kuśtrowski, Appl. Surf. Sci., 2016, 362, 297–303.
24. Z. M. Li, X. L. Guo, F. Tao and R. X. Zhou, RSC Adv., 2018, 8, 25283–25291.
25. C. Wang, C. H. Zhang, W. C. Hua, Y. L. Guo, G. Z. Lu, S. Gil and A. Giroir-Fendler, Chem. Eng. J., 2017, 315, 392–402.
26. S. Akram, Z. Wang, L. Chen, Q. Wang, G. L. Shen, N. Han, Y. F. Chen and G. L. Ge, Catal. Commun., 2016, 73, 123–127.
27. W. Abou Saoud, A. A. Assadi, M. Guiza, A. Bouzaza, W. Aboussaoud, A. Ouederni, I. Soutrel, D. Wolbert and S. Rtimi, Appl. Catal., B, 2017, 213, 53–61.
28. J. Van Durme, J. Dewulf, W. Sysmans, C. Leys and H. Van Langenhove, Chemosphere, 2007, 68, 1821–1829.
29. X. X. Xu, P. T. Wang, W. C. Xu, J. L. Wu, L. M. Chen, M. L. Fu and D. Q. Ye, Chem. Eng. J., 2016, 283, 276–284.
30. Y. F. Guo, D. Q. Ye and K. F. Chen, Plasma Chem. Plasma Process., 2006, 26, 237–249.
31. Y. F. Guo, D. Q. Ye, K. F. Chen, J. C. He and W. L. Chen, J. Mol. Catal. A: Chem., 2006, 245, 93–100.
32. T. Zhu, J. Li, W. Liang and Y. Jin, J. Hazard. Mater., 2009, 165, 1258–1260.
33. H. B. Huang, D. Q. Ye, D. Y. C. Leung, F. D. Feng and X. J. Guan, J. Mol. Catal. A: Chem., 2011, 336, 87–93.
34. H. T. Quoc An, T. Pham Huu, T. Le Van, J. M. Cormier and A. Khacef, Catal. Today, 2011, 176, 474–477.
35. H. T. Quoc An, T. Pham Huu, T. Le Van, J. M. Cormier and A. Khacef, Catal. Today, 2011, 176, 474–477.
36. X. X. Xu, J. L. Wu, W. C. Xu, M. L. He, M. L. Fu, L. M. Chen, A. M. Zhu and D. Q. Ye, Catal. Today, 2017, 281, 527–533.
37. X. Yao, J. Zhang, X. Liang and C. Long, Chemosphere, 2018, 208, 922–930.
38. M. Fernández-García, A. Martínez-Arias, A. Iglesias-Juez, C. Belver, A. B. Hungría, J. C. Conesa and J. Soria, J. Catal., 2000, 194, 385–392.
39. A. Buenolopez, I. Suchbasanez and C. Salinasmartinezdelecea, J. Catal., 2006, 244, 102–112.
40. J. E. Spanier, R. D. Robinson, F. Zhang, S. W. Chan and I. P. Herman, Phys. Rev. B: Condens. Matter Mater. Phys., 2001, 64, 245407.
41. X. Tu, H. J. Gallon and J. C. Whitehead, Catal. Today, 2013, 211, 120–125.
42. J. N. Xu, G. Z. Lu, Y. Guo, Y. L. Guo and X. Q. Gong, Appl. Catal., A, 2017, 535, 1–8.
43. S. Gil, J. Garcia-Vargas, L. Liotta, G. Pantaleo, M. Ousmane, L. Retailleau and A. Giroir-Fendler, Catalysts, 2015, 5, 671–689.
44. G. L. Zhou, H. Lan, T. T. Gao and H. M. Xie, Chem. Eng. J., 2014, 246, 53–63.
45. T. Cai, H. Huang, W. Deng, Q. G. Dai, W. Liu and X. Y. Wang, Appl. Catal., B, 2015, 166–167, 393–405.
46. W. J. Hong, S. Iwamoto, S. Hosokawa, K. Wada, H. Kanai and M. Inoue, J. Catal., 2011, 277, 208–216.
47. Z. P. Qu, K. Gao, Q. Fu and Y. Qin, Catal. Commun., 2014, 52, 31–35.
48. W. J. Shan, M. Fleys, F. Lapicque, D. Swierczynski, A. Kiennemann, Y. Simon and P. M. Marquaire, Appl. Catal., A, 2006, 311, 24–33.
49. X. B. Zeng, Y. Xu, B. Zhang, G. Q. Luo, P. Sun, R. J. Zou and H. Yao, Chem. Eng. J., 2017, 309, 503–512.
50. D. H. Mei, X. B. Zhu, Y. L. He, J. D. Yan and X. Tu, Plasma Sources Sci. Technol., 2014, 24, 015011.
51. X. Tu, H. J. Gallon and J. C. Whitehead, J. Phys. D: Appl. Phys., 2011, 44, 482003.
52. X. Tu and J. C. Whitehead, Appl. Catal., B, 2012, 125, 439–448.
53. H. H. Kim, J. H. Kim and A. Ogata, J. Phys. D: Appl. Phys., 2009, 42, 135210.
54. S. Y. Liu, D. H. Mei, L. Wang and X. Tu, Chem. Eng. J., 2017, 307, 793–802.
55. S. Y. Liu, D. H. Mei, M. A. Nahil, S. Gadkari, S. Gu, P. T. Williams and X. Tu, Fuel Process. Technol., 2017, 166, 269–275.
56. M. N. Lyulyukin, A. S. Besov and A. V. Vorontsov, Appl. Catal., B, 2016, 183, 18–27.
57. W. J. Liang, L. Ma, H. Liu and J. Li, Chemosphere, 2013, 92, 1390–1395.
58. X. B. Zhu, S. Y. Liu, Y. X. Cai, X. Gao, J. S. Zhou, C. H. Zheng and X. Tu, Appl. Catal., B, 2016, 183, 124–132.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c9ra00794f

---