High-Efficiency Catalytic Conversion of NO_x by the Synergy of Nanocatalyst and Plasma: Effect of Mn-Based Bimetallic Active Species

Abstract

Three typical Mn-based bimetallic nanocatalysts of Mn−Fe/TiO₂, Mn−Co/TiO₂, Mn−Ce/TiO₂ were synthesized via the hydrothermal method to reveal the synergistic effects of dielectric barrier discharge (DBD) plasma and bimetallic nanocatalysts on NO_x catalytic conversion. The plasma-catalyst hybrid catalysis was investigated compared with the catalytic effects of plasma alone and nanocatalyst alone. During the catalytic process of catalyst alone, the catalytic activities of all tested catalysts were lower than 20% at ambient temperature. While in the plasma-catalyst hybrid catalytic process, NO_x conversion significantly improved with discharge energy enlarging. The maximum NO_x conversion of about 99.5% achieved over Mn−Ce/TiO₂ under discharge energy of 15 W·h/m³ at ambient temperature. The reaction temperature had an inhibiting effect on plasma-catalyst hybrid catalysis. Among these three Mn-based bimetallic nanocatalysts, Mn−Ce/TiO₂ displayed the optimal catalytic property with higher catalytic activity and superior selectivity in the plasma-catalyst hybrid catalytic process. Furthermore, the physicochemical properties of these three typical Mn-based bimetallic nanocatalysts were analyzed by N₂ adsorption, Transmission Electron Microscope (TEM), X-ray diffraction (XRD), H₂-temperature-programmed reduction (TPR), NH₃-temperature-programmed desorption (TPD), and X-ray photoelectron spectroscopy (XPS). The multiple characterizations demonstrated that the plasma-catalyst hybrid catalytic performance was highly dependent on the phase compositions. Mn−Ce/TiO₂ nanocatalyst presented the optimal structure characteristic among all tested samples, with the largest surface area, the minished particle sizes, the reduced crystallinity, and the increased active components distributions. In the meantime, the ratios of Mn⁴⁺/(Mn²⁺ + Mn³⁺ + Mn⁴⁺) in the Mn−Ce/TiO₂ sample was the highest, which was beneficial to plasma-catalyst hybrid catalysis. Generally, it was verified that the plasma-catalyst hybrid catalytic process with the Mn-based bimetallic nanocatalysts was an effective approach for high-efficiency catalytic conversion of NO_x, especially at ambient temperature.

1. Introduction

Nitrogen oxides (NO_x) are regard as the main air pollutant contributing to acid rain, photochemical smog, greenhouse effects, and ozone depletion [1]. Selective catalytic reduction (SCR) of NO_x by NH₃ or urea is proposed to be the highly effective and completely developed method to eliminate NO_x pollution [2]. In coal fired power plants, the commercial catalyst of V₂O₅-WO₃(MoO₃)/TiO₂ is used for its excellent catalytic performance in the typical standard SCR reaction [3]:

$${4\mathrm{NO}+4\mathrm{NH}}_3{+\mathrm{O}}_2\to {4\mathrm{N}}_2{+6\mathrm{H}}_2\mathrm{O}$$

(1)

While the V₂O₅−WO₃(MoO₃)/TiO₂ catalysts demand a strict temperature window of 300–400 °C, which limit the arrangement flexibility of this kind of catalyst. The vanadium-based catalysts can not reach satisfactory efficiency of eliminating NO_x when the reaction temperature is lower than 250 °C. In recent years, the fast selective catalytic reduction (fast SCR) attracted the attention of many research groups due to its lower reaction temperature and higher reaction efficiency [4]:

$${\mathrm{NO}+\mathrm{NO}}_2{+2\mathrm{NH}}_3\to {2\mathrm{N}}_2{+3\mathrm{H}}_2\mathrm{O}$$

(2)

The catalysts appropriate to low temperature SCR are strongly desired, which could be located at downstream electrostatic precipitator and desulfurizer suitably [5]. However, the fast SCR still needs reaction temperature within 150–300 °C to achieve high efficiency of NO_x elimination [4,6]. Furthermore, the mole ratio of NO:NO₂ maintained at 1:1 is difficult in the real flue gas. Hence, it is necessary to develop an effective approach to eliminate NO_x with light concentration of NO₂ at low temperature region, which could be beneficial to the deNO_x device arrangement, as well as the SO₂ resistance.

Plasma-catalyst hybrid catalysis has been proved as an efficient technology to unite the high reactivity of plasma and the high selectivity of catalyst [7,8,9]. During the plasma-catalyst hybrid process, the plasma modifies not only the chemical properties and morphologies of the catalysts, but also changes the reaction pathway of an original catalytic process [10]. Plasma is confirmed to form an abundance of active species, such as O and O₃ radicals, which could oxidize NO into NO₂, further promoting catalysis via the fast SCR approach, especially at low temperature [4]. For the plasma-catalyst hybrid catalysis, the catalysts of V₂O₅−WO₃/TiO₂ [11], Ag/r-Al₂O₃ [12], Cu-ZSM-5 [13], and Mn−Ce/ZSM5−MWCNTs [4] have presented acceptable NO_x conversion efficiency under relatively low specific input energy. While the NO_x conversion maximum could still be further promoted at lower reaction temperature and smaller energy consume. Among the various transition metal elements applied in the catalysts for NO_x reduction, manganese displays superior activity especially at the low temperature, which can be attributed to the multifarious types of labile oxygen and high mobility of valence states [1]. Meanwhile, it has been found that iron, cobalt, and cerium species can combine with manganese to produce bimetallic catalysts, which contain abundant oxygen vacancies on the catalyst surface, forming strong interaction bands at atomic scale, such as Mn-O-Fe [14], Mn-O-Co [15], and Mn-O-Ce [16]. Moreover, the active metal species of FeO_x, CoO_x, and CeO_x are also regarded as the three typical promoters for NO_x conversion, which serve as core catalyst components of active metal oxides, supplying surface oxygen to accelerate NO_x elimination [14,15,17]. However, the effects of Mn-based bimetallic catalysts on the plasma-catalyst hybrid catalysis, especially the Mn−Fe/TiO₂, Mn−Co/TiO₂, and Mn−Ce/TiO₂ nanocatalysts have not been explored clearly.

In this study, we systematically synthesized three typical Mn-based bimetallic nanocatalysts of Mn−Fe/TiO₂, Mn−Co/TiO₂, and Mn−Ce/TiO₂. The synergistic effects of non-thermal plasma and Mn-based bimetallic nanocatalysts on NO_x catalytic conversion were investigated compared with the catalytic effects of plasma alone and nanocatalysts alone. Meanwhile, the influence factors of reaction temperature and discharge energy were taken into consideration during studying the synergetic mechanisms focusing on NO_x conversion of plasma and bimetallic nanocatalysts hybrid system. Furthermore, the physicochemical properties of these three typical Mn-based bimetallic nanocatalysts were analyzed by Brunauer-Emmett-Teller (BET), transmission electron microscopy (TEM), X-ray diffraction (XRD), H₂-temperature-programmed reduction (TPR), NH₃-temperature-programmed desorption (TPD) and X-ray photoelectron spectroscopy (XPS), in order to expose the relationship between structures and activities. The purpose of this work was mean to explore the synergistic reinforcement mechanism of plasma-catalysis hybrid catalytic process over Mn-based bimetallic nanocatalysts for NO_x elimination with high catalytic efficiency and satisfied catalytic selectivity, especially at atmospheric temperature.

2. Results and Discussion

2.1. NO_x Conversion of Catalyst Alone Catalytic Process

The NO_x catalytic conversion and the catalytic selectivity of three typical Mn-based bimetallic nanocatalysts of Mn−Fe/TiO₂, Mn−Co/TiO₂, and Mn−Ce/TiO₂ were exhibited in Figure 1, and the catalytic ability of Mn/TiO₂ catalyst was also depicted as a contrast. For all the tested Mn-based bimetallic nanocatalysts, the NO_x conversion increased significantly with the temperature rising from 25 °C to 250 °C and presented excellent performance (>90%, above 150 °C). Compared with the Mn/TiO₂ catalyst, the catalytic activities of Mn-based bimetallic nanocatalysts were remarkably improved at the whole temperature range, potentially due to the strong interaction of Mn−O−X bond (X refered to Fe, Co or Ce), the improvement of Brønsted acid sites and Lewis acid sites, and the enhancement of Eley-Rideal (E-R) mechanism reaction [18], which could be further verified by the following physicochemical properties. As shown in Figure 1a, Mn−Ce/TiO₂ nanocatalyst achieved higher catalytic activity than the other samples in the temperature range of 25~200 °C. The Mn−Fe/TiO₂ nanocatalyst showed the minimum NO_x conversion among these three Mn-based bimetallic nanocatalysts, while still much larger than that of Mn /TiO₂ sample. However, the catalytic selectivity of Mn−Fe/TiO₂ nanocatalyst was lower than that of Mn−Co/TiO₂ and Mn−Ce/TiO₂ within 175~250 °C, as exhibited in Figure 1b. Furthermore, it could be easy to find there was no obvious difference of NO_x conversion or catalytic selectivity over these three Mn-based bimetallic nanocatalysts at ambient temperature, which was proposed to be due to the low catalytic activities for all the tested catalysts.

2.2. NO_x Conversion of Plasma-Catalyst Hybrid Catalytic Process

The NO conversion and NO₂ concentration over three typical Mn-based bimetallic nanocatalysts were compared in Figure 2. The performance of all prepared nanocatalysts were measured in terms of various discharge energies to reveal the interaction of Mn−O−Fe, Mn−O−Co, and Mn−O−Ce. As shown in Figure 2a, both Mn−Co/TiO₂ and Mn−Ce/TiO₂ nanocatalysts could reach NO_x conversion maximum >99% within the discharge energy range of 18~24 W·h/m³. While the start discharge energy of Mn−Ce/TiO₂ nanocatalyst with superior SCR activities was much lower than that of Mn−Co/TiO₂nanocatalyst. The Mn−Ce/TiO₂ bimetallic nanocatalyst raised the optimal NO_x conversion to 93.3% with the relatively low discharge energy of 12 W·h/m³. For the other Mn-based bimetallic nanocatalysts, a lower NO_x elimination efficiency was achieved with NO_x conversion less than 85% at 15 W·h/m³ and the maximum obtained at 24 W·h/m³, which meant that the higher discharge energy was required to induce the plasma-catalyst catalytic process, and the narrower discharge energy window was limited to the hybrid catalytic reaction.

The N₂ and O₂ contained in the gas mixture were motivated to form N and O atoms via the collision of active electrons in the plasma-catalyst hybrid system. Compared to the chemical-bond dissociation energies of N₂ (945.33 kJ/mol), the O₂ was much easier to react with the energetic electrons for its lower chemical-bond dissociation energies of 498.36 kJ/mol. As a result, a high concentration of O radicals was produced in the plasma-catalyst system. The generated dominating O radicals and subordinate N radicals could react with NO/O₂/N₂/NH₃ gas mixture in the following reactions (3)~(8) [19]. The oxidation reactions (5) and (6) occurred between the radicals of O and O₃ and the NO molecules to generate NO₂ were regarded as the positive main steps to enhance NO conversion [7,9].

$${\mathrm{O}+\mathrm{O}}_2\to {\mathrm{O}}_3$$

(3)

$${\mathrm{O}}_3+\mathrm{NO}\to {\mathrm{NO}}_2{+\mathrm{O}}_2$$

(4)

$$\mathrm{O}+\mathrm{NO}\to {\mathrm{NO}}_2$$

(5)

$${\mathrm{O}+\mathrm{NO}}_2\to {\mathrm{NO}+\mathrm{O}}_2$$

(6)

$$\mathrm{O}+\mathrm{N}\to \mathrm{NO}$$

(7)

$$\mathrm{NO}+\mathrm{N}\to {\mathrm{N}}_2+\mathrm{O}$$

(8)

Thus, in the plasma-catalyst hybrid catalytic process, the catalytic reactions (2) and (3) become the predominant paths for NO elimination [4]. It had been testified that the reaction rate of fast SCR reaction (2) was more than 10 times larger than that of standard SCR reaction (1) below 200 °C [17].

Meanwhile, the NO₂ concentrations over Mn-based bimetallic nanocatalysts were relatively lower compared to that of plasma without catalyst assistance. Under discharge energy of 24 W·h/m³, more than 120 ppm NO₂ generated in the plasma-only catalytic process. However, the NO₂ concentration in Mn-based bimetallic nanocatalysts combining with plasma was no more than 20 ppm, which indicated that almost 100 ppm NO₂ took part in the catalytic reaction probably via the fast SCR reaction or the catalytic oxidation, as shown in Figure 2b. Therefore, it was believed that both the fast SCR and the standard SCR reactions occurred in the plasma-catalyst hybrid system simultaneity and the proportion of NO_x conversion via the fast SCR reaction improved with the discharge energy increasing. The N₂ selectivity over the Mn-based nanocatalysts was displayed in Figure 2c. The N₂ selectivity of the plasma-catalyst hybrid catalytic process was obviously larger than that of plasma-only process within discharge energy range of 0~24 W·h/m³, which was owing to the possibility of higher NO conversion and lower NO₂ formation, discussed above in Figure 2a,b. All test results presented a decreasing trend of N₂ selectivity with the discharge energy rising, which resulted from a great deal of N₂O produced in this reaction operation. It was proposed that the pivotal disadvantages of catalyzing NO_x by plasma were the low selectivity and the complex chemical productions that formed via diverse reaction pathways [10]. In order to verify the actual reactions during the plasma-catalyst hybrid process over the Mn-based bimetallic nanocatalysts, the NO_x conversion and the N₂ selectivity over Mn−Ce/TiO₂ sample in the balance gas of N₂ and Ar were tested, as shown in Figure 2d. It was obvious that the variation tendency of NO_x conversion obtained in the balance gas of N₂ and Ar were quite similar. While within the whole discharge energy range of 0~24 W·h/m³, the NO_x conversion in Ar was slightly higher. According to a previous report, under abundant O radicals or O₂, the N species is ten times more likely to react with O₂ than with NO [20]. Hence, almost N atoms produced from N₂ in the plasma transformed to NO via reaction (7), which was further oxidized into NO₂ and eliminated via fast SCR reactions immediately [7]. Therefore, in the balance gas of N₂, the NO concentration formed from N and O radicals was relatively small compare to the initial NO_x concentration, which caused little influence on the NO_x conversion during the plasma-catalyst hybrid process. Meanwhile, there was no obvious difference between the N₂ selectivity obtained in the balance gas of N₂ and Ar. The NO_x conversions over Mn−Ce/TiO₂ nanocatalyst with and without O₂ were analyzed as exhibited in Figure 2e. The NO_x conversion decreased drastically from 99.1% to 43% with the O₂ concentration dropping from 8% to 0%, which demonstrated the oxidation pathway for NO reduction by O species via reactions (3), (4), and (5) was dominant during the plasma-catalyst hybrid process. The NO_x conversions under O₂ 8% and 4% were almost the same, indicating the amount of oxygen excessive for NO_x redox reactions. Due to the dissociation energy of O₂ much smaller than that of N₂, the rate for dissociation of O₂ was much higher compared to the dissociation of N₂, which was the main reason for the remarkable promotion of O₂ on NO_x conversion [21].

The interaction effects of discharge energy and temperature on NO_x conversion in plasma with and without Mn-based bimetallic nanocatalysts reaction process were shown in Figure 3. The variation tendency of NO_x conversion in plasma-only process was opposite to that in plasma-catalyst hybrid process with reaction temperature increasing. As displayed in Figure 3a, high reaction temperature led to significant reduction in NO_x elimination, with the maximum catalytic conversion of 49.5% at 21 W·h/m³, 25 °C declining to 20.5% at 12 W·h/m³, 200 °C. In the plasma catalytic process, the NO₂ generation via interaction between the radicals of O and O₃ and the NO molecules was conducive to deNO_x as analyzed above. While with the formed NO₂ accumulation in the plasma-only process, the inhibition of reaction (6) on NO elimination progressively intensified. The concentration of O radical could be improved under high temperature, which could further promote reaction (5), (6), and (7). As a result, the temperature increase spurred the formation of NO and impeded NO oxidation into NO₂ [19]. Considering the energy consume during the plasma-catalyst process, the reaction temperature in the catalyst bed could be higher than the outside of nanocatalysts. In order to clearly realize the relation between reaction temperatures and plasma, an infrared thermometer was introduced to detect the specific temperature of discharge area during the plasma process. The test results were shown in Figure 3b. The plasma energy caused the temperature of the discharge area improved at different degrees and the largest temperature increase could reach 47 °C under the discharge energy of 24 W·h/m³. While the reaction temperature of gas mixture influenced by the plasma energy was relatively smaller with the Maximum temperature rise no larger than 13 °C, due to the short residence time of the gas mixture in the discharge area. Therefore, under the experiment conditions of this research, the plasma effects on NO conversion could be primarily analyzed by the discharge energy based on the gas mixture temperature.

It was apparent that the trends of NO conversion of these three Mn-based bimetallic nanocatalysts were consistent, as exhibited in Figure 3c–e). The NO conversion under different reaction temperatures and various discharge energies could be divided into three zones. In zone I, the NO_x conversion >90% only depended on the discharge energy and not affected by the reaction temperature. In zone II, the satisfied NO_x conversion (>90%) was achieved and both depended on the discharge energy and the reaction temperature. In zone III, it was impossibility to acquire a desired NO_x conversion. Mn−Ce/TiO₂ nanocatalyst presented superior catalytic property than Mn−Co/TiO₂ and Mn−Fe/TiO₂ samples with much broader zone I, which signified high NO_x conversion obtained with lower reaction temperature and the less discharge energy. A variety of previous works had revealed the optimal NO_x conversions obtained with the specific input energy varying from 4.7 to 40.3 W·h/m³ and the temperature changing from 25 to 350 °C, as shown in

Table 1. Plasma-catalyst performance in previous researches.

Samples	Specific Input Energy (W·h/m3)	NOx Conversion (%)	Temperature (°C)	Reductant	GHSV (h−1)	Gas Flow Rate (m3/h)	Ref
V2O5-WO3/TiO2	4.7 a	~76.5	170	NH3	--	31.8	[11]
H-mordenite	5	76	160	NH3	20,000	31	[7]
Ag/Al2O3	16.7 a	~91	350	C3H6	10,000	1.2	[12]
BaTiO3-Al2O3	~40.3 a	~61.5	150	CH3OH	11,000	--	[22]
Cu-ZSM-5	37.5 a	~90	25 b	C2H4	--	0.12	[13]
Co-ZSM-5	8.3 a	~70.6	150	C2H4+NH3	1000	0.12	[22]
Co-HZSM-5	38.3	~92	300	C2H2	12,000	0.03	[23]
Mn−Ce/ZSM5−MWCNTs	16.7 a	~85	25	NH3	60,000	0.12	[4]
Mn−Ce/TiO2	15	99.5	25	NH3	20,000	0.1	This study

^a calculated according to the data in the report (1 W·h/m³ = 3.6 J/L); ^b room temperature.

. In this study, Mn−Ce/TiO₂ sample exhibited the superior performance with NO_x conversion of 99.5% under 15 W·h/m³ at 25 °C, respectively, which was believed to be a potential excellent catalyst for the NO removal via the plasma-catalyst process.

2.3. Morphological Characterization

2.3.1. BET Measurements

In order to achieve the physical properties of these three typical Mn-based bimetallic nanocatalysts, the results of specific surface areas (S_BET), total pore volumes (V_total), and average pore diameters (D_p) were summarized in

Table 2. Physical properties of Mn-based bimetallic nanocatalysts.

Samples	SBET (m2·g−1)	Vtotal (cm3·g−1)	Dp (nm)
Mn−Ce/TiO2	239.7	0.527	17.57
Mn−Co/TiO2	189.9	0.531	33.06
Mn−Fe/TiO2	104.6	0.424	54.85

. It was evident that Mn−Ce/TiO₂ nanocatalyst obtained larger specific surface areas than Mn−Co/TiO₂ nanocatalyst and was more than twice as much as Mn−Fe/TiO₂ nanocatalyst, which was probable, owing to the Mn−Ce−O_x species better dispersed on the nanocatalyst surface. Meanwhile, there were noticeable changes of D_p, increasing from 17.57 nm in Mn−Ce/TiO₂ to 33.06 nm in Mn−Co/TiO₂ and further rising to 54.85 nm in Mn−Fe/TiO₂. It was proposed that the Mn−Ce−O_x species were more likely to promote nanocatalyst to form micropores compared with Mn−Co−O_x and Mn−Fe−O_x species [24,25]. However, the difference of total pore volumes among these three Mn-based bimetallic nanocatalysts was not obvious. The total pore volumes of Mn−Ce/TiO₂ and Mn−Co/TiO₂ samples centered on 0.53 cm³·g^-1, approximately. While the total pore volume of Mn−Fe/TiO₂ decreased to 0.424 cm³·g^-1 slightly, which was probable due to the mesoporosity formation that suppressed the micropore generation, resulting in the total pore volume reduced a little. Thereby, it was believed that Mn−Ce/TiO₂ nanocatalyst had superior physical properties than the other two samples with larger specific surface area, more micropores structure and satisfied total pore volumes, which coincided with catalytic performance of catalysts without plasma, as shown in Figure 1.

2.3.2. TEM Analysis

The morphological characterization and grain structure of these three typical Mn-based bimetallic nanocatalysts were collected by TEM analysis. From Figure 4a, it could be observed that Mn−Ce/TiO₂ nanocatalyst was constituted of fine uniform nanoparticles with narrow size distribution, smooth elliptic surfaces, and without evident agglomeration. The distinct and unbroken mesh structure of micropore was formed in the Mn−Ce/TiO₂ sample. According to the TEM images of Mn−Co/TiO₂ nanocatalyst, as shown in Figure 4b, there were some tightly aggregated metal oxide nanoparticles interfused into the smaller regular particles, which increased the average pore diameters and reduced the specific surface areas to some extent. However, from Figure 4c, a noticeable augment in the particle size was observed over Mn−Fe/TiO₂ nanocatalyst, which was consistent with Barrett–Joyner–Halenda (BJH) results. The nanoparticles were irregular, lots of which stacked on the catalyst surface with an abundant micropore structure collapsing and regional accumulations.

2.4. Structural Characterization

2.4.1. Textural Properties

Figure 5a exhibited the XRD spectra of Mn-based bimetallic nanocatalysts and the phases contained in the nanocatalyst samples were identified by the software of MDI Jade 6.5. Among all these three nanocatalysts, there were strong and distinguished diffraction peaks at about 2θ values of 25.3°, 37.8°, 48.0°, 53.9°, 62.7°, 68.8°, 70.3°, 75.1°, and 82.7° well matched the XRD pattern of anatase TiO₂ (ICDD PDF card # 71-1166) [26]. While the diffraction peaks for the structure of TiO₂ support were reserved completely, the diffraction angles of the matching peaks shifted at different degrees. In Mn−Ce/TiO₂ nanocatalyst, the anatase TiO₂ presented the lowest diffraction angle for every corresponding peak, which probably verified the interaction between MnCeO_x and anatase TiO₂ was stronger than that between MnCoO_x or MnFeO_x and anatase TiO₂. Comparing these three nanocatalysts, it could be found that the diffraction peaks of anatase TiO₂ in Mn−Ce/TiO₂ nanocatalyst were broader and weaker than that in the other two nanocatalysts, indicating the crystalline of TiO₂ reduced by the MnCeO_x loading. Meanwhile, there was no obvious characterization reflections for MnO_x or CeO_x in Mn−Ce/TiO₂ nanocatalyst that manifested the active species were finely dispersed on the nanocatalyst surface or the active species of MnO_x and CeO_x incorporated into TiO₂ lattice [27].

In the XRD patterns of Mn−Co/TiO₂ and Mn−Fe/TiO₂ nanocatalysts, the diffraction peaks accord with MnO_x were very complex due to the transformation among MnO₂, Mn₂O₃, Mn₃O₄, and MnO in the incomplete crystallization of manganese oxides. The diffraction peaks matched with MnO₂ exactly at 2θ = 22.10°, 35.19°, 36.96°, 38.72°, 47.86°, and 57.166°, corresponding to the crystallographic plane reflections of (110), (310), (201), (111), (311), and (420), respectively (ICDD PDF card # 82-2169) [28]. At the same time, the diffraction peaks of Mn₂O₃ and Mn₃O₄ were evident in Mn−Co/TiO₂ and Mn−Fe/TiO₂ nanocatalysts. The intensive and sharp characteristic peaks at 2θ values of 23.08°, 26.72°, 32.87°, and 56.89° could be primarily ascribed to Mn₂O₃ matching with the crystallographic plane reflections of (211), (220), (222), and (433), correspondingly (ICDD PDF card # 78-0390), and the distinct signals at 36.28°, 40.67°, 41.80°, 57.73°, and 64.17° could be assigned to Mn₃O₄ corresponding to the crystallographic plane reflections of (112), (130), (131), (115), and (063), respectively (ICDD PDF card # 75-0765) [28,29]. Comparing the pattern of Mn−Co/TiO₂ nanocatalyst, it could be noticed that the diffraction peaks of both Mn₂O₃ and Mn₃O₄ were remarkably decreased in Mn−Fe/TiO₂ nanocatalyst, simultaneously, the diffraction peaks matched anatase TiO₂ were also visibly weakened. These possibly suggested the addition of cobalt into manganese oxides had better effects than iron on diminishing the crystallization of MnO_x and TiO₂ at the same time. Furthermore, there were no obvious distinct diffraction peaks of CoO_x were observed in Mn−Co/TiO₂ nanocatalysts, which indicated the addition ratios of cobalt not only enhanced the dispersion of MnO_x, but also promoted the dispersion of CoO_x entirely on the nanocatalyst surface. A similar proposal could be obtained over Mn−Fe/TiO₂ nanocatalysts. Generally, among Mn−Ce/TiO₂, Mn−Co/TiO₂, and Mn−Fe/TiO₂ nanocatalysts, the MnCeO_x loading on anatase TiO₂ performed the superior properties with smaller the nanoparticle sizes, reducing the chemical compounds crystallinities and increasing the active species distributions, which were facilitated to the SCR reactions [30]. In order to confirm the presence of nitrates in the mixtures during the plasma-catalyst process, the Energy Dispersive Spectrometer (EDS) test was introduced to qualitatively analyze the elements changes, as exhibited in Figure 5b. It was apparent that the variation of nitrogen contents on the Mn−Ce/TiO₂ sample before and after the plasma-catalyst reaction was tiny, which indicated little deposition of nitrates on the catalyst surface.

2.4.2. Reducibility Properties

In order to explore the oxidation states and the reduction potentials of the active species contained in the Mn-based bimetallic nanocatalysts, H₂-TPR analysis was performed with the reduction peaks fitted by Gaussian functions, as exhibited in Figure 6. The H₂ consumptions together with all reduction temperature values were summarized in

Table 3. H₂-TPR quantitative analysis of Mn-based bimetallic nanocatalysts.

Samples	Temperature (°C)					H2 Consumption (mmol·g−1)
	R 1	R 2	R 3	R 4	R 5	R 1	R 2	R 3	R 4	R 5	Total
Mn−Ce/TiO2	261	336	367	423	717	2.52	1.24	0.36	0.67	0.07	4.86
Mn−Co/TiO2	218	327	418	517	--	1.14	0.21	2.12	0.96	--	4.43
Mn−Fe/TiO2	276	387	436	501	--	0.83	0.41	0.54	1.59	--	3.37

. On account of the support of anatase TiO₂ induced no noteworthy reduction peaks in the test temperature region, all the H₂ consumption peaks displayed in Figure 6 could be ascribed to the reduction reactions of diverse active species of MnO_x, CeO_x, CoO_x, and FeO_x. For Mn-based catalysts, the typical reduction peaks were regard as following the order of MnO₂ → Mn₂O₃ (Mn₃O₄) →MnO [31]. For Mn−Ce/TiO₂ nanocatalyst, as shown in Figure 6a, there were five main H₂ consumption peaks appearing within the temperature range of 50~850 °C. The initial dominating reduction peak (R1) at around 261 °C was mainly caused by the reduction reaction of the high oxidation state of Mn⁴⁺ reducing to Mn³⁺ [32]. The subsequent asymmetrical reduction peak from 260 °C to 410 °C could be further divided into two reduction peaks (R2 and R3), according to the two processes of Mn₂O₃ reducing to Mn₃O₄ and Mn₂O₃ reducing to MnO reported in previous literatures [28,31]. The converting from Mn₂O₃ to Mn₃O₄ preferred to occur on the primal amorphous Mn₂O₃ [33], which was consistent with appearance of R2 peak. While the transformation from Mn₂O₃ to MnO was apt to happen at higher reaction temperatures [34], well coinciding with the temperature value of R3 peak. For Ce-containing sample, the typical CeO_x reduction process usually presented two separated peaks, the one of CeO₂^s converting to Ce₂O₃^s on the catalyst surface occurred at about 450 °C, the other one of CeO₂^b transforming to Ce₂O₃^b in the catalyst bulk came up at 730 °C approximately [35]. Therefore, the fourth wide reduction peak (R4) in the Mn−Ce/TiO₂ nanocatalyst was related to the reduction processes of Mn₃O₄ to MnO and CeO₂^s to Ce₂O₃^s simultaneously, and the fifth peak (R5) at around 717 °C was potentially associated with the CeO₂^b reduction reaction. Among these three Mn-based bimetallic nanocatalysts, Mn−Ce/TiO₂ nanocatalyst displayed the highest low-temperature reducibility and exhibited a noticeable lack of high-temperature reduction peaks at the same time, which manifested the higher oxidation states of manganese ion (Mn⁴⁺ and Mn³⁺) constituted the dominating phase [34].

Comparing with Mn−Ce/TiO₂ nanocatalyst, the H₂-TPR curve of Mn−Co/TiO₂ nanocatalyst was conspicuously different in both the reduction temperatures and the peak intensions. For Mn−Co/TiO₂ nanocatalyst, the reduction peak of MnO₂ to Mn₂O₃ shifted toward lower temperature (218 °C) and weakened significantly. Meanwhile, the reduction peaks of Mn₂O₃ to Mn₃O₄ and Mn₂O₃ to MnO moved to higher temperatures and strengthened noticeably. The two reduction processes of Mn₂O₃ presented as a whole peak centered at about 418 °C. The reduction reaction of cobalt oxides exhibited two peaks at around 327 (R2) and 517 °C (R4), which could be ascribed to the transformation of Co³⁺ → Co²⁺ and Co²⁺ → Co⁰, respectively [15]. However, these two reduction peaks were overlapped with the MnO_x peaks in whole or partly. For Mn−Fe/TiO₂ nanocatalyst, considering the coexistence of FeO_x and MnO_x, the joint peaks (R2 and R3) from 330 °C to 530 °C were mainly attributed to the conversion of Mn₂O₃ to Mn₃O₄ combining with the transformation of Fe₂O₃ to Fe₃O₄. According to previous report [36], the majority of Fe₂O₃ (Fe₂O₃^m) was in the form of nanoparticles, oligomeric clusters or isolated ions locating at effortlessly reducible sites. After the Fe₂O₃^m reduction reaction, the reduction of residual Fe₂O₃ (Fe₂O₃^r) to Fe₃O₄ accomplished at the higher temperature [37]. The remarkable strong peak (R4) at about 501 °C was ascribed to the overlapped peaks of Mn₃O₄ to MnO and Fe₃O₄ to FeO.

As exhibited in Table 3, the total H₂ consumptions of Mn−Ce/TiO₂ and Mn−Co/TiO₂ nanocatalysts were 4.86 mmol·g^-1 and 4.43 mmol·g^-1, respectively, much larger than that of Mn−Fe/TiO₂ nanocatalysts. It was proposed that the peaks appearing at lower temperatures demonstrated superior catalytic activity in low temperature region [2]. While the starting reduction peak temperature of Mn−Co/TiO₂ nanocatalyst was the lowest at 218 °C, its total H₂ consumption was obvious smaller than that of Mn−Ce/TiO₂ nanocatalyst, which was regarded as a more important factor affecting the reducing capacity. Based on the H₂ consumption as a vital factor to the redox property of catalyst, it was reasonable that Mn−Ce/TiO₂ nanocatalyst presented the higher NO_x conversion with and without plasma than Mn−Co/TiO₂ and Mn−Fe/TiO₂ nanocatalysts.

2.4.3. Ammonia Adsorption Properties

Besides the redox property, the acid capacity on the catalyst surface was another crucial factor influencing the catalytic performance in SCR reactions [14,38]. NH₃-TPD and NO-TPD tests were introduced in order to establish the connection between the surface acidities and the SCR activities for the Mn-based bimetallic nanocatalysts. The test results were analyzed and compared in Figure 7 and

Table 4. Quantitative analysis of NH₃-TPD profiles.

Samples	Temperature (°C)				NH3 Composition (mmol·g−1)
	Peak 1	Peak 2	Peak 3	Peak 4	Peak 1	Peak 2	Peak 3	Peak 4	Total
Mn−Ce/TiO2	245	396	583	647	0.27	0.52	0.58	0.39	1.76
Mn−Co/TiO2	276	402	587	653	0.11	0.49	0.52	0.32	1.44
Mn−Fe/TiO2	209	398	585	649	0.14	0.24	0.39	0.20	0.97

, respectively. The NH₃-TPD curves for these three typical Mn-based bimetallic nanocatalysts were attributed to four desorption peaks of chemisorbed NH₃ within the temperature range of 150~750 °C. It was obvious that the first weak NH₃ desorption peak (P1) of Mn−Fe/TiO₂ nanocatalyst appeared at about 209 °C ascribed to the NH₃ desorption from the weak acid sites, which was deemed too weak to be stable bound NH₃ in the gas mix during the SCR reactions [14]. The second and the third successive desorption peaks (P2 and P3) located from 398 °C to 585 °C indicated the distribution of medium strong acid sites. Additionally, the final and relatively stronger peak (P4) at 649 approximately were attributed to strong acid sites, which could be regard as plenty of Lewis acid sites generated on the nanocatalyst surface with adsorbing a large amount of strongly bound ammonia [39]. For Mn−Co/TiO₂ nanocatalyst, the NH₃ desorption result demonstrated superior acidity capacity at medium and high temperatures, but an undesirable temperature shift towards higher temperature regions appeared at the same time. The desorption peak temperature value of weak acid sites (P1), medium strong acid sites (P2 and P3), and strong acid sites (P4) reached 276 °C, 402 °C, 587 °C, and 653 °C, respectively, which signified that it is more difficult for the chemisorbed NH₃ to desorb from the acid sites and participate in SCR reactions [40]. Compared with the NH₃-TPD curve of Mn−Co/TiO₂ nanocatalyst, the four desorption peaks of Mn−Ce/TiO₂ nanocatalyst shifted towards lower temperatures slightly and became much stronger. The enrichments of the weak acid sites, the medium strong acid sites and the strong acid sites were all positive to ammonia adsorption, which could undoubtedly form more abundant Brønsted acid sites and Lewis acid sites promoting NH₃ adsorption on the nanocatalyst surface [41,42]. This change was regard as an important reason for the outstanding catalytic performance of Mn−Ce/TiO₂ nanocatalyst with and without plasma.

In order to exactly confirm the total acid capacity, the quantitative analysis of the desorption peaks of Mn−Ce/TiO₂, Mn−Co/TiO₂ and Mn−Fe/TiO₂ nanocatalyst was performed and summarized in Table 4. Among these three nanocatalysts, the total NH₃ desorption of Mn−Ce/TiO₂ sample achieved the maximum value of 1.76 mmol·g^-1, which further verified the better promotion effects of MnCeO_x on the surface acidity than that of MnCoO_x and MnFeO_x. It was noteworthy that, although Mn−Fe/TiO₂ nanocatalyst exhibited the weak acid sites at the lowest temperature, the acid capacity of its weak acid sites was only 0.14 mmol·g^-1 too small to satisfy the NH₃ adsorption requirements during SCR recations. Therefore, Mn−Ce/TiO₂ nanocatalyst with NH₃ desorption of 0.27 mmol·g^-1 at around 245 °C was qualified for the best acidity properties at the low temperature. However, it had been revealed that NH₃ could block NO adsorption and activation onto the metal active sites on the catalyst surface via the undesired electronic contact between the adsorbed NH₃ and the metal sites [43]. As a result, NH₃ presented inhibiting effects on SCR reactions at low temperature. Hence, the acidity properties on the surface of the Mn-based bimetallic nanocatalysts were closely related to the redox performance. It was required to achieve an adequate equilibrium between the oxidation states of active species and the acidity properties of active metal compounds in order to develop the optimal catalyst.

2.4.4. Oxidation States of Active Species

The elemental valence states and the atomic concentrations on the surface of Mn-based bimetallic nanocatalysts were explored by XPS analysis for the purpose of a better insight into the metal oxidation states and the surface compositions. The XPS spectra of Mn 2p, O 1s, Ti 2p, Fe 2p, Co 2p, and Ce 3d in the nanocatalysts were exhibited in Figure 8. The valence state of every element was determined numerically according to Gaussian fitting, respectively. The specific binding energy and the individual element concentration in various valence states were summarized in

Table 5. Surface atomic compositions of the catalysts determined by XPS.

Samples	Binding Energy (eV)/Atomic Composition (%)
	Mn			Fe		Co		Ce		O
	Mn2+	Mn3+	Mn4+	Fe2+	Fe3+	Co2+	Co3+	Ce3+	Ce4+	Oα	Oβ
Mn−Ce/TiO2	640.9/ 9.3	642.4/ 34.2	644.0/ 56.5	-/-	-/-	-/-	-/-	885.6/ 40.4	882.4/ 59.6	531.2/ 40.4	530.2/ 59.6
Mn−Co/TiO2	641.2/ 13.8	642.6/ 39.4	644.1/ 46.8	-/-	-/-	779.6 39.3	782.1/ 60.7	-/-	-/-	531.4/ 33.7	530.3/ 66.3
Mn−Fe/TiO2	641.4/ 21.3	642.8/ 38.5	644.6/ 40.2	709.6/ 43.4	711.7/ 56.6	-/-	-/-	-/-	-/-	531.6/ 28.2	530.3/ 71.8

. Figure 8a displayed the XPS spectrum for Ti 2p of catalyst support, which comprised two peaks of Ti 2p_1/2 locating at about 464.3 eV and Ti 2p_3/2 situating at around 458.7 eV, respectively [44]. Comparing the three XPS spectrum of Ti, it could be easily found that, although the Mn-based bimetallic nanocatalysts were doped with different active elements of iron, cobalt, and manganese, no significant changes were observed in the Ti peaks. The +4 valence state of titanium on the catalyst surface was stabilized and dominating in Mn−Fe/TiO₂, Mn−Co/TiO₂, and Mn−Ce/TiO₂ nanocatalysts.

The XPS spectra of Mn 2p in these three typical Mn-based bimetallic nanocatalysts consisted of two characteristic peaks, assigned to Mn 2p_1/2 peak at around 653 eV and Mn 2p_3/2 peak at about 642 eV [45], as shown in Figure 8b. The asymmetrical Mn 2p_3/2 peak further verified the complicated manganese species in divers valence states coexisting in the nanocatalysts. The curve of Mn 2p_3/2 peak could be split into three peaks via multi-peaks gaussian fitting, the first peak at around 641.2 ± 0.3 eV was assigned to Mn²⁺, the second one at 642.6 ± 0.2 eV associated with Mn³⁺, and the third one at 644.1 ± 0.5 eV consistent with Mn⁴⁺, respectively [46]. The complex MnO_x including three valence states were apparently difficult to distinguish with the binding energy difference no more than 3.7 eV. In order to make an accurate identification of the atomic composition and Mnⁿ⁺ concentration on the nanocatalyst surface, a quantitative analysis was introduced based on the area covered under each separated peak, as listed in Table 4. According to previous studies [28,33,47], NO_x conversion over pure MnO_x could been ranked as MnO₂ > Mn₂O₃ > Mn₃O₄. Hence, the improved concentration of MnO₂ on the nanocatalyst surface was advantageous to SCR reactions [48]. Among the three Mn-based bimetallic nanocatalysts, the major amount of manganese primarily presenting +4 valence state on the catalyst surface dispersedly, as shown in Table 4. While Mn−Ce/TiO₂ sample presented the highest atomic composition of Mn⁴⁺/Mnⁿ⁺ (56.5%), which was much larger than 46.8% and 40.2% in Mn−Co/TiO₂ and Mn−Fe/TiO₂ nanocatalysts, respectively. It was regard as the reason for the higher catalytic activity of Mn−Ce/TiO₂ nanocatalyst, as exhibited in the above results.

For Mn−Fe/TiO₂ nanocatalyst, the Fe 2p XPS spectra was was presented in Figure 8c with two individual peaks attributed to Fe 2p_3/2 at about 710 eV, Fe 2p_1/2 at around 724 eV and a satellite peak of Fe³⁺ in Fe₂O₃ exhibited at 718.3 eV [49]. The broad peak of Fe 2p_3/2 was composed of two overlapped peaks, the one ascribed to Fe²⁺ locating at around 709.6 eV and the other one attributed to Fe³⁺ seated at about 711.6 eV. These two peaks confirmed the co-occurrence of iron in +2 and +3 valence states on Mn−Fe/TiO₂ nanocatalyst surface, as quantified in Table 5. The promotion effect of iron on Mn-based bimetallic nanocatalysts was ascribed to the interaction happening in the redox circulation: Fe³⁺ + Mn³⁺ ↔ Fe²⁺ + Mn⁴⁺ [50]. For Mn−Co/TiO₂ nanocatalyst, there were two main peaks in the Co 2p spectrum ascribed to Co 2p_3/2 at about 780.6 eV and Co 2p_1/2 at around 796.5 eV. Each of these two peaks was accompanied by an adjacent satellite peak at 786.8 eV and 803.1 eV correspondingly, as depicted in Figure 8d. The two broader and gentler satellite structures at the relatively higher binding energy region were caused by the metal-to-ligand charge transfer, also known as the shakeup process of cobalt in its high spin state. While this process can only be observed with the high spin state of Co²⁺ ion, but cannot be observed with the diamagnetic low-spin Co³⁺ ion [51]. The XPS spectra of Co 2p_3/2 scope could be further divided into Co³⁺ spectrum at binding energy of 780.0 eV and Co²⁺ spectrum at 781.6 eV. This test result showed the ions of Co²⁺ and Co³⁺ were co-existed on Mn−Co/TiO₂ nanocatalyst surface and the Co³⁺ exhibited a comparatively higher atomic composition of 60.7%. The Co³⁺ species existed in a relatively high valence state and gave rise to more anionic defects, generating abundant surface oxygen to enhance the process of adsorption and oxidation during SCR reactions [52]. For Mn−Ce/TiO₂ nanocatalyst, the Ce 3d spectra result was depicted in Figure 8e. The Ce 3d pattern was composed of u and v multi-peaks matching to the spin orbit split 3d_5/2 and 3d_3/2 core holes [53]. According to the binding energies of different peaks, the Ce 3d spectra could be elaborately separated into eight peaks, labeled as u, u′, u″, u‴ and v, v′, v″, v‴, respectively [54]. The u′ and v′ peaks were matched with the 3d¹⁰4f¹ electronic state of Ce³⁺, and the u, u″, u‴, v, v″, and v‴ peaks were ascribed to the 3d¹⁰4f⁰ electronic state of Ce⁴⁺ [55]. These distinctive peaks verified the coexistence of of Ce³⁺ and Ce⁴⁺ species on Mn−Co/TiO₂ nanocatalyst surface. The Ce³⁺ species were important incentives for the formation of unsaturated chemical bonds and the generation of electric charge balance. [56]. In the active compounds of manganese and cerium, the negative charge transferred from Mn²⁺ or Mn³⁺ to Ce⁴⁺ strengthening the interaction between manganese and cerium [1,2]. The oxygen circle of storing and releasing was easier for the Mn−Ce/TiO₂ nanocatalyst with the redox couple of Ce³⁺/Ce⁴⁺ to form more surface oxygen vacancies that were advantageous to oxygen adsorption and chemisorbed oxygen generation [57].

The O 1s spectra for Mn-based bimetallic nanocatalysts were displayed in Figure 8f. On the base of curve-fitting results, the O 1s spectra was divided into two peaks: The O_α peak ascribed to chemisorbed oxygen centered at binding energy of 531.2 ~ 531.6 eV, the O_β peak attributed to lattice oxygen appeared at binding energy of 530.2 ~ 530.3 eV. Compared with the O 1s spectra of nanocatalysts, it could be found that the binding energies of O_α shifted to lower values from 531.6 eV in Mn−Fe/TiO₂ to 531.4 eV in Mn−Co/TiO₂ and to 531.2 eV in Mn−Ce/TiO₂, and similar variation tendency occurred on the binding energies of O_β. Meanwhile, as shown in Table 4, the surface atomic composition of chemisorbed oxygen over Mn−Ce/TiO₂ nanocatalyst reached the maximum 40.4%, much higher than 33.7% on Mn−Co/TiO₂ and 28.2% on Mn−Fe/TiO₂ samples. The chemisorbed oxygen was the most energetic oxygen species due to its high mobility [58]. Therefore, these surface atomic composition of Mn−Ce/TiO₂ nanocatalyst were regard as another reason for its superior catalytic performance with and without plasma.

2.5. Reaction Mechanism Analysis

According to the catalytic performance of NO_x conversion over Mn-based bimetallic nanocatalysts with and without plasma and the physicochemical properties of these nanocatalysts presented above, the complex bimetallic oxides of MnFeO_x, MnCoO_x, and MnCeO_x affected the hybrid catalyst-plasma catalytic process obviously with the different redox characteristics of active chemisorbed sites. All the three bimetallic nanocatalysts enhanced the catalytic ability of manganese species by increasing the ratio of Mn⁴⁺/Mnⁿ⁺, generating more lattice oxygen and plenty of oxygen vacancy on the catalyst surface [2]. In the catalyst-plasma hybrid catalytic system, the plasma derivatives reformed the chemical compositions of the gas mix and modified the electronic states on the nanocatalyst surface. For Mn−Fe/TiO₂, Mn−Co/TiO₂ and Mn−Ce/TiO₂ nanocatalysts, a dynamic equilibrium was sustained on their surfaces with the electron transfer between Mn and Fe (Co or Ce) ions during the catalytic oxidation process, which could be expressed as Fe³⁺ + Mn³⁺ ↔ Fe²⁺ + Mn⁴⁺, Co³⁺ + Mn³⁺ ↔ Co²⁺ + Mn⁴⁺, Ce³⁺ + Mn³⁺ ↔ Ce²⁺ + Mn⁴⁺ and Ce⁴⁺ + Mn³⁺ ↔ Ce³⁺ + Mn⁴⁺, respectively.

Besides originally partial NO oxidation into NO₂ over the lattice oxygen of MnO_x, FeO_x, CoO_x and CeO_x, more NO was oxidized to NO₂ via the reaction (4) and (5) under the energetic particles. The valence state part of manganese cations increased, which was caused by the electron transition from Mn³⁺ to Mn⁴⁺ via lattice oxygen [59]. Mn⁴⁺ was more desirable for the oxidation of NO to NO₂ over Mn-based catalysts [16] and it was reduced to Mn³⁺ during the SCR reactions [58]. Under the plasma derived species, such as O₃ or O radicals, the Mn³⁺ could be fast re-oxidation into Mn⁴⁺, thus accelerating the catalytic oxidation process and the fast SCR reaction. Furthermore, the concentration of chemisorbed oxygen on the nanocatalyst surface was also improved in the catalyst-plasma hybrid catalytic system. More surface oxygen species could form via the direct interaction of MnO_x, FeO_x, CoO_x, and CeO_x with plasma excited oxygen species. Considering the inhibiting effect of NO on O₃ formation, the surface oxygen species were more likely to generate from O radicals. The adsorbed oxygen reacted with NO to form NO₂ according to the following reaction steps (9)~(11):

$$\mathrm{O}+\mathrm{M}\to {\mathrm{M}-\mathrm{O}}_{\mathrm{ads}}$$

(9)

$${\mathrm{NO}+\mathrm{M}-\mathrm{O}}_{\mathrm{ads}}\to {\mathrm{M}-\mathrm{O}-\mathrm{NO}}_{\mathrm{ads}}$$

(10)

$${\mathrm{M}-\mathrm{O}-\mathrm{NO}}_{\mathrm{ads}}\to {\mathrm{M}+\mathrm{NO}}_2$$

(11)

where M represented the active sites on the nanocatalyst surface, O_ads and NO_ads represented adsorbed NO and oxygen on the nanocatalyst surface, respectively. During this process, the NO_ads liberated electron to Mn⁴⁺ and the O_ads trapped electron from Fe²⁺ or Co²⁺ or Ce³⁺, respectively, which transform into absorbed NO⁺ and O⁻. Then formed NO⁺ further reacted with O⁻ to generate NO₂. Simultaneously, a part of NO was oxidized to NO₂ directly by the active oxygen produced from O₂ activation on the surface oxygen vacancies. The possible catalyst-plasma hybrid catalytic process of SCR reaction over Mn-based bimetallic nanocatalysts was exhibited in Figure 9.

3. Materials and Methods

3.1. Catalysts Preparation

The three typical Mn-based bimetallic nanocatalysts were prepared by hydrothermal method. Mn(NO₃)₂ (analytical pure 50%, Sinopharm, Shanghai, China), Fe(NO₃)₃·9H₂O (analytical pure 99.9%, Sinopharm, Shanghai, China), Co(CH₃COO)₂·4H₂O (analytical pure 99.9%, Kermel, Tianjin, China), and Ce(NO₃)₃·6H₂O (analytical pure 99.9%, Nanjing-reagent, Nanjing, China) were introduced as the precursors of MnO_x, FeO_x, CoO_x, and CeO_x, respectively. The tetrabutyl titanate was used as the precursors of TiO₂ for supporting the active metallic oxides. Mn(NO₃)₂ was added into deionized water at room temperature and then Fe(NO₃)₃·9H₂O was dissolved in the solution. Amount of glycol was added into the above mixture with magnetic stirring continuously. A Teflon-lined stainless steel autoclave was introduced to heat the homogeneous solution at 180 °C for 8 h. After the autoclave cooling down to the ambient temperature, tetrabutyl titanate was added into this solution and aged in the autoclave again at 180°C for 3 h. The mixture was collected by reduplicative centrifugation and wash. Finally, the precipitate was dried at 150 °C for 12 h and calcined in air at 500 °C for 4 h. The produced Mn-based bimetallic nanocatalysts were triturated and filtered with 60−80 mesh for activity tests and characterization analysis. The nanocatalyst was denoted as Mn−Fe/TiO₂ with the molar ratios of Mn:Fe:Ti = 2:1:7. The Mn−Co/TiO₂ and Mn−Ce/TiO₂ nanocatalysts were prepared under the same process with Co(CH₃COO)₂·4H₂O and Ce(NO₃)₃·6H₂O replacing Fe(NO₃)₃·9H₂O, respectively.

3.2. Catalysts Characterization

The Maxon Tristar II 3020 micropore-size analyzer (Maxon, Chicago, IL, USA) was used for testing N₂ adsorption isotherms of the prepared nanocatalysts at -196 °C. The surface areas and the pore-size distributions of the nanocatalysts were measured after the nanocatalysts degassing in vacuum at 350 °C for 10 h. BET plot linear portion was used to determine the nanocatalysts specific surface areas, and the desorption branch with Barrett–Joyner–Halenda (BJH) formula was introduced to calculate the pore-size distributions. The XRD data was captured by a Bruker D8 advance analyzer (Bruker, Frankfurt, Germany) with Mo K_α radiation, diffraction intensity from 10° to 90°, point counting time of 1s and point counting step of 0.02°. The element phases contained in the nanocatalys were distinguished by comparing characteristic peaks presented in the XRD patterns with the International Center for Diffraction Data (ICDD). The advanced microstructural image data and the surface element contents of the nanocatalysts were achieved by a high resolution transmission electron microscope JEOL JEM-2010 combined with EDS ((Japan electronics corporation, Tokyo, Japan). H₂-TPR and NH₃-TPD tests were performed with a Micromeritics Autochem II 2920 chemical adsorption instrument (Micromeritics, Houston, TX, USA). During H₂-TPR experiment, nanocatalysts were pretreatment in He at 400 °C for 1 h, and then cooled to environment temperature in H₂ and He gas mixture at 30 mL/min. The test temperature range of H₂ consumptions was from 50 °C to 850 °C with the heating rate of 10 °C/min. The operating process of NH₃-TPD test was similar to that of H₂-TPR test with NH₃ replacing H₂. XPS analysis was performed on a Thermo ESCALAB 250XI (Thermo Fisher, Boston, MA, USA) with pass energy 46.95 eV, Al K_α radiation 1486.6 eV, X-ray source 150 W and binding energy precision ± 0.3 eV. The C 1s line at 284.6 eV was measured as a reference.

3.3. Catalytic Performance Tests

The catalytic performance of Mn-based bimetallic nanocatalysts was explored in a catalyst-plasma hybrid system as shown in Figure 10. The dielectric barrier discharge (DBD) plasma reactor was comprised of two electrodes and a quartz tube. The high voltage electrode was a stainless-steel rod with diameter of 3 mm, installed inside the quartz tube coaxially. The ground electrode was a copper wire mesh wrapped outside the quartz tube tightly. The discharge energy was produced by an AC power transverter with a digital controller of voltage, electricity, and frequency. The quartz tube was in the height of 800 mm, outer diameter of 12 mm and thicknesses of 0.8 mm. 5 mL of nanocatalyst was filled in the discharge zone of plasma reactor. A resistance furnace was introduced to maintain the desired reaction temperature located upstream plasma reactor, connected to the temperature controller. The concentration of gas mixture was measured by German MRU MGA-5 analyzer (MRU, Berlin, Germany) joint with an external special detector for N₂O and NH₃. An Infrared Thermometer (HCJYET, HT-8872, Hongcheng, Shanghai, China) was introduced to detect the specific temperature of discharge area during the plasma process. During plasma-catalyst catalytic activity experiment, the inlet mixed gas included 300 ppm NO, 300 ppm NH₃, 8% O₂, ~0.1% H₂O and N₂ as balance gas. The gas hourly space velocity (GHSV) was about 20 000 h^-1. The NO_x conversion rate was calculated according to Equation (12), where [NO_x] = [NO] + [NO₂]. The N₂ selectivity was calculated by the concentrations of N₂O and NO_x, as shown in Equation (13). Each experiment was repeated three times to assure the results accuracy. The discharge energy density was defined as discharge power divided by the inlet gas flow rate [9], which was calculated using Equation (14) [60], where E (W·h/m³) was energy density, P (W) was discharge power, and Q (m³/h) was the gas flow rate. More basic data relating to the discharge energy was listed in Table 1.

$${\mathrm{NO}}_{\mathrm{x}}\mathrm{conversion}\mathrm{rate}=\left(\frac{{\left[{\mathrm{NO}}_{\mathrm{x}}\right]}_{\mathrm{in}}-{\left[{\mathrm{NO}}_{\mathrm{x}}\right]}_{\mathrm{out}}}{{\left[{\mathrm{NO}}_{\mathrm{x}}\right]}_{\mathrm{in}}}\right)\text{×}100\text{\%}$$

(12)

$${\mathrm{N}}_2\mathrm{selectivity}=1-\frac{2{\left[{\mathrm{N}}_2\mathrm{O}\right]}_{\mathrm{out}}}{{\left[{\mathrm{NO}}_{\mathrm{x}}\right]}_{\mathrm{in}}-{\left[{\mathrm{NO}}_{\mathrm{x}}\right]}_{\mathrm{out}}}\text{×}100\text{\%}$$

(13)

$${\mathrm{E}(\mathrm{W}\text{·}\mathrm{h}/\mathrm{m}}^3)=\frac{\mathrm{P}\left(\mathrm{W}\right)}{{\mathrm{Q}(\mathrm{m}}^3/\mathrm{h})}$$

(14)

4. Conclusions

The Mn-based bimetallic nanocatalysts of Mn−Fe/TiO₂, Mn−Co/TiO₂, Mn−Ce/TiO₂, synthesized by hydrothermal method, presented obvious synergistic effects on NO_x catalytic conversion via the plasma-catalyst hybrid catalytic process. In the catalytic process with catalyst alone, the NO_x conversions of all tested catalysts were lower than 20% at ambient temperature. While in the plasma-catalyst hybrid catalytic process, the catalytic activities for NO_x elimination improved significantly with discharge energy enlarging. The maximum NO_x conversion of about 99.5% achieved on Mn−Ce/TiO₂ with discharge energy of 15 W·h/m³ at ambient temperature. The reaction temperature had an inhibiting effect on plasma-catalyst hybrid catalysis.

Among these three Mn-based bimetallic nanocatalysts, Mn−Ce/TiO₂ displayed the optimal catalytic property with higher catalytic activity and superior selectivity in the plasma-catalyst hybrid catalytic process. Furthermore, based on the multiple characterizations performed on the Mn-based bimetallic nanocatalysts, it could be confirmed that the catalytic property of plasma-catalyst hybrid catalytic process was highly dependent on the phase composition of the catalyst. Mn−Ce/TiO₂ nanocatalyst presented the optimal structure characteristic among all tested samples, with the largest surface area, the increased active components distributions, the reduced crystallinity and the minished particle sizes. In the meantime, the ratios of Mn⁴⁺/(Mn²⁺ + Mn³⁺ + Mn⁴⁺) in the Mn−Ce/TiO₂ sample was the highest, which was beneficial to plasma-catalyst hybrid catalysis. Generally, it was believed that the plasma-catalyst hybrid catalytic process with the Mn-based bimetallic nanocatalyst was an effective approach for high-efficiency catalytic conversion of NO_x, especially at ambient temperature.