The preparation of γ-Al₂O₃ was given in Ref. ^[12].

Manganese acetate (West Sichuan Chemical Company, AR) and citric acid (Chongqing Oriental Reagent Factory, AR) were mixed in a molar ratio of 1:1 and the solution was adjusted to pH 8.5 by NH₃·H₂O under vigorous stirring at 70 °C to form a sol. The sol was kept overnight and then heated at 70 °C for 3 h in a water bath to eliminate the solvent. The resulting viscous gel was dried at 110 °C for 12 h, then crushed into a fine powder. The dried powder was divided into four parts which were heated in a muffle furnace for 4 h at 300, 400, 500, and 600 °C, respectively. The MnO_x samples were obtained. In addition, a mixture of the MnO_x and γ-Al₂O₃ in a mass ratio of 4:1 was ball-milled with distilled water, and then was dried at 110 °C for 6 h to prepare the MnO_x + γ-Al₂O₃.

An aqueous solution of Al(NO₃)₃ (Chongqing Southern China inorganic reagent factory, AR) was adjusted to pH 9 by NH₃·H₂O with vigorous agitation. An Al(OH)₃ precipitate was formed. Then 50% Mn(NO₃)₂ and NH₃·H₂O were added dropwise into the Al(OH)₃ precipitate at pH 9 under vigorous stirring. The final precipitate was aged at 70 °C for 3 h, filtered off, washed with distill water, and dried at 110 °C for 12 h. After that, the dried powder was divided into four parts and calcined for 6h at 300, 400, 500, and 600 °C, respectively. MnO_x/γ-Al₂O₃ powder containing MnO_x 80% and γ-Al₂O₃ 20% was obtained.

The Pd catalysts were prepared by wet impregnation. The loading of Pd was 1.6 wt% for all of the samples. (1). An amount of an aqueous solution of Pd(NO₃)₂ (Chengdu Guangming Equipment Company, AR) was impregnated in the MnO_x prepared by the sol-gel method and γ-Al₂O₃ in a mass ratio of 4:1. Then, the Pd/MnO_x and Pd/γ-Al₂O₃ powder catalysts were mixed together. (2). An amount of an aqueous solution of Pd(NO₃)₂ was impregnated in the MnO_x/γ-Al₂O₃ prepared by the sequential precipitation method. (3). An amount of an aqueous solution of Pd(NO₃)₂ was impregnated in γ-Al₂O₃. After impregnation, the Pd/MnO_x+Pd/γ-Al₂O₃, Pd/MnO_x/γ-Al₂O₃, and Pd/γ-Al₂O₃ catalysts were dried at 120 °C overnight, and then calcined at 200 °C for 3 h in air.

The above powder was ball-milled with distilled water to form a slurry. The slurry was coated onto a cordierite substrate of 0.28 cm³ (Coring Corporation in American, 400 pore·inch^-2, diameter 5 mm, length 14 mm) and excess slurry was blown off. The catalyst was dried at 110 °C overnight and calcined at 200 °C for 3 h in air to obtain the Pd/MnO_x+Pd/γ-Al₂O₃, Pd/MnO_x/γ-Al₂O₃, MnO_x+γ-Al₂O₃, MnO_x/γ-Al₂O₃, and Pd/γ-Al₂O₃ catalysts. The loading of catalyst on the cordierite substrate was 350 g·L^-1.

H₂-TPR was carried out using an automated instrument. In a typical experiment, 100 mg of sample was loaded in a U-shaped quartz microreactor. The sample was heated from room temperature to 550 °C at a heating rate 10 °C·min^-1 in a flowing 5%H₂/N₂ mixture gas (30 cm³·min^-1). Hydrogen consumption was monitored using a thermal conductivity detector. The XRD analysis was conducted on DX-1000 X-ray diffractometer using Cu K_α radiation (λ = 0.15406 nm) at 40 kV and 25 mA. The X-ray diffraction (XRD) data were recorded for 2θ values from 10° to 80° at a scan 0.05°/s. X-ray photoelectron spectra (XPS) of the samples were acquired at room temperature using a Vacuum Generator Scientific XSAM800 system (Kratos Co., UK). The spectra were recorded with Mg K_α (hν = 1253.6 eV) radiation. For the X-ray radiation, the electron energy analyzer was operated at 180 W (12 kV, 15 mA). The specific surface area and pore size of the catalysts were determined by N₂ adsorption-desorption at -196°C on a QUADRASORB SI Automated Surface Area and Pore Size Analyzer (Quantachrome Instruments). Before the measurement, the sample was degassed in vacuum at 350 °C for 1 h.

The catalyst tests were performed in a continuous flow tubular quartz reactor (inner diameter 10 mm) placed in a temperature programmed furnace. The catalyst temperature was controlled by a thermocouple mounted internally. Ozone gas was generated from an ozonizer (JY-3 type, Chengdu Qiangui Purification Equipment Company) and was fed from an independent mass flow controller. Air was from an air compressor, then was separated by an oil-water separator and dried over silica gel. Its flow rate was controlled using a rotameter. Measurement was performed at a relative humidity (RH) of 85%-90% using gas hourly space velocities (GHSV) of 380000, 450000, 510000, and 580000 h^-1. The feed gas consisted of 0.6 µL∙L^-1 of O₃ and balance air. The reactor outlet flow was analyzed using an ozone analyzer (measurement range 0-20 µL∙L^-1, error ±0.02 µL∙L^-1, Nanjing 8Shang Technology Co., Ltd.). The activity of the catalyst was calculated on the basis of the equation: O₃ conversion = (inlet concentration of O₃- outlet concentration of O₃)/inlet concentration of O_3. The initialconversion was referred to O₃ decomposed at 12 °C.

As shown in Fig. 1(a), the initial O₃ conversion at 12 °C was 90%, 70% and 30%, and the O₃ complete conversion temperature was 36, 44, and 62 °C, for the Pd/MnO_x+Pd/γ-Al₂O₃, MnO_x + γ-Al₂O₃ and Pd/γ-Al₂O₃ catalysts, respectively. The activities of the catalysts increased with rising reaction temperature, especially for the Pd/γ-Al₂O₃ sample. At 12-30 °C, the activities of the three catalysts were quite different Pd/γ-Al₂O₃ < MnO_x + γ-Al₂O₃ < Pd/MnO_x+Pd/γ-Al₂O₃. A synergetic effect between Pd and MnO_x was observed.

Fig. 1. conversion as a function of reaction temperature on different catalysts prepared by the sol-gel method (a) and sequential precipitation method (b). Reaction conditions: O3 inlet concentration = 0.6 ± 0.02 µL∙L-1, GHSV = 450000 h-1, RH = 85%-90%.

As shown in Fig. 1(b), the initial O₃ conversion at 12 °C was 77% and 56%, and the O₃ complete conversion temperature was 46 and 62 °C for the Pd/MnO_x/γ-Al₂O₃ and MnO_x/γ-Al₂O₃ catalysts, respectively. O₃ decomposition was enhanced with reaction temperature elevation. The loading of Pd improved the activity of the MnO_x/γ-Al₂O₃ sample due to its activity and resistance to humidity.

Figure 2(a) shows the H₂-TPR profiles of the Pd/γ-Al₂O₃, MnO_x+γ-Al₂O₃, and Pd/MnO_x+Pd/γ-Al₂O₃ catalysts prepared by the sol-gel method. A weak peak at 121 °C was present for the Pd/γ-Al₂O₃ catalyst. This was attributed to the reduction of the PdO species ^{[22, 23]}. For the MnO_x + γ-Al₂O₃ catalyst, broad peaks at 459 and 538 °C were observed, which were ascribed to the reduction of Mn₃O₄ to MnO ^[23]. For the Pd/MnO_x+ Pd/γ-Al₂O₃ catalyst, the peak at 116 °C was attributed to the reduction of PdO species ^[22], the peak at 187 °C was due to Mn₂O₃ reduced to Mn₃O₄ ^[23], and a minor peak at 407 °C belonged to Mn₃O₄ reduced to MnO ^[23]. The reduction temperature of MnO_x was lowered in the presence of Pd, implying that there was a strong interaction between Pd and MnO_x. Xu et al. ^[24] reported that the presence of Pd lowered the reduction temperature of MnO_x due to hydrogen spillover from Pd to the oxide. The interaction of Pd and MnO_x improved the reducibility of MnO_x. Dhandapani et al. ^[10] and Ren et al. ^[25] reported that the activity of MnO_x was related to its reducibility. Therefore, the activity of the Pd/MnO_x+Pd/γ-Al₂O₃ catalyst would be enhanced for the decomposition of O₃.

Fig. 2. H2-TPR profiles of different catalysts prepared by the sol-gel method (a) and sequential precipitation method (b).

Figure 2(b) exhibits H₂-TPR profiles of the MnO_x/γ-Al₂O₃ and Pd/MnO_x/γ-Al₂O₃ catalysts prepared by the sequential precipitation method. For the MnO_x/γ-Al₂O₃ catalyst, the peaks at 175 and 304 °C were attributed to MnO₂ reduced to Mn₃O₄ ^[23]. The peak at 460 °C was ascribed to Mn₃O₄ reduced to MnO ^[23]. For the Pd/MnO_x/γ-Al₂O₃ catalyst, the broad peak at 188 °C may be due to the reduction of PdO and MnO₂ to Mn₃O₄. The reduction temperature of MnO_x was decreased, which was due to the interaction of Pd and MnO_x. From Fig. 2, it was seen that the reduction of the Pd/MnO_x+Pd/γ-Al₂O₃ catalyst prepared by the sol-gel method was easier than that of the Pd/MnO_x/γ-Al₂O₃ prepared by the sequential precipitation method. This implied that the activity of the Pd/MnO_x+Pd/γ-Al₂O₃ catalyst was better than that of the Pd/MnO_x/γ-Al₂O₃ catalyst.

Figure 3 shows the catalytic activity for O₃ decomposition on Pd/MnO_x+Pd/γ-Al₂O₃ or Pd/MnO_x/γ-Al₂O₃ using MnO_x prepared by the sol-gel method and sequential precipitation methods and calcined at different temperatures. The experimental results showed that O₃ conversion at 12 °C was 86%, 90%, 88%, and 85%, and the O₃ complete conversion temperature was 40, 35, 38, and 42 °C over the samples using MnO_x prepared by the sol-gel method and MnO_x calcination temperatures of 300, 400, 500, and 600 °C, respectively. The calcination temperature of MnO_x had a slight impact on the catalytic activity of these catalysts. This indicated that the activity of the prepared catalysts was relatively steady. Ground-level O₃ can be completely decomposed at 35 °C at a space velocity of 450000 h^-1 over the catalyst prepared by using MnO_x calcined at 400 °C.

Fig. 3. O3 conversion on Pd/MnOx+Pd/γ-Al2O3 prepared by the sol-gel method (1,4) and Pd/MnOx/γ-Al2O3 prepared by the sequential precipitation method (2,3) with different calcination temperatures of MnOx. Reaction conditions: O3 inlet concentration = 0.6 ± 0.02 µL∙L-1, GHSV = 450000 h-1, RH = 85%-90%.

The O₃ conversion at 12 °C was 67%, 73%, 82%, and 75%, and the O₃ complete conversion temperature was 60, 55, 46, and 51 °C over the samples using MnO_x prepared by the sequential precipitation method and calcined at 300, 400, 500, and 600 °C, respectively. The effect of the calcination temperature of MnO_x on the catalytic activity was obvious.

As shown in Fig. 4, the phases of MnO_x were Mn₃O₄ and Mn₅O₈ when MnO_x was calcined at 300 °C. The phases of MnO_x were Mn₃O₄, Mn₅O₈ and a little Mn₂O₃ for calcination at both 400 and 500 °C. The phases of MnO_x were Mn₂O₃ and Mn₃O₄ for calcination at 600 °C. A portion of Mn₃O₄ was oxidized into Mn₂O₃ in air, and Mn₅O₈ was decomposed to Mn₂O₃ when the calcination temperature was increased from 300 to 400 °C. The diffraction peaks of Mn₃O₄, Mn₅O₈, and Mn₂O₃ were enhanced at 500 °C, and their crystalline gradually became excellent. Mn₅O₈ was decomposed completely at 600 °C, and the intensities of the diffraction peaks of Mn₃O₄ and Mn₂O₃ were stronger and the crystalline became better. Combining the phase of MnO_x with the activity data shown in Fig. 3, the activity of the catalyst constituted of Mn₃O₄, Mn₅O₈, and Mn₂O₃ was high for O₃ decomposition.

Fig. 4. XRD patterns of Pd/MnOx+Pd/γ-Al2O3 catalyst prepared by the sol-gel method and different calcination temperatures of MnOx.

According to Fig. 5, the phases of MnO_x were MnO₂ and Mn₅O₈ at the calcination temperatures of 300, 400, and 500 °C. The phases of MnO₂ were Pyrolusite MnO₂ and a little Akhtenskite MnO₂ at 300 and 400 °C. The phase of MnO₂ was only Akhtenskite MnO₂ at 500 °C. The phases of MnO_x were Mn₂O₃ and Mn₅O₈ at 600 °C. The diffraction peaks of MnO₂ and Mn₅O₈ were broad and diffused at 300 °C. The diffraction peaks of MnO₂ and Mn₅O₈ gradually became strong at 400 °C. Pyrolusite MnO₂ was completely transformed into Akhtenskite MnO₂ at 500 °C. MnO₂ disappeared and the diffraction peaks of Mn₂O₃ appeared at 600 °C.

Fig. 5. XRD patterns of Pd/MnOx/γ-Al2O3 prepared by the sequential precipitation method and different calcination temperatures of MnOx.

Combining with the Fig. 3, the activity of the catalysts followed the order: catalyst containing Akhtenskite MnO₂ and Mn₅O₈ > catalyst constituted of Mn₂O₃ and Mn₅O₈ > catalyst with mixed phases of Pyrolusite MnO₂, a little Akhtenskite MnO₂ and Mn₅O₈.

According to the result in Figs. 4 and 5, we concluded that the phase of MnO_x present was determined by the calcination temperature and preparation method. Mn has valence states of +2, +3, +4, and +7, and many phases as MnO, MnO₂, Mn₂O₃, Mn₃O₄, and Mn₅O₈. The excellent activity of the various catalysts differed due to their unique phase composition. For example, the phases of MnO_x were Mn₃O₄, Mn₅O₈, and Mn₂O₃ for the catalysts prepared by the sol-gel method. However, the phases of MnO_x were Akhtenskite MnO₂ and Mn₅O₈ for the catalysts prepared by the sequential precipitation method. This is an interesting and peculiar phenomenon for O₃ decomposition. Dhandapani et al. ^[10] reported that MnO₂ was active. Zhou et al. ^[12] reported that the mixed phase of MnO₂ and Mn₂O₃ in a proper proportion favored the decomposition of O₃. The catalytic activity was excellent on the catalyst constituted of Mn₃O₄, Mn₅O₈, and Mn₂O₃ prepared by the sol-gel method in the paper. Mn₃O₄ may be seen as 2(MnO)∙MnO₂, and Mn₅O₈ can be seen as 2(MnO)∙3(MnO₂). That is, the catalyst constituted of MnO, MnO₂, and Mn₂O₃ has the best activity. The mechanism will be further studied in the future.

According to Ref. ^{[26, 27]}, the binding energy of Mn 2p_3/2 was 640.8, 641.8, and 642.6 eV for Mn(II), Mn(III), and Mn(IV), respectively. From the XPS spectra of Mn 2p_3/2, it can be seen that the valence states of Mn were +2 and +4 on the surface of the catalyst using MnO_x calcined at 300 °C. The valence states of Mn were +2, +3,and +4 on the surface of the catalyst using MnO_x calcined at 400, 500, and 600 °C. The amounts of the various valence states of Mn are listed in Table 1. Combining with the XRD results, the amounts of Mn²⁺ and Mn⁴⁺ were decreased when calcination temperature increasing from 300 to 400 °C, and Mn³⁺ was formed. It was possible that Mn₃O₄ was partially oxidized into Mn₂O₃ (2Mn₃O₄ + 1/2O₂ → 3Mn₂O₃) ^[28], and Mn₅O₈ was decomposed into Mn₂O₃ (4Mn₅O₈ → 10Mn₂O₃ + O₂). The above reactions were sequential and the amount of Mn³⁺ was increased at 500 °C. Mn₅O₈ was completely transformed into Mn₂O₃ at 600 °C, and the amount of Mn³⁺ was increased considerably.

Table 1 XPS results of the Mn 2p3/2 peak for Pd/MnOx+Pd/γ-Al2O3 prepared by the sol-gel method and different calcination temperatures of MnOx. Temperature (°C) Mn2+ Mn3+ Mn4+ Position (eV) Concentration (at%) Position (eV) Concentration (at%) Position (eV) Concentration (at%) 300 640.6 41 — — 642.2 59 400 640.6 29 641.5 17 642.1 54 500 640.6 22 641.5 28 642.2 50 600 640.6 22 641.5 67 642.2 11

Table 1 XPS results of the Mn 2p3/2 peak for Pd/MnOx+Pd/γ-Al2O3 prepared by the sol-gel method and different calcination temperatures of MnOx.

In the XPS spectra of O 1s, the peak with a low binding energy (O_I: 529.7-530.1 eV) was attributed to lattice oxygen (O^2-), the peak with the middle binding energy (O_II: 531.0-531.6 eV) was ascribed to surface adsorbed oxygen (O₂^-/O^-), hydroxide and oxygen vacancy, and the peak with the high binding energy (O_III: 533.0 eV) was due to oxygen in the form of chemically adsorbed water ^{[27, 29]}. The amounts of surface adsorbed oxygen (O₂^-/O^-), hydroxide and oxygen vacancy were the highest on the catalyst using MnO_x calcined at 400 °C, as shown in Table 2, the following order is at 500 °C > 300 °C > 600 °C.

Table 2 XPS results for the O 1s peak for Pd/MnOx+Pd/γ-Al2O3 prepared by the sol-gel method and different calcination temperatures of MnOx. Temperature (°C) Concentration (at%) OI (at 529.4 eV) OII (at 530.6 eV) OIII (at 532.0 eV) 300 31 42 27 400 24 45 31 500 20 43 37 600 24 41 35

Table 2 XPS results for the O 1s peak for Pd/MnOx+Pd/γ-Al2O3 prepared by the sol-gel method and different calcination temperatures of MnOx.

For the Pd/MnO_x+Pd/γ-Al₂O₃ catalyst prepared by the sol-gel method, the activity of O₃ decomposition was the highest when Mn²⁺, Mn³⁺, and Mn⁴⁺ in a molar ratio of 1.7:1:3 coexisted and the amount of the surface active oxygen species was the highest.

From the XPS spectra of the Mn 2p_3/2 peak, it can be seen that the valence states of Mn were +2 and +4 on the surface of the catalysts using MnO_x calcined from 300 to 500 °C. Combining with the XRD results, the phase of MnO_x was Mn₅O₈ and MnO₂ for the catalysts using MnO_x calcined from 300 to 500 °C. Mn₅O₈ was constituted of Mn²⁺ and Mn⁴⁺ (2MnO·3MnO₂), in agreement with the report in literature ^[30]. The amounts of Mn²⁺ and Mn⁴⁺ were hardly changed for calcination from 300 to 500 °C listed in Table 3. Pyrolusite MnO₂ was gradually transferred into Akhtenskite MnO₂ in this range of temperatures. The valence states of Mn were +2, +3, and +4 on the surface of the catalyst using MnO_x calcined at 600 °C. The phases of MnO_x were Mn₅O₈ and Mn₂O₃, MnO₂ and some Mn₅O₈ were decomposed to Mn₂O₃.

Table 3 XPS results of the Mn 2p3/2 peak for Pd/MnOx/γ-Al2O3 prepared by the sequential precipitation method and different calcination temperatures. Temperature (°C) Mn2+ Mn3+ Mn4+ Position (eV) Concentration (at%) Position (eV) Concentration (at%) Position (eV) Concentration (at%) 300 641.0 36 — — 642.6 64 400 641.0 34 — — 642.5 66 500 641.0 34 — — 642.5 66 600 641.0 9 641.5 50 642.8 41

Table 3 XPS results of the Mn 2p3/2 peak for Pd/MnOx/γ-Al2O3 prepared by the sequential precipitation method and different calcination temperatures.

In the XPS spectra of the O 1s peak, the peak at 529.7 eV was attributed to lattice oxygen (O^2-), the peak at 531.0 eV was ascribed to surface adsorbed oxygen (O₂^-/O^-), hydroxide and oxygen vacancy, and the peak at 532.3 eV was due to oxygen in chemically adsorbed water ^{[27, 29]}. The amount of surface adsorbed oxygen (O₂^-/O^-), hydroxide and oxygen vacancy increased gradually for calcination temperature from 300 to 500 °C, and then was decreased slightly at 600 °C shown in Table 4.

Table 4 XPS results for the O 1s peak for Pd/MnOx/γ-Al2O3 prepared by the sequential precipitation method. Temperature (°C) OI OII OIII Position (eV) Concentration (at%) Position (eV) Concentration (at%) Position (eV) Concentration (at%) 300 529.7 25 531.0 44 532.3 31 400 529.7 31 531.0 48 532.3 21 500 529.7 30 531.0 51 532.3 19 600 529.7 26 531.0 49 532.2 25

Table 4 XPS results for the O 1s peak for Pd/MnOx/γ-Al2O3 prepared by the sequential precipitation method.

It is helpful for O₃ decomposition that Mn²⁺ and Mn⁴⁺ coexisted on the surface of the Pd/MnO_x/γ-Al₂O₃ sample, and the amount of surface active oxygen species was the highest.

Einaga et al. ^[31] suggested a mechanism for O₃ decom­position as O₃ + Mnⁿ⁺ → O^2- + Mn⁽ⁿ⁺²⁾⁺ + O₂; O₃ + O^2- + Mn⁽ⁿ⁺²⁾⁺ → O₂^2-+ Mn⁽ⁿ⁺²⁾⁺ + O₂; O₂^2- + Mn⁽ⁿ⁺²⁾⁺ → Mnⁿ⁺+O₂. This indicated that the active Mnⁿ⁺ participated in both oxidation and reduction reactions during O₃ decomposition, and it underwent a cycle of oxidation-reduction. Therefore, ozone can be decomposed effectively on the surface of catalysts on which MnO_x comprised Mn²⁺ (MnO), Mn³⁺ (Mn₂O₃), and Mn⁴⁺ (MnO₂).

Bulanin et al. ^[32] reported that first O₃ interacted with surface hydroxide to form a hydrogen bond. Secondly, it dissociated and adsorbed on Lewis sites to generate the active oxygen species O₂^-, O^-, and O₂^{2- [10, 32]}. Li et al. ^[33] put forward the mechanism of O₃ decomposition as O₃ + * → O₂ + O; O + O₃ → O₂ + O₂; O₂→ O₂ + (* denoted site). That is, the surface active oxygen species played an important role. Surface adsorbed oxygen, hydroxide, and oxygen vacancy are the active oxygen species in the samples. The catalytic activity of the samples was coincident with the amounts of the active oxygen species ^[10]. The more surface active oxygen species there is, the better is the catalytic activity, regardless of the preparation method and calcination temperature of the MnO_x.

As shown in Table 5, for the Pd-MnO_x/γ-Al₂O₃ samples with the MnO_x prepared by the sol-gel and sequential precipitation methods, the binding energies Pd 3d_5/2 were from 336.6 to 336.9 eV. The binding energy of Pd in PdO is 336.7 eV ^[34]. Therefore, Pd in the prepared samples was Pd²⁺. The binding energy of Pd in Table 5 was in the range of measurement error. This meant that the valence state of Pd was not obviously different on the surface of two samples. The role of Pd and the mechanism of the resistance to water vapor using Pd in the decomposition of O₃ will be studied in the future.

Table 5 XPS results for the Pd 3d5/2 and Pd 3d3/2 peaks for Pd-MnOx/γ-Al2O3 prepared by the sol-gel and sequential precipitation methods. MnOx calcination temperature (°C) Sol-gel method Sequential precipitation method Pd 3d5/2 Pd 3d3/2 Pd 3d5/2 Pd 3d3/2 300 336.6 342.0 336.6 341.9 400 336.9 342.2 336.8 342.2 500 336.9 342.2 336.7 342.1 600 336.7 342.0 336.9 342.3

Table 5 XPS results for the Pd 3d5/2 and Pd 3d3/2 peaks for Pd-MnOx/γ-Al2O3 prepared by the sol-gel and sequential precipitation methods.

The textural characteristics of the samples prepared using MnO_x prepared by different methods are present in Table 6. The specific surface area and total pore volume decreased gradually with elevating of the calcination temperature of MnO_x for the catalyst using MnO_x prepared by sol-gel method. Although the textural property of the catalyst using MnO_x calcined at 300°C was better than that of the catalyst using MnO_x calcined at 500 °C, its catalytic activity was low. The textural property and activity of the catalyst using MnO_x calcined at 400 °C was better than that of the catalyst using MnO_x calcined at 500 °C. This indicated that a better textural property favored O₃ decomposition for the same phase of MnO_x. The textural property has only a small impact on the activity of the catalyst prepared by the sol-gel method.

Table 6 Textural characteristics of Pd/MnOx+Pd/γ-Al2O3 and Pd/MnOx/γ-Al2O3 using MnOx calcined at different temperatures MnOx calcination temperature (°C) Sol-gel method Sequential precipitation method ABET/(m2·g-1) V/(cm3·g-1) D/nm ABET/(m2·g-1) V/(cm3·g-1) D/nm 300 100 0.21 7.7 133 0.23 6.9 400 88 0.19 7.6 128 0.24 6.6 500 72 0.12 5.8 107 0.22 8.2 600 70 0.10 6.0 81 0.21 10.8 ABET: BET surface area; V: Total pore volume; D: Average pore diameter.

Table 6 Textural characteristics of Pd/MnOx+Pd/γ-Al2O3 and Pd/MnOx/γ-Al2O3 using MnOx calcined at different temperatures

The specific surface area decreased sharply, total pore volume remained invariable, and pore diameter decreased slightly and then increased when the calcination temperature of MnO_x was increased from 300 to 600 °C for the catalysts that used MnO_x prepared by the sequential precipitation method. This showed that the surface area of the prepared catalyst was relatively sensitive to temperature variation. The large porous framework can be easily damaged during thermal treatment, subsequently resulting in a considerable aggregation of crystal grains and a sharp decrease in surface area ^[35]. The catalyst using MnO_x calcined at 500°C has the best performance of O₃ decomposition due to a large surface area and pore diameter. A large surface area helps the dispersion of Pd on the surface of catalyst, where Pd would suppress the adsorption of H₂O molecules ^[18]. Therefore, this catalyst would facilitate O₃ decomposition under a high relative humidity. A large pore favors O₃ mass transfer. It was obvious that the textural property has an obvious impact on the catalytic activity.

We can conclude that the O₃ decomposition was closely correlated to the physicochemical properties of the catalysts. The phase of MnO_x is important. However, it is not the only factor affecting the activity for O₃ decomposition. Actually, the MnO_x species, surface active oxygen species, and textural property together determine the catalytic activity.

O₃ conversion on the catalysts is illustrated in Fig. 6. It was seen that the O₃ initial conversion at 12 °C on the Pd/MnO_x+Pd/γ-Al₂O₃ catalyst using MnO_x prepared by the sol-gel method was 93%, 87%, 82%, and 74%, and the O₃ complete conversion temperature was 16, 35, 62, and 83 °C at GHSVs of 380000, 450000, 510000, and 580000 h^-1, respectively. The O₃ initial conversion at 12 °C on the Pd/MnO_x/ γ-Al₂O₃ catalyst using MnO_x prepared by the sequential precipitation method was 83%, 77%, 70%, and 61%, and the O₃ complete conversion temperature was 23, 46, 71, and 95 °C at GHSVs of 380000, 450000, 510000, and 580000 h^-1, respectively.

Fig. 6. O3 conversion on Pd/MnOx+Pd/γ-Al2O3 prepared by the sol-gel method (1,4) and Pd/MnOx/γ-Al2O3 prepared by the sequential precipitation method (2,3) at different space velocities. Reaction conditions: O3 inlet concentration = 0.6 ± 0.02 µL∙L-1, RH = 85%-90%.

Wu et al. ^[19] reported that the O₃ conversion decreased by 10% in the range of GHSVs of 132000-226000 h^-1, where the O₃ conversion rate was much faster than the rate of transporting the reactant to (or the product away from) the catalyst under the experimental conditions. The O₃ conversion rate approached the rate of O₃ mass transfer limitation. The O₃ decomposition declined by 20% when GHSV changed from 380000-580000 h^-1 in our experiments, where the O₃ conversion rate was slower than the rate of mass transfer of O₃ and O₂. O₃ contact with the active site was shortened on the surface of catalyst with the increase of GHSV, which decreased O₃ decomposition. According to the experimental results, O₃ was completely decomposed at room temperature (16-23 °C) at a low GHSV (such as 380000 h^-1). Therefore, O₃ would be decomposed completely when these catalysts were coated on air conditioner condensers. Also, O₃ would be decomposed completely if the catalysts were coated on vehicle radiators when the vehicle was cold started. Furthermore, the O₃ complete conversion temperature was 83 °C on the Pd/MnO_x+Pd/γ-Al₂O₃ catalyst prepared by the sol-gel method using MnO_x calcined at 400 °C at a higher GHSV (580000 h^-1). O₃ would be decomposed completely if the catalyst was coated on vehicle radiators in which the temperature ranged from 25 to 90 °C. Therefore, the prepared catalyst has potential application for O₃ abatement.

According to analysis above, the activities and physicochemical properties of the catalysts were significantly affected by the preparation method. The activity for O₃ decomposition was better on the Pd/MnO_x+Pd/γ-Al₂O₃ catalyst using MnO_x prepared by the sol-gel method.