赖潇潇 ，冯洁 ，周晓英 ，侯忠燕 ，林涛 ,＊，陈耀强 ,3,＊

1四川大学化学学院，成都 610064

2四川大学化学工程学院，成都 610064

3四川省环境催化材料工程技术中心，成都 610064

1 Introduction

Volatile organic compounds (VOCs), originating from industrial production and human life, not only pollute atmospheric environment but also endanger human health due to its toxicity1. Among them, toluene as a common organic solvent and industrial raw material is mainly responsible for photochemistry pollution1. Many treatment technologies,including absorption, thermal combustion and catalytic oxidation and so on, have been applied to eliminate the VOCs in recent years, and the catalytic combustion is a perfect pathway because it can be handled at low temperature and without secondary pollution2,3.

It is known to all that the supported precious metal catalysts(Pd, Pt)have excellent catalytic activities for the oxidation of VOCs, however the expensive price and their own shortcomings(e.g., volatility and sintering)prohibit their extensive application4,5. Recently, it is found that the cheap transition metal oxides, such as manganese oxides, copper oxide, and cobalt oxides, have good catalytic performance for VOCs oxidation at high temperature6-8, but are inferior to that of precious metals at low temperature9-11. Hence, it is critical to develop an effective metal oxide catalyst for VOCs oxidation at low temperature.

Among the transition oxide catalysts for VOCs oxidation,manganese based catalysts such as MnO12，Mn3O42and binary Mn oxides13and supported manganese oxides14are known to display good activity and are also regarded as environmentally friendly materials. Manganese oxides with different valence states and crystalline phases have different catalytic effects on VOC oxidation15,16. For supported metal oxide catalysts, the selection of an appropriate carrier is essential since the interaction of the support with the active component affects the relevant properties of the catalyst, such as low temperature activity and thermal stability17. As reported in the literature, the CeZrO2materials applied as the support could enhance the lowtemperature reducibility, leading to the improvement of catalytic activity for redox reaction, which is due to their thermal stability and redox property18.

In recent years, some researchers have discovered that alkali metals have played an active role in many reactions, and have been used as catalyst promoters for improving the performance and durability of the catalyst19-22. Xue and co-workers23have declared that alkali metals could act as electronic promoters to improve the reducibility of Co2+, which lead to an enhancement of the N2O decomposition performance over cerium-cobalt composite oxide catalyst. Bai and Li24have found that doping potassium into 3-D Ag/Co3O4can increase the hydroxyl species on the active crystal surface of Ag (111), which is conducive to the low temperature oxidation of formaldehyde. Additionally,the alkali metal could also change the intrinsic structure of metal oxides. Tang et al.21have discovered that the Mn-O octahedral structure of manganese oxides could have an aberrance with the influence of potassium ions. The addition of potassium is beneficial to the existence of manganese oxides on γ-Al2O3in α-MnO2instead of β-MnO2, leading to the improvement of catalytic activity in aromatic alcohols oxidation on KMn/γ-Al2O325. Moreover, the alkali metals also lead to the alteration of acidic and basic properties of the catalysts26.

In our previous work, we have studied the toluene catalytic oxidation over MnOxsupported on mesoporous material Ce0.65Zr0.35O2and we have found that the excellent catalytic activity could be ascribed to good redox performance and more available surface oxygen species27. In this work, the catalytic property for the oxidation of toluene over K+modified Mn supported on ceria-zirconium material was investigated. In addition, the structure, redox performance and chemical valence state of elements on catalyst surface were analyzed by a series of characterizations. And the in situ DRIFTs experiment was carried out to investigate the adsorption/desorption of toluene.

2 Experimental

2.1 Catalyst preparation

The co-precipitation method was used to prepare the Ce0.65Zr0.35O2composites27. Firstly, a certain stoichiometric ratio cerium nitrate hexahydrate (99.9%, purity)and zirconium nitrate (99.9%, purity)were dissolved in deionized water and concentrated nitric acid respectively, and then mixed with vigorous stirred to form a uniform solution. Then, the above mixed solution was titrated simultaneously with the ammonia solution to obtain an initial precipitates. The resulting precipitates were stirred vigorously in a boiling water bath and aged for 6 h. After aged, filtered and washed, the precipitates were dried at 90 °C for 24 h. Afterwards, the products were calcined at 600 °C for 3 h in a dynamic air atmosphere to obtain the final composites which were labeled as CZ.

Catalysts modified with K+were prepared by co-impregnation method with 10% (w)Mn and an amount of K loaded supported on Ce0.65Zr0.35O2. Firstly, manganese nitrate and potassium nitrate aqueous solutions with a molar ratio (K/Mn)of 0, 0.05,0.2 and 0.5 were impregnated on the cerium-zirconium support respectively. Then, the desired catalysts were obtained after dried at 100 °C and calcined at 500 °C for 3 h. All the powder catalysts were coated on honeycomb cordierite monoliths27,28.Finally, the prepared monolithic catalysts with approximately 160 g·L-1washcoat were marked as Mn/CZ-K-y (y = 0, 0.05, 0.2,0.5. where “y” indicated the molar ratio of K/Mn).

2.2 Catalyst characterization

The powder X-ray diffraction (XRD)patterns were recorded on a D/max-RA diffractometer (Rigaku, Japan)equipped with Cu Kαradiation (λ = 0.15406 nm, 40 kV, 40 mA). The catalysts were performed at a scanning speed of 0.03 (°)·s-1in the range of 20° ≤ 2θ ≤ 80°.

The UV Raman spectroscopy data of catalysts were obtained on in via reflex Raman (Reni Shaw, UK)with excitation wavelength of 325 nm. Visible Raman results of catalysts were recorded on SR-500i Raman spectrograph (Andor, UK), and scanned with an excitation wavelength of 532 nm.

The specific surface areas of the samples were carried out on an automatic surface area and pore size analyzer (Autosorb SI,Quantachrome, USA)by N2adsorption-desorption at -196 °C.

The temperature-programmed reduction and desorption experiments (H2-TPR/O2-TPD)were carried out by an auto adsorption analyzer. For the H2-TPR, 100 mg samples were pretreated in pure N2(20 mL·min-1)at 450 °C for 45 min, then cooled down to 50 °C. The catalysts were finally heated in a mixed-gas flow (5% H2balanced with N2, 25 mL·min-1)from 50 °C to 900 °C. The signals were monitored by a thermal conductivity detector. The experimental process of O2-TPD was similar to H2-TPR; 50 mg sample was saturated with 5% O2/N2mixture flow at 50 °C after pretreated. Then purging the excess oxygen, the catalyst was heated up to 800 °C in helium gas flow.The oxygen desorption signals were also monitored by TCD detector.

The X-ray photoelectron spectroscopy (XPS)results of the catalysts were acquired on an XSAM-800 electron spectrometer(Kratos, England)equipped with an Al Kαradiation. The binding energy of C 1s was set to 284.6 eV to correct the charge effect of the all samples.

The toluene-TPSR in N2was recorded on an in situ DRIFTs analyzer (Thermo Nicolet 6700)and detected with a MCT detector. The catalysts were pretreated in nitrogen for 30 min at 400 °C and cooled down to 50 °C. Then, the catalysts were exposed to a flow of toluene/N2(0.1%)until the samples were saturated with toluene. And the chamber is heated directly from 50 °C to 400 °C at a rate of 10 °C·min-1in a flow of N2. The spectrums at different temperature were recorded. The final infrared spectrograms were exactly obtained after subtracting the background under the corresponding conditions.

2.3 Catalytic activity evaluation

The catalytic activities were evaluated in a fixed-bed continuous flow reactor of quartz tube. Fig. 1 showed the experimental schematic drawing for the catalytic tests. The total flow rate (440 mL·min-1)was adjusted to a gas hourly space velocity (GHSV)of 12000 h-1through the reactor and the concentration of toluene was 0.1% (air as the balance).

The output flow from the reactor was detected by an on-line gas chromatograph (GC-2000, China)equipped with FID detector.In the present work, no additional products except CO2and H2O were detected under the current experiment conditions29,30. The calculation equation for the conversion of toluene (x)was as following:

where Cinwas the inlet toluene concentration at steady state and Coutwas the outlet toluene concentration.

Fig. 1 The schematic diagram of the experimental set-up for the catalytic combustion experiments (1)Air cylinder; (2)mass flow controller; (3)thermometer; (4)toluene saturator; (5)oven and reactor; (6)gas chromatograph with an FID detector.

3 Results and discussion

3.1 XRD characterization of the catalysts

The XRD patterns of catalysts were displayed in Fig. 2, and Fig. 2b exhibited the details of 2θ from 30° to 46°. As showed in Fig. 2a, the main diffraction peaks of all catalysts were located at 29.1°, 33.7°, 48.3° and 57.5°, which were corresponded to the(111), (200), (220)and (311)crystal faces of cubic fluorite fabric structure Ce0.6Zr0.4O2(PDF-ICDD 38-1439). This indicated the formation of Ce-Zr-O solid solution31.

As shown in Fig. 2b, a weak diffraction peak was detected at 37.5° in the XRD pattern of Mn/CZ, which was indexed to the MnO2structure (PDF-ICDD 44-0141). After introduction of potassium, the peak of MnO2became weaker and no the diffraction peak of MnO2were detected when the K/Mn molar ratios increased to 0.2. This experimental phenomenon might be ascribed to that the potassium could promote the dispersion of manganese oxides or act as electronic promoters and parts of Mn4+were reduced into Mn3+lead to the disappeared peak of MnO232.

Fig. 2 XRD patterns of all the catalysts (a)and the magnified diffraction peaks range from 30° to 46° of 2θ (b).

Fig. 3 Visible (λ = 532 nm)(a)and UV (λ = 325 nm)(b)Raman spectra of the catalysts.

3.2 Raman results of the catalysts

The visible and UV Raman spectra of all the catalysts were shown in Fig 3. It was observed that all the catalysts exhibit a characteristic Raman peak around 460 cm-1in the visible Raman spectra，which was attributed to the F2gactive mode of cubic fluorite CeO233. (Raman peak of carrier CZ around 468 cm-1was showed in the inside illustration in Fig. 3a). With the increasing of doping potassium，the F2gpeak became weakened and the Raman shift moved toward the high wavenumber,indicating that the structure of cubic fluorite of CeO2began to be changed. Compared to the inside illustration in Fig. 3a, all the catalysts had obviously Raman peaks across 500 cm-1to 700 cm-1except for CZ carrier. Based on the references16,25, the strong Raman peak at 643 cm-1was belonged to the Mn-O-Mn stretching mode of β-MnO2, and 508 cm-1, 578 cm-1and 620 cm-1to α-MnO2respectively. It could be observed that for Mn/CZ catalyst, the main crystalline phase of MnO2was β-MnO2. However, with increasing the K content, β-MnO2was changed into α-MnO2gradually, and the main component of manganese oxide was α-MnO2for Mn/CZ-K-0.5. As for the Mn/CZ-K-0.2, both β-MnO2and α-MnO2might be existed in the catalyst. As reported in previous study25, the presence of potassium could promote the crystalline transformation from β-MnO2phase to α-MnO2phase, and the co-existed of α and β-MnO2was conducive to the formation of lattice defects16.

Fig. 4 The I620/I460 value of all catalysts.

Since UV resonance Raman had strong interaction with the defect sites of ceric oxide, resulting in enhancing the Raman signal of defect sites31. Therefore, UV Raman could effectively investigate the defective structure. Fig. 3b showed the UV Raman patterns of catalysts, the characteristic peak at 620 cm-1was ascribed to defect space and 467 cm-1assigned to cubic fluorite CeO234. The relative concentration of oxygen vacancies of catalysts was depended on the intensity ratio of the peak at 620 cm-1to that at 467 cm-1(I620/I460)35. Fig. 4 showed the effects of potassium content on the I620/I460values for Mn/CZK-y catalysts. According to the pattern, the concentration trend of oxygen vacancy varied with the K/Mn ratio, following the sequence of Mn/CZ-K-0.2 ＞ Mn/CZ-K-0.05 ＞ Mn/CZ ＞ Mn/CZK-0.5. This result indicated that Mn/CZ-K-0.2 had the most oxygen vacancies among the catalysts, which was well content with the visible Raman results.

Table 1 Textural properties and H2-TPR results of catalysts.

3.3 Catalyst textural properties

The N2adsorption-desorption isotherms of supported manganese oxide catalysts were performed and the main textural properties of these catalysts were summarized in Table 1. From Table 1 it could be observed that the specific surface of the Mn/CZ decreased slightly after manganese were loaded on CZ.Compared with that of Mn/CZ, the BET surface area and the pore volume of all the K-modified samples decreased gradually with the potassium content increasing. This phenomenon could be due to that after introduction of potassium components, some micropores were blocked and the surface area and pore volume were declined36. However, there was no significant difference in textural properties of all the catalysts. Therefore, it was deduced that the textural properties of the catalysts had little effect on its catalytic performance in current work.

3.4 Results of H2-TPR analysis

The reduction properties of the samples were investigated by the H2-TPR experiments and the H2consumption profiles were shown in Fig. 5. It could be found that a wide peak ranged from 500 to 650 °C was observed in the curve of pure CZ support,which was attributed to the hydrogen consumption of surface and bulk oxygen of CeO237.

For the all the supported manganese oxide catalysts, it could be found that two overlapped reduction peaks appeared at the temperature range of 200-500 °C. According to the studies of Azalim and Zhou38,39, the first peak around 330 °C could be ascribed to the reduction step of MnO2/Mn2O3to Mn3O4while the second reduction peak around 420 °C could be assigned to the reduction step of Mn3O4to MnO. In addition, the broad peak more than 500 °C was belonged to the reduction of oxygen species of CeO2(Ce4+to Ce3+).

Fig. 5 H2-TPR profiles of the samples.

From Fig. 5 it could be observed that when small amount of K was co-impregnation with the manganese oxides, the peaks shifted to lower temperature, which suggested that the reducibility of catalysts could be enhanced by doping K. It was noteworthy that the Mn/CZ-K-0.2 exhibited the lowest reduction temperature, which indicated that the mole ratio of 5 : 1 (Mn/K)was optimum for improving of the reduction ability. However,as the potassium content continued to increase, the reduction peaks moved to higher temperature again. Jiratova has investigated the effect of potassium for Co-Mn-Al mixed oxide catalysts and they found similar phenomenon in their H2-TPR experiments26. According to literature40, the addition of potassium element could enhance the activation of oxygen in the Mn-O bond and promote the mobility of oxygen species in catalysts. Hence enhancing the redox performance of catalysts would lead to the improvement of its catalytic performance41.

In order to quantitatively analyze the TPR results of catalysts,the total hydrogen consumption experiments of all samples were also performed and the results were listed in Table 1. It was obvious that the hydrogen consumption of Mn/CZ increased greatly after manganese was loaded on CZ. Nevertheless, the hydrogen consumption of the catalysts was decreased after the doping of alkali metal potassium. This might be due to the potassium performed as an electron promoter to boost the conversion of partial high-valent manganese to the low valence,leading to a decrease in hydrogen consumption23. However,Mn/CZ-K-0.2 had the lowest reduce temperature, indicating that Mn/CZ-K-0.2 catalyst had the best low-temperature reducibility among all the catalysts.

3.5 O2-TPD results analysis

To research the influence of oxygen species on the catalytic performance of the catalysts, the O2-TPD experiments were measured and displayed in Fig. 6. Based on the desorption temperature of oxygen, it could be determined that there were two kinds oxygen species in the catalyst42. The O2desorption below 400 °C was ascribed to the surface adsorbed oxygen(noted as α)and the desorption peaks at middle temperature of 400-650 °C could be ascribed to the near-surface lattice oxygen(noted as β)and the peaks higher than 650 °C could be assigned to the lattice oxygen27,42. Generally, the near-surface oxygen is near the solid surface, while the lattice oxygen comes from the inner bulk oxygen vacancies. From Fig. 6 it could be observed that the O2desorption temperature of β was decreased by the addition of potassium, suggesting that the mobility of lattice oxygen was enhanced. Compared with the other samples,Mn/CZ-K-0.2 had the lowest desorption temperature, which indicated that Mn/CZ-K-0.2 would have the best mobility of lattice oxygen. The maximum oxygen desorption temperature(Td)and the amount of oxygen desorption of surface adsorbed oxygen (α)were calculated and exhibited in Table 2. An obvious result could be found in the Table 2 that the Mn/CZ-K-0.2 had the biggest desorption amount of oxygen species for 262.4 μmol·g-1catalysts with the lowest desorption temperature of 118°C. On the basis of Mars-van Krevelen mechanism26,43, the organic molecules might be oxidized by surface adsorbed oxygen species or near-surface lattice oxygen, then the lattice oxygen would move to the surface and supply the surface oxygen species and the lattice oxygen could be supplemented by gas phase oxygen. Therefore, the oxidation reaction was related to the participation of surface adsorbed oxygen and lattice oxygen. Mn/CZ-K-0.2 had the most surface oxygen species and more active lattice oxygen which were both benefitted to the toluene oxidation.

Fig. 6 O2-TPD patterns of as-prepared catalysts.

3.6 XPS analysis of the catalysts

The XPS technique was conducted to study the state of the surface elements of all the catalysts. Fig. 7 showed the XPS spectra of Ce 3d, Mn 2p and O 1s and the element relative contents were calculated in Table 3. As showed in Fig. 7a, letters U and V were denoted as the Ce 3d3/2and Ce 3d5/2, respectively.The U, U’, U’’ could be assigned to Ce4+3d3/2, whereas the V,V’, V’’ were ascribed to Ce4+3d5/2. And the V’ and U’ were attributed to homologous peaks of Ce3+, respectively. According to the works of Atribak44, the Ce3+would contribute to create oxygen vacancy. However, as seen from Table 3, the percentage content of Ce3+/Cetotalwhich were obtained by the whole peak areas of Ce3+to the total peak areas of Ce content were little changes from 14.7% to 15.6%, so the addition of K had a little effect on the ratio of Ce3+/Cetotal.

Table 2 Maximum oxygen desorption temperatures (Td)and the amount of oxygen desorption below 400 °C of the samples.

Fig. 7 XPS spectra of (a)Ce 3d, (b)Mn 2p and (c)O 1s of the catalysts.

As displayed in Fig. 7b, the Mn 2p3/2signals of all catalysts could be separated to three portions. The binding energies in the intervals 640.5-641.0 eV, 641.6-642.2 eV and 642.9-643.6 eV were attributed to the Mn2+, Mn3+and Mn4+respectively40,41.The relative content of Mn3+/Mn4+was also recorded in Table 3.It could be seen that with the addition of potassium, the Mn4+content decreased, while the Mn3+increased. These results were in accordance with the XRD experiment, the potassium could act as electron provider, which led to electronic change of manganese oxides and parts of Mn4+was transformed to the Mn3+.

Table 3 The XPS results of the catalysts.

The XPS spectra for O 1s were shown in Fig. 7c. There were two kinds of oxygen species fitted through deconvolution of the O 1s spectra. The binding energy at 529.5-530.1 eV was assigned to the lattice oxygen (labeled as Olatt), and at 531-531.9 eV was attributed to the surface oxygen (denoted as Osur)41,42.Table 3 exhibited the Osur/Ototalmolar ratio of each sample and it could be observed that with the addition of K, the molar ratio of Osurwas increased. Meanwhile the Mn/CZ-K-0.2 catalyst exhibited the largest (43.39%)molar ratio of the Osur/Ototal; this result implied that the Mn/CZ-K-0.2 sample possessed a massive amount of surface oxygen species. Generally, the Osurwas the primary active oxygen, which gave rise to the high catalytic performance of catalyst in the redox reaction43.

Based on the results of H2-TPR, it was deduced that potassium as an electronic promoter, could weak the Mn-O bond strength and enhanced the reduction ability. Furthermore, on the basis of XPS results, the K could affect the chemical environment of Mn ion, which promoted the content of surface oxygen species.

3.7 Catalytic activities of the catalysts

The catalytic activities for toluene combustion over all the catalysts and pure CZ were given in Fig. 8 and Table 4. From the graph, it can be seen the catalytic performance of pure CZ was evidently poorer than that of the manganese-based catalysts.Besides, the catalytic activities were improved apparently with addition of potassium, and the Mn/CZ-K-0.2 exhibited the best performance. While with further increasing the potassium content, the corresponding catalytic activity of Mn/CZ-K-0.5 was declined, indicating that the high content of potassium could restrain the activity of catalyst. As exhibited in the Table 4, the Mn/CZ-K-0.2 performed the best activity, of which the T50and T90(the temperatures at which the toluene conversion achieved 50% and 90%, respectively)was at 228 °C and 242 °C,respectively. Besides, the sequence of the catalytic performance followed the order of: Mn/CZ-K-0.2 ＞ Mn/CZ-K-0.05 ＞Mn/CZ ＞ Mn/CZ-K-0.5, which was similar with the sequence of oxygen vacancies concentration calculated from Raman spectrogram and the amounts of surface adsorbed oxygen (α)in O2-TPD experiments listed in Table 2.

Fig. 8 Catalytic activities for toluene oxidation.

Table 4 The temperatures of T50 and T90 over the catalysts.

Fig. 9 Stability test of Mn/CZ-K-0.2 catalyst for toluene oxidation at 242 °C.

In order to test the stability of the Mn/CZ-K-0.2 catalyst, the catalyst was continuously evaluated for 30 h at 90% conversion temperature. The Fig. 9 exhibited the result of stability test of catalyst. Judge from the curve, the toluene conversion of the Mn/CZ-K-0.2 catalyst at 242 °C was not decreased significantly in the process of testing, which indicated that the Mn/CZ-K-0.2 catalyst exhibited a good stability for toluene oxidation.

3.8 Toluene-TPSR in N2 with in situ DRIFTs analysis

In order to investigate the effect of potassium on catalytic properties for toluene oxidation, toluene-TPSR experiments in N2for Mn/CZ and Mn/CZ-K-0.2 catalysts were carried out with in situ DRIFTs. Fig. 10 exhibited the DRIFTS spectra of temperature programmed surface reaction of toluene at nitrogen atmosphere on Mn/CZ (A)and Mn/CZ-K-0.2 (B)respectively.As shown in the Fig. 10A, the bands detected at 1621 cm-1and 1454 cm-1could be attributed to aromatic rings characteristic vibration of toluene44. The multiple bands appeared at 1191 cm-1and 1126 cm-1could belong to the vibration of C-O,which implied that toluene was adsorbed on Mn/CZ mainly in the form of benzyl alkoxide species50. And the band at 1537 cm-1and 1394 cm-1were assigned to νas(COO)and νs(COO)of bridging benzoate complex45,46. In addition, the weak bands at around 2379 cm-1and 2316 cm-1were attributed to asymmetric stretching vibration of CO250. With rising of the temperature, the vibration bands of C-O became weaker and disappeared at 250 °C, meanwhile a new band at 1272 cm-1distributed to the vibration of adsorbed benzaldehyde appeared at 100 °C51,indicating the further oxidation of benzyl alkoxides. With continued heating up, the vibration peaks of toluene derivatives disappear and only the peaks of 1510 cm-1and 1394 cm-1which were assigned to maleate and carbonate species52were detected on the surface of the catalyst. These results demonstrated that in the absence of oxygen, toluene could be oxidized to alkoxide species and benzaldehyde by the lattice oxygen on Mn/CZ and further oxidized to benzoate. Subsequently, the benzoate was cracked at high temperature to convert to carbonate and maleate species adsorbed on the catalyst surface.

Fig. 10 DRIFT spectra of temperature programmed surface reaction of toluene in a flow of N2 for Mn/CZ (A)and Mn/CZ-K-0.2 (B).

In the Fig. 10B, the infrared spectrum of the adsorbed toluene on Mn/CZ-K-0.2 was quite different. Compared with Fig. 10A,it could be clearly seen that the band belonged to absorbed benzaldehyde (1251 cm-1)emerged at 50 °C, which suggested that the alkoxides was oxidized to form aromatic aldehyde at low temperature with the addition of K. The signal of adsorbed alkoxide species (1186 and 1120 cm-1)disappeared at 200 °C in advance, which was 50 °C lower than that of Mn/CZ,demonstrated that the presence of potassium was more conducive to the conversion of benzyl alkoxides to benzaldehyde. Moreover, it could be observed that there was only carbonate (1375 and 1303 cm-1)and formate species (1562 cm-1)52remained on the surface of the catalyst Mn/CZ-K-0.2 at 400 °C, and no maleate was detected, indicating that the toluene oxidation was performed more completely on Mn/CZ-K-0.2 compared to Mn/CZ.

Therefore, from the results of in situ DRIFTs analysis, it could be seen that with the addition of K, the lattice oxygen of Mn/CZK-0.2 was more availed and active than Mn/CZ, which was favor to the toluene oxidation reaction.

In general, for VOCs catalytic oxidation, the catalytic activity was closely linked to the oxygen species and the reducibility of the catalysts46. In this work, the monolith Mn/CZ catalyst was firstly prepared for toluene catalytic oxidation, and the activity was apparently increased by the addition of K with the K/Mn molar ratio of 0.2. In the XRD results, it could deduce that the MnOxmight be evenly distributed on the surface of the carrier or Mn4+be partly transformed into Mn3+by the addition of potassium. The Raman spectrum demonstrated that with increasing the K content, β-MnO2was changed into α-MnO2gradually, and two crystalline phases might coexist on the Mn/CZ-K-0.2 catalyst, thus increasing the defect sites on surface.From H2-TPR experiment, it was concluded that Mn/CZ-K-0.2 had the lowest reduction temperature and exhibited the good reducibility. Combined with the experimental results of O2-TPD,Mn/CZ-K-0.2 had the most surface oxygen species and better mobility of lattice oxygen which were both benefitted to the toluene oxidation, and thus facilitated the redox process and leaded to a better catalytic performance47. In addition, the XPS results suggested that the content of surface adsorbed oxygen species of Mn/CZ-K-0.2 catalyst was the highest among all samples; it might indicate that the surface oxygen species might have an effective role in the toluene combustion. In addition,toluene-TPSR in N2with in situ DRIFTs analysis was further demonstrated that more available lattice oxygen or oxygen defect site were existed in Mn/CZ-K-0.2 catalyst. Therefore,Mn/CZ-K-0.2 catalyst had the best redox property and oxygen mobility and it had the excellent activity to toluene oxidation.

4 Conclusions

In this work, the alkali metal potassium as a promoter was introduced by the co-impregnation method. The catalytic activity was promoted and the monolith Mn/CZ-K-0.2 catalyst performed the best activity with the T90of 242 °C at a GHSV of 12000 h-1. The addition of potassium promoted the reducibility of the catalyst and facilitated the formation of more surface adsorbed oxygen on the catalyst surface. Additionally, suitable amount of potassium could generate more oxygen vacancies, and enhanced the mobility of lattice oxygen. Therefore, the performance of Mn/CZ-K-0.2 catalyst for toluene oxidation was significantly improved due to the addition of potassium.