Catalytic Oxidation of NO over MnOx–CeO2 and MnOx–TiO2 Catalysts

A series of MnOx–CeO2 and MnOx–TiO2 catalysts were prepared by a homogeneous precipitation method and their catalytic activities for the NO oxidation in the absence or presence of SO2 were evaluated. Results show that the optimal molar ratio of Mn/Ce and Mn/Ti are 0.7 and 0.5, respectively. The MnOx–CeO2 catalyst exhibits higher catalytic activity and better resistance to SO2 poisoning than the MnOx–TiO2 catalyst. On the basis of Brunauer–Emmett–Teller (BET), X-ray diffraction (XRD), and scanning transmission electron microscope with mapping (STEM-mapping) analyses, it is seen that the MnOx–CeO2 catalyst possesses higher BET surface area and better dispersion of MnOx over the catalyst than MnOx–TiO2 catalyst. X-ray photoelectron spectroscopy (XPS) measurements reveal that MnOx–CeO2 catalyst provides the abundance of Mn3+ and more surface adsorbed oxygen, and SO2 might be preferentially adsorbed to the surface of CeO2 to form sulfate species, which provides a protection of MnOx active sites from being poisoned. In contrast, MnOx active sites over the MnOx–TiO2 catalyst are easily and quickly sulfated, leading to rapid deactivation of the catalyst for NO oxidation. Furthermore, temperature programmed desorption with NO and O2 (NO + O2-TPD) and in situ diffuse reflectance infrared transform spectroscopy (in situ DRIFTS) characterizations results show that the MnOx–CeO2 catalyst displays much stronger ability to adsorb NOx than the MnOx–TiO2 catalyst, especially after SO2 poisoning.


Introduction
Nitrogen oxides (NO x ) emitted from stationary and mobile sources are some of the main air pollutants, which cause a variety of serious environmental problems, such as photochemical smog, acid rain, and greenhouse effect [1]. Moreover, NO x are the primary precursors of haze occurring in China. Therefore, NO x removal has become the focus of recent environmental protection. The most effective and mature technology is the selective catalytic reduction using ammonia as a reducing agent (NH 3 -SCR). However, there still exist some problems, such as high reaction temperature, sophisticated system design, and high operation cost. Additionally, it is possible to cause secondary pollution due to the leakage of ammonia [2].
In order to solve the problems of NH 3 -SCR, much attention has been paid to the simultaneous removal of SO 2 and NO x by chemical absorption. For the absorption operation, the oxidation of NO with low water-solubility to NO 2 is a crucial process because NO accounts for about 95% of NO x . In general, the oxidation of NO to NO 2 can be realized through gas phase oxidation and liquid phase oxidation. The presence of SO 2 is disadvantageous to NO oxidation in the liquid phase because of the high solubility and oxidizability of SO 2 , whereas the oxidation rate of SO 2 is much lower than that of NO in the gas phase [3]. The gas phase oxidation is divided into homogeneous gas phase oxidation and heterogeneous gas-solid catalytic oxidation. Nowadays, catalytic oxidation of NO is potentially

Catalytic Activity Tests
The NO oxidation efficiencies over the MnO x -CeO 2 -x and MnO x -TiO 2 -y catalysts are shown in Figure 1a,b, respectively. It can be seen that TiO 2 shows negligible catalytic activity during the reaction temperature range, while CeO 2 has certain catalytic activity for NO oxidation. Nonetheless, the catalytic activity of CeO 2 is lower and the activity temperature is higher, compared with those of MnO x -CeO 2 -x catalysts. Therefore, MnO x was the main active component for the catalytic oxidation of NO. In the presence of SO 2 , the optimal molar ratio of Mn/Ce and Mn/Ti was 0.7 and 0.5, respectively. Meanwhile, the maximum NO oxidation efficiency of 72% over MnO x -CeO 2 -0.7 catalyst is obtained at 325 • C, while that of 62% over MnO x -TiO 2 -0.5 catalyst is obtained at 375 • C Therefore, the MnO x -CeO 2 -0.7 catalyst has better catalytic activity than the MnO x -TiO 2 -0.5 catalyst. Actually, the catalytic activities of MnO x -CeO 2 -0. 7 and MnO x -TiO 2 -0.5 catalysts were also investigated in the absence of SO 2 , and the results show that the maximum NO oxidation efficiency of MnO x -CeO 2 -0. 7 and MnO x -TiO 2 -0.5 catalysts are 91% and 86% at 300 • C, as shown in Figure 2. Clearly, the presence of SO 2 results in a decrease of NO oxidation efficiency and an increase of the active temperature, especially for the MnO x -TiO 2 -0.5 catalyst. The MnO x -CeO 2 -0.7 catalyst displays better resistance to SO 2 poisoning than the MnO x -TiO 2 -0.5 catalyst. The stability tests for NO oxidation over MnOx-CeO2-0.7 and MnOx-TiO2-0.5 catalysts were carried out under different temperatures. As shown in Figure 3a, the NO oxidation efficiency of the MnOx-TiO2-0.5 catalyst decreases much more rapidly than that of the MnOx-CeO2-0.7 catalyst. The catalytic activity of the MnOx-CeO2-0.7 catalyst gradually decreases at 300 °C, and maintains almost unchanged within 5 h at 350 °C while it decreases after 5 h. The stability tests without SO2 over two catalysts were also carried out at 300 °C, and no activity decrease is observed in 20 h (the results are not shown here), which convinces us that the deactivation in Figure 3 is caused by the presence of SO2. The on-off effect of SO2 for NO oxidation over MnOx-CeO2-0.7 catalyst was investigated. As shown in Figure 3b, when 100 ppm SO2 are added to the reactants, the NO oxidation efficiency decreases from the initial 80% to 27% after 10 h. After excluding SO2 from the flue gas, the NO oxidation efficiency only recovers to 32%, which indicates that the poisoning effect of SO2 is irreversible.   The stability tests for NO oxidation over MnO x -CeO 2 -0.7 and MnO x -TiO 2 -0.5 catalysts were carried out under different temperatures. As shown in Figure 3a, the NO oxidation efficiency of the MnO x -TiO 2 -0.5 catalyst decreases much more rapidly than that of the MnO x -CeO 2 -0.7 catalyst. The catalytic activity of the MnO x -CeO 2 -0.7 catalyst gradually decreases at 300 • C, and maintains almost unchanged within 5 h at 350 • C while it decreases after 5 h. The stability tests without SO 2 over two catalysts were also carried out at 300 • C, and no activity decrease is observed in 20 h (the results are not shown here), which convinces us that the deactivation in Figure 3 is caused by the presence of SO 2 . The on-off effect of SO 2 for NO oxidation over MnO x -CeO 2 -0.7 catalyst was investigated. As shown in Figure 3b, when 100 ppm SO 2 are added to the reactants, the NO oxidation efficiency decreases from the initial 80% to 27% after 10 h. After excluding SO 2 from the flue gas, the NO oxidation efficiency only recovers to 32%, which indicates that the poisoning effect of SO 2 is irreversible. The stability tests for NO oxidation over MnOx-CeO2-0.7 and MnOx-TiO2-0.5 catalysts were carried out under different temperatures. As shown in Figure 3a, the NO oxidation efficiency of the MnOx-TiO2-0.5 catalyst decreases much more rapidly than that of the MnOx-CeO2-0.7 catalyst. The catalytic activity of the MnOx-CeO2-0.7 catalyst gradually decreases at 300 °C, and maintains almost unchanged within 5 h at 350 °C while it decreases after 5 h. The stability tests without SO2 over two catalysts were also carried out at 300 °C, and no activity decrease is observed in 20 h (the results are not shown here), which convinces us that the deactivation in Figure 3 is caused by the presence of SO2. The on-off effect of SO2 for NO oxidation over MnOx-CeO2-0.7 catalyst was investigated. As shown in Figure 3b, when 100 ppm SO2 are added to the reactants, the NO oxidation efficiency decreases from the initial 80% to 27% after 10 h. After excluding SO2 from the flue gas, the NO oxidation efficiency only recovers to 32%, which indicates that the poisoning effect of SO2 is irreversible.    Figure 4 presents the XRD patterns of fresh and SO 2 poisoned catalysts. For MnO x -CeO 2 -0.7 catalyst and SO 2 poisoned MnO x -CeO 2 -0.7 catalyst (donated as MnO x -CeO 2 -0.7-S), crystalline phases of CeO 2 can be clearly observed, and very weak signals of Mn 2 O 3 are also detected, which indicates that Mn 2 O 3 exists in a poor crystal structure. For MnO x -TiO 2 -0.5 catalyst and SO 2 poisoned MnO x -TiO 2 -0.5 catalyst (donated as MnO x -TiO 2 -0.5-S), the stronger diffraction peaks of Mn 2 O 3 are observed besides crystalline phases of rutile and anatase TiO 2 , which suggests that Mn 2 O 3 exists in crystal structure. It is well know that the low crystallinity of MnO x is favorable for catalytic reaction [20]. Therefore, the higher activity of MnO x -CeO 2 -0.7 catalyst may be partly due to the well dispersion of MnO x . For all of the samples, the diffraction peaks almost do not change due to SO 2 poisoning.  Figure 4 presents the XRD patterns of fresh and SO2 poisoned catalysts. For MnOx-CeO2-0.7 catalyst and SO2 poisoned MnOx-CeO2-0.7 catalyst (donated as MnOx-CeO2-0.7-S), crystalline phases of CeO2 can be clearly observed, and very weak signals of Mn2O3 are also detected, which indicates that Mn2O3 exists in a poor crystal structure. For MnOx-TiO2-0.5 catalyst and SO2 poisoned MnOx-TiO2-0.5 catalyst (donated as MnOx-TiO2-0.5-S), the stronger diffraction peaks of Mn2O3 are observed besides crystalline phases of rutile and anatase TiO2, which suggests that Mn2O3 exists in crystal structure. It is well know that the low crystallinity of MnOx is favorable for catalytic reaction [20]. Therefore, the higher activity of MnOx-CeO2-0.7 catalyst may be partly due to the well dispersion of MnOx. For all of the samples, the diffraction peaks almost do not change due to SO2 poisoning.  The BET surface areas of the catalysts are summarized in Table 1. It can be seen that the specific surface areas of fresh MnOx-CeO2-0.7 and MnOx-TiO2-0.5 catalysts are 96.30 and 60.21 m 2 ·g −1 . Compared to catalytic performance, it is consistent with that of BET surface. Furthermore, it is worth noting that the BET specific surface areas of SO2 poisoned catalysts decrease to 67.92 (MnOx-CeO2-0.7) m 2 ·g −1 and 39.71 (MnOx-TiO2-0.5) m 2 ·g −1 , which may be caused by the formation of sulfate species.

Catalysts
Surface Area (m 2 /g)  The BET surface areas of the catalysts are summarized in Table 1. It can be seen that the specific surface areas of fresh MnO x -CeO 2 -0.7 and MnO x -TiO 2 -0.5 catalysts are 96.30 and 60.21 m 2 ·g −1 . Compared to catalytic performance, it is consistent with that of BET surface. Furthermore, it is worth noting that the BET specific surface areas of SO 2 poisoned catalysts decrease to 67.92 (MnO x -CeO 2 -0.7) m 2 ·g −1 and 39.71 (MnO x -TiO 2 -0.5) m 2 ·g −1 , which may be caused by the formation of sulfate species.

XPS Analysis
XPS analysis was performed to identify the surface component and chemical states of fresh and SO2 poisoned catalysts. Surface atomic concentration and ratio are summarized in Table 2, and XPS spectra of Mn 2p, O 1s, Ce 3d, and Ti 2p of all catalysts are displayed in Figure 6. Through the deconvolution of the spectra, two main peaks due to Mn 2p1/2 and Mn 2p3/2 are observed. The Mn 2p3/2 profiles are fitted with the Mn 2+ , Mn 3+ , and Mn 4+ , characterized by the binding energy at about 641.1 eV, 642.5 eV, and 645.1 eV [33], respectively. Previous studies [20,22,34] have shown that Mn2O3 has a higher catalytic activity than MnO2 for NO oxidation. Cimino et al. [35] attributed the higher activity of Mn 3+ than Mn 4+ for CO catalytic oxidation to the weaker Mn 3+ -O bond. Similarly, it can be deduced that the weaker Mn 3+ -O bonds will also favor the catalytic oxidation of NO since the Mn 3+ -O bond is easily broken, thus, promoting the generation and release of the NO2 oxidation product. As shown in Table 2, all catalysts contain high concentration of Mn 3+ . Corresponding to the high catalytic activity of catalyst, it, combining with XRD analysis results, can be also speculated that Mn 3+ has higher catalytic activity than Mn 2+ and Mn 4+ for NO oxidation. The Ce 3D XPS spectra can be separated into eight peaks: u0 (900.6 eV), u1 (902.4eV), u2 (907.9 eV), u3 (916.6 eV), v0 (881.9 eV), v1 (884.4 eV), v2 (889.1 eV), and v3 (898.1 eV) [36]. The bands labeled as u1 and v1 are attributed to Ce 3+ species, and the other six peaks are assigned to Ce 4+ species. The ratio of Ce 3+ /(Ce 3+ + Ce 4+ ) can be estimated by the formula [37]:

XPS Analysis
XPS analysis was performed to identify the surface component and chemical states of fresh and SO 2 poisoned catalysts. Surface atomic concentration and ratio are summarized in Table 2, and XPS spectra of Mn 2p, O 1s, Ce 3d, and Ti 2p of all catalysts are displayed in Figure 6. Through the deconvolution of the spectra, two main peaks due to Mn 2p 1/2 and Mn 2p 3/2 are observed. The Mn 2p 3/2 profiles are fitted with the Mn 2+ , Mn 3+ , and Mn 4+ , characterized by the binding energy at about 641.1 eV, 642.5 eV, and 645.1 eV [33], respectively. Previous studies [20,22,34] have shown that Mn 2 O 3 has a higher catalytic activity than MnO 2 for NO oxidation. Cimino et al. [35] attributed the higher activity of Mn 3+ than Mn 4+ for CO catalytic oxidation to the weaker Mn 3+ -O bond. Similarly, it can be deduced that the weaker Mn 3+ -O bonds will also favor the catalytic oxidation of NO since the Mn 3+ -O bond is easily broken, thus, promoting the generation and release of the NO 2 oxidation product. As shown in Table 2, all catalysts contain high concentration of Mn 3+ . Corresponding to the high catalytic activity of catalyst, it, combining with XRD analysis results, can be also speculated that Mn 3+ has higher catalytic activity than Mn 2+ and Mn 4+ for NO oxidation. The Ce 3D XPS spectra can be separated into eight peaks: u 0 (900.6 eV), u 1 (902.4eV), u 2 (907.9 eV), u 3 (916.6 eV), v 0 (881.9 eV), v 1 (884.4 eV), v 2 (889.1 eV), and v 3 (898.1 eV) [36]. The bands labeled as u 1 and v 1 are attributed to Ce 3+ species, and the other six peaks are assigned to Ce 4+ species. The ratio of Ce 3+ /(Ce 3+ + Ce 4+ ) can be estimated by the formula [37]: It is well known that Ce 3+ species can make charge imbalance and create oxygen vacancies via the shift from Ce 3+ to Ce 4+ , which leads to the increase of surface adsorbed oxygen (Ce [38]. For the catalytic oxidation of NO, surface adsorbed oxygen plays a significant role because of its mobility and redox performance [39]. As listed in Table 2, the Ce 3+ concentration can reach about 41.6%. Figure 6c displays the O 1s XPS spectra of all samples, two distinct bands are obtained. The one peak O β in the range of 528-530 eV belongs to lattice oxygen and the other peak O α with binding energy of 530-532 eV corresponds to weakly surface adsorbed oxygen [18]. From Table 2, it can be seen that the O α concentration over MnO x -CeO 2 -0.7 catalyst is higher than that over MnO x -TiO 2 -0.5 catalyst, which is attributed to the presence of Ce 3+ species. On the other hand, the Mn concentration of the MnO x -CeO 2 -0.7-S catalyst is almost the same to that of the fresh MnO x -CeO 2 -0.7 catalyst, while the Mn concentration of MnO x -TiO 2 -0.5 catalyst and the Ce concentration of MnO x -CeO 2 -0.7 catalyst decrease from 13.6% to 9.4% and from 26.4% to 19.7%, respectively, because of SO 2 poisoning, which is attributed that the MnO x over MnO x -TiO 2 -0.5-S catalyst and CeO 2 over MnO x -CeO 2 -0.7-S catalyst are partly covered with sulfate species [36]. Meanwhile, the ratio of Ce 3+ /(Ce 3+ + Ce 4+ ) of the MnO x -CeO 2 -0.7 catalyst also decreases from 41.6% to 25.2%, which indicates that cerium(IV) sulfate may be formed on the catalyst surface [40]. Therefore, we can deduce that SO 2 might be preferentially adsorbed to the surface of CeO 2 to form sulfate species, lessening the sulfation of MnO x active sites. It was also reported by Jin and co-workers [32] that the presence of CeO 2 might partially prevent MnO x active sites from being sulfated. Waqif [41] investigated the adsorption of SO 2 on CeO 2 -Al 2 O 3 , and concluded that ceria was a basic material for SO 2 adsorption. Figure 4d shows the Ti 2p XPS spectra, four peaks are formed, referred to as Ti 3+ at 458.3 eV, 464.1 eV, and Ti 4+ at 459.8 eV, 466.1 eV, respectively [23]. Though the Ti 3+ concentration is pretty high, it still cannot improve the resistance to SO 2 poisoning.  It is well known that Ce 3+ species can make charge imbalance and create oxygen vacancies via the shift from Ce 3+ to Ce 4+ , which leads to the increase of surface adsorbed oxygen (Ce 3+ → Ce 4+ + e − , O2 + e − → O2 − ) [38]. For the catalytic oxidation of NO, surface adsorbed oxygen plays a significant role because of its mobility and redox performance [39]. As listed in Table 2, the Ce 3+ concentration can reach about 41.6%. Figure 6c displays the O 1s XPS spectra of all samples, two distinct bands are obtained. The one peak Oβ in the range of 528-530 eV belongs to lattice oxygen and the other peak Oα with binding energy of 530-532 eV corresponds to weakly surface adsorbed oxygen [18]. From Table  2, it can be seen that the Oα concentration over MnOx-CeO2-0.7 catalyst is higher than that over MnOx-TiO2-0.5 catalyst, which is attributed to the presence of Ce 3+ species.
On the other hand, the Mn concentration of the MnOx-CeO2-0.7-S catalyst is almost the same to that of the fresh MnOx-CeO2-0.7 catalyst, while the Mn concentration of MnOx-TiO2-0.5 catalyst and the Ce concentration of MnOx-CeO2-0.7 catalyst decrease from 13.6% to 9.4% and from 26.4% to 19.7%, respectively, because of SO2 poisoning, which is attributed that the MnOx over MnOx-TiO2-0.5-S catalyst and CeO2 over MnOx-CeO2-0.7-S catalyst are partly covered with sulfate species [36]. Meanwhile, the ratio of Ce 3+ /(Ce 3+ + Ce 4+ ) of the MnOx-CeO2-0.7 catalyst also decreases from 41.6% to 25.2%, which indicates that cerium(IV) sulfate may be formed on the catalyst surface [40]. Therefore, we can deduce that SO2 might be preferentially adsorbed to the surface of CeO2 to form sulfate species, lessening the sulfation of MnOx active sites. It was also reported by Jin and co-workers [32] that the presence of CeO2 might partially prevent MnOx active sites from being sulfated. Waqif [41] investigated the adsorption of SO2 on CeO2-Al2O3, and concluded that ceria was a basic material for SO2 adsorption. Figure 4d shows the Ti 2p XPS spectra, four peaks are formed, referred to as Ti 3+ at 458.3 eV, 464.1 eV, and Ti 4+ at 459.8 eV, 466.1 eV, respectively [23]. Though the Ti 3+ concentration is pretty high, it still cannot improve the resistance to SO2 poisoning.

NO + O2-TPD and in Situ DRIFTS Analyses
The adsorption behavior of the catalyst is considered a crucial step in a catalytic oxidation reaction. Therefore, NO + O2-TPD experiments were conducted to explore the NOx adsorption ability over MnOx-CeO2-0.7 and MnOx-TiO2-0.5 catalysts. As shown in Figure 7a,b, the NO and NO2 curves over MnOx-CeO2-0.7 and MnOx-TiO2-0.5 catalysts are observed. For the MnOx-CeO2-0.7 catalyst, the desorption peak at about 240 °C is assigned to nitrosyl species [42], the desorption peak in the temperature range of 350-450 °C can be ascribed to the decomposition of strong adsorption species such as nitrate on catalyst surface [43]. For the MnOx-TiO2-0.5 catalyst, three major desorption peaks at 80, 180, and 320 °C are observed, which may be attributed to desorption of molecularly-adsorbed NO and NO2, nitrosyl species and desorption of nitrate species, respectively [42,44,45]. It is obvious that the total amount of NOx desorbed from MnOx-CeO2-0.7 catalyst is remarkably larger than that of the MnOx-TiO2-0.5 catalyst, indicating stronger adsorption and oxidation abilities on the surface of the MnOx-CeO2-0.7 catalyst.
In order to understand the NOx adsorption behaviors and SO2 poisoning process, in situ DRIFTS measurements over MnOx-CeO2-0.7 and MnOx-TiO2-0.5 catalysts were carried out at 350 °C. Figure  7c,d shows the NO-O2 co-adsorption accompanied by SO2 adsorption. After introducing NO + O2, for the MnOx-CeO2-0.7 catalyst, the bands at 1593, 1566, 1540, 1242, and 1212 cm −1 are detected. All of the bands' intensities gradually increase with the adsorption time until reaching their highest intensities and remain stable after about 40 min. The bands at 1566, 1540, and 1212-1242 cm −1 are assigned to bidentate nitrate, monodentate nitrate, and bridge nitrate, respectively [46]. A very weak band at 1593 cm −1 is due to the adsorption of NO2 [47]. For the MnOx-TiO2-0.5 catalyst, the bands attributed to monodentrate nitrate (1235 cm −1 ), bidentrate nitrate (1548 cm −1 ), and bridge nitrate (1608 cm −1 ) are observed [44]. The change trend of these bands' intensities is similar to those over the MnOx-CeO2-0.7 catalyst. However, it is obvious that all of the adsorption bands' intensities of the MnOx-TiO2-0.5 catalyst are significantly lower than those of the MnOx-CeO2-0.7 catalyst, which is probably one of the reasons that the MnOx-CeO2-0.7 catalyst has better activity than the MnOx-TiO2-0.5 catalyst in the absence of SO2.

NO + O 2 -TPD and In Situ DRIFTS Analyses
The adsorption behavior of the catalyst is considered a crucial step in a catalytic oxidation reaction. Therefore, NO + O 2 -TPD experiments were conducted to explore the NO x adsorption ability over MnO x -CeO 2 -0.7 and MnO x -TiO 2 -0.5 catalysts. As shown in Figure 7a,b, the NO and NO 2 curves over MnO x -CeO 2 -0.7 and MnO x -TiO 2 -0.5 catalysts are observed. For the MnO x -CeO 2 -0.7 catalyst, the desorption peak at about 240 • C is assigned to nitrosyl species [42], the desorption peak in the temperature range of 350-450 • C can be ascribed to the decomposition of strong adsorption species such as nitrate on catalyst surface [43]. For the MnO x -TiO 2 -0.5 catalyst, three major desorption peaks at 80, 180, and 320 • C are observed, which may be attributed to desorption of molecularly-adsorbed NO and NO 2 , nitrosyl species and desorption of nitrate species, respectively [42,44,45]. It is obvious that the total amount of NO x desorbed from MnO x -CeO 2 -0.7 catalyst is remarkably larger than that of the MnO x -TiO 2 -0.5 catalyst, indicating stronger adsorption and oxidation abilities on the surface of the MnO x -CeO 2 -0.7 catalyst.
In order to understand the NO x adsorption behaviors and SO 2 poisoning process, in situ DRIFTS measurements over MnO x -CeO 2 -0.7 and MnO x -TiO 2 -0.5 catalysts were carried out at 350 • C. Figure 7c,d shows the NO-O 2 co-adsorption accompanied by SO 2 adsorption. After introducing NO + O 2 , for the MnO x -CeO 2 -0.7 catalyst, the bands at 1593, 1566, 1540, 1242, and 1212 cm −1 are detected. All of the bands' intensities gradually increase with the adsorption time until reaching their highest intensities and remain stable after about 40 min. The bands at 1566, 1540, and 1212-1242 cm −1 are assigned to bidentate nitrate, monodentate nitrate, and bridge nitrate, respectively [46]. A very weak band at 1593 cm −1 is due to the adsorption of NO 2 [47]. For the MnO x -TiO 2 -0.5 catalyst, the bands attributed to monodentrate nitrate (1235 cm −1 ), bidentrate nitrate (1548 cm −1 ), and bridge nitrate (1608 cm −1 ) are observed [44]. The change trend of these bands' intensities is similar to those over the MnO x -CeO 2 -0.7 catalyst. However, it is obvious that all of the adsorption bands' intensities of the MnO x -TiO 2 -0.5 catalyst are significantly lower than those of the MnO x -CeO 2 -0.7 catalyst, which is probably one of the reasons that the MnO x -CeO 2 -0.7 catalyst has better activity than the MnO x -TiO 2 -0.5 catalyst in the absence of SO 2 . In the following, 100 ppm SO2 was added to the reaction system. It can be seen from Figure 7c that a new band at 1346 cm −1 appears over the MnOx-CeO2-0.7 catalyst and the intensity grows with time. Similarly, the new peaks at 1346 cm −1 and 1152 cm −1 are also observed over the MnOx-TiO2-0.5 catalyst and their intensities rise rapidly with the reaction time. The band at 1346 cm −1 is due to the (S=O) vibration of surface sulfate species, and the band at 1152 cm −1 can be ascribed to sulfate species [48]. Moreover, it can be noted that all the adsorption bands' intensities almost remain unchanged within 10 min and the bands' intensities of monodentrate and bidentrate nitrate slightly decrease for the MnOx-CeO2-0.7 catalyst. However, the bands' intensities at 1608 cm −1 and 1548 cm −1 drop rapidly with time and the peak of monodentrate nitrate almost vanishes after 60 min for the MnOx-TiO2-0.5 catalyst. The results confirm that SO2 has little influence on NOx adsorption over the MnOx-CeO2-0.7 catalyst, while there is strongly competitive adsorption between SO2 and NOx over the MnOx-TiO2-0.5 catalyst in a certain reaction time. Tang et al. [49] reported the mechanism of catalytic oxidation of NO over Mn-based catalysts that NO firstly adsorbed on Mn sites to form nitrosyls, and then were oxidized to nitrates, which decomposed to the final product, NO2.
According to the DRIFTS results and the mechanism, we further deduce that SO2 preferentially combines with CeO2 to form sulfate species, and MnOx active sites are exposed to the surface to adsorb NOx over the MnOx-CeO2-0.7 catalyst. Whereas MnOx active sites are sulfated so seriously that the MnOx-TiO2-0.5 catalyst has no ability to adsorb NOx, leading to low catalytic activity. The results are consistent with XPS analysis. Moreover, the formation of sulfate species is irreversible and sulfate species occupied the sites for NO oxidation permanently. Through the above analysis, it is sufficient to prove that the catalytic activity and resistance to SO2 poisoning of MnOx-CeO2 catalysts are better than MnOx-TiO2 catalysts. In the following, 100 ppm SO 2 was added to the reaction system. It can be seen from Figure 7c that a new band at 1346 cm −1 appears over the MnO x -CeO 2 -0.7 catalyst and the intensity grows with time. Similarly, the new peaks at 1346 cm −1 and 1152 cm −1 are also observed over the MnO x -TiO 2 -0.5 catalyst and their intensities rise rapidly with the reaction time. The band at 1346 cm −1 is due to the υ (S=O) vibration of surface sulfate species, and the band at 1152 cm −1 can be ascribed to sulfate species [48]. Moreover, it can be noted that all the adsorption bands' intensities almost remain unchanged within 10 min and the bands' intensities of monodentrate and bidentrate nitrate slightly decrease for the MnO x -CeO 2 -0.7 catalyst. However, the bands' intensities at 1608 cm −1 and 1548 cm −1 drop rapidly with time and the peak of monodentrate nitrate almost vanishes after 60 min for the MnO x -TiO 2 -0.5 catalyst. The results confirm that SO 2 has little influence on NO x adsorption over the MnO x -CeO 2 -0.7 catalyst, while there is strongly competitive adsorption between SO 2 and NO x over the MnO x -TiO 2 -0.5 catalyst in a certain reaction time. Tang et al. [49] reported the mechanism of catalytic oxidation of NO over Mn-based catalysts that NO firstly adsorbed on Mn sites to form nitrosyls, and then were oxidized to nitrates, which decomposed to the final product, NO 2 .
According to the DRIFTS results and the mechanism, we further deduce that SO 2 preferentially combines with CeO 2 to form sulfate species, and MnO x active sites are exposed to the surface to adsorb NO x over the MnO x -CeO 2 -0.7 catalyst. Whereas MnO x active sites are sulfated so seriously that the MnO x -TiO 2 -0.5 catalyst has no ability to adsorb NO x , leading to low catalytic activity. The results are consistent with XPS analysis. Moreover, the formation of sulfate species is irreversible and sulfate species occupied the sites for NO oxidation permanently. Through the above analysis, it is sufficient to prove that the catalytic activity and resistance to SO 2 poisoning of MnO x -CeO 2 catalysts are better than MnO x -TiO 2 catalysts.

Catalyst Preparation
A series of MnO x -CeO 2 -x and MnO x -TiO 2 -y catalysts, where x and y are the molar ratio of Mn/Ce and Mn/Ti, respectively, were prepared by homogeneous precipitation method. Take MnO x -CeO 2 -0.7 for example, 13.02 g Ce(NO 3 ) 3 ·6H 2 O and 7.15 g Mn(NO 3 ) 2 (50% solution) were firstly added to 100 mL deionized water and stirred for 2 h. Excessive urea aqueous solution was added into the mixed solution under stirring. Then, the mixed solution was stirred for 12 h at 90 • C. In order to make Mn precipitate completely, an appropriate amount of ammonia solution were added into the mixed solution until the pH value was 9.5. The precipitate was collected by filtration and washed with deionized water, followed by drying at 110 • C overnight and subsequently calcination at 500 • C for 4 h in the air atmosphere. MnO x -TiO 2 -y catalysts were prepared by similar process with MnO x -CeO 2 -x catalysts. The difference is that tetrabutyl titanate was firstly dissolved in ethanol. Finally, the catalysts were crushed and sieved to 40-60 mesh for activity test.

Catalytic Activity Measurement
The catalytic activity was evaluated in a quartz U-tube fixed-bed flow reactor (i.d. 13 mm) from 450-250 • C. The test data was recorded after the reaction for 40 min at each temperature. The reaction gas consisted of 400 ppm NO, 10% O 2 , 1% H 2 O, 100 ppm SO 2 (when used), and balanced with N 2 . The total flow rate was fixed at 2 L/min, which is corresponded to a gas hourly space velocity (GHSV) of 40,000 h −1 . The concentrations of NO, NO 2 , O 2 , and SO 2 were analyzed by a flue gas analyzer (Testo 350, Testo AG, Schwarzwald, Germany). The NO oxidation efficiency was calculated by the following equation:

Catalyst Characterization
The XRD patterns were recorded by powder X-ray diffractometer (XRD-600) with Cu Kα radiation (λ = 1.54 Å). The samples were scanned at 2θ ranging from 10 • to 80 • with a scan speed of 6 • min −1 . BET Surface areas of the catalysts were determined by N 2 adsorption-desorption isotherms at −196 • C using specific surface area and porosity analyzer (NOVA 2200, Quantachrome, Boynton Beach, FL, USA). The samples were degassed under vacuum at 300 • C for 4 h. The STEM-mapping analysis was performed using a transmission electron microscope (JEM-2100F, JEOL, Tokyo, Japan) to observe distribution of metal oxides. The surface chemical states of catalysts were tested by X-ray photoeletron spectra (PHI Quantro SXM TM , ULVAC-PHI, Kanagawa, Japan) using an Al Kα X-ray source (1486.7 eV) at 15 kV and 25 W with the binding energy calibrated by C 1s at 284.8 eV.
The NO + O 2 -TPD experiments were performed in a quartz reactor with a FTIR spectrometer (MultiGas TM 2030 HS). Prior to the tests, the samples (200 mg) were pretreated in 10% O 2 /N 2 (500 mL/min) at 500 • C 0.5 h followed by cooling down to 350 • C. The catalysts were exposed to 400 ppm NO, 10% O 2 , N 2 at 350 • C for 40 min, and then cooled down to 50 • C rapidly with N 2 purging. Subsequently, the catalysts were again heated from 50-600 • C with a rate of 10 • C/min in N 2 .
In situ DRIFTS investigations were performed using a Nicolet 6700 spectrometer at 4 cm −1 resolution with 64 co-added scans. Prior to adsorption experiments, the catalysts were pretreated at 500 • C for 0.5 h in N 2 (100 mL·min −1 ) to eliminate the physisorbed water and other impurities. Then the samples were cooled down to 350 • C. After the background was subtracted, the samples were firstly exposed to certain reaction gas mixtures containing 400 ppm NO, 10% O 2 , 1% H 2 O, and balanced with N 2 (total flow 100 mL·min −1 ) for 40 min. Subsequently, the catalysts were treated under 400 ppm NO, 10% O 2 , 1% H 2 O, 100 ppm SO 2 , and balanced with N 2 for 60 min, and the in situ DRIFTS spectra were recorded in the range of 4000-900 cm −1 .

Conclusions
In this work, catalytic oxidation of NO over MnO x -CeO 2 and MnO x -TiO 2 catalysts were studied in the absence or presence of SO 2 . The optimal molar ratio of Mn/Ce and Mn/Ti are 0.7 and 0.5, respectively. MnO x -CeO 2 catalyst gives the highest NO oxidation efficiency of 72% at 325 • C and the NO oxidation efficiency maintained unchanged in 5 h in the presence of 100 ppm SO 2 at 350 • C, while MnO x -TiO 2 catalyst only yields 62% NO oxidation efficiency at 375 • C, and exhibits poor catalytic activity below 325 • C. MnO x -CeO 2 catalysts exhibit better catalytic activity and resistance to SO 2 poisoning than that of MnO x -TiO 2 catalysts, which is attributed that MnO x -CeO 2 catalyst possesses higher surface area, better dispersion of MnO x and stronger NO x adsorption oxidation ability, offers the abundance of Mn 3+ and more surface adsorbed oxygen, and SO 2 might be preferentially adsorbed to the surface of CeO 2 to form sulfate species, lessening the sulfation of MnO x sites.