Mn-Ce-VWO x / TiO 2 SCR Catalysts : Catalytic Activity , Stability and Interaction among Catalytic Oxides

A series of Mn-Ce-V-WOx/TiO2 composite oxide catalysts with different molar ratios (active components/TiO2 = 0.1, 0.2, 0.3, 0.6) have been prepared by wet impregnation method and tested in selective catalytic reduction (SCR) of NO by NH3 in a wide temperature range. These catalysts were also characterized by X-ray diffraction (XRD), Transmission Electron Microscope (TEM), in situ Fourier Transform infrared spectroscopy (in situ FTIR), H2-Temperature programmed reduction (H2-TPR) and X-ray photoelectron spectroscopy (XPS). The results show the catalyst with a molar ratio of active components/TiO2 = 0.2 exhibits highest NO conversion value between 150 ◦C to 400 ◦C and good resistance to H2O and SO2 at 250 ◦C with a gas hourly space velocity (GHSV) value of 40,000 h−1. Different oxides are well dispersed and interact with each other. NH3 and NO are strongly adsorbed on the catalyst surface and the adsorption of the reactant gas leads to a redox cycle with the valence state change among the surface oxides. The adsorption of SO2 on Mn4+ and Ce4+ results in good H2O and SO2 resistance of the catalyst, but the effect of Mn and Ce are more than superior water and sulfur resistance. The diversity of valence states of the four active components and their high oxidation-reduction performance are the main reasons for the high NO conversion in this system.


Introduction
The selective catalytic reduction (SCR) of NO x with NH 3 in the presence of O 2 has been widely used to control the emissions of NO x from mobile and stationary sources, such as coal-fired power plants and automobiles [1][2][3][4].V 2 O 5 -WO 3 /TiO 2 is the most used commercial catalyst for SCR of NO x at a relatively high temperature platform of 300 • C to 400 • C [5][6][7].However, the drawbacks of this catalyst system cannot be ignored.The operating temperature window is narrow and it has low catalytic activity for deNO x at low temperature, for instance.Moreover, the catalytic performance can be seriously deactivated by H 2 O and SO 2 in the emission [8][9][10].Therefore, a low temperature catalyst with good resistance to SO 2 and H 2 O is urgently needed for the SCR system to improve the situation.
So far, a lot of research has been done.Manganese-containing catalysts have attracted much attention due to their relatively high catalytic activity for the conversion of NO x at low temperature, including Ni-MnO x /TiO 2 [6], MnO x -TiO 2 [11,12], Cr-MnO x [13], Ce-MnO x [14,15] and Ca-MnO x /TiO 2 [10].Fang et al. [16] have compared the conversion of NO on MnO x /TiO 2 , MnO x /CNT and nano-flaky MnO x /CNT and found that nano-flaky MnO x /CNTs presents favourable stability, H 2 O resistance and better NO-SCR activity at a more extensive operating temperature window between 150 • C to 300 • C.However, SO 2 leads to the irreversible deactivation of MnO x /TiO 2 [9].In recent years, it has been reported that Ce-based catalysts reveal excellent SCR activity in the presence of SO 2 at 300-400 • C [17][18][19].Yang and co-workers [18] have investigated the effect of SO 2 on the SCR reaction over CeO 2 .The results have indicated that the adsorption of NH 3 over CeO 2 is obviously promoted with the sulfation of CeO 2 , resulting in an obvious promotion of the Eley-Rideal mechanism.
In many cases, one or two active components can hardly handle all the situations of the SCR system.In this work, four elements of manganese, cerium, vanadium and tungsten have been used as the active components, with a reduction in the amount of vanadium.Then, a series of Mn-Ce-V-WO x /TiO 2 composite oxide catalysts with different molar ratios of active components/TiO 2 have been prepared by an impregnation method.In addition, X-ray diffraction (XRD), Transmission electron microscopy (TEM), in situ Fourier transform infrared (FT-IR) spectroscopy, H 2 temperature-programmed reduction (H 2 -TPR) and X-ray photoelectron spectroscopy (XPS) have been measured.The catalyst with a molar ratio of 0.2 exhibits the highest conversion of NO between 150 • C to 400 • C at a gas hourly space velocity (GHSV) of 40,000 h −1 .This catalyst shows long lifetime and superior resistance to H 2 O and SO 2 at 250 • C. It is remarkable to note that the catalyst is not deactivated at all.Moreover, the analysis results also prove that the oxides on the catalyst surface interact with each other and enhance the redox properties of the oxides and the adsorption of reaction gas, resulting in a redox cycle with the valence state change of the surface oxides.This is probably the main reason for its unique performance.

Catalytic Behavior
All SCR activity of catalysts have been tested several times and error analysis have been performed, shown in the form of error bars in the figures.Figure 1A shows the SCR performances of Mn-Ce-V-WO x /TiO 2 composite oxide catalysts with different molar ratios of active components/ TiO 2 = 0.1, 0.2, 0.3, 0.6.For comparison, the SCR performances of V 2 O 5 /TiO 2 , WO 3 /TiO 2 , MnO 2 /TiO 2 , CeO 2 /TiO 2 and TiO 2 are given in Figure 1B.Compared with the single-component catalysts, composite catalysts show better catalytic activity.The NO conversion of all single component catalysts is less than 40% before 250 • C, while all Mn-Ce-V-WO x /TiO 2 composite catalysts are more than 70% from 150 • C to 350 • C.Among them, the NO conversion of the catalyst with active components/ TiO 2 = 0.2 molar ratio is even above 90% from 150 • C to 400 • C. Obviously, the addition of various active components can significantly improve the catalytic activity.The similar condition has also been observed through evaluating the SCR performances of two/three-components catalysts (Figure S1 in the Supplementary Materials).The catalytic activity of the three-components of MnO 2 -V 2 O 5 -WO 3 /TiO 2 and CeO 2 -V 2 O 5 -WO 3 /TiO 2 is significantly higher than that of reference catalysts.Furthermore, remarkably, the performance of CeO 2 -V 2 O 5 -WO 3 /TiO 2 is better than MnO 2 -V 2 O 5 -WO 3 /TiO 2 , especially between 200 to 300 • C. It can be inferred that the effect of MnO 2 and CeO 2 on the V 2 O 5 -WO 3 /TiO 2 catalyst is different.Moreover, with temperature increasing, the SCR activity of the Mn-Ce-V-WO x /TiO 2 catalyst with the highest amount of loading decreases obviously, but others remain (shown in Figure 1A).It should also be mentioned that the catalyst with molar ratio of active components/TiO 2 = 0.2 has the highest catalytic performance at 150 • C. Thus, the optimal molar ratio (Mn-Ce-V-WO x /TiO 2 = 0.2) is obtained, and the catalytic performance will drop when the ratio is low or high, especially when it is high.These results may be due to the different dispersion of the active components on the support surface and the strong interaction among them.Further explanation will be elaborated in the following analyses.

XRD Analysis
The XRD patterns of the Mn-Ce-V-WOx/TiO2 catalysts are shown in Figure 2. The XRD diffractions of the catalyst with the molar ratio of Mn-Ce-V-WOx/TiO2 = 0.1 are well indexed to anatase TiO2 (JCPDS 21-1272) [20,21], indicating that active components are well dispersed on the surface of support.When the molar ratio reaches 0.2, the active components on the surface of the support begins to aggregate, but no strong diffraction peaks of crystal phase appear.If the molar ratio continues to be increased to 0.6, crystalline phases of all active components appear and TiO2 peaks are diluted.The results are consistent with the SCR performances.Catalytic activity could be improved with the transition of active components from highly dispersed state to slightly aggregated state, yet a large amount of components' aggregation will also decrease the catalytic activity [22].
XRD patterns of the reference catalysts are shown in Figure S2 in the Supplementary Materials.It can be seen that the reference catalysts with a single component only have the diffraction peak of anatase TiO2.Interestingly, weak diffraction peaks of V2O5 and WO3 are observed at approximately 23.2° and 34.3° in the spectrum of V2O5-WO3/TiO2 catalyst and 17.9°, 29.1°, 29.3° and 36.1° in the spectrum of MnO2-V2O5-WO3/TiO2 catalyst, but the peaks of these angles disappear in the cerium containing systems; even the peaks of CeO2 and MnO2 have not appeared with the same molar ratio components added.The addition of Ce can be well dispersed, meanwhile, it can make other active components disperse more uniformly and avoid the large aggregation of active components on the support surface, which can make it easier for all active components to contact the reaction gas and improve the catalytic activity of this catalyst system.

XRD Analysis
The XRD patterns of the Mn-Ce-V-WO x /TiO 2 catalysts are shown in Figure 2. The XRD diffractions of the catalyst with the molar ratio of Mn-Ce-V-WO x /TiO 2 = 0.1 are well indexed to anatase TiO 2 (JCPDS 21-1272) [20,21], indicating that active components are well dispersed on the surface of support.When the molar ratio reaches 0.2, the active components on the surface of the support begins to aggregate, but no strong diffraction peaks of crystal phase appear.If the molar ratio continues to be increased to 0.6, crystalline phases of all active components appear and TiO 2 peaks are diluted.The results are consistent with the SCR performances.Catalytic activity could be improved with the transition of active components from highly dispersed state to slightly aggregated state, yet a large amount of components' aggregation will also decrease the catalytic activity [22].

XRD Analysis
The XRD patterns of the Mn-Ce-V-WOx/TiO2 catalysts are shown in Figure 2. The XRD diffractions of the catalyst with the molar ratio of Mn-Ce-V-WOx/TiO2 = 0.1 are well indexed to anatase TiO2 (JCPDS 21-1272) [20,21], indicating that active components are well dispersed on the surface of support.When the molar ratio reaches 0.2, the active components on the surface of the support begins to aggregate, but no strong diffraction peaks of crystal phase appear.If the molar ratio continues to be increased to 0.6, crystalline phases of all active components appear and TiO2 peaks are diluted.The results are consistent with the SCR performances.Catalytic activity could be improved with the transition of active components from highly dispersed state to slightly aggregated state, yet a large amount of components' aggregation will also decrease the catalytic activity [22].
XRD patterns of the reference catalysts are shown in Figure S2 in the Supplementary Materials.It can be seen that the reference catalysts with a single component only have the diffraction peak of anatase TiO2.Interestingly, weak diffraction peaks of V2O5 and WO3 are observed at approximately 23.2° and 34.3° in the spectrum of V2O5-WO3/TiO2 catalyst and 17.9°, 29.1°, 29.3° and 36.1° in the spectrum of MnO2-V2O5-WO3/TiO2 catalyst, but the peaks of these angles disappear in the cerium containing systems; even the peaks of CeO2 and MnO2 have not appeared with the same molar ratio components added.The addition of Ce can be well dispersed, meanwhile, it can make other active components disperse more uniformly and avoid the large aggregation of active components on the support surface, which can make it easier for all active components to contact the reaction gas and improve the catalytic activity of this catalyst system.XRD patterns of the reference catalysts are shown in Figure S2 in the Supplementary Materials.It can be seen that the reference catalysts with a single component only have the diffraction peak of anatase TiO 2 .Interestingly, weak diffraction peaks of V 2 O 5 and WO 3 are observed at approximately 23.2 • and 34.3 • in the spectrum of V 2 O 5 -WO 3 /TiO 2 catalyst and 17.9 • , 29.1 • , 29.3 • and 36.1 • in the spectrum of MnO 2 -V 2 O 5 -WO 3 /TiO 2 catalyst, but the peaks of these angles disappear in the cerium containing systems; even the peaks of CeO 2 and MnO 2 have not appeared with the same molar ratio components added.The addition of Ce can be well dispersed, meanwhile, it can make other active components disperse more uniformly and avoid the large aggregation of active components on the support surface, which can make it easier for all active components to contact the reaction gas and improve the catalytic activity of this catalyst system.

TEM and HRTEM Analysis
TEM images of the catalyst with a molar ratio of 0.2 are presented in Figure 3.The diameter of the catalyst particle is about 20 nm (Figure 3a,b).It is similar to TiO 2 support.The HRTEM micrographs (Figure 3c,d) display better defined contours, exhibiting higher crystalline order (Figure 3e-h).The d-spacing is measured at ca. 0.352 nm, ascribable to the (101) crystal planes of anatase TiO 2 .The detected spacing from different positions is the same, matching well with the XRD pattern.The result further indicates that active components are well dispersed on the surface.The addition of four active components occupy more positions on the surface of support, resulting in an increase of acidic sites exposed on the surface, which improves the ability of NH 3 adsorption; this indirectly corresponds to the in situ FTIR test results of NH 3 adsorption characteristic peak occurring at high temperature in the following.

TEM and HRTEM Analysis
TEM images of the catalyst with a molar ratio of 0.2 are presented in Figure 3.The diameter of the catalyst particle is about 20 nm (Figure 3a,b).It is similar to TiO2 support.The HRTEM micrographs (Figure 3c,d) display better defined contours, exhibiting higher crystalline order (Figure 3e-h).The d-spacing is measured at ca. 0.352 nm, ascribable to the (101) crystal planes of anatase TiO2.The detected spacing from different positions is the same, matching well with the XRD pattern.The result further indicates that active components are well dispersed on the surface.The addition of four active components occupy more positions on the surface of support, resulting in an increase of acidic sites exposed on the surface, which improves the ability of NH3 adsorption; this indirectly corresponds to the in situ FTIR test results of NH3 adsorption characteristic peak occurring at high temperature in the following.
High Resolution Transmission Electron Microscopy (HRTEM) images of the reference catalysts with a molar ratio of 0.6 are presented in Figure S3 in the Supplementary Materials.Compared with the former, HRTEM images of the 0.6-ratio catalyst show different diffraction fringes with different distances.Only the spacing of the representative lattice fringes and the corresponding crystal planes are shown and corresponding crystals of active components are detected, indicating that mass aggregation of active components have appeared on the support surface.The results are similar to XRD analysis, and specific information is presented in Supplementary Materials.However, there is no characteristic fringes of TiO2, indicating that TiO2 may be completely covered by the active components.

Catalyst Stability and H2O/SO2 Resistance
The catalyst with the best SCR performance has been selected for the stability and H2O/SO2 resistance tests.From the above information of activity tests, catalysts begin to remain stable at 250 °C , and the reaction may reach balance at this temperature.So, H2O/SO2 resistance and the 100 h High Resolution Transmission Electron Microscopy (HRTEM) images of the reference catalysts with a molar ratio of 0.6 are presented in Figure S3 in the Supplementary Materials.Compared with the former, HRTEM images of the 0.6-ratio catalyst show different diffraction fringes with different distances.Only the spacing of the representative lattice fringes and the corresponding crystal planes are shown and corresponding crystals of active components are detected, indicating that mass aggregation of active components have appeared on the support surface.The results are similar to XRD analysis, and specific information is presented in Supplementary Materials.However, there is no characteristic fringes of TiO 2 , indicating that TiO 2 may be completely covered by the active components.

Catalyst Stability and H 2 O/SO 2 Resistance
The catalyst with the best SCR performance has been selected for the stability and H 2 O/SO 2 resistance tests.From the above information of activity tests, catalysts begin to remain stable at 250 • C, and the reaction may reach balance at this temperature.So, H 2 O/SO 2 resistance and the 100 h stability test of the catalyst have been performed at 250 • C, as Figure 4 and Figure S4 in the Supplementary Materials present, respectively.It can be seen that the activity of the catalyst remains at about 95.3% during 100 h with small fluctuations, and no decline is observed, which indicates that the catalyst has a longer service life.Furthermore, it is obvious that the SCR activity maintains a very high NO conversion with small fluctuations when 5 vol % H 2 O, 100 ppm SO 2 or both are introduced into the typical reactant gas.The same results have been obtained after repeating the test several times.It has been reported that MnO 2 , V 2 O 5 and WO 3 are easily irreversible deactivated by SO 2 ; only CeO 2 shows high activity in the presence of SO 2 [6,7,9].In this research, although MnO 2 may still be attacked by SO 2 , the adulteration of CeO 2 is significant to the catalyst for resisting the effect of SO 2 .The addition of these two elements makes V and W less affected by SO 2 in the system.The result also indicates that the Mn-Ce-V-WO x /TiO 2 catalyst with molar ratio of active components/TiO 2 = 0.2 has strong resistance to H 2 O and SO 2 .However, it cannot be assumed that the addition of Mn and Ce is simply attacked instead of V and W because the high NO conversion of the catalyst can still last a long time rather than being deactivated quickly in the presence of SO 2 and H 2 O.There is a stronger interaction among the four active ingredients and the role of each element is crucial.Further investigations on the adsorption of reactive molecules and the effects of SO 2 and H 2 O on the catalyst have been performed by in situ FTIR below.
Catalysts 2018, 8, x FOR PEER REVIEW 5 of 13 stability test of the catalyst have been performed at 250 °C, as Figure 4 and Figure S4 in the Supplementary Materials present, respectively.It can be seen that the activity of the catalyst remains at about 95.3% during 100 h with small fluctuations, and no decline is observed, which indicates that the catalyst has a longer service life.Furthermore, it is obvious that the SCR activity maintains a very high NO conversion with small fluctuations when 5 vol % H2O, 100 ppm SO2 or both are introduced into the typical reactant gas.The same results have been obtained after repeating the test several times.It has been reported that MnO2, V2O5 and WO3 are easily irreversible deactivated by SO2; only CeO2 shows high activity in the presence of SO2 [6,7,9].In this research, although MnO2 may still be attacked by SO2, the adulteration of CeO2 is significant to the catalyst for resisting the effect of SO2.The addition of these two elements makes V and W less affected by SO2 in the system.The result also indicates that the Mn-Ce-V-WOx/TiO2 catalyst with molar ratio of active components/TiO2 = 0.2 has strong resistance to H2O and SO2.However, it cannot be assumed that the addition of Mn and Ce is simply attacked instead of V and W because the high NO conversion of the catalyst can still last a long time rather than being deactivated quickly in the presence of SO2 and H2O.There is a stronger interaction among the four active ingredients and the role of each element is crucial.Further investigations on the adsorption of reactive molecules and the effects of SO2 and H2O on the catalyst have been performed by in situ FTIR below.

In Situ FTIR Analysis
The adsorption, transformation and desorption of the reactant gas on the catalyst's surface play an important role in the SCR reaction and they influence the reaction process.Under the condition of continuous exposure of typical reactant gas with or without SO2, the in situ FTIR spectra of the catalyst with a molar ratio of 0.2 are shown in Figure 5. Several bands at 975, 1208, 1437, 1596, 1670, 3164, 3260, 3353 and 3395 cm −1 are observed under the condition of continuous exposure of typical reactant gas without SO2 (shown in Figure 5a).The bands at 975 and 1670 cm −1 disappear over 150 °C, pointing to NH4 + species on Brønsted acid sites [23,24].Then, ν(N−H) and δ(N−H) bands of NH3 adsorbed on Lewis sites at 1208 cm −1 and between 3100 and 3400 cm −1 are observed [25,26].Additionally, the band at 1437 cm −1 weakens gradually with increasing temperature, but exists even at 400 °C.So, it should be assigned to the adsorption of the nitro specie on the catalyst surface [27].Moreover, the band at 1596 cm −1 is attributed to the characteristic band of bridged nitrate species and it is not detected when the temperature is higher than 150 °C due to its poor stability.It can be concluded that NH3 and NO are strongly adsorbed on the catalyst surface even when the temperature reaches 400 °C, and the adsorption of NH3 on Lewis acid sites and NO have strong interaction with catalyst surface oxides.The strong interaction on the support surface promotes the strong adsorption of NH3 and the formation of nitrate species.At 400 °C , the activity of the catalyst remains at 90%, indicating that the catalyst still has a SCR reaction at this temperature.The results correspond well

In Situ FTIR Analysis
The adsorption, transformation and desorption of the reactant gas on the catalyst's surface play an important role in the SCR reaction and they influence the reaction process.Under the condition of continuous exposure of typical reactant gas with or without SO 2 , the in situ FTIR spectra of the catalyst with a molar ratio of 0.2 are shown in Figure 5. Several bands at 975, 1208, 1437, 1596, 1670, 3164, 3260, 3353 and 3395 cm −1 are observed under the condition of continuous exposure of typical reactant gas without SO 2 (shown in Figure 5a).The bands at 975 and 1670 cm −1 disappear over 150 • C, pointing to NH 4 + species on Brønsted acid sites [23,24].Then, ν(N−H) and δ(N−H) bands of NH 3 adsorbed on Lewis sites at 1208 cm −1 and between 3100 and 3400 cm −1 are observed [25,26].Additionally, the band at 1437 cm −1 weakens gradually with increasing temperature, but exists even at 400 • C. So, it should be assigned to the adsorption of the nitro specie on the catalyst surface [27].Moreover, the band at 1596 cm −1 is attributed to the characteristic band of bridged nitrate species and it is not detected when the temperature is higher than 150 • C due to its poor stability.It can be concluded that NH 3 and NO are strongly adsorbed on the catalyst surface even when the temperature reaches 400 • C, and the adsorption of NH 3 on Lewis acid sites and NO have strong interaction with catalyst surface oxides.The strong interaction on the support surface promotes the strong adsorption of NH 3 and the formation of nitrate species.At 400 • C, the activity of the catalyst remains at 90%, indicating that the catalyst still has a SCR reaction at this temperature.The results correspond well to the previous SCR activity test.
When SO2 is introduced into the typical reactant gas, several new bands at 986, 1141, 1213, 1276 and 1363 cm −1 have appeared (Figure 5b).The band at 1141 cm −1 is attributed to the stretching motion of adsorbed sulfate on the surface of the catalyst [1,28].The bands at 986 and 1213 cm −1 are the shift of 975 and 1208 cm −1 , due to their disappearance above 150 °C.However, the band at 1276 cm −1 turns up at 200 °C, which can be assigned to the follow-up shift of 1213 cm −1 [25,29,30].At the same time, the band at 1363 cm −1 may stem from asymmetric vibration of S-O bands of ammonium sulfate ((NH4)2SO4) and grows in intensity over temperature [9,29].The results indicate that there is no competitive adsorption between SO2 and NO, and the adsorption of NH3 on the catalyst can be promoted by adsorbed sulfate.Nevertheless, though SO2 is surely adsorbed on the catalyst and sulfates are formed, the catalytic still maintains high activity, which is the same as before.It can be proposed that SO2 is only adsorbed on specific sites on the catalyst and Ce 4+ may be one of the specific sites according to a previous report [18].

H2-TPR Analyses
To investigate the redox performance of the catalysts, H2-TPR and H2-TPR peak-differentiationimitating of Mn-Ce-V-WOx/TiO2 catalyst profiles have been obtained, as Figure 6 shows.Figure 6A are the H2-TPR curves of the catalysts.V2O5/TiO2 shows a strong reduction peak at 428 °C, corresponding to the vanadia species reduction of V 5+ →V 3+ [31].WO3/TiO2 shows two reduction peaks at 521 °C and 690 °C, and both of them are assigned to the reduction of W 6+ →W 4+ [32].MnO2/TiO2 displays a wide reduction peak at 367 °C ascribed to the reduction of MnO2 to Mn2O3 and Mn2O3 to Mn3O4 and a sharp reduction peak at 504 °C corresponds to the reduction of Mn3O4 to MnO [32][33][34].The weak reduction peak for CeO2/TiO2 at 520 °C is caused by the reduction of low amount surface Ce 4+ →Ce 3+ [7,32,35].Though the reference catalysts show multiple peaks in the temperature range of 300 °C to 800 °C, only one wide peak appears in the H2-TPR profile of the catalyst with the molar ratio of 0.2 at 521 °C and peak shape is very wide, which is due to the overlapping of multiple peaks, even the possible interaction among oxides on the support surface.
To further investigate the important role of active components in the redox performance and the interaction among them, we have presented the H2-TPR peak-differentiation-imitating of Mn-Ce-V-WOx/TiO2 catalyst in Figure 6B.Hydrogen consumption of single component catalysts and corresponding percent in composite catalysts have been listed in Table 1.Considering that the peak temperature of the WO3 and CeO2 is very close, peaks of the two oxides are unified into one peak here.The reduction peaks of one-component catalysts of V2O5 and MnO2 are larger than those of WO3 and CeO2.However, in the composite catalyst, the single peak of MnO2 and the overlap peak of CeO2 and WO3 are larger than the peak of V2O5, which is different from the redox properties of the four active components in the corresponding single-component catalysts.It can be inferred that the effect When SO 2 is introduced into the typical reactant gas, several new bands at 986, 1141, 1213, 1276 and 1363 cm −1 have appeared (Figure 5b).The band at 1141 cm −1 is attributed to the stretching motion of adsorbed sulfate on the surface of the catalyst [1,28].The bands at 986 and 1213 cm −1 are the shift of 975 and 1208 cm −1 , due to their disappearance above 150 • C.However, the band at 1276 cm −1 turns up at 200 • C, which can be assigned to the follow-up shift of 1213 cm −1 [25,29,30].At the same time, the band at 1363 cm −1 may stem from asymmetric vibration of S-O bands of ammonium sulfate ((NH 4 ) 2 SO 4 ) and grows in intensity over temperature [9,29].The results indicate that there is no competitive adsorption between SO 2 and NO, and the adsorption of NH 3 on the catalyst can be promoted by adsorbed sulfate.Nevertheless, though SO 2 is surely adsorbed on the catalyst and sulfates are formed, the catalytic still maintains high activity, which is the same as before.It can be proposed that SO 2 is only adsorbed on specific sites on the catalyst and Ce 4+ may be one of the specific sites according to a previous report [18].

H 2 -TPR Analyses
To investigate the redox performance of the catalysts, H 2 -TPR and H 2 -TPR peak-differentiation-imitating of Mn-Ce-V-WO x /TiO 2 catalyst profiles have been obtained, as Figure 6 shows.Figure 6A are the H 2 -TPR curves of the catalysts.V 2 O 5 /TiO 2 shows a strong reduction peak at 428 • C, corresponding to the vanadia species reduction of V 5+ →V 3+ [31].WO 3 /TiO 2 shows two reduction peaks at 521 • C and 690 • C, and both of them are assigned to the reduction of W 6+ →W 4+ [32].MnO 2 /TiO 2 displays a wide reduction peak at 367 • C ascribed to the reduction of MnO 2 to Mn 2 O 3 and Mn 2 O 3 to Mn 3 O 4 and a sharp reduction peak at 504 • C corresponds to the reduction of Mn 3 O 4 to MnO [32][33][34].The weak reduction peak for CeO 2 /TiO 2 at 520 • C is caused by the reduction of low amount surface Ce 4+ →Ce 3+ [7,32,35].Though the reference catalysts show multiple peaks in the temperature range of 300 • C to 800 • C, only one wide peak appears in the H 2 -TPR profile of the catalyst with the molar ratio of 0.2 at 521 • C and peak shape is very wide, which is due to the overlapping of multiple peaks, even the possible interaction among oxides on the support surface.
The reduction of hydrogen consumption of V2O5 should not be attributed to the strong interaction among active components causing inhibition on V2O5 redox properties, but rather to the fact that in the composite catalyst, the amount of V2O5 is less than that in the other three active components.According to XPS test results, there is still a large amount of V2O5 that has been reduced in the composite catalyst.It can be inferred that they all play major roles in the redox reaction.
The results of H2-TPR suggest that active components of the composite catalyst reveal characteristic redox properties, which is different from that of being alone, and the results indicate that the existence of a strong interaction among the active components on the surface of support is conducive to the promotion of reduction performance.To further investigate the important role of active components in the redox performance and the interaction among them, we have presented the H 2 -TPR peak-differentiation-imitating of Mn-Ce-V-WO x /TiO 2 catalyst in Figure 6B.Hydrogen consumption of single component catalysts and corresponding percent in composite catalysts have been listed in Table 1.Considering that the peak temperature of the WO 3 and CeO 2 is very close, peaks of the two oxides are unified into one peak here.The reduction peaks of one-component catalysts of V 2 O 5 and MnO 2 are larger than those of WO 3 and CeO 2 .However, in the composite catalyst, the single peak of MnO 2 and the overlap peak of CeO 2 and WO 3 are larger than the peak of V 2 O 5 , which is different from the redox properties of the four active components in the corresponding single-component catalysts.It can be inferred that the effect of Mn and Ce are more than superior water and sulfur resistance in the catalytic system.The interaction among the four active components enhances the redox properties of the oxides of the two elements.This can also be evidenced by the valence change of elements on the surface of the composite catalyst after H 2 reduction in the XPS below.It is well known that WO 3 acts as a promoter in traditional V 2 O 5 /WO 3 -TiO 2 catalysts and contributes to the electron transfer among different valences of V element, which is also one of the main reasons for the high SCR activity of traditional catalysts; the oxide of Ce has a better ability to store and release oxygen.However, in the composite catalyst system, both the redox properties of Ce and W have been promoted, which shared part of the role of V 2 O 5 with better redox ability in the catalytic activity.Therefore, both Ce and W contribute to the high NO conversion of the composite catalyst.The areas of reduction peaks of all oxides are different from the corresponding peaks of catalysts with a single component, which appeared in the case of the composite catalysts.The absolute amount of the four active components in the composite catalyst is less than that in the corresponding single-component catalysts, thus the hydrogen consumption of the four active components in the composite catalyst decreases.Among them, the reduction peaks of MnO 2 and CeO 2 +WO 3 are strong but the reduction peak of V is relatively weak.The reduction of hydrogen consumption of V 2 O 5 should not be attributed to the strong interaction among active components causing inhibition on V 2 O 5 redox properties, but rather to the fact that in the composite catalyst, the amount of V 2 O 5 is less than that in the other three active components.According to XPS test results, there is still a large amount of V 2 O 5 that has been reduced in the composite catalyst.It can be inferred that they all play major roles in the redox reaction.The results of H 2 -TPR suggest that active components of the composite catalyst reveal characteristic redox properties, which is different from that of being alone, and the results indicate that the existence of a strong interaction among the active components on the surface of support is conducive to the promotion of reduction performance.

XPS Analysis
In the case of NO-SCR with NH 3 over the composite oxide catalyst, the catalyst will take part in the reaction, which has been reported extensively [13,36,37].In order to obtain the information about the oxidation states of the active components on the catalyst, V 2p, W 4f, Mn 2p, Ce 3d and O 1s XPS spectra of the catalysts are recorded as shown in Figure 7.According to the literature, two main peaks are attributed to V 2p 3/2 and V 2p 1/2 in the V 2p XPS spectra [38,39].However, in this work, only the V 2p 3/2 level can be used to distinguish vanadium oxide species in different chemical states, and the peak of V 2p 1/2 is very weak and hindered by O 1s satellites.The peaks of the catalysts are separated into three peaks at the binding energies of 515.6, 516.3 and 516.8 eV, assigned to V 3+ , V 4+ and V 5+ , respectively [40][41][42].Two main peaks in the W 4f XPS spectra are due to W 4f 7/2 and W 4f 5/2 , and the W 4f 7/2 peak was divided into 34.6 (W 5+ ) and 35.2 eV (W 6+ ) and the W 4f 5/2 peak is divided into 36.9(W 5+ ) and 37.7 eV (W 6+ ), respectively [43,44].Two main peaks assigned to Mn 2p 3/2 and Mn 2p 1/2 are observed, and the Mn 2p 3/2 peaks are separated into three peaks at the binding energies of 640.3 ± 0.2, 641.3 ± 0.2 and 642.6 ± 0.2 eV, corresponding to Mn 2+ , Mn 3+ and Mn 4+ , respectively [1,36,45].The Ce 3d XPS spectra can be fitted into ten peaks: 880.3, 885.9, 898.8 and 903.9 eV assigned to Ce 3+ and 882.5, 888.8, 898.4,901.0, 907.5 and 916.7 eV associated with Ce 4+ .The O 1s peak is fitted into two sub-bands, one at 529.6 eV and the other at 531.1 eV, which can be attributed to the lattice oxygen O 2− and the surface adsorbed oxygen such as O 2 2− , O − , O 2 − or OH − , respectively [13,46].Compared with the XPS spectra of V 2p, W 4f and O 1s, Mn 2p and Ce 3d XPS spectra with the low intensity occur in 8 h using the composite catalyst with the molar ratio of 0.2 (shown in Figure 7C,D).The smaller nano-size and relative amount can lower the intensity of XPS spectra [45, [47][48][49].In this case, the low intensity can be attributed to sulfates covered on Mn and Ce species of the catalyst.Thus, the above conjecture that SO 2 adsorbed only on specific sites on the catalyst is proved to some extent.The quantitative analyses of V, W, Mn, Ce and O species on the catalysts from XPS spectra have been listed in Table 2.For comparison, the quantitative analysis of each element after H 2 reduction has also been listed in Table 2. H 2 for the reduction of the catalyst is complete.The XPS analysis after reduction of H 2 shows that in the single-component catalysts, valence of V on the support surface is all converted from +5 to +3, and valence of Mn is also greatly converted from +4 to +2 and +3.Conversely, valence of W and Ce are relatively hard to be reduced, which proves the high redox properties of V and Mn compared to W and Ce as we know.However, the condition of valence change becomes different in the composite catalyst.It is easily determined that the proportion of different valence states of each element on a single active substance is quite different from those on the composite catalysts, which further indicates that the interaction exists among the oxides in the composite catalysts.The reduction of Mn increases, and the reduction of vanadium is only slightly reduced.It is worth noting that higher valence W and Ce are reduced, which is consistent with the result of H 2 -TPR above.Interestingly, compared to the single-component catalysts, the amount of oxygen adsorbed on the composite catalyst surface decrease.Through the TEM and HRTEM analysis, we know that the addition of four active components occupy more positions on the surface of support, resulting in an increase of acidic sites exposed on the surface, which can explain why the surface adsorption of oxygen decreases.Although there are references supporting that the increase of adsorbed oxygen can promote the SCR catalytic reaction, more importantly, the catalytic reaction involves a large number of electron transfer processes and it is clear that the diversity of valence states of the four active components and their high oxidation-reduction performance are the main reason for the high SCR performance in this system.
into three peaks at the binding energies of 515.6, 516.3 and 516.8 eV, assigned to V 3+ , V 4+ and V 5+ , respectively [40][41][42].Two main peaks in the W 4f XPS spectra are due to W 4f7/2 and W 4f5/2, and the W 4f7/2 peak was divided into 34.6 (W 5+ ) and 35.2 eV (W 6+ ) and the W 4f5/2 peak is divided into 36.9(W 5+ ) and 37.7 eV (W 6+ ), respectively [43,44].Two main peaks assigned to Mn 2p3/2 and Mn 2p1/2 are observed, and the Mn 2p3/2 peaks are separated into three peaks at the binding energies of 640.3 ± 0.2, 641.3 ± 0.2 and 642.6 ± 0.2 eV, corresponding to Mn 2+ , Mn 3+ and Mn 4+ , respectively [1,36,45].The Ce 3d XPS spectra can be fitted into ten peaks: 880.3, 885.9, 898.8 and 903.9 eV assigned to Ce 3+ and 882.5, 888.8, 898.4,901.0, 907.5 and 916.7 eV associated with Ce 4+ .The O 1s peak is fitted into two sub-bands, one at 529.6 eV and the other at 531.1 eV, which can be attributed to the lattice oxygen O 2− and the surface adsorbed oxygen such as O2 2− , O − , O2 − or OH − , respectively [13,46].Compared with the XPS spectra of V 2p, W 4f and O 1s, Mn 2p and Ce 3d XPS spectra with the low intensity occur in 8 h using the composite catalyst with the molar ratio of 0.2 (shown in Figure 7C,D).The smaller nano-size and relative amount can lower the intensity of XPS spectra [45, [47][48][49].In this case, the low intensity can be attributed to sulfates covered on Mn and Ce species of the catalyst.Thus, the above conjecture that SO2 adsorbed only on specific sites on the catalyst is proved to some extent.The quantitative analyses of V, W, Mn, Ce and O species on the catalysts from XPS spectra have been listed in Table 2.For comparison, the quantitative analysis of each element after H2 reduction has also been listed in Table 2. H2 for the reduction of the catalyst is complete.The XPS analysis after reduction of H2 shows that in the single-component catalysts, valence of V on the support surface is all converted from +5 to +3, and valence of Mn is also greatly converted from +4 to +2 and +3.Conversely, valence of W and Ce are relatively hard to be reduced, which proves the high redox properties of V and Mn compared to W and Ce as we know.However, the condition of valence  After the 100 h test, the concentration of O specie almost remains the same.Low valence state V and W species partly transform into the high valence state while high valence state Mn and Ce species partly convert into the low valence state.However, the change is faint, which corresponds with its superior stability.However, the concentrations of V, Mn, Ce and O species display a novel change compared with the fresh-catalyst after the 8 h H 2 O and SO 2 resistance test.The concentration of Ce 4+ decreases, which can be caused by SO 2 .As a reducing agent, SO 2 induces a transformation from Ce 4+ to Ce 3+ on the surface, resulting in the formation of Ce 2 (SO 4 ) 3 [20].The decline of Mn 2+ concentration may be owing to the sulfation on Mn 2+ bound by SO 4 2− [50].Furthermore, adsorption of H 2 O and SO 2 on the sample can also affect the concentration of adsorbed oxygen, which makes it increase.Moreover, the increasing of V 3+ concentration can be due to the enhanced adsorption of NH 3 by the sulfation on the catalyst surface.It can be inferred that the addition of manganese and cerium reduced the adverse effects of SO 2 on the catalytic activity owing to the result that SO 2 molecules are only adsorbed on Ce 4+ and Mn 2+ species on the composite catalyst.

NH 3 -SCR Activity Test
The SCR activity tests were carried out in a fixed-bed quartz reactor (0.3 mL catalyst; 40-60 mesh).The typical reactant gas composition contained: 1500 ppm NO, 1500 ppm NH 3 , 3% O 2 , 5 vol % H 2 O (when added), 100 ppm SO 2 (when added) and balance Ar.The total gas flow rate was 200 mL/min and regulated by mass flow controllers (Sevenstar D08 series Flow Readout Boxed, Beijing Sevenstar Electronics Co., Ltd., Beijing, China), corresponding to the gas hourly space velocity (GHSV) of about 40,000 h −1 .The activity tests were examined at the temperature range of 100-400 • C. The NO outlet concentration was continuously monitored by the ThermoStar Gas Analysis System GSD320 analyzer (Pfeiffer Vacuum GmbH, Berlin, Germany).

Catalyst Characterization
XRD patterns were obtained using a D/MAX-3A Auto X-ray diffractometer (Rigaku Corporation, Tokyo, Japan) with Cu Kα radiation.The X-ray source was operated at 40 kV and 40 mA.The diffraction patterns were taken in the 2θ range of 10-90 • at a scan speed of 15 • min −1 and a resolution of 0.02 • .TEM and HRTEM were performed on a FEI Teccai G2S-Twin electron microscope (PHILIPS, Amsterdam, The Netherlands).
X-ray photoelectron spectra were obtained with K-Alpha spectrometer (Thermo Fisher Scientific, Waltham, MA, America) using Al Kα (1486.7 eV) radiation as the excitation source with a precision of ±0.3 eV.All binding energies were referenced to the C 1s line at 284.6 eV.
H 2 -TPR were performed on a sp-6801 gas chromatograph analyzer (Shandong Lunan Ruihong Chemical Instrument Co., Ltd., Tengzhou, China) using 0.1 g catalyst.The sample was first pretreated in Ar (30 mL•min −1 ) at 50 • C for 1h and then heated up to 800 • C at a rate of 10 • C•min −1 under 5 vol % H 2 /Ar.The consumption of H 2 was measured by a thermal conductivity detector (TCD, BEIJING BUILDER ELECTRONIC TECHNOLOGY CO., LTD., Beijing, China).
In situ FTIR spectra were recorded by a Fourier transform infrared spectrometer (Nicolet 6700, Thermo Fisher Scientific, Waltham, MA, America) equipped with a smart collector and MCT detector cooled by liquid N 2 , collecting 32 scans with a resolution of 4 cm −1 .The catalysts were firstly treated at 500 • C in Ar for 1 h, then cooled down to 50 • C. Subsequently, the SCR reactant gas were introduced to the catalyst for 30 min, and then flushed with Ar for 10 min.The spectra were normally collected at temperatures ranging from 50 • C to 400 • C in a continuous NH 3 and NO flow.The background spectrum was recorded with the flowing of NH 3 and NO and subtracted from the sample spectrum.

Conclusions
Mn-Ce-V-WO x /TiO 2 composite oxide catalysts with the molar ratio of active components/ TiO 2 = 0.2, prepared by wet impregnation method, exhibit high NO conversion between 150 • C to 400 • C and good resistance to H 2 O and SO 2 at 250 • C with a GHSV value of 40,000 h −1 .Four active components are well dispersed on TiO 2 surface and no crystalline phase is formed, but they can aggregate slightly, which is beneficial to the promotion of catalytic activity.NH 3 and NO are strongly adsorbed on the catalyst surface even at 400 • C, indicating that the catalyst still has a SCR reaction at this temperature.SO 2 is only adsorbed on Mn 4+ and Ce 4+ in this catalyst system, resulting in the formation of sulfates.However, the effect of Mn and Ce are more than superior water and sulfur resistance in the catalytic system.In addition, the characteristic redox properties of the catalyst are due to the existence of interaction among the active components on the support surface, and the interaction among them also enhances the redox properties of Ce and W oxides.Thus, all active components play major roles in the redox reaction, and the diversity of valence states of the four active components and their high oxidation-reduction performance are the main reason for the high SCR performance in this system.

Figure 5 .
Figure 5.In situ Fourier Transform infrared spectroscopy (in situ FTIR) spectra of the catalyst with molar ratio of 0.2 under the condition of continuous exposure of typical reactant gas with or without SO2: (a) without SO2; (b) with SO2.

Figure 5 .
Figure 5.In situ Fourier Transform infrared spectroscopy (in situ FTIR) spectra of the catalyst with molar ratio of 0.2 under the condition of continuous exposure of typical reactant gas with or without SO 2 : (a) without SO 2 ; (b) with SO 2 .

Table 1 .
Hydrogen consumption of single component catalysts and corresponding percent in composite catalyst.

Table 1 .
Hydrogen consumption of single component catalysts and corresponding percent in composite catalyst.

Table 2 .
The percent of different valence states for V 2p, W 4f, Mn 2p, Ce 3d and O 1s.