An investigation on the promoting effect of Pr modification on SO2 resistance over MnOx catalysts for selective reduction of NO with NH3

Pr-modified MnOx catalyst was synthesized through a facile co-precipitation process, and the results showed that MnPrOx catalyst exhibited much better selective catalytic reduction (SCR) activity and SO2 resistance performance than pristine MnOx catalyst. The addition of Pr in MnOx catalyst led to a complete NO conversion efficiency in 120–220 °C. Moreover, Pr-modified MnOx catalyst exhibited a superior resistance to H2O and SO2 compared with MnOx catalyst. After exposing to SO2 and H2O for 4 h, the NO conversion efficiency of MnPrOx catalyst could remain to 87.6%. The characterization techniques of XRD, BET, hydrogen-temperature programmed reduction (H2-TPR), ammonia-temperature programmed desorption (NH3-TPD), XPS, TG and in situ diffuse reflectance infrared spectroscopy (DRIFTS) were adopted to further explore the promoting effect of Pr doping in MnOx catalyst on SO2 resistance performance. The results showed that MnPrOx catalyst had larger specific surface area, stronger reducibility, and more L acid sites compared with MnOx catalyst. The relative percentage of Mn4+/Mnn+ on the MnPrOx-S catalyst surface was also much higher than those of MnOx catalyst. Importantly, when SO2 exists in feed gas, PrOx species in MnPrOx catalyst would preferentially react with SO2, thus protecting the Mn active sites. In addition, the introduction of Pr might promote the reaction between SO2 and NH3 rather than between SO2 and Mn active sites, which was also conductive to protect the Mn active sites to a great extent. Since the presence of SO2 in feed gas had little effect on NH3 adsorption on the MnPrOx catalyst surface, and the inhibiting effect of SO2 on NO adsorption was alleviated, SCR reactions could still proceed in a near-normal way through the Eley-Rideal (E-R) mechanism on Pr-modified MnOx catalyst, while SCR reactions through the Langmuir-Hinshelwood (L-H) mechanism were suppressed slightly.


Introduction
Nitrogen oxides (NO x ) emitted from high-temperature combustion processes in diesel engines, power stations, and industrial heaters are one of the main atmospheric pollutants, which can cause several environmental problems, such as acid rain, ozone depletion, photochemical smog, and greenhouse effects (Du et al. 2020;Wang et al. 2019b). Over the past few decades, selective catalytic reduction (SCR) with NH 3 as reductant has become the most effective technology for abating NO x emission from mobile and stationary sources (Gong et al. 2020;Ma et al. 2020). Generally, catalysts play a crucial role in controlling the construction cost and creating efficient reactions in SCR systems ). V 2 O 5 -WO 3 (MoO 3 )/TiO 2 catalysts are often used in power plants due to their excellent catalytic performance in the temperature range of 300-400 °C. However, for some plants with lower temperature flue gas, flue gas should be reheated by a spare heater unit to meet the working temperature of commercial V-based catalysts. Undoubtedly, such heating process would cost more energy for industrial application. Therefore, it is of great significance to develop non-toxic low-temperature SCR catalysts at desired working temperatures.
Up to date, a great effort has been devoted to develop low-temperature catalysts. Especially, transition metal oxide catalysts exhibit outstanding catalytic activities, which are of great potential for industrial application. Among them, MnO x catalysts have attracted increasing attention due to their low-temperature catalytic activities and low cost (Lee and Bai 2019;Liu et al. 2020b;Xie et al. 2020;Yang et al. 2020;Zhang et al. 2019). Nevertheless, MnO x catalysts still suffer from some challenging problems, such as narrow operation window, low N 2 selectivity, and poor resistance to H 2 O and SO 2 , which restricts their commercial application to a great extent. Up to date, it still requires a great of research to improve the SO 2 tolerance of Mn-based catalysts (Khan et al. 2020;Kang et al. 2020).
Generally, there are mainly two causes for the deactivation of MnO x catalysts at low temperature when SO 2 exists in flue gas. On one hand, NH 3 reductant would react with SO 2 to form ammonium sulfate, which would cover catalyst active sites. On the other hand, Mn active sites would also directly react with SO 2 to form metal sulfates (Chang et al. 2013;Chen et al. 2021;Gao et al. 2017b;Jin et al. 2014;Shao et al. 2020). Previous works reported that doping of rare-earth element could improve the SO 2 resistance of Mn-based catalysts. Nd modification on Mn/TiO 2 catalyst could inhibit the sulfation of Mn active species through a preferential reaction between SO 2 and NdO x . The introduction of Ce could also inhibit the accumulation of ammonium sulfates over Mn/TiO 2 catalyst in SCR reactions . Zhang et al. reported that Ho modification was beneficial to suppress the competitive adsorption of SO 2 and NH 3 / NO on the surface of Mn/TiO 2 catalyst, thus improving SO 2 resistance ). The above-mentioned studies indicate that the introduction of rare-earth elements is an effective way to enhance the SO 2 resistance of Mn-based catalysts.
Due to incompletely occupied 4f and empty 5d orbitals, rare-earth element praseodymium (Pr) has been widely applied as an additive in various fields Mamidi et al. 2018;Qi and Wang 2020;Yan et al. 2020). According to the theory of ionic polarization, SO 2 is more liable to react with PrO x rather than MnO x , making PrO x acting as a sacrifice additive to protect Mn active sites. In view of this, some research work has been done to introduce Pr in Mn-based catalysts for enhancing SCR activity, SO 2 and H 2 O tolerance. The previous work demonstrated that the addition of Ce and Pr in MnO x /SAPO-34 catalysts could obviously enhance the SCR performance and SO 2 resistance ). The addition of a suitable amount of Pr in Fe-Mn/TiO 2 catalysts also could improve the SCR performance and SO 2 resistance (Hou et al. 2020). MnO x @ PrO x catalysts with a hollow urchin-like core-shell structure had been synthesized, which exhibited excellent SCR activity and resistance to SO 2 and H 2 O . Pr-modified MnO x catalysts were also prepared through coprecipitation method, and the effect of Pr doping amount on SCR activity at low temperature was also explored in detail (Zhai et al. 2021). However, to the best of our knowledge, it still lacks some systematic research focusing on the enhancement effect and relevant reaction mechanism of Pr modification on MnO x catalysts on SO 2 tolerance.
In this work, pristine MnO x catalyst and Pr-modified MnO x catalyst with Pr/Mn molar ratio of 0.1 were synthesized via co-precipitation method, and the SCR activity test results showed that the introduction of Pr in MnO x catalyst had distinctively improved the resistance to SO 2 and H 2 O. A series of characterization techniques, including XRD, BET, H 2 -TPR, NH 3 -TPD, XPS, TG and in situ DRIFTS, had been adopted to explore the promoting effect of Pr modification on SO 2 tolerance, together with the corresponding reaction mechanism in depth. A possible reaction mechanism was also proposed to explain the reaction pathways for Pr-modified MnO x catalysts with superior SO 2 tolerance.

Catalyst preparation
All reagents were analytical grade and used without further purification. Pr-modified MnO x catalyst with Pr/Mn molar ratio of 0.1 was prepared using co-precipitation method. Firstly, 9.036 g Mn(NO 3 ) 2 ·4H 2 O and 1.566 g Pr(NO 3 ) 3 ·6H 2 O were dissolved in 50 mL deionized water successively and stirred at room temperature for 30 min. Then, ammonia solution (25 wt.%) was added dropwise to the resultant solution to initiate the co-precipitation reaction until the solution pH reached 10. Next, the solution was continuously stirred for 2 h, followed by aging at room temperature for 3 h. Finally, the precipitates were separated by filtrate, dried at 120 °C overnight, and calcined at 450 °C for 3 h in air. Here, the obtained Pr-modified MnO x catalyst was designated as MnPrO x . For comparison, MnO x catalyst with no Pr modification was also prepared by using the same method without the addition of Pr precursor. The MnO x and MnPrO x catalysts that suffered from SO 2 resistance tests were denoted as MnO x -S and MnPrO x -S, respectively. MnPrO x catalyst subjected to SO 2 and H 2 O resistance test was denoted as MnPrO x -SH.

Catalyst characterization
The X-ray diffraction (XRD) patterns of the prepared catalysts were obtained by an X-ray diffractometer (Empyrean, PANalytical) with Cu Kα as radiation source in a 2θ range of 5-70° at a scanning step size of 0.02°. N 2 adsorption-desorption isotherms of the prepared catalysts were tested by a physisorption meter (ASAP 2020 PLUS, Micromeritics). The catalyst sample was degassed under vacuum at 200 °C for 3 h before each test. The N 2 adsorption-desorption data were obtained at liquid nitrogen temperature (− 196 °C).
H 2 -temperature programmed reduction (H 2 -TPR) tests were carried out on a chemisorption analyzer (Autochem II 2920, Micromeritics). Before each test, the catalyst sample was pretreated in He stream (50 mL/min) at 200 °C for 1 h and then cooled down to 50 °C. Finally, the sample was heated from 50 to 800 °C at a heating rate of 10 °C/min in 10 vol.% H 2 /Ar stream.
NH 3 -temperature programmed desorption (NH 3 -TPD) experiments were conducted using a chemisorption analyzer (Autochem II 2920, Micromeritics) with a U-tube quartz reactor. Prior to each test, the sample was pretreated at 200 °C for 1 h in He stream (50 mL/min), cooled down to 100 °C, and then saturated with NH 3 for 30 min. Next, the sample was flushed at 100 °C for 2 h in pure He stream (50 mL/min) to remove physically absorbed NH 3 . Finally, the TPD profile was collected by heating the sample from 100 to 400 °C with a ramping rate of 10 °C/min in He stream (50 mL/min).
X-ray photoelectron spectroscopy (XPS) measurements were performed on a surface analysis photoelectron spectrometer (AXIS-ULTRA DLD-600 W, Shimadzu) using Al Kα as radiation source at 300 W. Binging energy was referred to the C 1 s peak at 284.6 eV as standard.
Thermogravimetric (TG) analysis tests were carried out on a thermogravimetric analyzer (TGA-50, Shimadzu) with the heating rate of 10 °C/min from 20 to 1000 °C in N 2 atmosphere.
In situ DRIFTS measurements were carried out on a FTIR spectrometer (Nicolet iS50, Thermo Fisher Scientific). The DRIFT spectra were recorded by accumulating 64 scans and 4 cm −1 resolution. Before each test, the catalyst sample was pretreated at 350 °C for 30 min in N 2 flow and then cooled down to the targeted reaction temperature to collect the background.

Catalyst activity
For NH 3 -SCR activity test, 1 mL of catalyst sample was filled into a fixed-bed quartz reactor (ID 6 mm). Firstly, the catalyst sample was pretreated at 200 °C for 1 h in N 2 stream and then cooled down to 60 °C. Then, a gas mixture including 500 ppm NO, 500 ppm NH 3 , 5 vol.% O 2 , 100 ppm SO 2 (when used), 5 vol.% H 2 O (when used), and N 2 as balance gas was fed into the reactor with a total flow rate of 500 mL/ min. The corresponding gas hourly space velocity (GHSV) was 30,000 h −1 . The reaction temperature increased from 60 to 300 °C. The concentrations of NO, NO 2 , N 2 O, SO 2 , NH 3 and H 2 O, were monitored continuously by a FTIR spectrometer (Antaris IGS, Thermo Fisher Scientific) equipped with a MCT detector. NO conversion efficiency was calculated according to the following equation:

Results and discussion
NH 3 -SCR performance Fig. 1a illustrated the NH 3 -SCR activities of MnO x and MnPrO x catalysts in the temperature range of 60-300 °C. It could be observed that NO conversion efficiency of pristine MnO x catalyst increased gradually from 60 to 180 °C with the rising of reaction temperature, and it reached ~ 100% at 180 °C. But when further increasing reaction temperature from 180 to 300 °C, NO conversion efficiency decreased slowly. By contrast, the addition of Pr element in MnO x catalyst led to an obvious enhancement on NO conversion performance. NO conversion efficiency of MnPrO x catalyst could maintain at ~ 100% in the temperature range of 120-220 °C. This result demonstrated that Pr modification on MnO x catalyst was helpful to greatly improve low-temperature activity and widen operation temperature window.
In practice, a certain amount of H 2 O and SO 2 always exist in flue gas, which may lead to the deactivation of SCR catalysts (Lee et al. 2013). In this regard, it was crucial to perform H 2 O and SO 2 resistance tests with the prepared catalysts.
As shown in Fig. 1b, with the introduction of 5 vol.% H 2 O in feed gas at 200 °C, NO conversion efficiency of pristine MnO x catalyst decreased quickly down to ~ 60% and then remained stable when further increasing the reaction duration. But when stopping H 2 O injection, NO conversion rate would recover to ~ 97% within half an hour. It suggested that the deactivation of MnO x catalyst by water vapor was reversible. It was interesting that, during the whole reaction process, the injection of H 2 O imposed little effect on NO conversion efficiency of MnPrO x catalyst, indicating that Pr modification on MnO x catalyst could enhance H 2 O resistance significantly.
The results of SO 2 resistance tests were presented in Fig. 1c. As 100 ppm SO 2 was introduced in feed gas at 200 °C, NO conversion efficiency of pristine MnO x catalyst decreased quickly down to 17.8% within 1 h, then further decreasing slowly to 2.7% after extra 4 h. In addition, there was no sign of recovery in NO conversion efficiency after shutting off SO 2 gas, suggesting that the deactivation of MnO x catalyst by SO 2 poisoning was irreversible. Generally, SO 2 resistance performance of pristine MnO x catalysts is a little poor, and our result is consistent with the previous studies (Fan et al. , 2020 According to the above-mentioned results, one might conclude that Pr modification on MnO x catalyst could promote the SCR activity and SO 2 /H 2 O resistance. Considered that it was relatively difficult to reveal the promotional effect of Pr doping in MnO x catalyst on H 2 O resistance, the subsequent work mainly focused on the exploration of the influence of Pr modification on SO 2 resistance. Fig. 2 displayed the XRD patterns of fresh and SO 2 -deactivated catalysts. It could be observed that several diffraction peaks attributed to Mn 3 O 4 clearly appeared in the XRD spectrum of fresh MnO x catalyst, indicating that Mn 3 O 4 was the main crystalline phase in MnO x catalyst. As to MnO x -S catalyst, the characteristic peaks corresponding to the crystalline phase of Mn 3 O 4 could also be observed, but the peak intensities decreased largely. On one hand, a certain amount of ammonium sulfates would be formed unavoidably due the reactions between NH 3 and SO 2 at 200 °C, which would deposit on the surface of MnO x -S catalyst. On the other hand, a portion of manganese oxides at the surface might be sulphated to MnSO 4 in SO 2 resistance test. The reason for the decrease in diffraction peak intensities could be further confirmed by the following TG analysis results.

XRD
As to MnPrO x and MnPrO x -S catalysts, almost no obvious diffraction peaks appeared in the corresponding XRD patterns. It was possible that the introduction of Pr imposed a great impact on the crystalline phase species of MnO x in the preparation process and resulted in fine nanocrystalline MnO x . N 2 adsorption/desorption Fig. 3 showed N 2 adsorption-desorption isotherms of fresh and SO 2 -deactivated catalysts. It could be seen that all the adsorption-desorption isotherms of these catalysts belonged to type IV, implying that SO 2 resistance tests had not changed the mesoporous structures of the catalysts . Fig. S1 illustrated the curves of pore diameter distribution of fresh and SO 2 -deactivated catalysts. It could be seen that the centralized pore diameter of MnO x catalyst shifted from 52.88 to 54.53 nm due to SO 2 resistance test. As for MnPrO x catalyst, it shifted from 21.62 to 22.92 nm.
The specific surface areas, pore diameters, and pore volumes of these catalysts were summarized in Table 1. Obviously, the averaged pore diameter of SO 2 -deactivated catalyst was much larger than the corresponding fresh catalyst. It was ascribed to the sulfate species generated in SO 2 resistance test, which would partially block the fine mesopores of the catalysts . The specific surface area of MnO x -S catalyst decreased down to 11.88 m 2 /g compared with MnO x catalyst (14.03 m 2 /g). As for MnPrO x catalyst, it decreased greatly from 64.76 to 40.76 m 2 /g before and after SO 2 resistance test. Since the specific surface area of MnPrO x catalyst was much higher than that for MnO x catalyst, it seemed normal that the specific surface area of MnPrO x -S catalyst was still much higher than that for MnO x -S catalyst. The high specific surface area was considered as a significant cause for superior SO 2 resistance of MnPrO x catalyst.

H 2 -TPR
The redox properties of fresh and SO 2 -deactivated catalysts were investigated via H 2 -TPR, and the results were shown in Fig. 4. H 2 consumption values and the reduction peak temperatures were listed in Table 2. It could be seen that there were three reduction peaks in the curves of MnO x and MnPrO x catalysts. The first reduction peak could be ascribed to the reduction of MnO 2 to Mn 2 O 3 , the second peak ascribed to the reduction of Mn 2 O 3 to Mn 3 O 4 , and   . Compared with the fresh catalysts, the reduction peak temperatures of corresponding SO 2 -deactivated catalysts had shifted to higher temperature zone, suggesting the decrease in redox properties of SO 2 -deactivated catalysts. However, compared with MnO x -S catalyst, MnPrO x -S catalyst still possessed a higher H 2 consumption value, and its low-temperature reduction peak appeared at a relatively lower temperature. It indicated that Pr modification on MnO x catalyst was conductive to maintain its redox ability in the presence of SO 2 in feed gas. In view of this, the high redox ability might be another reason for MnPrO x catalyst with excellent SO 2 resistance.

XPS
X-ray photoelectron spectra (XPS) were performed to obtain the information about the valence states of active species and atomic concentrations on the catalyst surface. The photoelectron spectra of Mn 2p, O 1 s, and S 2p were presented in Fig. 5, and the analysis results were shown in Table S2. Fig. 5a showed the Mn 2p spectra of fresh and SO 2 -deactivated catalysts. There were two main peaks ascribed to Mn 2p 3/2 and Mn 2p 1/3 for these catalysts. The Mn 2p 2/3 peaks could be further split into three sub-peaks, which could be attributed to Mn 2+ (639.7 ~ 640.5 eV), Mn 3+ (641.1 ~ 641.4 eV), and Mn 4+ (642.8 ~ 643.6 eV), respectively Meng et al. 2015). Here, Mn n+ was denoted as the sum of Mn 2+ , Mn 3+ , and Mn 4+ . As shown in Table S2, compared with fresh catalysts, the relative percentages of Mn 4+ /Mn n+ on the surface of the corresponding SO 2 -deactivated catalysts decreased slightly. Note that the relative percentage of Mn 4+ /Mn n+ on the MnPrO x -S catalyst surface was much higher than those of MnO x and MnO x -S catalysts. Generally, it was considered that a relatively high Mn 4+ proportion played a key role in enhancing the SCR activity at low temperature. This result demonstrated that Pr modification was in favor of maintaining a high level of Mn 4+ proportion in Mn-based catalysts, thus improving the SCR performance and SO 2 resistance. In addition, Mn atomic concentration on the surface of MnPrO x catalyst was 21.50% which was obviously lower than that of MnPrO x catalyst (28.39%). Though the introduction of Pr would result in the decrease of Mn atomic concentration on the surface of MnPrO x catalyst, the SCR performance had been improved greatly. It seemed that Pr modification rather than Mn atomic concentration in MnPrO x catalyst was the key factor for improving SCR activity. Fig. 5b showed the O 1s spectra of fresh and SO 2 -deactivated catalysts. Each O 1 s spectrum could be split into two peaks. One peak at 528.2 ~ 528.6 eV could be assigned to the lattice oxygen species O 2− (denoted as O β ) at low binding energy. The other peak at 530.1 ~ 530.2 eV could be attributed to the chemisorbed oxygen species at high binding energy (denoted as O α ), including O − (defectoxide) and OH − (hydroxyl-like group) (Boningari et al. 2015;Wu et al. 2020). As shown in Table S2, the SO 2 resistance tests resulted in an increase of O α /(O α + O β ) ratio, and the O α /(O α + O β ) ratio of MnPrO x -S catalysts was much higher than the other catalysts. The increase in O α /(O α + O β ) ratio might be ascribed to the existence of hydroxyl-like groups offered by S = O (Tang et al. 2018). Since the hydroxyl-like groups were in favor of forming more acid sites on the catalyst surface, it might impose a positive impact on the SCR activity of MnPrO x catalyst.   Fig. 5c showed the S 2p spectra of MnO x -S and MnPrO x -S catalysts. Each S 2p spectrum could be split into two peaks, which could be assigned to SO 4 2− (168.8 eV) and HSO 4 − (167.3 eV), respectively (Arfaoui et al. 2018;Sheng et al. 2018). The results indicated that sulfate species had been formed on the surface of MnO x -S and MnPrO x -S catalysts in SO 2 resistance tests, possibly as ammonium or Mn salts. As shown in Table S2, S atomic ratio of MnPrO x -S catalyst was 3.48%, which was much higher than that of MnO x -S catalyst (2.58%), indicating that more sulfate species had been formed or deposited on the surface of MnPrO x -S catalyst compared with MnO x -S catalyst. It implied that the enhancement effect of Pr modification on SCR performance might result from the sacrifice reaction between PrO x and SO 2 , rather than inhibiting the formation or deposition of sulfate species on the catalyst surface.

TG
In order to further distinguish the sulfate species on the surface of SO 2 -deactivated catalysts, thermogravimetric (TG) analysis of MnO x -S and MnPrO x -S catalysts was performed, and the TG-DTG curves were shown in Fig. S3. The weight loss in a temperature range of room temperature to 200 °C (step 1) could be attributed to the evaporation of physically absorbed water. The weight loss in a temperature range of 200-400 °C (step 2) could be ascribed to the decomposition of (NH 4 ) 2 SO 4 or NH 4 HSO 4 . The weight loss in a temperature range of 400-700 °C (step 3) could be related to the phase transformation of metal oxides. The weight loss at > 700 °C (step 4) could be assigned to the decomposition of metal sulfates (Sun et al. 2018;Tang et al. 2018).
It could be seen from Fig. S3(a) that, for MnO x -S catalyst, there was no obvious weight loss in the temperature range of 200-400 °C (step 2), indicating that little (NH 4 ) 2 SO 4 or NH 4 HSO 4 could be found on the surface of MnO x -S catalyst. The sharp declines in weight loss at 560 °C and 795 °C could be ascribed to the thermal transformation of MnO 2 to Mn 2 O 3 and the decomposition of MnSO 4 , respectively . In view of the above-mentioned XPS result, one could infer that the sulfate species on the surface of MnO x -S catalyst might be metal sulfates. It indicated that the rapid deactivation of pristine MnO x catalyst was possibly due to the sulfation of Mn active sites.
As shown in Fig. S3(b), a weight loss of 0.916% in MnPrO x -S catalyst could be observed in step 2, indicating that the introduction of Pr in MnO x catalyst would promote the formation of (NH 4 ) 2 SO 4 or NH 4 HSO 4 on the catalyst surface. The sharp declines in weight loss at 540 °C and 630 °C could be ascribed to the thermal transformation of MnO 2 to Mn 2 O 3 and PrO 2 to Pr 2 O 3 , respectively . At high temperature above 700 °C, the weight loss owing to the decomposition of manganese sulfates on MnPrO x -S catalyst surface was only 1.153%, which was much less than that for MnO x -S catalyst (3.394%). Namely, much less manganese sulfates had been formed on the MnPrO x -S catalyst surface compared with MnO x -S. Here, the sharp decline in weight loss at 910 °C could be ascribed to the decomposition of praseodymium sulfates . The above results suggested that, when SO 2 exists in feed gas, PrO x species in MnPrO x catalyst would preferentially react with SO 2 , thus protecting Mn active sites. On the other hand, the introduction of Pr might promote the reaction between SO 2 and NH 3 rather than between SO 2 and Mn active sites, which was also conductive to protect Mn active sites to a great extent.

In situ DRIFTS
In order to further investigate the adsorption behavior and SCR reaction mechanism in the presence of SO 2 at low temperature, a series of in situ DRIFTS tests were performed at a relatively low temperature (200 °C). Fig. 6 showed in situ DRIFT spectra of SO 2 + O 2 adsorption over MnO x and MnPrO x catalysts. For each test, the catalyst sample was pretreated by N 2 purging for 30 min, then 100 ppm SO 2 and 5 vol.% O 2 were introduced with a flow rate of 100 mL/min for 30 min, and IR spectra were recorded continuously.

SO 2 + O 2 adsorption on the catalysts
As shown in Fig. 6a, with the introduction of SO 2 in feed gas, several bands at 1617,1550,1423,1336,1273,1251,1148 and 1052 cm −1 appeared clearly in the IR spectrum of MnO x catalyst. The band at 1617 cm −1 could be attributed to the adsorbed HSO 4 -, while the band at 1423 cm −1 could be assigned to the symmetrical stretching vibration of the adsorbed SO 2 . The bands at 1336, 1273 and 1251 cm −1 were assigned to the asymmetric stretching vibration of O = S = O (Tang et al. 2020). The band at 1148 cm −1 was attributed to bulk metal sulfate, and the band at 1052 cm −1 was assigned to the S-O stretching vibrations of bidentate sulfates located on the catalyst surface (Gao et al. 2019). With the increase of SO 2 + O 2 exposure duration, most of the band intensities increased slightly. This result demonstrated that the presence of SO 2 had quickly led to the formation of sulfates on the MnO x catalyst surface, namely the sulfation of Mn active species, resulting in the instant deactivation of MnO x catalyst.
As shown in Fig. 6b, with the introduction of SO 2 in feed gas, several bands at 1641, 1535, 1146 and 1043 cm −1 appeared obviously in the IR spectrum of MnPrO x catalyst. The band at 1636 cm −1 could be assigned to the adsorbed H 2 O (Wang et al. 2019a). The band at 1146 cm −1 could be attributed to bulk metal sulfate, and the band at 1043 cm −1 could be ascribed to the S-O stretching vibrations of bidentate sulfates located on the catalyst surface (Gao et al. 2019). Similarly, the band intensities increased obviously with the rising of SO 2 + O 2 exposure time. Since the band intensities corresponding to sulfates on the MnPrO x catalyst surface were much higher than those on the MnO x catalyst surface within the same SO 2 + O 2 exposure time, it indicated that much more metal sulfates had been formed on the MnPrO x catalyst surface. This result agreed well with the above-mentioned XPS and TG-DTG analysis results. It demonstrated that PrO x rather than MnO x was apt to react with SO 2 and PrO x acted as a sacrifice additive to protect Mn active species effectively. It was considered as the major cause for the enhancement effect of Pr modification on SO 2 resistance.

Effect of SO 2 on NH 3 adsorption on the catalysts
In situ DRIFTS experiments were conducted to investigate the effect of SO 2 on NH 3 adsorption on the surface of MnO x Fig. 6 In situ DRIFT spectra of (a) MnO x and (b) MnPrO x catalysts exposed to 100 ppm SO 2 + 5 vol.% O 2 at 200 °C and MnPrO x catalysts. After pretreatment with N 2 purging for 30 min, each catalyst sample was exposed to 500 ppm NH 3 (100 mL/min) for 30 min, followed by introducing 100 ppm SO 2 in NH 3 /N 2 mixed gas for 30 min, and the IR spectrum was recorded with the increase of reaction time.
As shown in Fig. 7a, after NH 3 adsorption for 30 min, a saturated adsorption of NH 3 on MnO x catalyst surface would be reached, and several bands could be observed in the spectrum. The bands at 1192 and 1603 cm −1 could be assigned to coordinated NH 3 on Lewis acid sites. The band at 1427 cm −1 could be ascribed to NH 4 + species on Brønsted acid sites . Note that, before the introduction of SO 2 in NH 3 /N 2 mixed gas, NH 3 coordinated on Lewis acid sites and NH 4 + species on Brønsted acid sites existed on the MnO x catalyst surface at a relatively low level. With the introduction of SO 2 in the feed gas, a new band (1678 cm −1 ) ascribed to NH 4 + species on Brønsted acid sites appeared. With the increase of SO 2 injection duration, the intensities of bands at 1427 and 1678 cm −1 increased obviously, indicating that a large amount of NH 4 + species had been formed on Brønsted acid sites. That was because SO 2 facilitated the generation of NH 4 + species on Brønsted acid sites . At the beginning of SO 2 injection in mixed gas, the band (1192 cm −1 ) assigned to coordinated NH 3 species on Lewis acid sites disappeared within 1 min. It suggested that NH 3 adsorbed on Lewis acid sites had been consumed completely due to the reactions between SO 2 and activated NH 3 species. Meanwhile, the bands (1268, 1144 and 1029 cm −1 ) assigned to sulfate species appeared, and their intensities increased gradually with the rising of SO 2 injection duration. It implied that more and more SO 2 had been reacted with NH 3 and Mn active species to generate sulfates on the catalyst surface, thus resulting in the deactivation of MnO x catalyst. Fig. 7b showed the change in absorbance bands of MnPrO x catalyst with SO 2 exposure duration. It could be seen that, after saturated adsorption of NH 3 for 30 min, several bands could also be observed in the spectrum of MnPrO x catalyst. The bands at 1181 and 1607 cm −1 might be assigned to coordinated NH 3 on Lewis acid sites. The bands at 1270 and 1421 cm −1 might be ascribed to NH 4 + species on Brønsted acid sites. Similar to MnO x catalyst, NH 3 coordinated on Lewis acid sites and NH 4 + species on Brønsted acid sites existed at a relatively low level on the MnPrO x catalyst surface before the introduction of SO 2 in NH 3 /N 2 mixed gas. With the increase of SO 2 exposure duration, the absorbance intensities of the bands (1181 and 1607 cm −1 ) assigned to coordinated NH 3 species on Lewis acid sites increased gradually. Previous study reported that adsorbed NH 3 species on Lewis acid sites would play a key role in SCR reactions at low temperature (Zuo et al. 2014). The increase in band intensities of adsorbed NH 3 species on Lewis acid sites demonstrated that the introduction of Pr was beneficial to suppress the comparative adsorption of NH 3 and SO 2 on the catalyst surface, thus imposing a positive effect on SO 2 tolerance and SCR activity. Meanwhile, the new bands (1036 and 1122 cm −1 ) assigned to metal sulfates had also appeared, and their intensities increased gradually with the rising of SO 2 injection duration. Obviously, the band intensities corresponding to sulfates on the MnPrO x catalyst surface were much higher than those on the MnO x catalyst surface; it implied that much more metal sulfates had been formed on the MnPrO x catalyst surface. It suggested that PrO x was apt to react with SO 2 in the presence of NH 3 , resulting in the formation of Pr sulfates rather than Mn sulfates. Thus, Pr worked well as a sacrifice additive to protect Mn active sites.

Fig. 7
In situ DRIFT spectra of (a) MnO x and (b) MnPrO x catalysts exposed to 500 ppm NH 3 followed by the addition of 100 ppm SO 2 at 200 °C

Effect of SO 2 on NO + O 2 adsorption on the catalysts
In situ DRIFTS experiments were performed to investigate the effect of SO 2 on NO adsorption on the surface of MnO x and MnPrO x catalysts. Firstly, the catalyst was pretreated with 500 ppm NO and 5 vol.% O 2 for 30 min, and then 100 ppm SO 2 was introduced in the (NO + O 2 )/N 2 mixed gas.
As shown in Fig. 8a, after the introduction of NO + O 2 for 30 min, several bands (1627, 1545, 1345 and 1266 cm −1 ) corresponding to the adsorbed NO species appeared in the spectrum of MnO x catalyst. These bands could be assigned to adsorbed NO 2 species (1627 cm −1 ), bidentate nitrate (1545 cm −1 ), free NO 3 − (1345 cm −1 ), monodentate nitrate (1266 cm −1 ), respectively Gao et al. 2017a). It could be seen that, with the introduction of SO 2 , the absorbance intensities of these bands began to decrease quickly. Meanwhile, several new bands (1616,1337,1274,1152 and 1057 cm −1 ) assigned to sulfate species appeared after the injection of SO 2 in feed gas, and their absorbance intensities increased gradually. It indicated that there was obvious competitive adsorption of SO 2 and NO on the MnO x catalyst surface. In other words, SO 2 would restrain NO adsorption on the catalyst surface, further inhibiting the SCR reactions occurred on MnO x catalyst via the L-H mechanism. Langmuir-Hinshelwood(L-H) mechanism: NO is adsorbed on the active site adjacent to the active site of the adsorbed NH 3 , and the transition intermediate state is generated through the reaction, and then decomposed into N 2 and H 2 O. This might be the direct reason for the fast decline in NO conversion efficiency when MnO x catalyst was exposed to SO 2 (Jin et al. 2014).
As shown in Fig. 8b, the inhibiting effect of SO 2 on NO adsorption on the MnPrO x catalyst surface was similar to that for MnO x catalyst. But it was worth noting that the band (1545 cm −1 ) attributed to bidentate nitrate on the MnO x catalyst surface disappeared just after exposing to SO 2 for 5 min. By contrast, the corresponding band (1557 cm −1 ) for MnPrO x catalyst could also be seen even after the injection of SO 2 for 30 min. It suggested that the addition of Pr in MnO x catalyst could restrain the comparative adsorption of SO 2 and NO on the catalyst surface, which was undoubtedly beneficial to maintain SCR reactions occurring through the L-H mechanism. This was a vital factor influencing the SCR activity and SO 2 resistance performance of Pr-modified MnO x catalyst.

The promoting effect of Pr modification on SO 2 and H 2 O resistance
As shown in Table S2, S and N atomic ratios of MnPrO x -SH catalyst were 3.13% and 8.07%, respectively, which were a little lower than those of MnPrO x -S catalyst (3.48% and 8.35%). It implied that less sulfate species had been formed on the MnPrO x -SH catalyst surface compared with MnPrO x -S catalyst. As shown in Fig. S3, the weight loss in MnPrO x -SH catalyst in step 2 was 1.264%, which was obviously more than that for MnPrO x -S catalyst (0.916%). Here, it was considered that, in the presence of H 2 O and SO 2 in feed gas, it might be comparatively easier for PrO x species in MnPrO x catalyst to promote the reaction between the adsorbed SO 2 /H 2 O and NH 3 species. Thus, the introduction of Pr in MnO x catalyst could preferably protect Mn active species when H 2 O and SO 2 coexisted in feed gas, resulting in an excellent resistance to H 2 O and SO 2 . At high temperature above 700 °C, the weight loss in MnPrO x -SH catalyst due to the decomposition of manganese sulfates was 1.045%, which was slightly less than that for MnPrO x -S catalyst (1.153%). It implied that, to some extent, the coexistence of H 2 O and SO 2 in feed gas might be beneficial to decrease the consumption Fig. 8 In situ DRIFT spectra of (a) MnO x and (b) MnPrO x catalysts exposed to 500 ppm NO + 5 vol.% O 2 followed by the addition of 100 ppm SO 2 at 200 °C of PrO x species which acted as a sacrifice material to react with the adsorbed SO 2 . In other words, PrO x in MnPrO x catalyst might suppress SO 2 adsorption on the catalyst surface in the presence of H 2 O, resulting in forming less sulfate species and consuming less PrO x , thus further enhancing the SCR performance of MnPrO x catalyst.

The possible SCR reaction mechanism
In this work, Pr-modified MnO x catalyst exhibited a superior SO 2 resistance over pristine MnO x catalyst. According to TG and in situ DRIFTS analysis results, it suggested that the presence of SO 2 would quickly lead to the sulfation of Mn active species, resulting in a fast deactivation of MnO x catalyst. However, Pr modification on MnO x catalyst would promote the reaction between SO 2 and NH 3 species rather than between SO 2 and Mn active sites. Besides, PrO x additive might preferentially react with SO 2 when SO 2 was injected in feed gas, thus protecting Mn active sites. For MnO x catalyst, there was strong competitive adsorption between SO 2 and NH 3 /NO on the catalyst surface. Since the adsorption of NH 3 and NO on MnO x catalyst would be severely suppressed in the presence of SO 2 , the relevant SCR reactions following either L-H or E-R mechanisms would be inhibited obviously. Eley-Rideal(E-R) mechanism: Gaseous NO reacts with adsorbed activated NH 3 to form a transitional intermediate state, which is further decomposed into N 2 and H 2 O. But for Pr-modified MnO x catalyst, the presence of SO 2 in feed gas had little effect on NH 3 adsorption on the catalyst surface. In addition, the inhibiting effect of SO 2 on NO adsorption on MnPrO x catalyst was not so severe as the case for MnO x catalyst though competitive adsorption between SO 2 and NO could also be observed on the surface of Pr-modified MnO x catalyst. To the end, SCR reactions could still proceed in a near-normal way through the E-R mechanism on Pr-modified MnO x catalyst, while the SCR reactions through the L-H mechanism might be suppressed slightly. According to the above-mentioned discussion, the promoting effect of Pr modification on MnO x catalyst on SO 2 resistance and the possible reaction mechanism for NH 3 -SCR of NO might be illustrated in Fig. 9.

Conclusions
In this work, the promoting effect and reaction mechanism of Pr modification on SO 2 resistance over MnO x catalyst had been investigated systematically. The results showed that MnPrO x catalyst exhibited much better SCR activity and resistance to H 2 O and SO 2 than pristine MnO x catalyst. A nearly complete NO conversion efficiency of MnPrO x catalyst could be obtained in the temperature range of 120-220 °C. In the presence of H 2 O, NO conversion efficiency of MnPrO x catalyst could maintain at ~ 100% at 200 °C. As to pristine MnO x catalyst, NO conversion efficiency decreased quickly down to ~ 20% after exposing to 100 ppm SO 2 for 1 h. But for Pr-modified MnO x catalyst, NO conversion efficiency just decreased down to 90.8% after exposing to SO 2 for 2.5 h. The characterization results showed that, compared with MnO x catalyst, MnPrO x catalyst exhibited much larger specific surface area, stronger reducibility, and more L acid sites, which was beneficial to improve the SCR activity. XPS analysis results showed that the relative percentage of Mn 4+ /Mn n+ on the MnPrO x -S catalyst surface was much higher than those of MnO x and MnO x -S catalysts. It demonstrated that Pr modification was in favor of maintaining a high level of Mn 4+ proportion in Mn-based catalysts, thus improving the SCR performance and SO 2 resistance. TG analysis results showed that the rapid deactivation of pristine MnO x catalyst was possibly due to the sulfation of Mn active sites. But for Fig. 9 The promoting effect and reaction mechanism of Pr modification on SO 2 resistance over MnO x catalyst MnPrO x catalyst, PrO x species would preferentially react with SO 2 in the presence of SO 2 in feed gas, thus protecting the Mn active sites. In addition, the introduction of Pr might promote the reaction between SO 2 and NH 3 rather than between SO 2 and Mn active sites, which was also conductive to protect the Mn active sites to a great extent. In situ DRIFTS results confirmed that the presence of SO 2 had quickly led to the sulfation of Mn active species. As to MnPrO x catalyst, PrO x rather than MnO x was apt to react with SO 2 , thus acting as a sacrifice additive to protect Mn active species effectively. Since the adsorption of NH 3 and NO on MnO x catalyst would be severely suppressed in the presence of SO 2 , the relevant SCR reactions following either L-H or E-R mechanisms would be inhibited obviously. But for Pr-modified MnO x catalyst, the inhibiting effect of SO 2 on NO adsorption was not so severe as the case for MnO x catalyst. More importantly, the presence of SO 2 in feed gas had little effect on NH 3 adsorption on the catalyst surface. Therefore, SCR reactions could still proceed in a near-normal way through the E-R mechanism on Pr-modified MnO x catalyst, while SCR reactions through the L-H mechanism might be suppressed slightly.