Insights into Reaction Conditions for Selective Catalytic Reduction of NO x with a MnAl Oxide Catalyst at Low Temperatures

MnO x catalysts have been proven to have superior activity at low temperatures (LT) for selective catalytic reduction (SCR) of NO x . In this study, spherical mesoporous MnAl x catalysts with different molar ratios were prepared using a simple precipitation method. In addition, their NO x catalytic performance, along with relevant main factors (reaction temperature, molar ratio, particle size, space velocity, and O 2 content), was investigated in a simulated fixed bed reactor. The relationship between the main factors and DeNO x efficiency was evaluated by analyzing two parameters: the NO x conversion rate and N 2 selectivity. The results showed that the MnAl 0.1 catalyst had the optimal low temperature DeNO x performance and had a relatively stable NO x conversion rate of more than 92% in a space velocity range of 14331 h – 1 to 23885 h – 1 , where smaller catalyst particle size led to higher NO x conversion rate and N 2 selectivity as high as 100%. In addition, the appropriate oxygen concentration was found to increase NO x conversion and N 2 selectivity. This work would be beneficial to further optimization of LT SCR catalyst systems for practical applications.


INTRODUCTION
Nitrogen oxides (NOx) as air pollutants are primarily emitted from coal-fired power plants and automobiles and are regarded as the major source of many environmental and health problems, including photochemical smog, acid rain, greenhouse gases, and PM2.5 (Paolucci et al., 2017;Han et al., 2019). In response to environmental protection requirements, stricter emission standards have been adopted by governments around the world to control NOx emissions. Alternative postcombustion NOx reduction techniques, including selective catalytic reduction and selective noncatalytic reduction (SNCR), have been used for NOx elimination. Among them, the SCR process using NH3 as the reductant is the most widely applied in post-NOx abatement and has been shown to have high NOx removal efficiency (Forzatti et al., 2009;Imanaka and Masui, 2012). Although commercial vanadium (V)-based oxide catalysts have been widely studied and used, the main drawbacks (narrow temperature window and the toxicity of V components) limit their usage (Brandenberger et al., 2008;Kang et al., 2018). Also, V-based catalysts are not suitable for glass, cement, and steel plants due to the low-temperature flue gas (< 250°C). Recently, a great deal of effort has been made to explore LT V-free SCR high efficient catalysts.
MnOx catalysts exhibit excellent activity at low temperature owing to their excellent redox properties and variable valence states. However, poor SO2 tolerance and low N2 selectivity relatively restrict their practical applications. Mixing or doping MnOx with other metal oxides to obtain polymetal composites with various excellent properties is an effective method to overcome these shortcomings. A great deal of effort has been devoted to constructing bimetallic Mn-based catalysts, including MnCe, MnCo, MnFe, MnNb, and MnNi, due to the synergistic effects resulting from their complementary advantages Lei et al., 2014;Lian et al., 2014;Fan et al., 2017). Catalysts with high catalytic performance have also been obtained by adjusting the morphology and pore structure of the catalysts and promoting the dispersion of active components. Al2O3 is widely used in SCR catalysts because of its high thermal stability, moderate acidity, and high specific surface area, which can prevent the catalyst from sintering and improve the adsorption of NOx (Wang and Lin 1998;Liu et al., 2016). Zhao et al. (2014) found that Al2O3 modified CeO2/TiO2 can improve the acidity of the catalyst, thereby reducing SO2 adsorption. Mejía-Centeno et al. (2015) reported that Al2O3 doped TANs enhances activity at high temperatures due to the Lewis acidity and the multiwalled morphology of the TNTs.
Since Mn-based bimetal oxide catalysts are more active and selective than other alternatives, Al2O3-supported Mn-based oxide catalysts have been intensively reported. Singoredjo et al. (1992) first prepared MnOx/Al2O3 as SCR catalysts, and the results showed that Mn induced better dispersion on an Al2O3 support, resulting in higher catalyst activity. Kapteijn et al. (1994) studied the surface acidity of an MnOx-Al2O3 catalyst and the adsorption capacity of NOx and NH3 on a surface. Cao et al. (2014Cao et al. ( , 2015 found that the high surface area of Al2O3 was the key to increasing the number of active catalytic sites by modifying MnOx-Al2O3 with well-dispersed Fe, Ag, and Zr. The appropriate introduction of Al2O3 into MnOx can reduce the thermal stability of the adsorbed SO2 species and facilitate the decomposition of NH4HSO4 (Fan et al., 2020). Kijlstra et al. (1997) proposed that the Lewis acid center is Mn 3+ dispersed on the surface of MnOx/Al2O3, where NO or an activated nitrite intermediate adsorbed on the surface reacts with NH2 − on these acid centers, namely, two types of mechanisms (E-R and L-H) work together.
Previous reports have shown that Mn-based catalysts have high NOx conversion rates, but the working conditions of catalysts are complex and changeable, which creates more requirements for these catalysts in practical application. Thus, in this study, we prepared a novel spherical mesoporous MnAlx catalyst and aimed to explore the effects of various technological conditions on the NOx conversion performance of the catalysts. MnAlx oxide catalysts with different molar ratios were prepared using a simple precipitation method, and their NOx catalytic performance was investigated in the simulated fixed bed reactor, along with relevant main factors (reaction temperature, molar ratios, particle sizes, space velocity, and O2 content). The relationship between the main factors and DeNOx efficiency was evaluated by analyzing the NOx conversion rate and N2 selectivity. This work should be beneficial to the further optimization of LT SCR catalyst systems developed for practical applications.

Preparation of Catalysts
The MnAlx mixed oxides were synthesized using a simple precipitation method. A specific amount of Mn(NO3)2 and Al(NO3)3·9H2O at a total mole amount of 5 mmol was dissolved in 30 mL deionized water under magnetic stirring. After that, an aqueous (NH4)2CO3 solution containing solute of over-stoichiometric ratio to the above metal ions were added dropwise and stirred for 4 h. The obtained flocculent precipitate was subsequently washed with water and ethanol three times and dried at 120°C overnight, calcined in a tube furnace at 300°C for 1 h, and then heated to 500°C for 6 h to obtain MnAlx oxides (x = 0.1, 0.3, 0.5,0.7 and 0.9, where x is the mole ratio of Al/(Mn + Al)).

Catalyst Characterization
X-ray diffraction (XRD) was carried out with a BRUKER-AXS D8 Advanced X-ray Diffractometer with Cu Ka radiation at a scanning angle (2θ) ranging from 10°-90° and a scan rate of 8° min -1 . Scanning electron microcopy (SEM) and corresponding energy dispersive spectroscopy (EDS) mapping analyses were carried out using a JEOL SU8200 microscope at an acceleration voltage of 200 kV. The surface properties (BET specific surface area and pore size) of the samples were measured with a Micromeritics ASAP 2020 surface area analyzer. Temperature-programmed reduction (H2-TPR) tests were performed on a gas chromatograph (SP-6800, Shandong Lunan Ruihong, China) with a thermal conductivity detector (TCD) under a 5% H2/N2 gas flow (50 mL min -1 ) at a rate of 10°C min -1 up to 800°C. Temperature-programmed desorption of NH3 (NH3-TPD) was carried out using the same instrument as that used for the H2-TPR. 100 mg of the sample saturated with adsorbed NH3 was monitored at temperatures ranging from 20-700°C in a He flow.

Activity Measurement
The catalytic activity measurement was performed in a fixed-bed quartz tube reactor with an internal diameter of 8 mm. The catalyst was placed in the reactor and pretreated at 200°C for 1 h in an N2 atomosphere. The feed gas mixture contained NO, NH3, and O2, with He as the balance. When the reaction reached a stable state at each temperature, the concentrations of NO and NO2 were continually monitored with the TH-990S NO and NO2 analyzers, and N2O in the outlet gas was detected with a Antaris™ IGS Gas Analyzer from Thermo Fisher Scientific Inc. The N2 concentration was analyzed using a thermal conductivity detection method. The NOx conversion rate and N2 selectivity were calculated as follows: where the subscripts "in" and "out" refer to the inlet concentration and outlet concentration at steady state, respectively.

XRD
In order to obtain the crystal structure information for the MnAlx catalysts, XRD measurements were carried out, for which the results are shown in Fig. 1. It can be observed that the MnAl0.1 oxide mainly exhibited the characteristic peaks of Mn2O3 (JCPDS PDF: 73-1826). The main diffraction peak located at 32.9° was ascribed to the reflection of (222) plane in Mn2O3. Furthermore, the reflection peaks at 23.1°, 55.1°, and 65.8°, corresponded to the (211), (440) and (662)  Al doping amount, which could be obviously observed in the MnAl0.5 sample. In addition, when the Al doping amount reached 0.7, the peak generated at 36.9° attributed to Mn5O8 became the main diffraction peak. These results showed that the amount of Al doping adjusted the crystal growth due to the synergistic interaction between Mn oxide and Al oxide. It should also be noted that no clear Al2O3 crystalline peaks appeared in the XRD patterns of these catalysts even with Al doping amounts as high as 0.9, indicating that the doped Al oxide may have selectively interacted with the MnOx in the mixed oxides, where the amorphous Al2O3 acted as a carrier to stabilize and disperse the Mn active sites (Li et al., 2016).

SEM Analysis
The microstructure of the representative catalyst was further evaluated using SEM. Fig. 2 shows the morphologies of the MnAl0.1. The SEM images (Figs. 2(a-c)) revealed that the MnAl0.1 oxide exhibited a regular, spherical morphology. The diameters of the spherical particles in the MnAl0.1 sample were ca. 500-1000 nm. As can be seen in EDS elemental mapping images (Fig. 2(d)) of single spherical particle, the spherical particles were mainly composed of Mn and O species. The Al elements were highly dispersed in the MnAl0.1 at the same locations, demonstrating that the Al species were incorporated into the MnOx, corresponding with the results of the XRD patterns. In addition, the highly amorphous Al species also distributed homogeneously around the microsphere, which confirmed that Al2O3 can act as a carrier to stabilize the Mn active sites.

Pore Structure and Surface Area
The N2 adsorption-desorption isotherms of the MnAl0.1, MnAl0.3, and MnAl0.5 catalysts are plotted in Fig. 3. Table 1 summarizes the corresponding results for the textural properties. As shown in Fig. 3(a), the isotherms of the catalysts were displayed as type IV isotherms with H2-type hysteresis loops based on the IUPAC classification, indicating the existence of a mesoporous structure (Arandiyan et al., 2013;Aziz et al., 2014). This was confirmed by the pore size distribution centered at 4-10 nm shown in Fig. 3(b). There was an interesting phenomenon where MnAl0.1 exhibited a specific surface area similar to that of MnAl0.5. However, their pore volumes obviously increased from 0.292 cm 3 g -1 (MnAl0.1) to 0.549 cm 3 g -1 (MnAl0.5), and the corresponding pore diameter decreased from 6.5 nm to 3.8 nm. This demonstrated that the increase in Al doping may have the contribution to generating a greater number of small-diameter mesopores in the catalysts.   In general, for gas-solid heterogeneous catalysis, a high surface area and developed pore structure can facilitate easy access of a reactant gas to the active sites and the diffusion of products from the catalyst (Yu et al., 2006), which is favorable for catalytic activity. Summarizing the analyses discussed above, the as-prepared MnAlx composite oxide showed a regular micro-sphere morphology with a mesoporous structure.

H2-TPR
Reducibility is an important factor by which to evaluate the activity of SCR catalysts. To better understand the effect of Al elements on catalytic activity, H2-TPR of the prepared MnAlx catalysts with different molar ratios was carried out. As can be seen in Fig. 4, MnAl0.1 exhibited two successive reduction peaks at 320°C and 450°C, which were believed to be associated with the two-step reductions of Mn 4+ via Mn 3+ to Mn 2+ , concurring with previous literature suggesting that the reduction of MnO2 and/or Mn2O3 occur via a well-defined two-step reduction profile (Trawczyński et al., 2005;Wan et al., 2014). It can be seen that the reduction peaks obviously shifted to lower temperature region with an increase in Al from 0.1 to 0.3. Meanwhile, the intensities of the twostep reduction peak gradually increased, which could be observed on the MnAl0.5, indicating an improvement in the redox ability. This promoted reducible nature of MnAlx was attributed to the interaction between manganese species and carriers brought about by the doping of Al cations. It should be noted that the reduction peaks shifted to a higher temperature, and the corresponding peak intensity continually decreased when the Al doping content exceeded 0.7. Specifically, MnAl0.9 exhibited a very weak peak at about 470°C. This result indicates that Al species are not active substances, where the introduction of excessive Al inhibits the synergistic interaction between manganese species and Al species, thereby weakening the reducibility.

NH3-TPD
It is known that the surface acidity of a catalyst is an important index of the SCR activity because it is responsible for the adsorption and activation of NH3 (Ali et al., 2018;Wu et al., 2021). NH3-TPD was carried out to explore the surface acid properties of the prepared catalysts. As shown in Fig. 5,  Fig. 4. H2-TPR profiles of the MnAlx catalysts. two distinct desorption processes were exhibited in the NH3 desorption profiles. The peaks at low temperatures (100−200°C) originated from weak acid sites, and the peaks at higher temperatures (> 500°C) were due to strong acid sites (Fan et al., 2017). It is widely believed that the thermal stability of NH3 adsorbed at Brønsted acid sites is lower than that of the NH3 molecules coordinated on Lewis acid sites (Kang et al., 2019). Thus, the peaks at temperatures above 300°C were attributed to desorption of NH3 coordinated at the Lewis acid sites, and the peaks at low temperatures were attributed to desorption of the NH4 + ion from the Brønsted acid sites. All of the MnAlx micro-spheres showed that the intensity of the NH3 desorption peak centered at low temperatures were much stronger than those at high temperature, indicating that Brønsted acid could be the main acid species. The Brønsted acid sites of the catalysts increased to various degrees, and the Lewis acid sites increased first and then decreased with increases in the Al doping amount. This result confirmed that Al plays a positive role in regulating acidic sites on the mixed oxide.

The Effect of Reaction Temperature
The catalytic activity of Mn-Al0.3 at different temperatures was tested. As shown in Fig. 6(a), the catalytic activity curve changed slightly over time. To obtain accurate experimental data for the stable catalyst, the activity experiment was carried out for 180 min at each corresponding reaction temperature. As shown in Fig.6(b), the DeNOx activity increased with increases in the temperature. When the bed reaction temperatures were 80°C, 100°C, 120°C, and 150°C, the corresponding NOx conversion rates were 78.1%, 89.2%, 98.2% and 99.5%, respectively. However, the N2 selectivity showed a trend toward a sudden change, for which the values were 58.1%, 69.5%, 85.9% and 56.8%, respectively, with increasing temperatures. The selective catalytic reduction efficiency of this catalyst was the highest at 120°C. Although the NOx conversion rate was as high as 99.5% at 150°C, the N2 selectivity decreased to 56.8%, which was due to the fact that the oxidation reaction of NH3 was promoted with increases in the reaction temperature, causing the DeNOx efficiency to decrease. Madia et al. (2002) suggested that at low temperatures, N2O formation is probably derived from nitroamine or ammonium nitrate intermediate species.
They showed that the thermal decomposition of the generated ammonium nitrate may form either N2O + 2H2O or HNO3 + NH3 at fast heating rates, and the absence of H2O could favor the formation of N2O. Also, Tang et al. (2010) reported that NH3 can be changed into adsorbed N species on β-MnO2, which reacted with gaseous NO to form N2O. It can be seen that the MnAl0.3 catalyst had good LT catalytic activity, but the N2 selectivity requires improvement.

The Effect of Mole Ratios
In order to determine the effects of the Al doping amount on the LT activity, the catalytic activity of the MnAlx catalysts at different molar ratios was investigated at 80°C, as shown in Fig. 7(a). Fig. 7(b) shows that increases in the Al doping amounts cause the catalytic activity to develop a wave-like trend, where the NOx conversion rates were 97.4%, 72.3%, 78.1%, 79.5% and 50.9% for MnAl0.1, MnAl0.3, MnAl0.5, MnAl0.7, and MnAl0.9, respectively, and where the selectivity of N2 was 64.2%, 58.1%, 57.9%, 72%, and 47.7%, respectively. Interestingly, the N2 selectivity also exhibited variations similar to those in the NOx conversion rate. This may have been due to the fact that although Al oxides with moderate acidity increase the number of acid sites in a catalyst, where more Mn active sites can disperse easily on Al oxides, the Al oxides were the non-catalytic active species, thereby leading to a decrease in catalytic activity. It should be noted that MnAl0.1 showed the highest NOx conversion rate among these catalysts. Combined with the previous H2-TPR and NH3-TPD results, an increase in the Al doping led to the redox performance and acidity first increasing and then decreasing. These results indicated that the active sites played a decisive role in the low-temperature SCR reaction for the obtained catalysts, and the other properties played a certain regulatory role.

The Effect of Space Velocity
In practical applications, space velocity is the ratio of the volumetric flue gas flow to the reference volume of catalysts (He et al., 2015), which is an important factor for NOx removal.   .1 catalyst at different space velocities. As the space velocity was increased from 14331 h -1 to 23885 h -1 , the DeNOx rate only dropped slightly from 95.3% to 92.1%. When the space velocity was increased to 35828 h -1 , the catalytic activity drastically dropped to 56.8%. This may have been due to the fact that as the space velocity increases, the external diffusion resistance decreases, which facilitates NOx adsorption. However, in the meantime, the residence time of the gas on the surface of the catalyst becomes shorter, and the adsorbed NOx cannot fully react with NH3, resulting in a lower DeNOx rate. Increases in the space velocity led to decreases in the N2 selectivity. When the space velocity was 14331 h -1 , 17914 h -1 , 23885 h -1 , and 35828 h -1 , the corresponding N2 selectivity was 76.13%, 64.2%, 60.5%, and 43.5%, respectively. The space velocity showed a similar trend of influence on the NOx conversion rate, supporting the conclusions discussed above.

The Effect of Particle Size
The size of the catalyst particles affects the catalytic performance, so the catalyst was sieved into particles of different sizes to study its denitration performance. Fig. 9(a) shows the NOx conversion curve as a function of time. As the catalyst particles became smaller, the NOx removal rate increased. When the catalyst particles were 20-40 mesh, 40-60 mesh, and 60-100 mesh, the corresponding removal efficiencies were 56.7%, 71.2%, 86.4%, respectively, and the N2 selectivities were 43.5%, 93.1%, and 100%, respectively. In the case of a gas-solid heterogeneous reaction, after the catalyst particles are reduced to a certain degree, due to an increase in residence time and exposure to more reaction sites, the reaction gas can be more efficiently adsorbed onto the surface active sites of the catalyst, resulting in more NOx being fully reduced by NH3 instead of directly escaping or generating by-products from an incomplete reaction. As shown in Fig. 9, a smaller the catalyst particle size led to a higher NOx conversion rate and higher N2 selectivity. In turn, in practical applications, a catalyst with a small particle size will increase the gas flow resistance and easily cause blockage of the gas path, thereby affecting the continuous, effective progress of subsequent reactions. Fig. 10 shows the relationship between the O2 content and the catalytic activity of the MnAl0.1 catalyst. It can be seen that the MnAl0.1 catalyst exhibited poor catalytic activity with a NOx conversion rate of 26.0% in the absence of O2. The NOx conversion rate changed significantly in the presence of 4% O2, and when the O2 concentration was further increased to 8%, the conversion rate increased slightly to 87.6%. With the further increase in the O2 content, the catalytic activity remained basically unchanged. In the absence of O2, the reaction had almost no N2 selectivity. As the O2 content was increased, the N2 selectivity first increased and then stabilized at about 90.1%. Based on the "Standard SCR" reaction mechanism combined with the "Fast SCR" proposed by Arnarson et al. (2017), two cycles, including an NO activation cycle and a "Fast SCR" cycle, were suggested, which shared the same reduction cycle. Thus, the oxidation reaction of NO to NO2 does not easily occur in the absence of O2, where only isolate NO could react slowly with NH3. NO2 has been proven to be a more effective oxidant than O2 and NO (Zhu et al., 2017). In the presence of O2, more NO2 is generated and participates in the reduction reaction, resulting in a "Fast SCR." However, when the O2 content is further increased to 12%, no more NO2 can be produced because the oxidation reaction is also limited by the reaction temperature. It is worth noting that the N2 selectivity increased and tended to be stable with increases in the O2 concentration, indicating that a "Fast SCR" reaction is beneficial for NOx to be more fully reduced by NH3.

In situ DRIFTS
In situ DRIFTS were carried out to further explore the SCR reaction mechanism by clarifying the adsorbed species and intermediates over the MnAl0.1 catalyst surface and the results were shown in Fig. 11.   Fig. 11. DRIFT spectra of (a) NH3 adsorption (b) NO + O2 adsorption (c) the reaction of pre-adsorbed NH3 with NO + O2 and (d) the reaction of preadsorbed NO + O2 with NH3 on the MnAl0.1 catalyst at 150°C. The reactant gases contained 500 ppm NH3, N2 balance (a, d) and a 500 ppm NO + 5% O2, N2 balance in (b, c).
The DRIFT results for NH3 adsorption on the MnAl0.1 catalyst recorded at 150°C over time are illustrated in Fig. 11(a). In the range of 1000-1650 and 3200-3700 cm -1 , several bands were observed after introducing NH3. With increases in adsorption time, these bands became more prominent, resulting from the generating of more adsorbed NH3 species. The bands at 1609 cm -1 and 1192 cm -1 were ascribed to the asymmetric and symmetric NH3 stretching mode adsorbed on the Lewis acid sites, respectively (Ramis et al., 1995;Zhan et al., 2014). The bands at 1547 and 1515 cm -1 were ascribed to the -NH2 originating from the deprotonation of ammonia via partial oxidation of NH3 (Fan et al., 2017;Ma et al., 2018), and the bands at 1429 cm -1 were ascribed to the ionic NH4 + formed on the Brønsted acid sites (Liu et al., 2012;Liu et al., 2014). The bands at 3361 and 3260 cm -1 were attributed to the N-H stretching vibration of coordinated NH3 on the Lewis acid sites, and the negative bands around 3624 were regarded as the hydroxyl consumption by NH3 (Wu et al., 2007;Ding et al., 2016). Fig. 11(b) shows the DRIFT results of NO + O2 co-adsorption on the catalyst. Two strong bands centered at 1565 and 1283 cm -1 were observed for the MnAl0.1 catalyst after the NO adsorbed, which were attributed to bridging nitrate and monodentate nitrate species (Ramis et al., 1990;Larrubia et al., 2001;Ma et al., 2018). In addition, these two nitrate species had similar peak areas, and the adsorption capacity increased with time.
The reaction between the preadsorbed NH3 and NO + O2 species on MnAl0.1 was performed at 150°C is shown in Fig. 11(c). Compared with the spectra obtained for the NH3 adsorption, similar spectra were observed while the band intensity was slightly reduced after N2 purging. After NO + O2 was introduced, the bands at 1609, 1547, and 1515 cm -1 still existed, but the peak area was further decreased. The bands at 1429 and 1192 cm -1 ascribed to NH3 adsorbed on the Lewis acid sites and the ionic NH4 + formed on the Brønsted acid sites disappeared, implying that these adsorbed species could be the significantly reactive intermediate species participating in the SCR reaction. Meanwhile, a new band at 1320 cm -1 appeared and increased over time, which was attributed to monodentate nitrate species. This indicates that NH3 first adsorbed on both the Brønsted and Lewis acid sites and subsequently reacted with gas-phase NO following an Eley-Rideal (E-R) mechanism. Fig. 11(d) shows the reaction between the preadsorbed NO + O2 species and NH3 on MnAl0.1 catalyst. After introducing NH3 into the IR cell, the bridging nitrate at 1565 cm -1 quickly disappeared, and the band at 1283 cm -1 assigned to monodentate nitrate species barely changed, meaning that the bridging nitrate was more active than the monodentate nitrate. The peak at 1600 cm -1 became more prominent, and a round peak appeared at 1192 cm -1 , which indicated that the asymmetric and symmetric stretching modes of NH3 were adsorbed on the Lewis acid sites. In addition, the N-H stretching vibrations of coordinated NH3 appeared at 3361 and 3260 cm -1 . The negative bands around 3624 cm -1 ascribed to the hydroxyl consumption were also found. These results indicated that the adsorbed NOx species were reduced by the adsorbed NH3 species to N2 and H2O following a Langmuir-Hinshelwood (L-H) mechanism.
Based on the above analysis and catalytic activity results, it was deduced that the SCR reaction could simultaneously follow the E-R and L-H mechanisms. Both NO + O2 and reductant NH3 can be adsorbed on the surface of MnAlx oxide catalysts. Gaseous NH3 can be adsorbed on the surface of MnAlx oxide catalysts or react with hydroxyl species to form NH4 + , which can react with gaseous NOx to produce N2. On the other hand, NO can react with O2 to form NO2, and the adsorbed NO2 can be further converted into various nitrate and nitrite species, which can easily react with the adsorbed NH3 species to generate N2.

CONCLUSIONS
In this study, spherical mesoporous MnAlx catalysts with different molar ratios were prepared using a simple precipitation method. Their NOx catalytic reduction performance and relevant main factors were investigated in a simulated fixed bed reactor. The MnAl0.1 catalyst exhibited the optimal low temperature DeNOx performance, where the highest catalytic efficiency was obtained at a temperature of 120°C. The MnAl catalyst had a relatively stable NOx conversion rate of more than 92% at a space velocity range of 14331 h -1 to 23885 h -1 . A smaller catalyst particle size led to a higher NOx conversion rate, with N2 selectivity as high as 100%. The appropriate oxygen concentration is beneficial to increasing NOx conversion and N2 selectivity. The NH3-SCR process carried out on the MnAlx oxide reflected both the E-R and LH mechanisms. This work should be beneficial to the further optimization of LT SCR catalystic system developed for practical applications.