Effects of Synthesis Conditions on Rare Earth Doped Iron Oxide Catalyst for Selective Catalytic Reduction of NO x with NH 3

A series of rare earth-doped Fe-based oxide catalysts were prepared by co-precipitation method as Selective Catalytic Reduction (SCR) catalysts. The effects of the various rare earth species, doping amount of Sm, calcination temperature and the kind of precipitant on the deNO x activity of the catalysts were systematically investigated. The SO 2 resistance performance was tested on the optimal catalyst. The catalysts have been characterized by X-ray diffraction (XRD), The X-ray photoelectron spectra (XPS), scanning electron microscopy (SEM) and Brunner Emmet Teller (BET). The results showed that the doping of Sm significantly improves the removal efficiency of Fe-based oxides. Sm 0.075 Fe 0.925 catalyst showed the optimal deNO x performance and excellent resistance to SO 2 . At the optimal doping rate (0.075), the denitrification rate was close to 100% between 200 and 250 ° C. The calcination temperature has a significant effect on the catalyst. The order of catalytic activity for different calcination temperatures was 350 ° C ≈ 400 ° C > 450 ° C > 500 ° C. The Sm 0.075 Fe 0.925 achieved 100% the de-NOx efficiencies at calcination temperatures of 350–400 ° C. It was also found that the deNO x performance of the catalyst prepared by using NH 3 ·H 2 O as the precipitating agent was better than the catalyst prepared by using (NH 3 ) 2 CO 3 or NaOH as the precipitating agent. Normally a small amount of SO 2 would render the catalyst inactive, but the Sm 0.075 Fe 0.925 catalyst was basically regenerated after 0.05% SO 2 removal in this resistance test.


INTRODUCTION
Nitrogen oxides (NOx) are recognized as one of the main atmospheric pollutants, which are not only the main component of acid rain, but also the culprit in the formation of photochemical smog, acid rain, greenhouse effect and ozone depletion (Eigenmann et al., 2006;Wei, 2014).In terms of economic and technological efficiency, SCR with ammonia is currently the most widely used method for eliminating NOx from stationary sources (France et al., 2017;Chen et al., 2021a).The V2O5-WO3/TiO2 has been extensively used as a commercial SCR catalyst in the electric plant due to its excellent catalytic activity in the temperature range of 300-400°C (Kroecher and Elsener, 2008;Kompio et al., 2012).However, some shortcomings of this type of catalysts, such as narrow operation temperature window, biological toxicity of active component (V) and poor SO2 resistance, constrain their application (Kobayashi and Miyoshi, 2007;Liu et al., 2009).Therefore, the development of eco-friendly SCR catalyst with high efficiency and high SO2 tolerance is further imperative.
Iron is a low economic cost, and non-toxic transition metal and Fe2O3-based catalysts have attracted wide attention in NH3-SCR process because of good medium-high SCR activity, satisfactory N2 selectivity, excellent thermal stability and prominent SO2-tolerant activity at high temperatures (> 300°C), but the operating temperature window of pure FeOx is narrow (Chen et al., 2018;Han et al., 2019).In recent years, many efforts have been made to develop composite catalyst systems due to the synergetic effects.Therefore, a number of Fe-based catalysts doped with other metal components have been reported.For instance, Yang et al. (2011) prepared nanosized Fe-Mn oxide as a super catalyst for the low temperature SCR of NOx with NH3.Fan et al. (2017) rationally designed the porous nanoneedle-like MnOx-FeOx catalyst derived by heat-treating MOFs, which showed excellent NH3-SCR activity in low temperatures.Fe-Ce mixed oxide catalyst was prepared and the SCR activity at low-temperature over iron oxide was improved significantly by adding a small amount of cerium (Xiong et al., 2013).Besides, the W-modified Fe2O3 catalyst can improve the activity and N2 selectivity in a wider temperature window by tuning surface acidity and redox property (Xin et al., 2018;Wang et al., 2020a;Zhang et al., 2020).The modification of Fe2O3 catalyst with Ti species can inhibit the transformation of γ-Fe2O3 to α-Fe2O3 and weaken the oxidation ability of Fe2O3 for the oxidization of NH3 (Liu et al., 2008;Yang et al., 2013).Besides, many researchers have paid great attention to Fe-exchanged zeolites, i.e., Fe/ZSM-5 (Brandenberger et al., 2011;Ruggeri et al., 2015), Fe/BEA (Shwan et al., 2015(Shwan et al., , 2012)), Fe/SBA-15 (Li et al., 2018).
Rare earth is an important mineral, especially for China.The 5d orbits of electronic structure for most rare earth elements under normal conditions are empty, which can be used as an electron transfer station for catalytic reaction (Da Silva et al., 2016;Xue et al., 2017).Thus, rare earth elements and their oxides have high catalytic activity.Rare earth catalytic materials have been crucial in a number of industries over the years, including petrochemical catalysis, vehicle exhaust catalysis, fuel cell catalysis, catalytic combustion, hydrogen production catalysis, and others (Guo and Lu, 2007;Zhan et al., 2014b).Ce is one of the most common rare earth metals and has been wildly reported as promoter in SCR reaction due to the high oxygen storage/release capacity and superior redox (Shan et al., 2012;Tan et al., 2021).As an important rare earth metal, Sm has also been explored in SCR reaction.Meng et al. (2015Meng et al. ( , 2016) ) reported that incorporating Sm into the MnOx catalyst can significantly improve catalytic activity for NH3-SCR while also improving sulfur and H2O resistance.Sun et al. (2021) found that the incorporation of Sm enhanced adsorption and activation of NH3 and facilitated the decomposition of NH4HSO4.
As well known, the morphology, crystal structure, and nanostructure of oxide catalysts are important to improve the SCR activity, which peculiarly depends on the regulation of their synthesis conditions.In addition, there are a few reports on Sm doped FeOx catalyst.In this work, rare earth-doped Fe-based oxide catalysts were prepared by the coprecipitation method.The influence of catalyst preparation factors (various rare earth species, doping amount of Sm, calcination temperature and the kind of precipitant) on NH3-SCR activity were evaluated.In addition, the ability of catalysts for sulfur resistance was also investigated.Various characterizations were used to investigate the effect of Sm addition on the structure and physicochemical properties of FeOx.These results have important reference value for the optimization of catalyst preparation in further applications.

Preparation of Catalysts
The rare earth-doped iron oxide catalysts were prepared by the co-precipitation method.Take SmFe oxide synthesis as an example.Typically, Fe(NO3)3•9H2O and Sm(NO3)3•6H2O with different stoichiometric ratios were dissolved into deionized water to obtain a uniform solution with a mixed metal cation concentration of 0.2 mole L -1 , and then the pH of the solution was adjusted to 8.5-9.5 using alkaline solution (2 mol L -1 ) containing precipitant (NH3•H2O, (NH4)2CO3 or NaOH) under vigorous stirring.To remove anion impurities from the formed slurry, it was filtered and washed several times.The crystalline product was dried at 105°C for 12 h and calcined at different temperatures (350, 400, 450 and 500°C) for 5 h in a muffle furnace.The obtained product was denoted as SmxFe1-x (x = 0, 0.025, 0.05, 0.075, 0.1).Samples doped with different rare earths were denoted as Sm0.05Fe0.95,Nd0.05Fe0.95,Gd0.05Fe0.95 and Pr0.05Fe0.95.The pure FeOx catalyst was prepared with the same method by comparison.
A Micromeritics Tristar 3020 surface area analyzer was used to measure the surface areas and pore characterization of the catalysts at liquid N2 temperature (-196°C).The specific surface area, pore size and pore volume were calculated by the multi-point BET method and the BJH model, respectively.SEM(JSM-6700F) analysis was done on the catalysts' morphological characteristics.
Using Al-Ka radiation, the catalysts were subjected to XPS examination on a scanning X-ray microprobe (PHI5300X).As a standard, the binding energies were calibrated using the C 1s peak (BE = 284.8eV).

Activity Measurements
An 8 mm internal diameter fixed-bed quartz tube reactor was used to test the evaluation of catalytic activity.The sample (0.5 g, 40-60 mesh) was fixed with quartz wool in a reactor before being placed in an electric tubular furnace that was temperature-controlled.A simulated reactant gas mixture containing 1000 ppm NO, 1100 ppm NH3, 6% O2, 500 ppm SO2 (when used), and balance N2 with a gas hourly space velocity (GHSV) of 15,000 h -1 .When the reaction reached a stable state at various temperatures, the associated activity statistics were recorded.The TH-990S NO and NO2 analyzers were used to continuously monitor the amounts of NO and NO2.The catalytic performance of NH3-SCR reaction was presented in terms of NOx conversion calculated as follows: where the subscripts "in" and "out" represent the steady-state inlet and output concentrations, respectively.

XRD Analysis
Fig. 1 shows the XRD patterns of Fe-based catalysts with different additions of samarium oxide.All the diffraction peaks observed on FeOx can be assigned to α-Fe2O3 (Jayashree et al., 2019).After the doping of Sm on FeOx, the intensity of the Fe2O3 diffraction peaks in the catalyst gradually decreased with the increase of Sm and the diffraction peak of α-Fe2O3 disappeared when the mole ratio of Sm was up to 0.075.It is difficult to observe the diffraction peaks of samarium oxide on all the mixed samples.This phenomenon suggests that samarium oxide and iron oxide interacted vigorously.The addition of Sm affected the nucleation process of iron oxide grains and inhibited the growth of iron oxide particles, reducing the crystallinity of α-Fe2O3.This means that the iron and samarium components form a solid solution with amorphous state.
Fig. 2 shows the XRD patterns of Sm0.075Fe0.925catalysts calcined at different temperatures.For the catalysts calcined at 350°C and 400°C, no diffraction peaks of iron oxide and samarium oxide were found.It can be seen that the diffraction peaks of Fe2O3 can be observed when the calcination temperature rises to 450°C.The diffraction peak of Fe2O3 at 500°C was slightly wider than it was at 450°C, indicating that the higher calcination temperature caused the Sm0.075Fe0.925catalyst to generate more Fe2O3 grain crystals.Furthermore, the interaction between samarium oxide and iron oxide was weakened, affecting the formation of iron and samarium solid solution.

Porosity and Surface Area Analysis
Table 1 shows the results of the specific surface area and pore volume of iron-based catalysts with different Sm amounts calcined at 400°C.It can be seen that the doping of Sm significantly   Intensity (a.u) increased the specific surface area and pore volume of the iron-based catalyst.The specific surface area of pure FeOx catalyst was only 39 m 2 g -1 and with the increase of doping amount, the surface area of the catalyst Sm0.1Fe0.9increased to 170 m 2 g -1 .In addition, the pore volume of the catalyst gradually increased as the increases of Sm amount.Simultaneously, more mesoporous pores with smaller pore diameter generated, which could be speculated that the substitution of smaller Fe by Sm with a larger cation radius reduced the pore diameters of the catalysts.To our best knowledge, high surface area and pore volume would greatly benefit the activity of catalysts, due to offering more active sites, efficient channels for mass transport and short diffusion length of reaction gases (Zhan et al., 2014a;Parlett et al., 2013).

SEM
In order to investigate the effect of different precipitating agents on the morphology and particle size of the SmFe composite oxide, SEM was carried out on the Sm0.075Fe0.925catalyst prepared with NH3•H2O and (NH4)2CO3 as the precipitating agent.It can be seen from the comparison of the low-magnification images in Figs.3(A) and 3(B) that the catalyst presented large agglomeration, and the surface of the block was relatively smooth.As shown in the high-magnification images of Figs.3(C) and 3(D) that the catalyst particles prepared with different precipitating agents were all composed of nanoparticles, and the catalyst particles prepared with NH3•H2O were smaller.In general, the dispersion of the catalyst prepared by the two precipitating agents was not uniform, the catalyst prepared with NH3•H2O had smaller particle size and lighter agglomeration, which may be related to the nucleation rate of the precipitate is faster compared to that prepared with (NH4)2CO3.

XPS
Considering the influence of valence state of catalyst surface element on the deNOx activity, XPS of FeOx and Sm0.075Fe0.925 was performed as shown in Fig. 4. Fig. 4(A) shows the Fe 2p spectra.Three prominent peaks over FeOx catalyst were observed in the binding energy range of 705-735 eV.The peak at 709.0 eV could be attributed to the binding energy of Fe 2+ two and the peaks at 711.6 and 724.0 eV are the binding energy of Fe 3+ (France et al., 2017;Yang et al., 2011), indicating that Fe 2+ and Fe 3+ coexisted on the surface of FeOx catalyst.While the less content of Fe 2+ resulted  in the failure to observe divalent iron oxides in the XRD pattern (Fig. 1).After the addition of Sm, the Fe 2p XPS Gaussian fitting curve of Sm0.075Fe0.925 was similar to that of FeOx, indicating that there were Fe 2+ and Fe 3+ on the surface.The binding energy for Sm0.075Fe0.925shifted to lower values of about 0.4 eV, indicating that the addition of Sm increased the density of the electron cloud around Fe atoms, and there is a strong interaction between Sm and Fe.The O1s spectra of the catalyst are shown in Fig. 4(B).Three Gaussian simulation peaks were observed at 527.4, 530.2 and 532.2 eV on FeOx.The peak at 527.4 and 530.2 eV were attributed to the lattice oxygen and the peak at 532.2 eV was assigned to surface adsorbed oxygen (Kang et al., 2007;Parlett et al., 2013).Compared with FeOx, the chemically adsorbed oxygen at of 532.2 eV disappeared over Sm0.075Fe0.92,indicating that the addition of Sm reduced the content of chemically adsorbed oxygen on the surface.Fig. 4(C) shows the Sm3d XPS curve of Sm0.075Fe0.925.The peak at 1081.2 eV corresponds to the 3d5/2 structure of Sm 2+ , and the peak at about 1110.0 eV is the 3d3/2 peak of Sm 3+ (Sone et al., 2015;Liu et al., 2019).The weaker accompanying peak at 1097.4 eV further proved the existence of two valence states of Sm.However, the Sm 3+ at 1084.8 eV moved about 1.6 eV to higher binding energy than the standard spectrum Sm 3+ (1083.2eV), indicating that the electron density around Sm atoms was decreased.This phenomenon was mutually corroborated by the increase in electron density around Fe atoms in the catalyst.This finding was related to the increase of the surrounding electron density around Fe atoms in the catalyst and further confirmed the existing strong interaction between samarium and iron.3.5 Catalytic Performance 3.5.1 Effect of different rare earth doping on SCR activity Fig. 5 shows the effect of different rare earth-doped iron-based oxides on the NOx conversion rate.As shown in Fig. 5, for pure FeO, the NOx removal rate was only 27.2% at 100°C, reached a maximum NOx conversion rate of 87.4% at 250°C, and then rapidly decreased.After adding four different rare earth oxides, the similar conversion curves can be seen.The NOx conversion rate was about 50.0% at 100°C and the maximum removal rate was nearly 100% between 200-250°C.The doping of rare earth oxides significantly increased NOx conversion rate over iron-based oxide and broadened the whole temperature window, which could be due to the existing synergistic interaction between doping and Fe species for the bi-metal oxides.In detail, the addition of rare earth oxides increased the specific surface area of the catalyst, which is conducive to the exposure of more active sites and increases the oxygen storage capacity (Qi and Yang, 2003;Tang et al., 2006).Thus, more reactive gases can be quickly adsorbed on the catalyst surface and fully reacted with more active sites exposed, thereby improving the deNOx activity of the catalysts.

Effect of Sm amount on SCR activity
Fig. 6 shows the effect of the amount of Sm addition on the NOx conversion rate of the ironbased catalyst.The doping in rare earth Sm oxide had a significant effect of improving the activity of the catalyst.Compare to the catalyst FeOx, the NOx conversion rate of the catalyst added a small amount of samarium oxide was significantly improved.The deNOx performance of the catalyst first increased and subsequently dropped as the ratio of doped Sm grew steadily from 0 to 0.1; the ideal doping ratio was 0.075.As the increase of Sm amount, the specific surface area and pore volume of the catalyst gradually increased (Table 1) and the optimum molar ratio of Sm doping with the best activity is 0.075.At the optimal doping rate, the deNOx rate is close to 100% between 200 and 250°C.Compared with the NO conversion (90% and 92%) of the same type of catalyst (Mn@ZIF-8 and MnOx) in recent years, the conversion is higher and the reaction temperature is lower (Chen et al., 2021b;Zhu et al., 2021).This means structural properties are not single decisive factors affecting catalytic activity, and the composition of the catalyst would also affect the catalytic activity due to the interaction of Sm and Fe species.

Effect of different precipitants on SCR activity
The effect of different precipitants (NH3•H2O, (NH4)2CO3 and NaOH) on SCR activity over Sm0.075Fe0.925 was investigated.As shown in Fig. 7, in the temperature range of 150-250°C,  the catalyst produced with NH3•H2O as precipitant had the highest activity with a maximum NOx conversion exceeding 93%.There was no significant difference in the activity below 250°C for the catalyst prepared with the (NH4)2CO3 and NaOH as the precipitating agent.In addition, at 300°C the catalyst prepared with (NH4)2CO3 exhibited a slightly better NOx conversion (66.7%) than the catalyst prepared with the other two precipitants, which may be due to the difference in the particle size and dispersity of catalyst.From the SEM analysis in Fig. 3, the catalyst prepared with NH3•H2O as the precipitant had a small particle size, loose structure and good particle dispersion, meaning that particle size and dispersity could affect the SCR activity.

Effect of calcination temperature on SCR activity
Fig. 8 shows the effect of calcination temperature on the catalytic activity of Sm0.075Fe0.925catalyst.The order of catalytic activity at different calcination temperatures is 350°C ≈ 400°C > 450°C > 500°C, the Sm0.075Fe0.925achieved 100% the de-NOx efficiencies at calcination temperatures of 350-400°C.The catalyst's catalytic activity rapidly reduced as the calcination temperature rose.This might be because differing calcination temperatures resulted in the presence of active substances on the catalyst surface in various forms.According to the XRD pattern of the Sm0.075Fe0.925catalyst that was calcined at various temperatures (Fig. 2), the diffraction peaks of Fe2O3 were not visible at 350°C and 400°C whereas they were present at 450°C.It should be noted when the calcination temperature raised to 500°C, the diffraction peaks of Fe2O3 were more obvious.A negative correlation was observed between the crystalline phase strength of Fe2O3 on the surface of the catalysts and their catalytic activity.Based on the previous reports, the low crystallization or amorphous phase for composite oxides catalysts is responsible for SCR activity due to short-range ordered structure in the location of active sites (Li et al., 2012;Xin et al., 2018).Thus, a reasonable explanation can be drawn that as the calcination temperature increased, finer Fe2O3 phases were precipitated on the catalysts, which decreased the amount of solid solution formed by Fe and Sm components and weakened the interaction between samarium oxide and iron oxide, reducing the activity of the catalysts.

Effect of SO2 on SCR activity
A tiny quantity of SO2 is typically present in the flue gas during industrial application, which causes the catalyst to become inactive.So, at 200°C, the impact of SO2 on the SCR activity of Sm0.025Fe0.975 was examined.As shown in Fig. 9, when 0.05% SO2 was added to the reaction gases, the NOx conversion slowly decreased and then stabilized.The NOx conversion decreased from 99% of the initial value to about 86% then recovered to about 94% after SO2 was removed about 1 h, which was slightly lower than the initial value, indicating that the catalyst was partially deactivated.According to the literature (Wang et al., 2020b), this might be a result of the catalyst developing ammonium sulfate or ammonium bisulfate on its surface, which clogged the pores and took up the catalyst's active sites, decreasing its activity.

CONCLUSIONS
(1) The doping of rare earth oxides significantly improved the iron-based oxides with similar NOx removal efficiency.(2) With the increase of rare earth Sm, the deNOx efficiency of the catalyst first increased and then decreased.Sm0.075Fe0.925Oxcatalyst exhibited the highest activity, resulting from the improved structural properties and component interaction.(3) The catalyst activity was significantly impacted by the calcination temperature.The order of catalytic activity is 350°C ≈ 400°C > 450°C > 500°C.This is due to the decreased solid solution formed by Fe and Sm components and weakened interaction between samarium oxide and iron oxide, reducing the activity of the catalysts.(4) The catalyst prepared by using NH3•H2O as the precipitating agent exhibited better activity than that of the catalyst prepared by (NH4)2CO3 and NaOH.(5) When adding 0.05% SO2, the deNOx efficiency was decreased to 86% from 98% and recovered to 93% after SO2 was removed, indicating the Sm0.075Fe0.925catalyst can be basically regenerated.

Fig. 5 .
Fig. 5. Effect of different rare earth-doping (Sm, Nd, Gd and Pr) on the NH3-SCR activity of iron oxide catalysts.

Fig. 6 .
Fig. 6.Effect of the amount of Sm on the NH3-SCR activity of iron oxide catalyst.

Table 1 .
BET surface area, and pore diameter/volume of samples calcined at 400°C.