Effect of Ce/Y Addition on Low-Temperature SCR Activity and SO₂ and H₂O Resistance of MnO_x/ZrO₂/MWCNTs Catalysts

Abstract

The effects of SO₂ and H₂O on the low-temperature selective catalytic reduction (SCR) activity over MnO_x/ZrO₂/MWCNTs and MnO_x/ZrO₂/MWCNTs catalysts modified by Ce or Y was studied. MnCeZr and MnYZr catalysts reached nearly 100% and 93.9% NO_x conversions at 200 °C and 240 °C, respectively. They displayed a better SO₂ tolerance, and the effect of H₂O was negligible. The structural properties of the catalysts were characterized by XRD, H₂-TPR, XPS, and FTIR before and after the reaction. The results showed that Ce could increase the mobility of the oxygen and improve the valence and the oxidizability of manganese, while the effect of Y was the opposite. This might be the main reason why the catalytic activity of MnCeZr was better than MnYZr in the presence or absence of SO₂ and H₂O. Doping Ce or Y broadened the active temperature window. Ce or Y, which existed in the catalysts with a high dispersion or at the amorphous state, preferred to react with SO₂ to form sulfate species, and protected the manganese active sites from combing with SO₂ to some extent, which coincided with the theory of ionic polarization.

1. Introduction

Selective catalytic reduction (SCR) technology is widely used to remove nitrogen oxides from flue gas and has the advantages of stability and efficiency [1,2]. Catalysts are a crucial factor in SCR technology and directly determine the efficiency of the system. However, conventional commercial catalysts (V₂O₅-WO₃(MoO₃)/TiO₂) have a relatively narrow temperature window of 300–400 °C, and they are not active at the typical temperature of flue gases (≤250 °C) in the catalytic bed located downstream of the desulfurizer and particulate removal device, and will be deactivated by SO₂ and H₂O [1,3,4]. Therefore, there is great interests in developing new catalysts which are active at relatively low temperatures and have better resistance to SO₂ and H₂O. Some transition metal oxides (Fe, V, Cr, Cu, Mn) supported on catalysts have been found with high activities at low temperatures [5,6,7,8,9]. Among them, Mn-based catalysts have been investigated in depth via several means because manganese oxides contain various kinds of labile oxygen, which is beneficial for the fulfillment of the catalytic cycle [9,10,11,12,13,14,15,16]. MnO₂ is the most active species among MnO₂, Mn₅O₈, Mn₂O₃, and Mn₃O₄ [17].

Carbon nanotubes (CNTs) have been extensively studied in low temperature SCR reactions based on their unusual physical and chemical proprieties, such as their unique electronic properties, special structure, higher surface areas, and so on [18,19]. In our previous study, Zr and Mn supported on multi-wall carbon nanotubes (MWCNTs) for NH₃-SCR reactions showed high activity at low temperatures. The modification of zirconium could increase the dispersion of MnO_x on the MWCNTs, the specific surface area, the total pore volume and the average pore size of the catalysts. The modification of zirconium could also enhance the atomic concentration of chemisorbed oxygen on the catalyst surface and promote the conversion of Mn³⁺ to Mn⁴⁺. Therefore, the amounts of the active sites on the surface and the redox ability of the catalysts were improved. Additionally, the amounts and strength of acid on the surface of the catalysts were increased. However, their SCR activity was reduced in the presence of SO₂ and H₂O. Catalysts were deactivated by the formation of (NH₄)₂SO₃ and NH₄HSO₄ which did not decompose at low temperatures and blocked the pore channels of the catalysts or the competitive adsorption with NO [20,21,22,23,24,25,26]. H₂O inhibited the SCR reaction due to its competitive adsorption with NH₃ on the Lewis sites [27,28].

According to the theory of ionic polarization [29], the formation of a metal sulfate can be regarded as the inverse process of metal sulfate decomposition (Equation (1)). If MeSO₄ is harder to decompose, MeSO₄ is more stable and it is easier for Me and SO₃ to form MeSO₄. If the decomposition temperature of MeSO₄ is higher than that of manganese sulfate, Me prefers to combine with SO₃ over Mn. The decomposition temperature of ceric sulfate (973 °C) and yttrium sulfate (976 °C) were higher than manganese sulfate (814–928 °C), which indicates that ceric sulfate and yttrium sulfate are more easily formed than manganese sulfate, and manganese sites over these catalysts can be protected. Therefore, Ce and Y were selected to improve the tolerance of catalysts caused by SO₂ and H₂O.

MeSO₄→MeO(s) + SO₃

(1)

MnO_x-CeO_x/ZrO₂/MWCNTs(MnCeZr) and MnO_x-YO_x/ZrO₂/MWCNTs(MnYZr) catalysts were prepared in this study by using the step impregnation method. The role of Ce/Y, the influences of SO₂ and H₂O over these MnO_x/ZrO₂/MWCNTs(MnZr), MnCeZr, and MnYZr, and finally the structural property and reducibility of the catalysts are discussed. X-ray powder diffraction (XRD), H₂ temperature programmed reduction (H₂-TPR), X-ray photoelectron spectroscopy (XPS), and Fourier transform infrared (FTIR) spectroscopy were used in this study.

2. Results and Discussion

2.1. SCR Performance

NH₃-SCR activity experiments over the MnZr, MnCeZr, and MnYZr catalysts were carried out, as depicted in Figure 1. The results showed that the activities of the catalysts were significantly improved and had obvious differences in NO_x conversion with the increasing temperature. Nearly 100% of NO_x conversion over the MnCeZr catalyst occurred at 200 °C, while the maximum NO_x conversions over MnZr and MnYZr were 93.6% at 200 °C and 99.3% at 240 °C, respectively. The NO_x conversions remained above 90% over MnZr, MnCeZr, and MnYZr catalysts in the temperature range of 190–220 °C, 170–270 °C, and 220–270 °C, respectively. According to the experimental data, the addition of Ce on MnZr catalysts improved the catalytic activity significantly while Y gave the opposite effect. However, they both enlarged the active temperature windows of the catalysts.

2.2. Effect of SO₂ on SCR Catalytic Activity

Figure 2 shows the effect of SO₂ on the NO_x conversion over MnZr, MnCeZr and MnYZr catalysts. The NO_x conversion decreased as SO₂ was introduced into the system. The NO_x conversion of MnZr showed a relatively faster deactivation, and decreased from 88.6% to 37.4% after 20 min in the presence of SO₂. In contrast, the NO_x conversions of MnCeZr and MnYZr decreased to 85% and 74.8%, respectively. Their NO_x conversion declined slowly and were maintained above 60% for the next 105 min, respectively. It can be deduced from Figure 2 that the poisoning of SO₂ on the SCR activity of the catalysts was most likely caused by the sulfation of the active sites. This coincides with the work of Pan [24]. The addition of Ce or Y on MnZr slowed the sulfation progress and resulted in the improvement of SO₂ resistance. Furthermore, it was evident that the catalytic activity of MnCeZr was better than that of MnYZr in the reactions. When SO₂ was absent from the feed gas, the NO_x conversion could not be recovered. This indicated that the deactivation caused by SO₂ was irreversible.

2.3. Effect of H₂O on SCR Catalytic Activity

Figure 2 shows that when 10 vol. % of H₂O was continuously introduced into the system, the NO_x conversions on MnZr, MnCeZr, and MnYZr had a significant drop in a short time and then remained stable at this level. The NO_x conversions on MnCeZr and MnYZr decreased to about 70% and 60%, respectively. When H₂O was completely removed from the feed gas, the activity could almost completely recover to its original level. H₂O caused a decline in activity and the inhibiting effect of H₂O on the SCR activity of the catalysts was reversible.

2.4. Effect of H₂O and SO₂ on SCR Catalytic Activity

Figure 2 further shows that when 0.01 vol. % of SO₂ and 10 vol. % of H₂O were introduced into the system, a sharp decline of the NO_x conversion from 98.91% to 50.24% on MnCeZr was observed, and then the NO_x conversion dropped slowly in the rest time. According to the experimental data, the coexistence of SO₂ and H₂O deactivated the catalysts more seriously than either in the presence of SO₂ or H₂O alone. Moreover, the activity was not recovered after cutting off H₂O + SO₂, implying that the coexistence of H₂O and SO₂ had a synergistic poisoning effect on the SCR activity.

2.5. XRD Results

Figure 3 displays the XRD patterns of different catalysts at wide diffraction angles. The XRD profiles of MnCeZr and MnYZr remained almost unchanged. ZrO₂ and MnO_x were the major species after the introduction of Ce or Y, respectively, which indicated that the loading of Ce or Y did not change the structure of MnZr. The only exception was that the XRD pattern of MnCeZr showed a weak peak at 47.5°, representing CeO₂ (PDF#34-0394). No other characteristic peaks of Ce and Y compounds over MnCeZr and MnYZr were observed, indicating that Ce and Y were well incorporated and dispersed in the catalysts, or that the formed compounds were not large enough to be detected.

Figure 3 displays the XRD patterns of the fresh and used catalysts with/without loading Ce or Y on MnZr. The peak that appeared at 42.7° can be assigned to the phases of MnSO₄ (PDF#35-0751) formed. With the SCR activity results (Figure 2), the newly formed crystal phases MnSO₄ on MnZr decreased the SCR activity. However, there were no detectable peaks of MnSO₄ for the used MnCeZr and MnYZr, indicating that these sulfate species may exist as amorphous species or do not reach the detection limit.

2.6. H₂-TPR Results

Figure 4 shows the H₂-TPR results of various fresh catalysts. The three catalysts all present three redox peaks in the temperature ranges of 100–200 °C, 250–400 °C, and 450–700 °C, which correspond to the reduction of the surface oxygen, MnO₂ to Mn₃O₄ and Mn₃O₄ to MnO, respectively [9,30]. The peak at 325 °C which corresponded to the reduction of MnO₂ to Mn₃O₄ played an important role in the reaction over MnZr [17,31]. It shifted to a lower temperature at 299 °C and a higher temperature at 344 °C over MnCeZr and MnYZr, respectively. Considering the SCR activity (Figure 1), this may have been due to the addition of Ce to MnZr, which improved the mobility of the oxygen as well as the valence and oxidization properties of manganese, while the addition of Y gave the opposite effect [32,33].

Figure 4 displays the H₂-TPR results of the fresh and used catalysts when SO₂ was introduced to the system. In the case of the MnZr catalysts, the peak of MnO₂ to Mn₃O₄ at 325 °C disappeared while two new peaks at 416 °C and 453 °C appeared after the reaction under SO₂ flow. Based on the activity test and characterization results, the poisoning effect of SO₂ on the catalysts was directly related to the newly formed sulfate species, which decreased the mobility of the oxygen that led to a decrease in SCR activities [34]. With the addition of Ce or Y to MnZr, the reduction peak of MnO₂ to Mn₃O₄ did not disappear, but became weaker after the reaction under SO₂ flow. It may be deduced that the addition of Ce or Y inhibited the formation of sulfate species and protected the active sites to some extent.

2.7. XPS Results

The surface atomic compositions by XPS analysis are given in

Table 1. Surface atomic concentration of catalysts before and after the reaction in the presence of SO₂.

Catalyst	Surface Atomic Concentration (%)
	C	O	Mn	Zr	Ce	Y	S	N
Fresh MnZr	88.25	8.14	1.6	2.01	-	-	-	-
Fresh MnCeZr	86.98	9.22	1.32	1.72	0.76	-	-	-
Fresh MnYZr	89.91	7.23	0.75	1.89	-	0.22	-	-
Used MnZr	88.82	7.54	1.38	1.61	-	-	0.14	0.51
Used MnCeZr	88.66	7.78	1.09	1.56	0.52	-	0.15	-
Used MnYZr	88.53	7.71	0.71	1.66	-	0.18	0.14	1.07

. It was noted that surface oxygen on the fresh MnZr catalysts was 7.96% and increased to 9.22% with the addition of Ce while the surface oxygen decreased to 7.23% with Y addition. This is in accordance with H₂-TPR results, which indicated that Ce improves the oxygen storage capacity of the catalysts while the effect of Y was opposite. Compared to the fresh catalysts, the manganese concentration in the catalysts declined by 31.3% (MnZr), 18.2% (MnCeZr), and 5.3% (MnYZr) after the reaction under SO₂ flow, respectively. The formation of sulfate species on the active sites reduced the amount of surface manganese, which was in accordance with the results of Luo’s study [34]. The results also confirmed that the sulfation of the active sites was inhibited to some extent after loading Ce or Y. Naturally, some sulfur and nitrogen atoms were detected under the XPS measurement on the surface of the used catalysts, which confirmed that sulfate species were formed on the surface of the catalysts.

The XPS spectra for the O 1s species is shown in Figure 5. The peaks at 529.8 and 531.8 eV correspond to the lattice oxygen (O_β) and the surface oxygen (O_α), respectively [35]. From the calculations, the ratio of (O_α)/(O_α + O_β) in MnZr (46.1%) was lower than that in MnCeZr (52.4%) and MnYZr (72.6%), respectively. This is owing to the fact that loading Ce or Y increased the defects of metal oxides and oxygen vacancies. [36]. Figure 6 shows the XPS spectra for Mn 2p of the fresh and used catalysts after reacting with SO₂. Compared to the fresh catalysts, the Mn 2p_1/2 and 2p_3/2 peaks of the used MnZr catalysts (Figure 6a) moved from 653.4 eV and 642.2 eV to 654.1 eV and 642.7 eV, respectively. This was due to the sulfation of the manganese, which made the Mn 2p_3/2 peak of the Mn/TiO₂ catalysts shift towards a higher binding energy (approximately 0.7 eV after SCR with SO₂), in accordance with the study of Jin [37]. When loading Ce or Y, the peaks were not significantly varied. The results also further confirmed that the sulfation of manganese was inhibited to some extent after loading Ce or Y. The Ce 3d XPS pattern (Figure S1) indicated that Ce existed as Ce⁴⁺ and Ce³⁺. The characteristic peaks u, u'', u''', v, v'', and v''' corresponded to Ce⁴⁺, and peaks u' and v' were attributed to Ce³⁺ [31]. From the calculations, the ratio of Ce⁴⁺/Ce³⁺ in the used MnCeZr catalysts was lower (from 6.87% to 3.28% after SCR with SO₂), which indicated that Ce⁴⁺ was partially reduced to Ce³⁺. The Y 3d XPS pattern (Figure S2) of MnYZr remained almost unchanged after reacting with SO₂, which may be interpreted as an indication that the valence of Y was not varied.

The sulfation of the active sites on the catalysts was the process of SO₂ combining with the surface oxygen, which was accompanied by changes in the surface oxygen. According to the results of Dupin’s study [38], the binding energy of 527.7–530.6 eV corresponded to the lattice oxygen; 530.6–531.1 eV was ascribed to OH⁻; and 531.1–532 eV was denoted as O⁻ which could supply the defects of lattice oxygen, such as the oxygen in Mn–O–Ce, Ce=O, and Mn–O–Y species. More than 532 eV was attributed to the weakly adsorbed oxygen in the surface. Surface oxygen of the fresh and used catalysts by XPS analysis is shown in Figure 7. According to the results of XPS, it was observed that the binding energy of 529.8 eV corresponding to the lattice oxygen exhibited the largest peak areas for all samples. Combined with the results of XRD indicating that there were many crystalline phases of ZrO₂, it was implied that the peak was attributed to the lattice oxygen of ZrO₂. The peaks at 531.2 eV and 532.5 eV of the MnZr catalysts were ascribed to O⁻ and the surface weakly adsorbed oxygen, respectively. Due to the sulfate ions formed after reacting with SO₂, these two peaks moved to 531.6 eV and 532.8 eV, respectively, while a new peak appeared at 530.6 eV, corresponding to OH⁻ at the same time. This was caused by the decomposition of the formed sulfate ions. This phenomenon was also observed on the catalysts after loading Ce or Y. The peak at 529.2 eV, which was ascribed to the disappearance of the oxygen of CeO₂ over the MnCeZr catalysts, indicated that a large number of CeO₂ was consumed after SCR with SO₂.

Table 2. Proportions of different O species obtained by the peak fitting of the O 1s curves.

Catalyst	O 1s 1	O 1s 2	O 1s 3	O 1s 4
MnZr	0.62	0.21	0.17	-
MnZr-SO2	0.43	0.26	0.18	0.13
MnCeZr	0.15	0.32	0.24	0.29
MnCeZr-SO2	0.29	0.20	0.28	0.23
MnYZr	0.33	0.36	0.30	-
MnYZr-SO2	0.27	0.19	0.26	0.28

shows the proportions of different O species obtained by the peak fitting of the O 1s curves. The O 1s peaks were numbered as O 1s 1, O 1s 2, O 1s 3, and O 1s 4, from low to high binding energy. According to the experimental data, the lattice oxygen and the surface oxygen showed a significant decrease after reacting with SO₂, while OH⁻ and O⁻ increased. This may be caused by the hydrolyzation of the sulfate ions.

According to the results of XRD, H₂-TPR, and XPS, it was concluded that SO₂ combined with Ce or Y easily compared with Mn on the catalysts. Ce or Y inhibited the formation of manganese sulfate and protected the manganese active sites to some extent.

2.8. FTIR Analysis

To understand the role of Ce or Y better, FTIR experiments were performed. As shown in Figure 8, the spectrum of all fresh catalysts exhibited three characteristic peaks at 3440 cm⁻¹, corresponding to the vibration of –OH units [39,40,41]; at 1645 cm⁻¹, which was ascribed to the stretching vibration of C=C on the surface of MWCNTs [42]; and a broad peak in the 1000–400 cm⁻¹ range, which was attributed to the stretching vibration of metallic oxide groups (Me–O–Me) [43,44]. Compared to the fresh catalysts, a peak around 1102 cm⁻¹, assigned to the SO₄²⁻ species, appeared after reacting with SO₂ [6,45]. From the partial magnification of corresponding FTIR spectra, the peak became weaker after loading Ce or Y, indicating that the addition of Ce or Y may prevent the manganese active sites from being sulfated to some extent [46].

3. Materials and Methods

3.1. Pretreatment of MWCNTs

The MWCNTs were purchased from Shenzhen Nanotech Port Co. Ltd. (Shenzhen, China) According to Wang et al. [9], the MWCNTs were pretreated in HNO₃ solution by ultrasound for 30 min and heated at 100 °C for 4 h in a water-bath. Subsequently, MWCNTs were treated with filtration and washed by deionized water until the pH was neutral. The last sample was dried at 100 °C overnight for further use.

3.2. Preparation of MnO_x/ZrO₂/MWCNTs

The catalysts were prepared by step impregnation methods. Zirconium was supported on multi-wall carbon nanotubes (MWCNTs) to obtain zirconium-modified MWCNTs. The precursor was zirconium nitrate solution and zirconium loading was 30 wt. %. The sample were calcinated at 400 °C in an N₂ atmosphere for 2 h and denoted as ZrO₂/MWCNTs. Subsequently, to obtain MnO_x/ZrO₂/MWCNTs, 10 wt. % Mn was loaded on ZrO₂/MWCNTs by using the same method, for which the precursor was manganese nitrate. The detailed process has been previously described in Reference [47].

3.3. Preparation of Mn-CeO_x(YO_x)/ZrO₂/MWCNTS

MnO_x/ZrO₂/MWCNTs were introduced into a beaker containing 20 mL ethanol, and then stirred magnetically for 2 h. Cerium nitrate solution, used as the Ce precursor and controlled at 6 wt. % Ce loading with MWCNTS, was gradually added to the abovementioned mixture at room temperature and stirred for 3 h. The mixture was held under ultrasound for 30 min, then dried in the oven at 60 °C for 12 h. The powder was transferred to a tube furnace and calcined at 400 °C under N₂ for 2 h to obtain a Mn-CeO_x/ZrO₂/MWCNTS composite (denoted as MnCeZr). For comparison, Mn-YO_x/ZrO₂/MWCNTS (denoted as MnYZr) was prepared using yttrium nitrate as the Y precursor and controlling to 2 wt. % Y loading with MWCNTS by a similar method.

3.4. Catalytic Evaluations

The activity evaluation was carried out in a fixed-bed quartz continuous flow reactor under atmospheric pressure. 150 mg of the MnCeZr and MnYZr catalysts were used in the tests, respectively. The feed gas was composed of 0.06% NO, 0.06% NH3, 0.01% SO₂ (when added), 10% H₂O (when added) and 3.0% O₂ balanced by Ar. The total flow rate was 600 mL/min, corresponding to a gas hourly space velocity (GHSV) of 40,000 h⁻¹. The inlet and outlet concentrations of NO/NO₂ were analyzed by an online chemiluminescence NO-NO₂-NO_x analyzer (Thermal Scientific, model 42i-HL). The activity data were recorded when the reactions reached the steady state condition at each temperature. The reaction temperature was controlled from 80 to 300 °C with an isotherm step of 20 °C. The heating rate was about 5 °C/min. NO_x conversions were obtained by the following equation:

$$\mathsf{\delta }\left({\mathrm{NO}}_x\right)=\frac{\varnothing {\left({\mathrm{NO}}_x\right)}_{\mathrm{in}}-\varnothing {{(\mathrm{NO}}_x)}_{\mathrm{out}}}{\varnothing {\left({\mathrm{NO}}_x\right)}_{\mathrm{in}}}\times 100\%$$

(2)

where δ(NO_x) represents the NO_x conversion; and ϕ(NO_x)_in and ϕ(NO_x)_out correspond to the inlet and outlet volume fraction of NO_x, respectively.

3.5. Catalyst Characterization

X-ray powder diffraction (XRD) patterns were recorded on a D8 Advace diffractometer (Bruker, Karlsruhe, Germany) with Cu Kα radiation (40 kV, 40 mA). Data were collected between 2θ = 10°–90° with 0.02° steps, and the XRD phases were identified by comparison with the reference data files from the Joint Committee on Power Diffraction Standards (JCPDS).

Hydrogen temperature-programmed reduction (H₂-TPR) of the catalyst was carried out on an Auto Chem II (Micromeritics, Gwinnett County, GA, USA). The samples were first flushed with Ar (30 mL/min) at 350 °C for 1 h. After that, the experiments were carried out at a heating rate of 10 °C/min from 60–700 °C in 10% H₂/Ar (30 mL/min). The H₂ consumption was recorded by a Thermal Conductivity Detector (TCD).

The chemical state and surface composition of the catalysts were analyzed by X-ray photoelectron spectroscopy (XPS, Axis Ultra DLD, Kratos, UK). All spectra were acquired at a basic pressure 3.33 × 10⁻⁶ Torr with Al Kα radiation (hv = 1253.6 eV) at 15 kV. The binding energy calibration was performed using the C 1s peak in the background as the reference energy (284.6 eV).

A Fourier transform infrared (FTIR) spectroscope (Nicolet 6700, Thermo Fisher Scientific, Waltham, MA, USA) was used to analyze the functional groups and determine the composition of the samples. Samples were achieved by pressing the mixture of catalysts and KBr. The weight ratio of catalysts: KBr was about 200. The FTIR measurements were performed at 4 cm⁻¹ resolution with an accumulation of 64 scans.

4. Conclusions

The addition of Ce or Y loading on the MnO_x/ZrO₂/MWCNTs catalysts were prepared using the step impregnation method. Under simulated SCR reaction conditions, Ce improved the activity of the catalysts while the effect of Y was the opposite, however, they both enlarged the active temperature windows. The inhibition effect of H₂O on SCR activity was reversible, while the poisoning of SO₂ was more serious and irreversible when Ce or Y existed in the catalysts. This study found that the catalytic activity of MnCeZr was better than MnYZr in the abovementioned reaction conditions. Furthermore, Ce could improve the H₂O resistance of the catalysts to some extent, while the effect of Y was not obvious. For the MnCeZr catalysts, the coexistence of SO₂ and H₂O deactivated the catalysts more seriously than that in the presence of SO₂ or H₂O alone, implying that the co-existence of H₂O and SO₂ had a synergistic inhibition effect on SCR activity. Based on the results of XRD, H₂-TPR, XPS, and FTIR, Ce or Y existed in the catalysts in high dispersion or amorphous form. Ce could increase the mobility of the oxygen as well as the valence and oxidization of manganese while the effect of Y was the opposite, which may have been the main reason for the better catalytic activity of MnCeZr compared with MnYZr. Ce or Y combined with SO₂ easily to form sulfate species on the catalysts compared with Mn, then inhibited the formation of sulfate species and protected the manganese active sites to some extent, which was in accordance with the theory of ionic polarization.