Improving the Performance of Gd Addition on Catalytic Activity and SO₂ Resistance over MnO_x/ZSM-5 Catalysts for Low-Temperature NH₃-SCR

Abstract

SO₂ poisoning is a great challenge for the practical application of Mn-based catalysts in low-temperature selective catalytic reduction (SCR) reactions of NO_x with NH₃. A series of Gadolinium (Gd)-modified MnO_x/ZSM-5 catalysts were synthesized via a citric acid–ethanol dispersion method and evaluated by low-temperature NH₃-SCR. Among them, the GdMn/Z-0.3 catalyst with the molar ratio of Gd/Mn of 0.3 presented the highest catalytic activity, in which a 100% NO conversion could be obtained in the temperature range of 120–240 °C. Furthermore, GdMn/Z-0.3 exhibited good SO₂ resistance compared with Mn/Z in the presence of 100 ppm SO₂. The results of Brunauer–Emmett–Teller (BET), X-ray photoelectron spectroscopy (XPS), temperature-programmed reduction of H₂ (H₂-TPR) and temperature-programmed desorption of NH₃ (NH₃-TPD) illustrated that such catalytic performance was mainly caused by large surface area, abundant Mn⁴⁺ and surface chemisorbed oxygen species, strong reducibility and the suitable acidity of the catalyst. The in situ diffuse reflectance infrared Fourier transform spectra (DRIFTS) results revealed that the addition of Gd greatly inhibited the reaction between the SO₂ and MnO_x active sites to form bulk manganese sulfate, thus contributing to high SO₂ resistance. Moreover, in situ DRIFTS experiments also shed light on the mechanism of low-temperature SCR reactions over Mn/Z and GdMn/Z-0.3, which both followed the Langmuir–Hinshelwood (L–H) and Eley–Rideal (E–R) mechanism.

1. Introduction

Nitrogen oxides (i.e., NO_x, mainly in the forms of NO and NO₂) emitted from stationary sources have caused serious environmental problems, including photochemical smog, acid rain, ozone depletion and the greenhouse effect, thus they are a kind of extremely harmful air pollutants to human beings [1,2,3,4]. Selective catalytic reduction (SCR), using ammonia as a reductant, is considered to be one of the most prospective technologies in the abatement of NO_x from stationary sources, which is attributed to the core of catalysts with economic and technological efficiency. Although the commercial catalysts (V₂O₅-WO₃(MoO₃)/TiO₂) in traditional SCR techniques made great contributions in the removal of NO_x, they suffered extreme challenges, such as a narrow temperature window (300–400 °C), the biological toxicity (from vanadium) of the active component, the over-oxidation of SO₂ or NH₃, the deactivation in the high concentration of dust and SO₂ poisoning [5,6]. Therefore, some vanadium-free catalysts with high activity and good stability, as well as SO₂-resistance ability for NO_x removal in the low-temperature SCR (<250 °C) process should be urgently explored.

In the past few decades, plenty of transition metal oxides (FeO_x, MnO_x, CuO_x and CoO_x, etc.) supported on different carriers were widely studied for the SCR of NO_x [1,2,3]. Among them, Mn-based catalysts are universally considered as suitable alternatives for substituting vanadium-based SCR catalysts due to the properties of variable valence and redox [7], such as MnO_x/SAPO-34 [8], MnO_x/ZSM-5 [9], MnO_x/TiO₂ [10,11], MnO_x/Al₂O₃ [12] and MnO_x loaded on carbon nanotubes [13]. Particularly, owing to the high surface area, abundant acid sites and the thermal stability, ZSM-5 zeolite has attracted great attention in Mn-based SCR reactions [14,15,16]. Specifically, ZSM-5 can facilitate the adsorption of NH₃ and activate the adsorbed NH₃ to intermediate species in SCR reactions, which is beneficial to the NH₃-SCR [17]. Besides, both the large surface area and abundant porous structure contribute to the high dispersion of active components. For example, Shi et al. reported that the Fe-ZSM-5 catalyst exhibited the superior performance, i.e., widened active temperature window, good SO₂ resistance and hydrothermal stability, which is ascribed to the good dispersion of active iron species on ZSM-5 [18]. Therefore, ZSM-5 is considered as a promising support candidate for enhancing SCR performance. However, SO₂ poisoning has been a critical issue for the Mn-based SCR catalysts, which limited their practical application and greatly drew our interest [19]. Research on promoting the SO₂ resistance of catalysts can be divided into three aspects, including reducing the adsorption of SO₂, blocking the oxidization of SO₂ to SO₃ and lowering the deposition of active metal sulfate or ammonium sulfate [20]. Currently, the introduction of many rare earth metals into Mn-based catalysts exhibits positive effects on improving SO₂ resistance, which can be attributed to the strong interaction between the additives and active components to suppress SO₂ poisoning [21,22,23,24]. You et al. reported that Ce-modified MnO_x supported by graphene presented good activity and high-resistance ability to SO₂ and H₂O [25]. Liu et al. found that the introduction of Eu into a Mn/TiO₂ catalyst was advantageous to prevent the formation of sulfate species, and thus contributed to an excellent SO₂ tolerance for the SCR reaction process [26]. According to the ion polarization theory, Yu et al. revealed that the introduction of Pr into MnO_x/SAPO-34 could enhance SO₂ resistance, which made the sample difficult to be sulfated and restrained the generation of sulfates compared with Ce-doping [27]. Nevertheless, the enhanced mechanism of Mn-based catalysts modified by the different additives on catalytic activity and SO₂ resistance is not very clear, and few studies have been done on the mechanism and the promotion of SO₂ resistance in low-temperature SCR reactions over Mn-based catalysts modified by rare earth metals. Therefore, it is worth exploring further.

Inspired by the superior properties of rare earth metals, we were very interested in investigating the effect of their addition on activity and SO₂ resistance of low-temperature SCR catalysts. So far numerous studies have been developed to improve the catalytic performance of the catalysts, but explorations of the addition of Gd into MnO_x/ZSM-5 catalysts (Gd-MnO_x/ZSM-5) are still rare [28]. Herein, MnO_x/ZSM-5 and Gd-MnO_x/ZSM-5 catalysts with different molar ratios of Gd/Mn were synthesized and used for NO removal in the presence of SO₂ in low-temperature NH₃-SCR. In comparison with MnO_x/ZSM-5, the addition of an appropriate amount of Gd (Gd/Mn = 0.3) could not only improve the activity, but also significantly promote the SO₂ tolerance of the sample. Meanwhile, the relationship between material properties and catalytic performance was studied through a series of characterizations. Moreover, the key role of Gd in catalytic activity and SO₂ resistance was further investigated by in situ diffuse reflectance infrared Fourier transform spectra (DRIFTS) experiments, providing powerful support for further improvement and optimization of catalysts in the future.

2. Results

2.1. Low-Temperature NH₃-SCR Performance

The SCR activity of GdMn/Z-x (Gd-modified MnO_x/ZSM-5 denoted as GdMn/Z-x) catalysts was evaluated for NO conversion and the results are shown in Figure 1. All the synthesized catalysts presented excellent activity in low-temperature NH₃-SCR. Notably, the NO conversion over the GdMn/Z-0.3 catalyst was higher than that of Mn/Z (MnO_x/ZSM-5 denoted as Mn/Z) in the range of 80–140 °C, and nearly 100% conversion was obtained at as low as 120 °C. In particular, the catalytic activity of the GdMn/Z-0.3 was competitive or even better than previously reported catalysts under similar condition [29,30], as 90% NO conversion was obtained below 120 °C in this work. The NO conversion decreased with the increase of Gd/Mn, and the worst performance was achieved at 0.6, because the main active sites were covered by the excess GdO_x particles on the surface of the catalyst [31]. Moreover, the N₂ selectivity (Figure 1b) and stability tests (Figure S1) were also performed to evaluate the practical value of the GdMn/Z-0.3 catalyst. The results displayed that GdMn/Z-0.3 consistently presented above 96% N₂ selectivity in low temperatures, and 97%–99% NO conversion was retained in 24 h at 180 °C. It could be concluded that among all the as-prepared catalysts for the NH₃-SCR of NO, the GdMn/Z-0.3 catalyst exhibited the optimal catalytic performance with a broadening active temperature window and high N₂ selectivity, as well as good stability.

2.2. Physicochemical Properties of the Fresh Catalysts

2.2.1. X-ray Diffraction (XRD) and Brunauer–Emmett–Teller (BET) Analyses

The characteristic diffraction peaks of all the as-prepared catalysts are shown in Figure 2, and all these peaks corresponded well to ZSM-5 zeolite with the standard card of PDF#44-0638. Notably, no obvious diffraction peaks of GdO_x or MnO_x could be observed, suggesting that the components might be highly dispersed or in the amorphous state [15,32]. Moreover, with the increase of Gd/Mn, the diffraction intensities in the XRD patterns of the catalysts became weakened and followed the order: Mn/Z > GdMn/Z-0.1 > GdMn/Z-0.3 > GdMn/Z-0.6. These results revealed that the addition of Gd could inhibit the formation of large MnO_x bulks to some extent and cause the poor crystallization of the Mn/Z catalyst [33]. The textural properties of the samples are summarized in

Table 1. Texture properties of the catalysts.

Catalysts	Specific Surface Area (m2g−1)	Total Pore Volume (cm3g−1)	Average Pore Size (nm)
Mn/Z	266.8	0.11	2.98
GdMn/Z-0.1	241.6	0.20	3.27
GdMn/Z-0.3	273.4	0.25	3.47
GdMn/Z-0.6	259.2	0.19	3.40

. It has been universally acknowledged that the high specific surface area greatly contributes to the adsorption of reactants, further resulting in high activity by offering more active sites [34]. The GdMn/Z-0.3 catalyst had the largest specific surface area of 273.4 m²/g, corresponding to the optimal low-temperature NH₃-SCR activity (Figure 1). It was interesting that all the catalysts possessed the mesopores structure, indicating that the inherent micropores of ZSM-5 disappeared. This might be due to the fact that after the loading process, the surface structure could be reconstructed. The results revealed that the Mn/Z catalyst modified by a certain amount of Gd (Gd/Mn = 0.3) could increase the specific surface area and pore volume as well as average pore size, which was advantageous to improve the catalytic performance in low-temperature SCR reactions.

2.2.2. Microstructure Analysis

Scanning electron microscopy (SEM) images of the catalysts are shown in Figure S2. It can be seen that all catalysts mainly exhibited the typical morphology of the ZSM-5 zeolite. With the addition of Gd, the morphology of the Mn/Z (Figure S2a) catalyst was changed significantly. The original hexahedral block structure of Mn/Z was decomposed into a polyhedral block structure with different sizes (Figure S2c,e,g), while the smaller hexahedral block morphology could still be observed in all three Gd-containing samples, which might be related to the decrease of crystallinity, as observed in the XRD patterns (Figure 2). In order to further characterize the morphology and microstructure of the active component loaded on the support of Mn/Z and optimal GdMn/Z-0.3 catalysts, transmission electron microscopy (TEM) analysis was performed, and the results are displayed in Figure 3. It was apparent that many small component particles were well dispersed on the surface or edge of the support for Mn/Z (Figure 3a,b) and GdMn/Z-0.3 (Figure 3d,e). In contrast, the particles of active components on Mn/Z were bigger and more dispersed than those on GdMn/Z-0.3. It can be clearly seen from the high-resolution TEM (HR-TEM) image of Mn/Z (Figure 3c) that the lattice fringe of Mn/Z and the corresponding lattice space were 0.24, 0.31, 0.33 and 0.49 nm, which could be ascribed to the (101) plane, (111) plane of MnO₂, (220) plane of Mn₂O₃ and (101) facet of Mn₃O₄, respectively [35,36,37], confirming the presence of MnO₂, Mn₂O₃ and Mn₃O₄. As for the GdMn/Z-0.3 (Figure 3f), there were more amorphous and fuzzy areas, except for a small area where the lattice fringe corresponding to the (101) facet of MnO₂ was clear. Additionally, the energy dispersive spectroscopy (EDS) elemental mappings over the GdMn/Z-0.3 catalyst are also shown in Figure 3h–k. The results confirmed that Gd and Mn species were dispersed uniformly on the surface of the support. Moreover, it was possible that Gd tended to enrich on the surface and decrease the particle size of MnO_x.

2.2.3. X-ray Photoelectron Spectroscopy (XPS) Analysis

The surface atomic information for the as-prepared catalysts were further analyzed by XPS. Figure 4 exhibits the XPS spectra of Mn 2p and O 1s. The Mn 2p mainly had two peaks of Mn 2p_1/2 (649–657 eV) and Mn 2p_3/2 (636–646 eV), and the Mn 2p_3/2 peak could be deconvoluted to three valence state peaks in Figure 4a, which were assigned to Mn²⁺ (641 ± 0.5 eV), Mn³⁺ (642 ± 0.5 eV) and Mn⁴⁺ (643 ± 0.5 eV), respectively [38,39]. The relative concentration of Mn²⁺/Mnⁿ⁺, Mn³⁺/Mnⁿ⁺ and Mn⁴⁺/Mnⁿ⁺ calculated from the XPS spectra are listed in

Table 2. XPS results for the surface atomic composition and concentrations of the catalysts.

Samples	Surface Atomic Concentration (%)				Relative Concentration Ratios (%)
	Mn	O	Gd	others a	Mn4+	Mn3+	Mn2+	OS	OL
Mn/Z	5.09	77.19	—	17.72	28.5	35.7	35.8	72.8	27.2
GdMn/Z-0.1	4.47	64.59	1.24	29.70	27.2	34.6	38.1	71.9	28.1
GdMn/Z-0.3	4.90	66.46	3.22	25.42	34.1	42.2	23.7	74.4	25.6
GdMn/Z-0.6	4.08	60.97	5.12	29.83	31.9	36.8	31.3	69.8	30.2

Other ^a elements include Si, Al and C.

. Obviously, GdMn/Z-0.3 possessed the highest relative ratios of Mn⁴⁺ and Mn³⁺ species. It has been known that Mn⁴⁺ is the most active species to accelerate the fast SCR reaction, owing to the oxidation of NO to NO₂, which contributes to the excellent activity in low-temperature SCR [24,40]. The highest concentration of Mn⁴⁺ in GdMn/Z-0.3 could be invoked to explain its superior activity compared to the other catalysts with the proper amount of Gd addition. In Figure 4b, two types of surface oxygen were observed in the XPS spectra of O 1s on the catalysts. The low-binding-energy peaks at 529.5 ± 0.5 eV could be attributed to the lattice oxygen O²⁻ (O_L), while the high-binding-energy peaks at 532.0 ± 0.5 eV could be attributed to the surface chemisorbed oxygen originating from defect oxide and hydroxyl-like groups, such as O₂²⁻ and O⁻ (O_S), respectively [41]. Generally, the surface chemisorbed oxygen species were highly active species compared with lattice oxygen [31]. GdMn/Z-0.3 exhibited a relatively high ratio of Os/(O_L + O_S) among all the catalysts and further resulted in a fast SCR reaction, which was beneficial for the low-temperature SCR activity [42]. Consequently, the excellent SCR activity of GdMn/Z-0.3 was ascribed to the high concentration of active Mn⁴⁺ and abundant surface chemisorbed oxygen species affected by the addition of Gd.

2.2.4. Temperature-Programmed Reduction of H₂ (H₂-TPR) and Temperature-Programmed Desorption of NH₃ (NH₃-TPD) Analyses

The redox properties of the catalysts were characterized by H₂-TPR experiments, and the results are shown in Figure 5a and

Table 3. H₂-TPR and NH₃-TPD data of the catalysts.

Catalysts	Weak Acid (mmol·g−1)	Medium-Strong Acid (mmol·g−1)	Total Acid Amount (mmol·g−1)	H2-TPR Curve Area
Mn/Z	0.605	0.304	0.909	0.31435
GdMn/Z-0.1	0.789	0.465	1.254	0.25803
GdMn/Z-0.3	0.702	0.542	1.244	0.69167
GdMn/Z-0.6	0.638	0.588	1.226	0.40507

. The area of the reduction peak was correlated to the consumption of H₂, and the results (Figure 5a) suggested that the GdMn/Z-0.3 catalyst possessed high-reduction capability. Both the Mn/Z and GdMn/Z-x catalysts mainly exhibited one wide reduction peak in the temperature range of 260–420 °C, which could be ascribed to the two main reduction steps of MnO₂ and Mn₂O₃ to Mn₃O₄ [9,43]. Compared with the Mn/Z catalyst, the reduction peak of GdMn/Z-0.3 obviously shifted to a lower temperature, corresponding to the excellent low-temperature SCR activity. Meanwhile, combined with the results of XPS, the rich and high-valence state of Mn species, such as Mn⁴⁺ species, could play a significant role in the redox reaction on the GdMn/Z-0.3 catalyst surface because of the strong interaction between MnO_x and GdO_x [25,44]. Therefore, a certain amount of Gd could improve the redox ability of Mn/Z catalysts.

The surface acidity of a catalyst is an essential factor for NH₃ adsorption, thus contributing to the high activity of the catalyst in the SCR reaction process [20]. The amount of acid sites and the acid strength of all the catalysts were investigated by NH₃-TPD, and the results are shown in Figure 5b. It can be seen that NH₃-TPD profiles of the catalysts presented two wide desorption peaks. The desorption peaks in the low-temperature range of 100–300 °C could be assigned to the adsorption of NH₃ on weak acid sites, while the desorption peaks in the high-temperature range of 350–650 °C could be ascribed to the medium-strong acid sites [45,46]. As shown in Table 3, the total acid amount of the catalysts was ranked by Mn/Z < GdMn/Z-0.6 < GdMn/Z-0.3 < GdMn/Z-0.1. Both the weak acid and medium-strong acid of Mn/Z could be improved by the introduction of Gd. Interestingly, although the total acid amount of the GdMn/Z-0.3 catalyst was not the maximum, it still exhibited excellent activity. The results indicated that a suitable amount of weak acid and medium-strong acid were conducive to the adsorption and activation of NH₃, which could enhance the catalytic performance of catalysts in NH₃-SCR reactions.

2.3. In Situ DRIFTS Studies for NH₃-SCR

2.3.1. Adsorption of NH₃

Figure 6 exhibits the in situ DRIFT spectra of the Mn/Z and GdMn/Z-0.3 catalysts during the reaction of 800 ppm NH₃ at 180 °C as a function of time. Several bands were presented at 3660, 3605, 3369, 3279, 3181, 1729, 1606, 1579, 1508, 1433, 1299, 1185 and 1040 cm⁻¹ for NH₃ adsorption over the Mn/Z catalyst, as shown in Figure 6a. On the basis of the previous literature [47,48], the negative bands at 3660 and 3605 cm^-1 were in the O–H stretching vibration modes, owing to the interaction of the surface hydroxyls with NH₃. The bands at 3369–3181 cm⁻¹ were derived from the N–H stretching vibration modes of coordinated NH₃ on Lewis acid sites. The features at 1729 and 1433 cm⁻¹ could be due to the asymmetric and symmetric bending vibrations of NH₄⁺ species on Brønsted acid sites, respectively [24,49]. The bands at 1606, 1579, 1299, 1185 and 1040 cm⁻¹ could be attributed to the coordinated NH₃ on Lewis acid sites [49,50,51], while the band at 1508 cm⁻¹ could be ascribed to the scissoring vibration mode of amide (−NH₂) species [50]. It could be seen that all these bands were strengthened along with the increase of NH₃ feeding time, and reached a maximum value in 15 min. Similarly, NH₃ adsorption on Brønsted acid sites (1737, 1468 and 1436 cm⁻¹) and Lewis acid sites (3356–3187, 1605, 1568, 1273, 1181 and 1045 cm⁻¹) were also observed on the GdMn/Z-0.3 catalyst (Figure 6b). All these NH₃ adsorption bands reached a maximum value at around 15 min. Differently, the bands’ intensities of NH₃ adsorption on the GdMn/Z-0.3 catalyst were slightly stronger than those on the Mn/Z catalyst, demonstrating that the introduction of Gd into the Mn/Z catalyst enhanced the surface acidity of the Mn/Z catalyst, which was also evidenced by the NH₃-TPD results.

2.3.2. Co-Adsorption of NO and O₂

In order to investigate the NO_x adsorption states on the as-prepared catalysts, in situ DRIFTS of the Mn/Z and GdMn/Z-0.3 catalysts during 800 ppm NO and 5% O₂ at 180 °C were performed. As displayed in Figure 7, the NO_x absorption bands emerged immediately after feeding the NO and O₂ at 180 °C for 1 min and strengthened with the continuous introduction of NO and O₂. On the surface of Mn/Z (Figure 7a), the adsorption band of gaseous or weakly adsorbed NO (1885 cm⁻¹) was observed [52]. In the typical scope of surface nitrates species (1500–1700 cm⁻¹), the bands at 1686, 1637, 1594 and 1523 cm⁻¹ could be assigned to the nitro species, adsorbed NO₂ molecules, bridged nitrate and bridged monodentate nitrate, respectively [47,52,53]. While the peaks centered at 1348, 1228 and 1085 cm⁻¹ were separately ascribed to the free nitrate, bridged nitrate and linear nitrite species [54,55]. For the GdMn/Z-0.3 catalyst, the linear nitrite (1098 cm⁻¹), bridged nitrate (1282 cm⁻¹), monodentate nitrate (1546 cm⁻¹) and gaseous NO₂ molecules (1629 cm⁻¹) were observed at 180 °C (Figure 7b) [52,53,54,55]. The bands’ intensities belonging to the adsorbed NO_x species over the GdMn/Z-0.3 catalyst were much stronger than those on the Mn/Z sample, which might be due to the strong redox on the GdMn/Z-0.3 catalyst, facilitating the NO oxidation and NO_x adsorption.

2.3.3. Reaction between NO and O₂ and Pre-Adsorbed NH₃

In consideration of the distinct surface acidity and NO_x adsorption capacity of the Mn/Z and GdMn/Z-0.3 catalysts, the surface intermediates and the reaction pathways on the two catalysts might potentially differ during the NH₃-SCR process. In situ DRIFTS experiments were performed to study the evolution of surface species during de-NO_x. The spectra of Figure 8a were recorded after adsorption of 800 ppm NH₃ for 30 min over the Mn/Z catalyst, with Ar purge for 30 min at 180 °C to remove any gaseous residues. It could be clearly seen that some adsorbed ammonia species included NH₄⁺ ions on Brønsted acid sites (1729 and 1433 cm⁻¹), coordinated NH₃ on Lewis acid sites (1606, 1579, 1299, 1185 and 1040 cm⁻¹), as well as amide (−NH₂) species (1508 cm⁻¹) that appeared on the Mn/Z catalyst [47,49,56]. The evolution of the surface species was monitored after the inlet gas was switched to 800 ppm NO and 5% O₂. The peak intensities of the adsorbed ammonia species gradually decreased, and some bands disappeared in the end, while some ammonia species still existed after introducing NO and O₂ for 30 min. Moreover, two new bands corresponding to the nitro and nitrate species (1675 and 1633 cm⁻¹) were generated on the Mn/Z catalyst [47,52]. This phenomenon indicated that adsorbed NH₃ species could react with adsorbed NO_x or gaseous NO₂ to form N₂ and H₂O [20]. Importantly, amide (-NH₂) species were important intermediates in the Eley–Rideal (E–R) reaction mechanism, signifying that adsorbed NH₃ species were first activated to -NH₂ species, and then consumed by gaseous NO₂ (1633 cm⁻¹). As for the GdMn/Z-0.3 catalyst (Figure 8b), a similar phenomenon occurred during the procedure of reaction between NO, O₂ and pre-adsorbed NH₃. Differently, the adsorbed ammonia species on the GdMn/Z-0.3 catalyst disappeared absolutely after feeding NO and O₂ for 15 min. Meanwhile, some new bands ascribed to the nitrate species appeared (1629, 1546, 1326 and 1223 cm⁻¹), which were stronger than those on the Mn/Z catalyst. The phenomena indicated that the adsorbed ammonia species were easily consumed by the adsorbed NO_x species, and NO_x was easily adsorbed on the GdMn/Z-0.3 catalyst because of its intrinsic redox property, which resulted in the higher low-temperature SCR activity.

2.3.4. Reaction between NH₃ and Pre-Adsorbed NO and O₂

Reversing the order of reactants, as shown in Figure 9, after switching the gas to NH₃, an apparent difference in the reaction process was observed over the Mn/Z and GdMn/Z-0.3 catalysts. For Mn/Z sample, some nitrate/nitrite species appeared, such as NO₂ molecules (1637 cm⁻¹), bridged nitrate (1594 cm⁻¹ and 1228 cm⁻¹), monodentate nitrate (1523 cm⁻¹), free nitrate species (1348 cm⁻¹) and linear nitrite (1085 cm⁻¹), which was mainly resulted from the pre-adsorbed NO and O₂ as recorded in Figure 9a [52,53,54,55]. Upon switching the gas to NH₃, all these bands got weaker and eventually disappeared for a short time of 3 min. Next, characteristic peaks belonging to ammonia species (3369–3181, 1723, 1606, 1579, 1502, 1433, 1305, 1185 and 1040 cm⁻¹) were observed, indicating that adsorbed NO_x could react with adsorbed NH₃ [47,49,50,51,56], and finally ammonia species accumulated on the catalyst surface, which followed the Langmuir–Hinshelwood (L–H) mechanism. Besides, some similar phenomena were also observed on GdMn/Z-0.3 (Figure 9b) catalyst. However, the difference was that the NO_x adsorption amount on GdMn/Z-0.3 catalyst was significantly higher than that on Mn/Z sample, and the adsorbed NO_x disappeared within the same time of NH₃ introduction, indicating that the reaction between NH₃ and pre-adsorbed NO and O₂ on GdMn/Z-0.3 was faster than that on Mn/Z sample.

2.3.5. In Situ DRIFTS of Standard NH₃-SCR Reaction on the Catalysts

Figure 10 presents the in situ DRIFTS experiments of the co-adsorption of 800 ppm NO, 800 ppm NH₃ and 5% O₂ for Mn/Z and GdMn/Z-0.3 at different times. The bands belonging to coordinated NH₃ species (3369, 3276, 3188, 1609, 1308, 1186 and 1039 cm⁻¹), and nitro/nitrate species (1685, 1637 and 1584 cm⁻¹) were observed simultaneously on the surface of the Mn/Z sample at 180 °C (Figure 10a). This fact demonstrated that the NH₃-SCR reaction over the Mn/Z catalyst followed a L–H mechanism due to the simultaneous adsorption of these reactant molecules (NH₃, NO and O₂) on the surface of the catalyst. More specifically, the adsorbed NO₂ molecules (1637 cm⁻¹) and bridged nitrates (1584 cm⁻¹) vanished within 15 min, while the nitro species appeared at 1685 cm⁻¹, slightly increased after exposure to NO, NH₃ and O₂. Moreover, all these peaks ascribed to the adsorbed NH₃ species were strengthened with the time. It was clearly proven that the nitro species located at 1685 cm⁻¹ was more stable than the NO₂ molecules (1637 cm⁻¹) and bridged nitrates (1594 cm⁻¹). As for the GdMn/Z-0.3 (Figure 10b) sample, it could be found that the spectra of NO, NH₃ and O₂ co-adsorption were different from those of NO and O₂ adsorption and NH₃ adsorption over the GdMn/Z-0.3 catalyst, as shown in Figure 6 and Figure 7. In particular, only one peak belonging to the stable nitro species (1685 cm⁻¹) could be detected apart from the adsorbed NH₃ species (3362–3178, 1732, 1613, 1473, 1428, 1273, 1175 and 1043 cm⁻¹), and all these band intensities were weaker than those observed in Figure 6 and Figure 7, which is because of the occurrence of the NH₃-SCR reaction during the co-adsorption of NO, NH₃ and O₂ process. In other words, the NH₃-SCR reaction route over the GdMn/Z-0.3 catalyst not only followed an L–H mechanism, but also an E–R mechanism, which manifested that the introduction of Gd only altered the reaction rate of the SCR reaction, and did not influence the reaction mechanism of NH₃-SCR over both catalysts.

2.4. SO₂ Resistance Studies

The impact of SO₂ on the low-temperature NH₃-SCR reaction with the initial NO conversion of nearly 100% and 80% over the Mn/Z and GdMn/Z-0.3 catalysts were tested respectively, and the results are shown in Figure 11. It could be clearly seen that the NO conversion of the Mn/Z catalyst decreased dramatically from nearly 100% to 32% during the 3 h test after the SO₂ stream was added in. Compared with Mn/Z, the GdMn/Z-0.3 catalyst presented excellent SO₂ resistance of about 85% NO conversion after the 3 h SO₂ test. Interestingly, when the SO₂ steam was cut off, GdMn/Z-0.3 restored to nearly 97% NO conversion while Mn/Z was almost unchanged. Notably, the SO₂ resistance of the GdMn/Z-0.3 catalyst in this work was competitive or even better than that reported in the literature under similar condition [57,58]. For example, the high NO conversion of GdMn/Z-0.3 was almost unaffected in the presence of SO₂ for 1 h, which was much better than previous studies in which NO conversion decreased from 100% to 60% [46]. In addition, it could be seen that the catalytic activity of the GdMn/Z-0.3 catalyst decreased slowly from the initial 80% NO conversion to 63% when SO₂ was introduced, while the Mn/Z catalyst dropped sharply to nearly 25%. It is known that the decline of catalytic activity in the presence of SO₂ could be attributed to the competitive adsorption of NO and SO₂ [59], indicating that Gd could inhibit the adsorption of SO₂ and the reaction between SO₂ and active MnO_x, which could be confirmed by in situ DRIFTS results. Furthermore, in comparison with the sole SO₂ at 180 °C, the NO conversion of two catalysts decreased obviously in the presence of the SO₂ and H₂O stream (Figure S3). Inspiringly, the initial NO conversion of GdMn/Z-0.3 only decreased to nearly 80%, while Mn/Z sharply fell to nearly 14%, suggesting that Gd could not only significantly enhance the low-temperature catalytic activity, but also improve the SO₂ and/or H₂O resistance of the Mn/Z catalyst.

2.5. XRD, SEM and XPS Analyses of the Used Samples

Both the properties of Mn/Z and GdMn/Z-0.3, after being treated at 180 °C in the presence of 100 ppm SO₂ (denoted as Mn/Z-S and GdMn/Z-0.3-S), were further investigated by XRD, SEM and XPS analyses. Figure S4 exhibits the XRD patterns of the catalysts after the SO₂ poisoning, and there were no characteristic peaks of sulfate species observed on Mn/Z-S and GdMn/Z-0.3-S. Compared with the fresh catalysts, the intensities of the two treated catalysts slightly decreased, indicating that some new amorphous species possibly formed and covered on the surface below the detection limit. SEM was performed to investigate the changes of the catalysts’ morphology after the reaction with SO₂. As shown in Figure 12, relatively large deposited particles or agglomerations were observed on the surface of the Mn/Z-S sample after the 100 ppm SO₂ test. In contrast, there was an almost smooth surface over the GdMn/Z-0.3-S sample, suggesting that metal sulfate or ammonium sulfate species were not easily formed after the addition of Gd, and consequently, the GdMn/Z-0.3 catalyst exhibited outstanding SO₂ resistance.

XPS experiments were carried out to investigate the changes of the elements’ existing states on the two catalysts in the presence of SO₂, and the results are exhibited in Figure 13. After a 3 h SO₂ test, the atomic ratio of Mn⁴⁺ and Mn³⁺ on Mn/Z-S and GdMn/Z-0.3-S was less than that of the fresh catalysts, but the proportion of Mn⁴⁺ and Mn³⁺ over GdMn/Z-0.3-S was still more than that of Mn/Z-S, demonstrating that SO₂ could affect the oxidation state of Mn species. In addition, more Mn³⁺ located on the surface of GdMn/Z-0.3-S might become the Lewis acid center for NH₃ adsorption in low-temperature SCR reactions, which was attributed to the high NO removal efficiency in the presence of 100 ppm SO₂ [54]. Compared with the fresh catalysts, the ratio of O_S/(O_L + O_S) of the used catalysts obviously decreased, which might account for the falling activity after the SO₂ poisoning. Particularly, the concentration of O_S on GdMn/Z-0.3-S was higher than on Mn/Z-S, corresponding to the better NO conversion in the SO₂ stream.

2.6. In Situ DRIFTS Studies for NH₃-SCR with SO₂

2.6.1. SO₂ Adsorption

In order to study the adsorption capacity of SO₂ on the two catalysts, the DRIFTS of 100 ppm SO₂ and 5% O₂ co-adsorption at 180 °C were conducted to investigate the existence states of sulfates. As shown in Figure 14a, several distinct peaks at 1635, 1589, 1471, 1431, 1343, 1298, 1186 and 1040 cm⁻¹ were detected on the Mn/Z catalyst after introducing SO₂ and O₂ for 30 min. As reported in the previous literature [60,61,62], the bands at 1040 cm⁻¹ could be attributed to the symmetric stretching vibrations of O–S–O species. The bands at 1298 and 1343 cm⁻¹ were ascribed to the asymmetric stretching frequencies of O–S–O species and coordinated sulfate species, respectively. The bands at 1431 and 1471 cm⁻¹ were usually assigned to the symmetrical stretching vibration of gaseous SO₂, while the bands at 1589 and 1635 cm⁻¹ were associated with the adsorbed HSO₄⁻ generated by the interaction between SO₂ and surface hydroxyl groups. A weak peak at 1186 cm⁻¹ over Mn/Z appeared, which was possibly because of the bulk-like sulfates in MnO₂ (Mn(SO₄)₂) [62,63]. Notably, only three weak bands belonging to the surface sulfate species (1630 and 1295 cm⁻¹) and gaseous SO₂ (1473 cm⁻¹) formed on the GdMn/Z-0.3 catalyst (Figure 14b), demonstrating that Gd as a sacrificial site potentially greatly inhibited the reaction between the SO₂ and MnO_x active sites to form bulk Mn(SO₄)₂ or Mn₂(SO₄)₃ on the catalyst surface, and further reduced the sulfurization rate [62]. It was reported that the reaction between SO₂ and metal oxide active sites could inactivate these active sites, resulting in the catalyst deactivation [23]. Fortunately, the introduction of Gd effectively suppressed the generation of Mn(SO₄)₂, which contributed to the excellent SO₂ resistance of the GdMn/Z-0.3 catalyst.

2.6.2. Reaction between NH₃, NO and O₂ (SO₂) and Pre-Adsorbed SO₂ (NH₃, NO and O₂)

The influence of sulfur species on the adsorption of NH₃ and NO_x species was further investigated over the Mn/Z and GdMn/Z-0.3 catalysts at 180 °C. As shown in Figure 14 and Figure 15, several peaks belonging to various sulfur species were detected on the Mn/Z and GdMn/Z-0.3 catalysts after adsorption of 100 ppm SO₂ and 5% O₂ for 30 min [60,61,62]. All these bands remained almost unchanged for both catalysts after Ar purging for 30 min (Figure 15). Compared with the above DRIFT spectra of the NH₃, NO and O₂ co-adsorption depicted in Figure 10, the adsorption capacity of NO_x on Mn/Z obviously decreased after SO₂ adsorption (Figure 15a), indicating competitive adsorption between NO_x and SO₂. The latter had a remarkably stronger adsorption capacity than the former. Thus, the low-temperature NH₃-SCR reaction between the adsorbed NO_x and NH₃ species following the L–H mechanism was repressed by the presence of SO₂, causing the serious deactivation of Mn/Z under the atmosphere of SO₂. As for the GdMn/Z-0.3 catalyst (Figure 15b), the DRIFT spectra were similar with those exhibited in Figure 10 conducting the standard NH₃-SCR reaction, demonstrating that the low-temperature NH₃-SCR reaction route over GdMn/Z-0.3 could remain unchanged in the presence of SO₂, owing to the small amount of adsorption of sulfur species on the GdMn/Z-0.3 catalyst. Herein, the GdMn/Z-0.3 catalyst exhibited excellent SO₂ resistance.

Notably, the attribution of the peaks in Figure 16 was consistent with that in Figure 6 and Figure 7. It could be seen that the original adsorption peaks of NO_x and NH₃ species located at a low wavenumber (1000–2000 cm⁻¹) over the Mn/Z catalyst almost remained unchanged after feeding SO₂, while the peaks at a high wavenumber (3100–3400 cm⁻¹) weakened gradually. In combination with Figure 6, Figure 7 and Figure 14, it could be concluded that the peaks of the sulfur species on the Mn/Z catalyst almost overlapped with the peaks of the adsorbed NH₃ and NO_x species (Figure 16a). Thus, the strength of these peaks was almost unchanged in the whole process, which was probably caused by the adsorption of the sulfur species. In addition, the adsorbed NH₃ species on GdMn/Z-0.3 gradually decreased with the time of SO₂ feeding (Figure 16b), and the adsorption peaks of SO₂ finally appeared at 1473 and 1295 cm⁻¹, which were weaker than those on Mn/Z. It could be seen from Figure 16 that the adsorption of SO₂ on GdMn/Z-0.3 was still weaker than that on Mn/Z, even in the case of NH₃ and NO_x pre-adsorption.

2.6.3. Co-Adsorption of SO₂, NH₃, NO and O₂

When SO₂ was introduced into the standard SCR reaction atmosphere, the position of the bands in the Mn/Z catalyst (Figure 17a) was similar to that described in the standard SCR reaction over the Mn/Z catalyst (Figure 10a). Differently, the bands located at 1583 and 1635 cm⁻¹, corresponding to adsorbed NO_x species, did not disappear within 30 min, while they were almost consumed within 10 min in the standard NH₃-SCR reaction without SO₂, indicating the presence of SO₂ could lead to a decrease in the reaction rate, which was also the reason for its decreased activity. In addition, it was surprising that the peak at 1583 cm⁻¹ attributed to bridged nitrates increased slightly with the presence of SO₂ in the initial period, owing to the overlapping of the adsorbed sulfates. As for the GdMn/Z-0.3 catalyst (Figure 17b), no bands ascribed to sulfate species were observed, while NO_x (1685 cm⁻¹) and NH₃ adsorption (1726 cm⁻¹) were strengthened. Compared with the standard NH₃-SCR reaction conditions, the reaction rate was hardly changed. Therefore, it could be speculated that the presence of SO₂ in the reaction atmosphere had little influence on the reaction path over the GdMn/Z-0.3 catalyst, contributing to its outstanding SO₂ resistance.

3. Materials and Methods

3.1. Catalyst Preparation

The Gd-modified MnO_x/ZSM-5 catalysts were prepared using a citric acid–ethanol dispersion method. All the reagents were of the required analytical grade and had no further purification. ZSM-5 zeolite (Si/Al = 38, XFNANO Company, Nanjing, China) was used as support material. Manganese acetate tetrahydrate (Mn(CH₃COO)₂·4H₂O, 99.9%, Aladdin, Shanghai, China) and gadolinium nitrate hexahydrate (Gd(NO₃)₃·6H₂O, 99.9%, Aladdin, Shanghai, China) were used as the precursors of the manganese oxides and gadolinium oxides, respectively. First of all, a certain amount of Mn(CH₃COO)₂·4H₂O (Mn loading = 15 wt%) and Gd(NO₃)₃·6H₂O were dissolved in 120 mL ethanol and then stirred vigorously for 30 min to dissolve precursors completely under room temperature. Next, a certain amount of ZSM-5 zeolite was slowly added to the above mixture and stirred continuously. Then, the mixed solution was sonicated for 45 min, followed by the addition of citric acid, and further stirred in an 80 °C water bath until the solvent was completely evaporated to obtain a colloidal solid. Finally, the product was put in a 110 °C oven to dry for 12 h, and then calcined in air at 400 °C for 4 h. The synthesized catalysts were denoted as GdMn/Z-x, where x represented the mole ratio of Gd and Mn (x = 0, 0.1, 0.3 and 0.6). For comparison, the MnO_x/ZSM-5 catalyst was also prepared using the same method and denoted as Mn/Z.

3.2. NH₃-SCR Activity Test

The NH₃-SCR activity evaluation was carried out in a fixed-bed quartz flow tube reactor (12 mm inside diameter and 600 mm in length) under atmospheric pressure in the reaction temperature range of 80–240 °C. Before the tests, all the catalysts were crushed and sieved as 40–60 mesh. The feed gases were designed as follows: 800 ppm NO, 800 ppm NH₃, 5 vol.% O₂, 100 ppm SO₂ (when needed) and 10% H₂O (when needed), and balanced by Ar. In order to obtain a gas hourly space velocity (GHSV) of 40,000 h⁻¹, the rate of the total flow was 600 mL min⁻¹, and each sample volume was 0.9 mL. The inlet and outlet concentrations of NO_x (NO and NO₂) from the fixed-bed quartz flow tube reactor were measured by a NO–NO₂–NO_x analyzer (model 42i-HL, Thermal Scientific, Waltham, MA, USA). At the same time, the outlet concentration of the product N₂ from the reactor was also monitored by a gas chromatograph (GC9560, Huaai, Shanghai, China) with a thermal conductivity detector (TCD) and 5A columns. The conversion of NO and the selectivity of N₂ in the low-temperature NH₃-SCR reaction were calculated by the following formulas:

$$\mathrm{NO}\mathrm{conversion}=\frac{{\mathrm{C}}_{\mathrm{NO}}^{\mathrm{in}}{-\mathrm{C}}_{\mathrm{NO}}^{\mathrm{out}}}{{\mathrm{C}}_{\mathrm{NO}}^{\mathrm{in}}}\times 100\%$$

(1)

$${\mathrm{N}}_2\mathrm{selectivity}=\frac{{\mathrm{C}}_{{\mathrm{N}}_2}^{\mathrm{out}}}{{\mathrm{C}}_{{\mathrm{NO}}_{\mathrm{x}}}^{\mathrm{in}}{-\mathrm{C}}_{{\mathrm{NO}}_{\mathrm{x}}}^{\mathrm{out}}}\times 100\%$$

(2)

where ${\mathrm{C}}_{{\mathrm{NO}}_{\mathrm{x}}}^{\mathrm{in}}$ is the inlet concentration (ppm) of NO and NO₂, while ${\mathrm{C}}_{{\mathrm{NO}}_{\mathrm{x}} }^{\mathrm{out}}\mathrm{and} {\mathrm{C}}_{{\mathrm{N}}_2}^{\mathrm{out}}$ correspond to the outlet concentration (ppm) of NO, NO₂ and N₂, respectively.

3.3. Catalyst Characterization

XRD patterns were recorded by a D8 Advance diffractometer (Bruker, Karlsruhe, Germany) with Cu-Kα radiation ($\lambda$ = 1.5418 Å). N₂ adsorption−desorption experiments were performed by an ASAP 2460 analyzer (Micromeritics, Norcross, GA, USA) at liquid N₂ temperature (−196 °C). Prior to the measurements, the samples were degassed in a vacuum at 300 °C for 4 h. The specific surface area was calculated using the BET method at P/P₀ = 0.01–0.3. The total pore volume was determined from the desorption branch isotherms by the Barrett–Joyner–Halenda (BJH) method and the average pore size was determined from the desorption average pore diameter (4V/A by BET). The morphology of samples was characterized by field-emission SEM with an instrument of ZEISS Merlin (Carl Zeiss AG, Jena, Germany). TEM and HR-TEM were performed on a JEM-2100F equipped with an EDS analysis to further obtain microstructure information of the samples. XPS with Al-Kα radiation (hυ = 1253.6 eV, Axis Ultra DLD, Kratos, UK) was conducted to analyze the composition and valences of elements on the surface of all catalysts. Both H₂-TPR and NH₃-TPD analyses were conducted by an Auto ChemII 2920 instrument (Micromeritics, Norcross, GA, USA).

In situ DRIFTS experiments were performed on a Fourier transform infrared spectrometer (FTIR, Nicolet iS50, Thermo Scientific, Waltham, MA, USA) equipped with an MCT/A detector and a Harrick DRIFT chamber in the range of 600–4800 cm⁻¹ at a resolution of 4 cm⁻¹. Before the single-beam background collection, the catalyst was purged in Ar gas at 300 °C for 1 h and then cooled to 180 °C for the aimed tests. A total flow of 100 mL min⁻¹ of the gas stream was balanced with Ar, and the test conditions of the adsorption and the reaction are described in detail in the results section. All the DRIFTS spectra were recorded through subtraction of the corresponding background reference by collecting 64 scans.

4. Conclusions

In summary, several GdMn/Z-x catalysts were synthesized by a citric acid–ethanol dispersion method and used for the NH₃-SCR of NO removal in the presence of SO₂. Among all the catalysts, the activity of NO conversion followed the order of GdMn/Z-0.3 > GdMn/Z-0.6 > Mn/Z > GdMn/Z-0.1. The results exhibited that the addition of a proper amount of Gd (Gd/Mn = 0.3) into the Mn/Z catalyst could remarkably enhance its catalytic activity and SO₂ resistance for NH₃-SCR. The excellent performance of GdMn/Z-0.3 could be attributed to the existence of a large surface area, abundant Mn⁴⁺ and chemisorbed oxygen species, strong reducibility and suitable surface acidity. In addition, the Mn/Z catalyst showed extremely poor SO₂ resistance due to its strong SO₂ adsorption ability. A large number of stable sulfate species, especially bulk manganese sulfate, were formed on the Mn/Z surface, which reduced the NO adsorption ability and covered the active sites. However, Gd significantly inhibited SO₂ adsorption on Mn/Z, and the bulk manganese sulfate was not observed in the DRIFT spectra of SO₂ adsorption, thus leading to excellent SO₂ resistance over GdMn/Z-0.3. In situ DRIFTS experiment results revealed that the reaction mechanism of low-temperature SCR over Mn/Z and GdMn/Z-0.3 followed the L–H and E–R mechanism simultaneously, while the reaction path over Mn/Z was greatly affected by SO₂, resulting in the poor activity. Fortunately, the reaction path on GdMn/Z-0.3 was almost unaffected by SO₂, thus possessing excellent SO₂ resistance. Consequently, GdMn/Z-0.3 as a high-effective potential catalyst can be applied in the remediation of harmful NO_x pollutants.