Low-Temperature NH₃-SCR on Ce_x-Mn-Ti_y Mixed Oxide Catalysts: Improved Performance by the Mutual Effect between Ce and Ti

Abstract

A series of Ce_x-Mn-Ti_y catalysts were synthesized using the coprecipitation method, and sodium carbonate solution was used as a precipitant. The various catalysts were assessed by selective catalytic reduction of NO_x with NH₃, and characterized by X-ray diffraction, Raman spectroscopy, H₂ temperature-programmed reduction, NH₃ temperature-programmed desorption, and X-ray photoelectron spectroscopy to investigate the physicochemical properties, surface acidity, and redox abilities of the Ce_x-Mn-Ti_y catalysts. The Ce_0.1-Mn-Ti_0.1 catalyst exhibited the best catalytic performance (more than 90% NO_x conversion in the range of 75 to 225 °C), as a result of proper redox ability, abundant acid sites, high content of Mn⁴⁺ and Ce³⁺, and surface-adsorbed oxygen (O_S). The results of in situ DRIFT spectroscopy showed that the NH₃-SCR reaction followed both the E-R and L-H paths over the Ce_0.1-Mn-Ti_0.1 catalyst, and it occurred faster and more sharply when it mainly abided by the E-R mechanism.

1. Introduction

The combustion of fossil fuels, especially at thermoelectric power plants as stationary sources and in diesel vehicle emissions as mobile sources [1], produces excessive nitrogen oxides (including NO, NO₂, and N₂O), causing the greenhouse effect, acid rain, and photochemical smog [2,3,4]. In order to deal with the serious harm caused by NO_x emissions, it is of great importance to find an effective approach to eliminate NO_x.

The selective catalytic reduction of NO_x with ammonia (NH₃-SCR) is considered to be one of the most efficient ways to eliminate emissions of NO_x [5]. As a reduction reagent, NH₃ is used to reduce NO_x, forming nitrogen and water. Owing to variable valence states, vanadium-based catalysts are currently the most commonly used catalysts in the industrial application of NH₃-SCR. However, vanadium-based catalysts have some inevitable disadvantages, such as their narrow active temperature window (300–400 °C) and the toxicity of V₂O₅ [6]. In addition, under real operating conditions, the existence of sulfur and water in the flue gas of power plants can cause the deactivation of the catalyst. Therefore, a catalyst with excellent activity and resistance to SO₂ and H₂O is urgently needed.

In addition to vanadium-based catalysts, transition metal oxide catalysts, especially Mn oxides, exhibit great SCR activity, even at temperatures below 200 °C. However, pure MnO_x catalysts show a narrow operating temperature window and poor tolerance to SO₂ and H₂O, restraining practical applications. Much effort has been devoted to improving the performance of the MnO_x catalysts, for example, doping with other metals (M = Fe [7,8], Co [9,10], Cu [11,12], Sm [13,14], Ti [15,16], etc.) to optimize the redox capacity and acidity of the catalyst by forming M–Mn composite oxides. Nevertheless, it is hard to simultaneously optimize the redox capacity and acidity through single-metal doping, since the redox capacity and acidity restrict one another. Liu et al. prepared a series of Mn_yTi_1-yO_x composite oxides to adjust the surface acidity while the reducibility declined with a decrease in Mn content [17]. Conversely, when two metals are introduced into pure MnO_x, the modified catalyst may possess a greater possibility of optimizing acidity and redox ability. Cu_wMn_yTi_1-yO_x catalysts were synthesized to improve the acidity and redox ability simultaneously, and showed superior SCR performance and SO₂ resistance [18]. Concurrently, Ce is known for its excellent oxygen storage and release capacity, which can significantly improve the oxidation of MnO_x [19,20]. Previous studies [21] have found that the incorporation of Ti can efficiently tune the MnO_x phase and construct surface acidity to promote the adsorption of reactants.

In this study, Ce_x-Mn-Ti_y mixed-oxide catalysts were prepared via the coprecipitation method to investigate their catalytic performance, sulfur dioxide/water resistance, and structure–activity relationships. Mn⁴⁺ forms the main active sites for the adsorption of NO and NH₃. Due to their high oxygen storage capacity and excellent redox properties, Ce and Mn cations can form a redox cycle [22]. The existence of Ti can efficiently decrease the MnO_x crystallinity and construct surface acid sites to promote the adsorption of reactants. Ce_x-Mn-Ti_y composite oxides show the best low-temperature activity at 75–225 °C, and better SO₂/H₂O resistance under a mixed flow of 2% H₂O and 50 ppm of SO₂ for 8 h at 180 °C. The effects of adding Ce and Ti to the structure and physicochemical properties of MnO_x were systematically investigated through various characterizations, and the reaction mechanism on the Ce_x-Mn-Ti_y catalyst was clarified.

2. Results and Discussion

2.1. Effects of Ce and Ti on Catalytic Activity and SO₂/H₂O Resistance

The low-temperature activity of Ce_0.1-Mn-Ti_0.1, Ce_0.1-Mn, Mn-Ti_0.1, and Ce_0.1-Ti_0.1 for NH₃-SCR at different GHSVs (50,000, 100,000, and 150,000 h⁻¹) is shown in Figure 1. The results showed that the Ce_0.1MnTi_0.1 catalyst exhibited the best catalytic performance, delivering ~90% NO_x conversion at 75–225 °C. Meanwhile, the Ce_0.1-Ti_0.1 catalyst was almost inactive in the temperature range 50–350 °C, and the Mn-Ti_0.1 catalyst exhibited significantly worse NH₃-SCR activity than the Ce_0.1-Mn-Ti_0.1 and Ce_0.1-Mn catalysts. The Ce_0.1-Mn-Ti_0.1 and Ce_0.1-Mn catalysts showed similar activity curves at 50,000 h⁻¹. As the GHSV increased from 50,000 to 150,000 h⁻¹, the catalytic performance of the four catalysts decreased to some extent. However, the activity of the Ce_0.1-Mn-Ti_0.1 catalyst only slightly declined (125–225 °C with NO_x conversion > 90%); in contrast, the activities of the Ce_0.1-Mn and Mn_0.1-Ti catalysts dropped significantly, seldom achieving 90% NO_x conversion. Clearly, the Ce_0.1-Mn-Ti_0.1 catalyst exhibited the best catalytic activity. Hence, we conjectured that the synergistic effects between Ce and Mn as well as Mn and Ti [23] existed for the Ce_0.1-Mn-Ti_0.1 catalyst, dramatically enhancing its catalytic performance. NO_x conversion per unit of S_BET for 150,000 h⁻¹ of the Ce_x-Mn-Ti_y catalysts is shown in Figure S1 from Supplementary Materials.

Due to the poor activity of the Ce_0.1-Ti_0.1 catalyst, it was of little significance in testing the N₂ selectivity. Thus, e Ce_0.1-Ti_0.1 catalyst was not included in the subsequent experiments. Therefore, Figure 1D only shows the N₂ selectivity of the Ce_0.1-Mn-Ti_0.1, Ce_0.1-Mn, and Mn-Ti_0.1 catalysts, with a GHSV of 50,000 h⁻¹. The N₂ selectivity of the Ce_0.1-Mn-Ti_0.1 catalyst was almost above 95%, and the curve was similar to that of the Ce_0.1-Mn catalyst, and higher than that of the Mn-Ti_0.1 catalyst, at moderate and high temperatures (200–330 °C). Due to the relatively lower NO_x conversion, the smallest amount of byproduct (N₂O) was produced with the Mn-Ti_0.1 catalyst.

The influence of H₂O and SO₂ is exhibited in Figure 2 for the Ce_0.1-Mn-Ti_0.1, Ce_0.1-Mn, and Mn_0.1-Ti catalysts. Firstly, the catalysts were kept in ideal conditions for 4 h at 180 °C. The Ce_0.1-Mn-Ti_0.1 and Ce_0.1-Mn catalysts maintained complete conversion, while the Mn_0.1-Ti catalyst only achieved 86% NO_x conversion. When introducing 2 vol.% H₂O into the feed gas for 8 h, the NO_x conversion of the Ce_0.1-Mn-Ti_0.1, Ce_0.1-Mn, and Mn_0.1-Ti catalysts declined quickly, to approximately 93%, 85%, and 32% in almost an hour, respectively, and then remained stable. After 4 h of H₂O removal, the catalytic activity was recovered to 96%, 96%, and 78%, respectively, illustrating that H₂O poisoned the Ce_x-Mn-Ti_y catalysts and caused irreversible deactivation. This may be attributable to the partial dissociation of H₂O at 180 °C, and the –OH formed on the catalysts being difficult to desorb or decompose. Then, 2 vol.% H₂O and 50 ppm of SO₂ were introduced simultaneously for 10 h, and the conversion gradually dropped to 75%, 67%, and 19%, respectively, implying that the formed sulfates covered part of the active sites, resulting in the decrease in activity. Ce can act as a sacrificial site on the catalyst, and CeO₂ preferentially reacts with SO₂, alleviating the sulfation of the active Mn oxide phase. The introduction of Ce may induce the deposition of bulk sulfate on CeO₂, thereby inhibiting the SO₂ poisoning of the active MnO_x sites. Once H₂O and SO₂ were removed, the conversion rebounded to about 96%, 90%, and 22%, respectively. It is worth noting that, after 10 h of the injection of 2 vol.% H₂O and 50 ppm of SO₂, the NO_x conversion of Ce_0.1-Mn showed a downward trend, while that of Ce_0.1-Mn-Ti_0.1 tended to stabilize. The overall performance within 2 vol.% H₂O and the H₂O/SO₂ resistance of the Ce_x-Mn-Ti_y catalysts at 100 °C are shown in Figures S2 and S3 from Supplementary Materials, respectively. The Ce-Mn-Ti catalyst exhibited a better catalytic performance at 2 vol.% H₂O in a wide temperature range of 50 to 350 °C and a better H₂O/SO₂ resistance at 100 °C. To summarize, the Ce_0.1-Mn-Ti_0.1 catalyst possessed a better H₂O/SO₂ resistance than the Ce_0.1-Mn and Mn-Ti_0.1 catalysts. Hence, the introduction of small amounts of Ce and Ti led to the improvement of catalytic performance and SO₂/H₂O resistance.

2.2. Physicochemical Properties of Ce_x-Mn-Ti_y Catalysts

2.2.1. X-ray Diffraction

Figure 3A shows the XRD patterns of all catalysts. In the XRD pattern of Ce_0.1-Mn-Ti_0.1, the diffraction peaks of MnO₂, Mn₅O₈, and Mn₂O₃ are the main ones that can be seen [24,25]. For Ce_0.1-Mn, however, in addition to the diffraction peaks of MnO₂, Mn₅O₈, and Mn₂O₃, the diffraction peaks of CeO₂ can also be seen [3,23], which indicates the enrichment of the Ce_0.1-Mn surface by CeO₂. Compared with Ce_0.1-Mn, the XRD pattern of the Ce_0.1-Mn-Ti_0.1 catalyst showed weaker peak intensities and a smaller FWHM, ascribed to the MnO₂ and Mn₅O₈. Owing to the similar ionic radius of Mn⁴⁺ and Ti⁴⁺, Ti⁴⁺ could be doped into the MnO_x lattice [21], thereby inhibiting the crystallization of MnO_x. The XRD pattern of Mn-Ti_0.1 shows the diffraction peaks of MnO₂, Mn₅O₈, Mn₂O₃, and TiO₂ [26]. Comparing the XRD patterns of the Ce-Mn-Ti and Mn-Ti catalysts, the peak ascribed to TiO₂ was not observed on Ce-Mn-Ti, while it appeared on the Mn-Ti catalyst. Thus, it can be inferred that the introduction of Ce can promote the dispersion of TiO₂ on the Ce-Mn-Ti catalyst. The XRD patterns of the Ce_0.1-Ti_0.1 catalyst are displayed in Figure S4 from Supplementary Materials, showing only the diffraction of CeO₂. These XRD results illustrated that the incorporation of Ti, inhibiting the MnO_x crystallization, resulted in a low crystallinity of MnO_x species, and Ce doping promoted the dispersion of Ti on the Ce_0.1-Mn-Ti_0.1 catalyst.

2.2.2. Raman Spectroscopy

Raman spectroscopy was used to further explain the composition and crystal phase of Ce_x-Mn-Ti_y. The bands at 639, 642, 653, and 677 cm⁻¹ were assigned to MnO_x, and the bands at 150 and 456 cm⁻¹ belonged to TiO₂ and CeO₂ [17,23], respectively (Figure 3B). It can be said that the Raman data agreed with the XRD results perfectly. In more detail, the bands at 639 and 642 cm⁻¹ were assigned to MnO₂ for the Ce_0.1-Mn-Ti_0.1 and Ce_0.1-Mn catalysts, with a shift to a higher wavenumber [27]. Meanwhile, the intensity of the band at 639 cm⁻¹ was much higher than that at 642 cm⁻¹, representing more high-valence Mn. In addition, the band at 653 cm⁻¹ was derived from the overlapping of the MnO₂ and Mn₂O₃ phases, leading to a strong intensity. The band at 653 cm⁻¹ with a shoulder band at 677 cm⁻¹ was well-matched with the standard spectrum of Mn₂O₃ on the Mn-Ti_0.1 catalyst [27]. That is, the Mn-Ti_0.1 catalyst was more likely to exist in the phase of Mn₂O₃. Thus, the Ce_0.1-Mn-Ti_0.1 catalyst may have the highest concentration of high-valence Mn, which was later confirmed by XPS.

2.2.3. Scanning Electron Microscopy and N₂ Adsorption–Desorption

The surfaces of the Ce_0.1-Mn-Ti_0.1, Ce_0.1-Mn, and Mn-Ti_0.1 catalysts were characterized by SEM. As shown in Figure 4A, Ce_0.1-Mn-Ti_0.1 was composed of many spherical particles, with a diameter of 1.5–2 μm. The shape of Ce_0.1-Mn was similar to that of Ce_0.1-Mn-Ti_0.1, except that the particles were larger (Figure 4B). For Mn-Ti_0.1, no spherical particles could be observed due to sintering (Figure 4C). These results implied that the incorporation of Ce could promote the dispersion of metals, thereby preventing the catalyst from sintering, in accordance with the XRD results. The pore structures and BET surface areas of the catalysts are given in

Table 1. BET surface area and pore structure of the Ce_x-Mn-Ti_y catalysts.

Sample	SBET (m2/g)	Pore Size (nm)	Pore Volume (cm3/g)	Mole Ratio a
				Ce/Mn	Ti/Mn
Ce0.1-Mn-Ti0.1	52.46	17.7	0.23	0.12	0.12
Ce0.1-Mn	46.61	16.9	0.20	0.12	-
Mn-Ti0.1	29.45	25.1	0.18	-	0.12

^a Molar ratio of Ce/Mn and Ti/Mn according to ICP-AES.

and Figure S5 from Supplementary Materials. The surface area of Ce_0.1-Mn-Ti_0.1 (52 m²/g), as well as its pore volume (0.23 cm³/g), were much larger than those of Ce_0.1-Mn and Mn-Ti_0.1. The mole ratios of Ce/Mn and Ti/Mn of the corresponding catalysts from ICP-AES were both 0.12—slightly higher than the theoretical value (0.1).

2.2.4. X-ray Photoelectron Spectroscopy

XPS was used to further analyze the valence states of Mn, Ce, O, and the surface atom concentrations. Figure 5 shows the XPS spectra of Mn 2p, Ce 3d, and O 1s of the Ce_x-Mn-Ti_y catalysts. The surface atom concentrations are listed in

Table 2. The XPS results and atomic molar ratios of the Ce_x-Mn-Ti_y catalysts.

Sample	Surface Atom Concentration b
	Ce/Mn	Ti/Mn	Mn4+/(Mn4+ + Mn3+)	OS/(OS + OL)	Ce3+/(Ce4+ + Ce3+)
Ce0.1-Mn-Ti0.1	0.29	0.14	60.1%	16.6%	14.7%
Ce0.1-Mn	0.32	-	58.9%	12.6%	6.9%
Mn-Ti0.1	-	0.16	55.7%	13.2%	-

^b Surface atom concentration detected by XPS analysis.

. The Mn 2p spectra could be decomposed into three spin–orbit doublets, which are located at 642.2 ± 0.1, 641.3 ± 0.1, and 644 ± 0.4 eV, representing Mn⁴⁺, Mn³⁺, and satellites [28], respectively (Figure 5A). According to Table 2, the surface atom ratio of Mn⁴⁺/(Mn⁴⁺+Mn³⁺) on the Ce_0.1-Mn-Ti_0.1 catalyst was the highest (60.1%) among the three catalysts, which is consistent with the Raman results. Compared with Mn-Ti_0.1, the concentration of Mn⁴⁺ atoms increased significantly after Ce doping of the catalyst. The order of Mn⁴⁺ /(Mn⁴⁺+Mn³⁺) was Ce_0.1-Mn-Ti_0.1 > Ce_0.1-Mn > Mn-Ti_0.1, which was consistent with the performance of NH₃-SCR.

As shown in Figure 5B, the Ce 3d spectra could be decomposed into five spin–orbit doublets, denoted as u₀, u, u′, u′′, u′′′ and v₀, v, v′, v′′, v′′′ [29,30], where u₀/v₀ and u′/v′ belonged to Ce³⁺ and the others were assigned to Ce⁴⁺. It can be observed from Table 2 that the Ce³⁺/(Ce⁴⁺+Ce³⁺) atom ratios on the Ce_0.1-Mn-Ti_0.1 and Ce_0.1-Mn catalysts were 14.7 and 6.9%, respectively. The O 1s peaks displayed in Figure 5C could be split into two peaks, assigned to lattice oxygen (O_L) at 529.7 ± 0.2 eV and surface active oxygen (O_S) at 531.4 ± 0.1 eV [31]. It has been reported [32] that chemically adsorbed oxygen (O_S) with higher mobility is more active than lattice oxygen (O_L), and is a key factor affecting the catalytic activity. O_S can oxidize NO to NO₂, thus promoting the “Fast SCR” process [33]. Table 2 shows that the O_S/(O_S+O_L) atomic ratio on the Ce_0.1-Mn-Ti_0.1 catalyst (16.6%) was the highest among the catalysts. As shown in Figure 5D, the binding energy of Ti 2p_3/2 for Ce_0.1-Mn-Ti_0.1 and Mn-Ti_0.1 was 458.3 and 458.1 eV, respectively, which can be ascribed to Ti⁴⁺ [18].

The above results illustrate that Ce_0.1-Mn-Ti_0.1 possessed the highest Mn⁴⁺/(Mn⁴⁺+Mn³⁺) and O_S/(O_S+O_L) atomic ratios. As a consequence, the Ce_0.1-Mn-Ti_0.1 catalyst exhibited the best NH₃-SCR performance. In addition, the XPS data analysis resulted in an atomic ratio of Ce/Mn and Ti/Mn. The atomic ratios of Ce/Mn on the Ce_0.1-Mn-Ti_0.1 and Ce_0.1-Mn catalysts were 0.29 and 0.32, respectively, while those for Ti/Mn on the Ce_0.1-Mn-Ti_0.1 and Mn-Ti_0.1 catalysts were 0.14 and 0.16, respectively—higher than the molar ratio tested with ICP, suggesting that Ce and Ti were abundant on the surface of the Ce_x-Mn-Ti_y catalysts.

2.3. Redox Ability and Acidity of the Ce_x-Mn-Ti_y Catalysts

Both redox and acidic sites play important roles in the NH₃-SCR reaction. Generally, the redox properties determine the low-temperature activity, while the acidity governs the high-temperature activity.

2.3.1. H₂ Temperature-Programmed Reduction and NO Oxidation Reaction

H₂-TPR tests were performed to judge the redox ability of the Ce_x-Mn-Ti_y catalysts. As shown in Figure 6A, the three catalysts all exhibited two reduction peaks, denoted as α and β, respectively. The curve of the Ce_0.1-Mn-Ti_0.1 catalyst displays two apparent peaks at 338 and 467 °C, which were attributed to MnO₂/Mn₂O₃ → Mn₃O₄ (α), Mn₅O₈/Mn₃O₄ → MnO, and CeO₂ → Ce₂O₃ (β) [34,35,36], respectively (

Table 3. H₂-TPR results of the Ce_x-Mn-Ti_y catalysts.

Sample	α (T/°C)	Curve Area β (T/°C)	α/β	H2 Consumption (mmol/g)
Ce0.1-Mn-Ti0.1	2.79 (338)	2.56 (467)	1.08	5.35
Ce0.1-Mn	3.65 (328)	2.91 (434)	1.25	6.56
Mn-Ti0.1	3.46 (351)	3.12 (469)	1.11	6.58

). The curves of Ce_0.1-Mn and Mn-Ti_0.1 were fairly similar to that of Ce_0.1-Mn-Ti_0.1. However, the reduction peaks of Ce_0.1-Mn shifted to lower temperatures (328 and 434 °C), while those of Mn-Ti_0.1 moved to higher temperatures (351 and 469 °C). Furthermore, the similar amounts of H₂ consumption for the Mn-Ti_0.1 and Ce_0.1-Mn catalysts (6.58 and 6.56 mmol/g, respectively) were larger than that of the Ce_0.1-Mn-Ti_0.1 catalyst (5.35 mmol/g). H₂ consumption originated from the reduction of Mn⁴⁺ and Ce⁴⁺. However, the Ce_0.1-Mn-Ti_0.1 catalyst is a tri-metal mixed oxide, and its Mn content was the lowest among the bimetal oxide catalysts with the same quality, contributing to a lower H₂ consumption. In more detail, the area ratios of peaks α and β were ranked as Ce_0.1-Mn > Mn-Ti_0.1 > Ce_0.1-Mn-Ti_0.1, which seemed to be inconsistent with the results of XPS. The reason for the discrepancy between H₂-TPR and XPS was that XPS characterized the surface atom concentration, while H₂-TPR represented the overall reduction of the metal oxides. Consequently, though possessing the highest surface Mn⁴⁺ concentration (60.1%), the α/β of the Ce_0.1-Mn-Ti_0.1 catalyst was the lowest, implying that its surface was rich in Mn⁴⁺. The results obtained based on Table 3 indicated that Ce_0.1-Mn showed the strongest reducibility, with lower reduction temperatures and a relatively larger amount of H₂ consumption.

Figure 6B displays the NO oxidation capability of the Ce_x-Mn-Ti_y catalysts under the reaction conditions of 500 ppm of NO and 5 vol.% O₂. The order of NO oxidation capability of the three catalysts was Ce_0.1-Mn > Ce_0.1-Mn-Ti_0.1 > Mn-Ti_0.1, completely consistent with the reduction temperatures of H₂-TPR, as shown in the illustration. Combined with H₂-TPR, the redox ability of the Ce_0.1-Mn catalyst was higher than that of the Ce_0.1-Mn-Ti_0.1 and Mn-Ti_0.1 catalysts. In general, excessively strong redox properties led to non-selective catalytic oxidation of NH₃ and the generation of N₂O as a byproduct [37]. Nevertheless, the order of redox ability was inconsistent with the catalytic performance, implying that redox ability was not the key factor determining the NH₃-SCR performance of the catalysts. Moderate redox ability has a beneficial effect on the SCR reaction, while excessive redox ability would be detrimental [23,38].

2.3.2. NH₃/NO Temperature-Programmed Desorption

The adsorption capacity of NO and NH₃ is an essential factor for evaluating catalysts. In order to characterize the adsorption capacity of NO and NH₃, NH₃/NO-TPD was carried out to analyze the properties of the Ce_x-Mn-Ti_y catalysts.

Since the capacity for NH₃ adsorption is one of the determinants for NH₃-SCR catalysts, the surface acidity is of great importance for an excellent NH₃-SCR catalyst. The area and position of the desorption peaks reflect the number and strength of acid sites, respectively. Before NH₃-TPD was performed, the catalysts were pretreated at 400 °C in a flow of Ar for 1 h. The NH₃-TPD profiles of all catalysts are shown in Figure 7A. In the NH₃-TPD profiles of the Mn-Ti_0.1 and Ce_0.1-Mn catalysts, there was one weak desorption peak at 106 and 127 °C, respectively, and for the Ce_0.1-Mn-Ti_0.1 catalyst there was an obvious desorption peak that appeared at 89 °C. Compared with the Mn-Ti_0.1 and Ce_0.1-Mn catalysts, the desorption peak of the Ce_0.1-Mn-Ti_0.1 catalyst shifted slightly to a lower temperature, and the peak area per unit of S_BET was high (183), as listed in

Table 4. NH₃-TPD and NO-TPD results.

Sample	Curve Area of NH3-TPD × 103	Area/SBET(NH3)	Curve Area of NO-TPD × 107	Area/SBET(NO) × 105
Ce0.1-Mn-Ti0.1	9.6	183	1.8	3.5
Ce0.1-Mn	6.0	129	2.0	4.3
Mn-Ti0.1	3.5	118	1.6	5.3

, suggesting that the number of acid sites on the Ce_0.1-Mn-Ti_0.1 catalyst was larger than that of the Mn-Ti_0.1 and Ce_0.1-Mn catalysts. The number of acid sites clearly decreased in the order Ce_0.1-Mn-Ti_0.1 > Ce_0.1-Mn > Mn-Ti_0.1, which was consistent with the order of catalytic performance. It was speculated that the main factors affecting the number of acid sites may be the content of Mn⁴⁺ and the introduction of TiO₂. Mn with a high valence state exhibits strong electron-adsorbing ability, and TiO₂ is known to have abundant acidic sites [39]. Thus, combined with the high surface concentration of Mn⁴⁺ (60.1%) and the introduction of TiO₂, the number of acidic sites on the Ce_0.1-Mn-Ti_0.1 catalyst increased significantly, improving its ability to adsorb NH₃ and, finally, promoting the SCR activity of Ce_0.1-Mn-Ti_0.1.

NO-TPD was conducted on an activity evaluation device, which was used to quantitatively analyze the NO adsorption capacity of the Ce_x-Mn-Ti_y catalysts. When the SCR reaction mainly follows the Langmuir–Hinshelwood (L-H) mechanism, the NO adsorption capacity plays a vital role in NH₃-SCR. In the NO-TPD curves of the Ce_x-Mn-Ti_y catalysts, there were two overlapped NO desorption peaks in two temperature regions of 150–250 and 250–500 °C (Figure 7B). Generally, the low-temperature desorption peak was attributed to desorption of physically adsorbed NO and decomposition of nitrite species, and the high-temperature desorption peak was ascribed to the decomposition of bridged nitrate species and bidentate nitrate species with higher thermal stability [27]. Furthermore, in comparison with the Mn-Ti_0.1 and Ce_0.1-Mn catalysts, the two desorption peaks of the Ce_0.1-Mn-Ti_0.1 catalyst shifted to a lower temperature, illustrating that the ability of NO to bind to adsorbed sites for the Ce_0.1-Mn-Ti_0.1 catalyst was weaker. To summarize, more gaseous NO on the Ce_0.1-Mn-Ti_0.1 catalyst can desorb and then be involved in the SCR reaction at lower temperatures, thereby improving the catalytic performance. Nevertheless, as listed in Table 4, the order of the adsorption capacity was Ce_0.1-Mn (2.00 × 10⁷) > Ce_0.1-Mn-Ti_0.1 (1.82 × 10⁷) > Mn-Ti_0.1 (1.57 × 10⁷), indicating that the adsorption of NO occurs mainly over both Ce and Mn active sites. NO adsorbed on basic surfaces of Mn⁴⁺/Ce⁴⁺, forming NO⁻ intermediate species and Mn³⁺/Ce³⁺, while O₂ adsorbed on Ce³⁺/Mn³⁺, generating active O⁻ and Ce⁴⁺/Mn⁴⁺. Then, NO⁻ intermediate species reacted with active O⁻, forming adsorbed nitrate/nitrite species. The integrated areas of desorption peaks per unit of S_BET were ranked as follows: Mn-Ti_0.1 (5.3 × 10⁵) > Ce_0.1-Mn (4.3 × 10⁵) > Ce_0.1-Mn-Ti_0.1 (3.5 × 10⁵). These results indicate that the adsorption capacity per unit of specific surface area is not the decisive factor affecting the activity. In this work, acidity played a key role in catalytic activity.

2.4. In Situ DRIFT Spectroscopy

2.4.1. Adsorption of NO + O₂ Followed by NH₃

The Ce_x-Mn-Ti_y catalysts were pretreated with Ar at 400 °C for 1 h to eliminate the surface-adsorbed H₂O and CO₂, and then NO + O₂ co-adsorption in situ DRIFT spectroscopy analysis was carried out on the Ce_x-Mn-Ti_y catalysts at 100 °C using 500 ppm of NO and 5 vol.% O₂ to further investigate the adsorbed NO_x species. In Figure 8A, it is clear that 1 min after the introduction of mixed gas (500 ppm of NO, 5% O₂ and Ar for balance), there were two prominent vibration bands on Ce_0.1-Mn-Ti_0.1 at 1259 and 1217 cm⁻¹, which were both assigned to bridged nitrates [29,40]. The bands at 1608 and 1394 cm⁻¹, corresponding to bridged nitrates and trans-N₂O₂²⁻ species [9,41], appeared after 5 min. Moreover, with the increase in the adsorption time, the band intensity enlarged gradually after 60 min. Figure 8B displays several bands at 1568, 1456, 1396, 1338, and 1207 cm⁻¹, showing that NO_x species adsorbed on the surface of the Ce_0.1-Mn catalyst. The band at 1207 cm⁻¹ appeared after 1 min and disappeared after 10 min, assigned to bridged nitrates. The bands at 1396 and 1568 cm⁻¹ could be assigned to trans-N₂O₂²⁻ and bidentate nitrates, while the bands at 1456 and 1338 cm⁻¹ belonged to monodentate nitrates [27,42,43]. Referring to the Mn-Ti_0.1 catalyst, it can be clearly seen from Figure 8C that several bands appeared. The bands at 1599, 1267, and 1225 cm⁻¹ were recognized as bridged nitrates, and the bands at 1475, 1348, and 1055 cm⁻¹ were related to monodentate nitrates [27]. The band at 1383 cm⁻¹ was assigned to trans-N₂O₂²⁻ species. The band intensity increased rapidly on Ce_0.1-Mn-Ti_0.1, implying that adding Ti and Ce could enhance the adsorption ability of NO + O₂ at 100 °C.

Prior to the introduction of 500 ppm of NH₃, it is necessary to purge with Ar (50 mL/min) for 60 min to completely remove the residual mixed gas. After purging with Ar for 60 min, most strongly adsorbed bands remained. Figure 9A shows the in situ DRIFT spectra of the Ce_0.1-Mn-Ti_0.1 catalyst under NO + O₂ adsorption conditions, followed by the introduction of NH₃. The bands at 1259 and 1217 cm⁻¹ assigned to bridged nitrates declined rapidly when continuously exposed to NH₃, implying that bridged nitrates could take part in the NH₃-SCR reaction. Surprisingly, a few new bands appeared; the band at 1660 cm⁻¹ was assigned to NH₃ coordinated to B acid sites, and the bands at 1595 and 1576 cm⁻¹ were ascribed to NH₃ coordinated on L acid sites [44,45]. The band at 1386 cm⁻¹ was stronger than before, which may have been caused by the overlap of the bands. As mentioned in the literature [9], the band at 1380–1400 cm⁻¹ could be NH₃ adsorbed on B acid sites. The bands ascribed to trans-N₂O₂²⁻ species and NH₃ adsorbed on B acid sites overlapped at 1386 cm⁻¹. The in situ DRIFT spectra of Ce_0.1-Mn are shown in Figure 9B; the intensity of adsorbed bands at 1568, 1456, and 1338 cm⁻¹ all became weaker, illustrating that the ads-NO_x species were slowly consumed by NH₃. The band at 1575 was ascribed to NH₃ coordinated on L acid sites, and the bands at 1647 and 1471 cm⁻¹ were related to NH₃ bonded to B acid sites [46,47]. Similarly, as for the Mn-Ti_0.1 catalyst, the bands at 1599, 1571, 1377, and 1352 cm⁻¹ after NH₃ adsorption for 60 min were still observed (Figure 9C), implying that the ads-NO_x species did not completely react with NH₃. In addition, the band at 1599 cm⁻¹ became stronger due to the NH₃ adsorbed on L acid sites. As can be seen in Figure 8C and Figure 9C, the bands at 1267 and 1225 cm⁻¹ disappeared after exposure to Ar for 60 min, demonstrating the weak adsorption of bridged nitrates. When NH₃ and NO both adsorb on the surface and then react to generate N₂ and H₂O, the reaction path is called the L-H mechanism. Therefore, it can be inferred from the consumption of ads-NO_x by NH₃ that the Ce_0.1-Mn-Ti_0.1 catalyst exhibited the best catalytic performance at 100 °C, following the L-H path.

2.4.2. Adsorption of NH₃ Followed by NO + O₂

Similar to Section 2.4.1, after Ar pretreatment at 400 °C for 60 min to remove the interference of H₂O and CO₂, 500 ppm of NH₃ was introduced to the Ce_x-Mn-Ti_y catalysts to analyze the reactively adsorbed NH₃ species in more detail. In Figure 10A, the band at 1572 cm⁻¹ was assigned to NH⁴⁺ on L acid sites, while the bands at 1658 and 1469 cm⁻¹ were attributed to NH₃ corresponding to B acid sites of Ce_0.1-Mn-Ti_0.1. Additionally, the bands at 933 and 966 cm⁻¹ were correlated with gaseous NH₃ or weakly adsorbed NH₃ [27]. The bands at 1377 and 1352 cm⁻¹ were assigned to the oxidized species of the adsorbed ammonia species and amide (⁻NH₂) species bonded to L acid sites [48,49]. The peak intensities increased with the extension of adsorption time. In Figure 10B, for the Ce_0.1-Mn catalyst, a series of peaks were observed at 1660, 1577, 1373, 964, and 924 cm⁻¹. Similar to the Ce_0.1-Mn-Ti_0.1 catalyst, the peaks at 964 and 924 cm⁻¹ were related to weakly adsorbed NH₃ and gaseous NH₃, respectively. The peaks at 1660 and 1577 cm⁻¹ were related to NH₄⁺ on B acid sites and NH₃ coordinated on L acid sites, respectively, while the peak at 1373 cm⁻¹ was the same as the peak at 1377 cm⁻¹. As seen in Figure 10C, similar peaks appeared at 1655, 1599, 1577, 1468, 1377, 1352, 1176, 966, and 930 cm⁻¹. The peaks at 1655 and 1468 cm⁻¹ were assigned to NH₄⁺ species on B acid sites, while the peaks at 1599, 1577, and 1176 cm⁻¹ were attributed to NH₃ on L acid sites [50]. Considering the three spectra comprehensively, the intensities of the NH₃ adsorption peaks on Ce_0.1-Mn-Ti_0.1 were the largest—much higher than those of the Ce_0.1-Mn and Mn-Ti_0.1 catalysts—implying the strongest ability to adsorb NH₃. With the increase in adsorption time, the intensities of the adsorption peaks reached the maximum value when injecting NH₃ for 60 min.

Figure 11A–C show the in situ DRIFT spectroscopy spectra of Ce_0.1-Mn-Ti_0.1, Ce_0.1-Mn, and Mn-Ti_0.1 catalysts with pre-adsorbed NH₃ followed by NO + O₂, respectively. As shown in Figure 11A, the intensities of the peaks at 1658, 1572, 1469, 1377, and 1352 cm⁻¹ declined gradually when exposed to NO + O₂ for 60 min, suggesting that the above adsorbed NH₃ species were active in the NH₃-SCR reaction at 100 °C. A new peak appeared at 1273 cm⁻¹ at 5 min, which was assigned to bridged nitrates, and increased with the exposure time, implying that NO could be adsorbed on the extra sites of the Ce_0.1-Mn-Ti_0.1 surface. When it came to the Ce_0.1-Mn catalyst, the three peaks at 1660, 1574, and 1373 cm⁻¹ were weakened, as shown in Figure 11B, meaning that the ads-NH₃ reacted with gaseous NO. For the Mn-Ti_0.1 catalyst, the intensities of the peaks at 1655, 1599, and 1572 cm⁻¹ were slightly decreased, while the peaks at 1379 and 1352 cm⁻¹ remained the same, as shown in Figure 11C. Additionally, there was a new weak peak at 1273 cm⁻¹, attributed to bridged nitrates at 5 min. As adsorbed NH₃ species reacting with gaseous NO, the reaction path is the E-R mechanism. Therefore, it can be concluded that the Ce_0.1-Mn-Ti_0.1, Ce_0.1-Mn, and Mn-Ti_0.1 catalysts all follow the E-R mechanism.

Moreover, in situ DRIFT spectroscopy was performed on Ce_x-Mn-Ti_y catalysts at 50 °C, as displayed in Figures S6–S9, where a similar phenomenon was found. Combined with the results of both pre-adsorbed NH₃ followed by NO + O₂ and pre-adsorbed NO + O₂ followed by NH₃ at 100 °C, the NH₃-SCR reaction is more likely to take place on the Ce_0.1-Mn-Ti_0.1 catalyst, thanks to its greater adsorption capacity for NH₃ and NO + O₂. On the one hand, in the process of pre-adsorbed NO and O₂ followed by NH₃, the disappearance of adsorbed NO_x species (1259 and 1217 cm⁻¹) demonstrated that the reaction between the active NO_x species and the adsorbed NH₃ followed the L-H mechanism. On the other hand, during the procedure of pre-adsorbed NH₃ followed by injecting NO + O₂, the peaks related to ads-NH₃ (1658, 1572, 1469, 1377, and 1352 cm⁻¹) were all obviously reduced as soon as NO + O₂ was introduced into the feed, implying that the reaction can follow the E-R mechanism as well.

3. Materials and Methods

3.1. Catalyst Preparation

The Ce-Mn-Ti mixed-oxide catalysts were prepared using the coprecipitation method, as previously reported [27]. All the reagents used in the experiment were purchased from Sinopharm Corporation in Shanghai, China. MnSO₄, Ce(NO₃)₃·6H₂O, and Ti(SO₄)₂ were dissolved in 50 mL of deionized water at room temperature. After stirring for 30 min, the mixed solution and 0.2 M Na₂CO₃ were dropped simultaneously into a 500 mL beaker kept at pH 9 for precipitation. After this solution was stirred for 6 h, the precipitate was obtained by filtration and washing with H₂O. The filter cake was dried overnight at 80 °C and, finally, calcined at 550 °C for 4 h. The prepared catalysts were named Ce_x-Mn-Ti_y, where x represents the theoretical molar ratio of Ce/Mn (x = 0, 0.1), and y represents the theoretical molar ratio of Ti/Mn (y = 0, 0.1). Ce_0.1-Mn, Mn-Ti_0.1, and Ce_0.1-Ti_0.1 were all synthesized by the same method.

3.2. Catalyst Characterization

The XRD measurement was conducted using a Brook/D8 diffractometer (Brooks Automation, Chelmsford, MA, USA) employing Cu Kα radiation. The diffraction patterns were obtained in the 2θ range of 10 to 80°, with a step size of 0.02°. The crystalline phase was confirmed through comparison with the reference data from ICDD files. Raman spectroscopy was performed using a Via Reflex spectrophotometer (Renishaw, Gloucestershire, UK). The N₂ adsorption–desorption isotherms were measured with a TriStar instrument (Micromeritics Instrument Corp.; Norcross, GA, USA). Before the measurements, all samples were degassed at 180 °C until a stable vacuum of ca. 5 mTorr was reached. The specific surface area was calculated from desorption data via the Brunauer–Emmett–Teller (BET) method. The X-ray photoelectron spectroscopy (XPS) spectra were obtained using a Thermo ESCALAB 250Xi spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) with Al Kα radiation. The binding energy of C 1s was 284.8 eV, which was defined as the reference.

Temperature-programmed desorption of NH₃ (NH₃-TPD) was conducted on a PX200 apparatus (Tianjin Pengxiang Technology Corporation, Tianjin, China) with a thermal conductivity detector (TCD). The catalyst (50 mg) was added to the quartz reactor and then pretreated at 400 °C in a flow of Ar (50 mL/min) for 1 h. When being cooled to room temperature, the sample was exposed to a flow of 10% NH₃/Ar (50 mL/min) for 30 min. After the sample was purged with Ar (50 mL/min) for 1 h, NH₃-TPD was carried out by heating the catalysts in Ar (50 mL/min) from room temperature to 400 °C at a rate of 10 °C/min.

Temperature-programmed desorption of NO (NO-TPD) was performed via the activity evaluation technique with a Thermo Fisher 42i-HL-NO-NO_x analyzer (Thermo Fisher Scientific, Waltham, MA, USA) as the detector. The sample was pretreated in Ar (300 mL/min) at 400 °C for 1 h, and then cooled to room temperature. Then, the sample was exposed to a flow of 500 ppm of NO/Ar (300 mL/min) for 1 h (when the NO_x on the samples reached a saturated state), followed by pure Ar (300 mL/min) purging for 1 h. Finally, NO-TPD was performed by heating the sample in Ar (300 mL/min) from room temperature to 500 °C at 10 °C/min.

Temperature-programmed reduction of H₂ (H₂-TPR) experiments were also conducted, using a VDSorb-92i-TPR apparatus (Tianjin Pengxiang Technology Corporation, Tianjin, China) with a TCD. The 50 mg sample was directly heated from room temperature to 800 °C at a rate of 10 °C/min, in a flow of 10 vol.% H₂/N₂ (40 mL/min). The hydrogen consumption was quantitatively evaluated by the TCD signal.

In situ DRIFT measurements were performed using a Nicolet 6700 FTIR spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) with an MCT detector. The DRIFT cell contained ZnSe windows and a connected gas flow system. The sample was pretreated at 400 °C in Ar for 1 h, and then cooled to 50 °C in Ar. The background spectra were collected after the temperature became stable, and the background spectra were subtracted from the sample spectra accordingly.

3.3. Catalytic Activity Testing

The catalytic activities of the Ce_x-Mn-Ti_y catalysts for NH₃-SCR in excess oxygen were investigated at atmospheric pressure in a fixed-bed continuous-flow quartz reactor (inner diameter 6 mm). The amount of catalyst (40–60 mesh) used was 300 mg. The reactant gas was composed of 500 ppm of NO, 500 ppm of NH₃, 5% O₂, 2% H₂O (when used), 50 ppm of SO₂ (when used), and Ar as the balance gas. The gas hourly space velocity (GHSV) was about 50,000 h⁻¹. The concentrations of NO and NO₂ remaining in the reaction gases were analyzed using the NO_x analyzer (Thermo Fisher 42i-HL-NO-NO_x analyzer). Since the oxidation of ammonia in the converter of the NO/NO_x analyzer may cause unexpected generation of NO_x, an ammonia trap containing phosphoric acid solution was installed prior to the analyzer. N₂O and NH₃ were monitored using a Nicolet 6700 FT-IR spectrometer with an MCT detector. NO_x conversion (x(NO_x)) and N₂ selectivity (S(N₂)) were calculated as follows:

$$\mathrm{x}\left(\mathrm{NOx}\right)=\frac{\mathrm{C}{\left({\mathrm{NO}}_{\mathrm{x}}\right)}_{\mathrm{in}}-\mathrm{C}{\left({\mathrm{NO}}_{\mathrm{x}}\right)}_{\mathrm{out}}}{\mathrm{C}{\left({\mathrm{NO}}_{\mathrm{x}}\right)}_{\mathrm{in}}}\times 100\%$$

(1)

$$\mathrm{S}\left({\mathrm{N}}_2\right)=\frac{\mathrm{C}{\left({\mathrm{NO}}_{\mathrm{x}}\right)}_{\mathrm{in}}+\mathrm{C}{\left({\mathrm{NH}}_3\right)}_{\mathrm{in}}-\mathrm{C}{\left({\mathrm{NO}}_{\mathrm{x}}\right)}_{\mathrm{out}}-\mathrm{C}{\left({\mathrm{NH}}_3\right)}_{\mathrm{out}}-2\mathrm{C}{\left({\mathrm{N}}_2\mathrm{O}\right)}_{\mathrm{out}}}{\mathrm{C}{\left({\mathrm{NO}}_{\mathrm{x}}\right)}_{\mathrm{in}}+\mathrm{C}{\left({\mathrm{NH}}_3\right)}_{\mathrm{in}}-\mathrm{C}{\left({\mathrm{NO}}_{\mathrm{x}}\right)}_{\mathrm{out}}-\mathrm{C}{\left({\mathrm{NH}}_3\right)}_{\mathrm{out}}}\times 100\%$$

(2)

where C(NO_x)_in and C(NO_x)_out are the concentrations of NO_x in the inlet and outlet, respectively, C(NH₃)_in and C(NH₃)_out are the concentrations of NH₃ in the inlet and outlet, respectively, and C(N₂O) is the concentration of N₂O generated during the reaction process.

4. Conclusions

The Ce_0.1-Mn-Ti_0.1 catalyst with abundant acidic sites and proper redox ability reveals an outstanding ability to adsorb NH₃ and NO, which gives rise to significant low-temperature activity (more than 90% NO_x conversion in the range of 75 to 225 °C) and improved SO₂/H₂O resistance. Compared to the Ce_0.1-Mn and Mn-Ti_0.1 catalysts, the incorporation of Ti can restrain the MnO_x crystallization, while Ce doping can promote the dispersion of Ti and further increase the specific surface area of the Ce_0.1-Mn-Ti_0.1 catalyst. On the basis of the results of XPS, the surface of the Ce_0.1-Mn-Ti_0.1 catalyst possessed the greatest contents of high-valence Mn ions (Mn⁴⁺), active adsorbed oxygen species (O_S), and Ce³⁺. In addition, charge transfer can occur between Mn and Ce cations to promote the redox cycle. The results of in situ DRIFT spectroscopy reflected that the NH₃-SCR reaction can take place simultaneously according to both the L-H and E-R paths. Hence, the Ce_0.1-Mn-Ti_0.1 catalyst can be regarded as a potential low-temperature catalyst for the industrial application of NH₃-SCR.