The Excellent Anti-K Ability of CeSbTi Oxide Catalysts for Selective Catalytic Reduction of NO with NH₃

Abstract

A novel K-resistant CeSbTi mixed oxide catalyst was prepared by co-precipitation method for ammonia selective catalytic reduction (NH₃-SCR) of NO_x. The experimental results show that the introduction of Sb₂O₅ can significantly improve the catalytic activity of the CeTi catalyst. The modulated CeSbTi catalyst has good resistance to K, and the NO_x conversion rate was as high as 95% after K poisoning. Its superior catalytic activity could be ascribed to the large specific surface area with increased acid sites and more oxygen defects and Ce³⁺ species after the introduction of Sb₂O₅, which prompt NH₃ adsorption and activation. In addition, NH₃-SCR reaction over CeSbTi and K/CeSbTi catalysts follows the E-R mechanism. The introduced Sb-O bond as the base capture site preferentially binds to potassium and releases part of the active Ce sites, thus retaining more acid sites and oxygen defects to a certain extent.

1. Introduction

Nitrogen oxides (NO_x) are a major contributor to environmental issues including acid rain, stratospheric ozone depletion, and photochemical smog [1]. These pollutants pose serious risks to human health and ecosystems. Currently, selective catalytic reduction (SCR) is the most widely used and effective technology for NO_x removal. However, in practical applications, SCR catalysts are easily deactivated by fly ash components in flue gas [2]. The flue gas generated from coal- and biomass-fired power plants contains substantial amounts of alkali or alkaline earth metals (e.g., K, Na, Ca) [3,4], which can damage the surface acidity and redox ability of SCR catalysts, leading to catalyst deactivation [5,6].

Antimony oxide has been employed as a promoter in some NH₃-SCR catalysts due to its favorable surface acidity. For instance, Kwon et al. introduced Sb and Mo into a VTi catalyst and observed that the resulting VMoSbTi catalyst exhibited enhanced catalytic activity, with the Mo–Sb addition suppressing the reaction between terminal V=O and SO₂ (g) [7]. Similarly, Wu et al. synthesized Sb-doped CeVO₄ catalysts via a hydrothermal method and found that Sb species reduced the crystallization of CeVO₄ while improving surface acidity and reducibility [8]. These studies suggest that Sb₂O₅ may serve as a promising additive for CeTi catalysts to optimize alkali metal tolerance, owing to its ability to enhance reducibility and surface acidity, as well as the possible synergistic interaction between Sb and Ce species.

In this work, we developed a CeSbTi catalyst with improved resistance to potassium poisoning. Through a series of characterizations, we investigated the influence of Sb₂O₅ addition on the physicochemical properties of CeSbTi. The positive effect on alkali metal tolerance was explored, and the surface reaction behavior before and after alkali metal poisoning was examined.

2. Results and Discussion

2.1. Catalyst Performance

As shown in Figure 1a, the NH₃-SCR performance of CeSbTi catalysts with different Sb feeding ratios was evaluated. The catalyst with a Ce:Sb:Ti molar ratio of 1:1.5:20 exhibited the best SCR performance, achieving over 90% NO conversion in the temperature range of 240–450 °C, which is similar to that of a commercial V₂O₅–WO₃/TiO₂ catalyst [6]. The introduction of Sb broadens the active temperature window of the CeTi catalyst. In contrast, the SbTi catalyst (without Ce) showed only about 20% NO conversion in the range of 150–350 °C (Figure S1). The alkali resistance of CeTi and CeSbTi (Ce:Sb:Ti = 1:1.5:20) was then evaluated (Figure 1b). After K poisoning, K/CeSbTi maintained nearly 90% NO conversion in the temperature window of 300–400 °C, while the maximum NO conversion of K/CeTi was only 52.9%. These results indicate that Sb improves both low-temperature catalytic activity and alkali resistance. The stability and N₂ selectivity of NH₃-SCR catalysts are critical for practical applications. As shown in Figure 1c, the introduction of Sb improves the N₂ selectivity of CeTi; CeSbTi exhibits nearly 100% N₂ selectivity in the range of 150–450 °C. Regarding stability (Figure S2), when CeSbTi and K/CeSbTi were operated at 250 °C for 24 h, their NO conversion remained unchanged, indicating excellent stability.

The flue gas produced by biomass fuel combustion also contains substantial amounts of alkali metals, alkaline earth metals, and heavy metals such as Na, Ca, Mg, and Pb. Therefore, the tolerance of CeSbTi to these metals was also tested (Figure 1d). CeSbTi shows good tolerance to all of them. The order of toxicity to CeSbTi is: K > Na > Ca > Pb > Mg.

2.2. Crystal Structure, Micro Morphology and Surface Analysis

XRD patterns of the catalysts are shown in Figure 2a. XRD patterns of all catalysts contain diffraction peaks of anatase TiO₂ (PDF #21-1272). A low-intensity peak at 31° is observed in all samples, which can be assigned to brookite TiO₂ (PDF # 76-1936).The average crystallite sizes of anatase TiO₂ were calculated using the Scherrer equation based on the (101) diffraction peak. The grain size decreases from 42 nm (CeTi) to 31 nm (CeSbTi), indicating that the addition of Sb₂O₅ inhibits the growth of TiO₂ crystallites [9]. This suggests that Sb species act as a structural promoter, limiting particle sintering and maintaining smaller crystallite sizes. After potassium loading, the grain sizes of both catalysts decrease slightly from 42 nm to 38 nm for K/CeTi, and from 31 nm to 25 nm for K/CeSbTi. No diffraction peaks corresponding to CeO_x, SbO_x, or K-containing species are detected in any of the catalysts. This indicates that Ce, Sb and K species are present as amorphous phases or are highly dispersed on the TiO₂ surface [10].

Raman spectra of the catalysts are presented in Figure 2b. All catalysts exhibit five characteristic peaks of anatase TiO₂ at approximately 144 (Eg), 197 (Eg), 397 (B₁g), 515 (A₁g + B₁g), and 639 cm⁻¹ (Eg) [11,12], which is consistent with the XRD results. In the magnified view of the 100–200 cm⁻¹ region, the K/CeTi catalyst shows an increased band intensity and a slight blue shift relative to CeTi, indicating that the introduction of K affects the lattice structure of TiO₂. For the CeSbTi catalyst, the main Eg peak exhibits a red shift relative to CeTi, which is ascribed to the incorporation of Sb into the TiO₂ lattice. After K poisoning, the peak position of K/CeSbTi remains almost unchanged compared to CeSbTi, indicating that the presence of Sb suppresses the K-induced lattice distortion of TiO₂.

The morphology of all catalysts was observed by SEM, revealing uniformly distributed nanoparticles (Figure S3). No significant morphological difference was observed between the fresh and K-poisoned samples. EDS mapping analysis (Figure S4) confirmed that all elements (Ce, Ti, Sb, and K where applicable) are evenly distributed on the catalyst surfaces. To determine the BET surface areas of the catalysts, N₂ adsorption–desorption experiments were conducted (Figure S5 and

Table 1. Physical properties of four catalysts.

Sample	BET Surface Area (m2/g)	Mean Pore Diameter (nm)	Total Pore Volume (cm3/g)
CeTi	47.16	7.8	0.11
K/CeTi	56.03	8.0	0.13
CeSbTi	111.27	5.4	0.19
K/CeSbTi	75.85	8.4	0.20

). Among the fresh catalysts, CeSbTi exhibits a higher specific surface area than CeTi, indicating that the introduction of Sb optimizes the textural properties of the CeTi catalyst. A larger BET surface area provides more active sites for SCR reactions, which is favorable for gas adsorption and activation [13]. After potassium loading, the specific surface area of the CeTi catalyst increases unexpectedly. This may be attributed to K₂O-induced changes in the crystallinity or structural reorganization of the CeTi catalyst during thermal treatment [14]. After K loading, the total pore volume of CeTi and CeSbTi both increased; this may be attributed to the deposition of toxic species primarily on the outer surface of the catalysts without affecting the interior, which simultaneously increases the number of accumulated pores on the material surface, thereby increasing the pore volume of the materials [15]. In contrast, the specific surface area of the CeSbTi catalyst decreases after K loading, which is likely due to a surface covering effect of K₂O species [16].

In the UV-vis DRS spectrum (Figure 2c), the light absorption edge of TiO₂ is around 400 nm [17], and the light absorption intensity is weak in the visible light region. Compared with TiO₂, the CeTi catalyst exhibits an absorption edge at 460 nm (band gap of ~2.70 eV), extending into the visible light region and resulting in a red-shift phenomenon. This red-shift is attributed to the formation of Ti–O–Ce bonds upon Ce introduction, which facilitates electron transfer and shifts the Fermi level. When Sb is further introduced, the absorption edge of CeSbTi shifts to 442 nm (band gap of ~2.81 eV), indicating a slight blue shift relative to CeTi. This blue shift suggests that the addition of Sb reduces the polymerization degree of CeO_x species and improves their dispersion, thereby promoting the reduction performance of the catalyst. After potassium poisoning, a slight blue shift in the absorption edge is observed, which may be due to the weakening of the Ce–Ti interaction by K.

XPS characterizes the valence states of different elements on the catalyst surface. In Figure 2d, the Ce 3d spectrum is deconvoluted into eight characteristic peaks [18,19]. The Ce³⁺ ratios of the CeTi, K/CeTi, CeSbTi, and K/CeSbTi catalysts were 34.7%, 29.8%, 47.2%, and 44.3%, respectively. In principle, compared with Ce⁴⁺, Ce³⁺ ions possess an electronic density of states closer to the Fermi level, which is known to enhance O₂ activation and promote the generation of reactive oxygen species, thereby improving the redox ability. This is consistent with the H₂-TPR results [20,21]. The introduction of Sb can form a redox cycle (Sb³⁺ + 2Ce⁴⁺ ↔ Sb⁵⁺ + 2Ce³⁺) [22], which induces the formation of Ce³⁺ and enhances the redox capacity. For K-poisoned catalysts, the Ce³⁺ ratios of K/CeTi and K/CeSbTi decreased, indicating that K inhibited the redox ability of the catalysts. As summarized in

Table 2. XPS analysis results.

Sample	Ce (at.%)	O (at.%)	Ti (at.%)	Sb (at.%)	K (at.%)	Ce3+/(Ce3+ + Ce4+) (%)	Sb5+/(Sb3+ + Sb5+) (%)
CeTi	2.3	65.5	32.2	-	-	34.7	-
K/CeTi	2.1	65.8	30.6	-	1.5	29.8	-
CeSbTi	1.9	72.0	24.3	1.8	-	47.2	79.4
K/CeSbTi	1.9	71.2	23.4	1.7	1.8	44.3	70.1

, the surface element distributions of fresh and poisoned catalysts are presented. The Ce content on the CeTi catalyst surface decreased after potassium poisoning, which may be caused by K covering the Ce active sites. In contrast, the Ce content on the CeSbTi surface remained almost unchanged after potassium poisoning, while the Sb content decreased. This suggests that Sb–O sites preferentially bind to potassium ions, thereby protecting the Ce active sites.

The Sb 3d XPS spectra of CeSbTi catalysts divided into two photoelectron peaks at ~530.5 eV and ~540.2 eV, which could be ascribed to the Sb 3d_5/2 and Sb 3d_3/2 sublevels, respectively [23]. Since the Sb 3d_5/2 sublevels were completely overlapped with the O 1s spectra, only the Sb 3d_3/2 spectra were fitted in order to avoid the interference of the O 1s core-level. The Sb 3d_3/2 spectrum is divided into two peaks at ~539.5 eV and ~540.6 eV, Sb⁵⁺ and Sb³⁺. The ratio of Sb is shown in Table 2. After the CeSbTi catalyst was poisoned by K, the proportion of Sb⁵⁺ decreased, and the binding energy also shifted to a lower position, this may be due to interaction of K and Sb. In the XPS spectrum of Ti 2p_3/2, for the poisoned catalyst, the binding energy of K/CeTi shifted to a lower position, indicating that K inhibited the interaction of Ce and Ti. While K/CeSbTi exhibits a slightly smaller offset, it shows a higher Ti 2p_3/2 binding energy than the K/CeTi catalyst, indicating that Sb species can reduce the negative impact of K poisoning on the Ce–Ti interaction.

2.3. Surface Acidity and Redox

The surface acidity of a catalyst plays an important role in the SCR reaction; therefore, NH₃-TPD was employed to evaluate the NH₃ adsorption/desorption behavior of the catalysts, and the results are shown in Figure 3a. All catalysts exhibit a continuous desorption band in the range of 100–500 °C, which corresponds to acid sites with different thermal stabilities, including physisorbed/weakly adsorbed NH₃ (I), medium-strength acid sites (II), and strong acid sites (III) [24,25]. These three types of acid sites are observed on CeTi, CeSbTi, and K/CeSbTi catalysts, whereas K/CeTi lacks strong acid sites. The integrated NH₃-TPD peak area reflects the relative amount of acid sites on the catalyst surface [26]. The integrated NH₃ desorption peak areas are summarized in

Table 3. NH₃-TPD analysis results.

Sample	NH3 Desorption Peak Area (a.u.)			Total NH3 Desorption Peak Area (a.u.)	Retention Ratio (Poisoned Peak Area/Fresh Peak Area, %)
	SI	SII	SIII	SI + SII + SIII
CeTi	78.9	218.6	76.5	374.0	-
K/CeTi	35.8	87.6	-	123.5	33.0%
CeSbTi	143.8	159.7	240.2	543.7	-
K/CeSbTi	81.9	126.7	52.8	261.4	48.1%

for semi-quantitative comparison. CeSbTi exhibits the largest normalized peak area (543.7), indicating that it possesses the most acid sites, which can be attributed to the introduction of Sb. After K poisoning, the surface acidity of both CeTi and CeSbTi catalysts is impaired to varying degrees. As shown in Table 3, strong acid sites exhibit the most pronounced decline in peak area contribution after poisoning, suggesting that K⁺ preferentially binds to strong acid sites. Moreover, the peak area retention ratio of K/CeSbTi (48.1%) is higher than that of K/CeTi (33.0%), indicating that the introduction of Sb can partially preserve the acid sites of the CeSbTi catalyst against K poisoning.

The reduction behavior of the catalyst was tested by H₂-TPR (Figure 3b). For the CeTi catalyst, the reduction peaks at 384 °C are attributed to the reduction in surface-capping oxygen, the reduction peaks at 560 °C are attributed to the reduction in surface lattice oxygen and the reduction peaks at 719 °C are attributed to the reduction in bulk lattice oxygen. For the CeSbTi catalyst, the decrease in the reduction temperature after the adding of Sb indicates that Sb improves the reducibility of the catalyst. In addition, CeSbTi and K/CeSbTi catalysts have higher and sharper reduction peaks than CeTi and K/CeTi catalysts, due to the high dispersion and high reducibility of Ce species. Compared with the fresh catalyst and the poisoned catalyst, the reduction peak of the poisoned catalyst moved to a higher temperature, which indicates that the K poisoning weakened the reducibility of the catalyst. The reduction temperature of K/CeSbTi catalyst is lower than that of CeTi catalyst, indicating that CeSbTi catalyst still has good reducibility after poisoning.

The oxygen adsorption performance and surface acidity are two important factors that affect catalytic activity. The oxygen adsorption properties of the catalyst surface were investigated by O₂-TPD. As shown in Figure 3c, three O₂ desorption peaks are observed in the desorption profile. The peak at 50–300 °C is attributed to chemically adsorbed oxygen, the desorption peak at 300–700 °C corresponds to oxygen chemically adsorbed on oxygen vacancies, and the peak above 700 °C is assigned to lattice oxygen. The relative amounts of adsorbed oxygen, defect oxygen, and lattice oxygen can be estimated by curve integration for semi-quantitative comparison [27]. As presented in

Table 4. O₂-TPD analysis results.

Sample	Adsorbed Oxygen Peak Area (a.u.)	Defect Oxygen Peak Area (a.u.)	Lattice Oxygen Peak Area (a.u.)	Total Peak Area (a.u.)	Retention Ratio (Poisoned Total Peak Area /Fresh Total Peak Area, %)
CeTi	0.82	0.96	0.34	2.12	-
K/CeTi	0.69	0.30	0.41	1.39	65.8%
CeSbTi	0.95	1.12	0.36	2.43	-
K/CeSbTi	0.63	0.88	0.56	2.07	85.4%

, CeSbTi exhibits a higher defect oxygen peak area due to the addition of SbO_x, and a greater amount of oxygen vacancies facilitates oxygen adsorption; consequently, the CeSbTi catalyst displays higher catalytic activity. After K poisoning, the peak areas of adsorbed oxygen and defect oxygen decrease, while that of lattice oxygen increases, indicating that K inhibits oxygen adsorption, which in turn adversely affects the SCR reaction. Furthermore, the total oxygen peak area retention ratio of K/CeSbTi (85.4%) is higher than that of K/CeTi, suggesting that the introduction of Sb can partially preserve the oxygen species.

2.4. Catalytic Mechanism

2.4.1. NH₃ and NO + O₂ Adsorption

The in situ DRIFTS spectra are presented in Kubelka–Munk units. This transformation converts diffuse reflectance spectra into linearized absorbance units, which corrects for scattering effects and enhances the contrast of weak bands, thereby facilitating the identification and comparison of adsorbed species [28].

To further determine the adsorption of NH₃ by the catalysts, in situ DRIFTS experiments of NH₃ adsorption were carried out at elevated temperatures. The in situ DRIFTS spectra of NH₃ adsorbed on CeSbTi and K/CeSbTi are shown in Figure 4a,b. For the CeSbTi catalyst, after one hour of NH₃ adsorption, characteristic peaks corresponding to symmetric and asymmetric bending vibrations of N-H on Lewis acid sites (1593 and 1186 cm⁻¹) and absorption peaks of NH₄⁺ species on Brønsted acid sites (1670, 1463, and 1436 cm⁻¹) appeared [29,30]. After CeSbTi was poisoned by K, the number and strength of Brønsted acid sites (1474 cm⁻¹) on K/CeSbTi decreased, while the impact on Lewis acid sites (1608, 1400, 1306, and 1172 cm⁻¹) was less pronounced. This indicates that K preferentially binds to Brønsted acid sites on CeSbTi, resulting in their significant reduction [31,32]. As the temperature increased, the absorption peaks of NH₃ species on both CeSbTi and K/CeSbTi gradually decreased. At 300 °C, only the adsorption peaks of Lewis acid sites remained, while those of Brønsted acid sites disappeared. This observation indicates that Lewis acid sites on CeSbTi are much more stable than Brønsted acid sites, suggesting that Brønsted acid sites play a more important role in the low-temperature NH₃-SCR process. Therefore, it can be inferred that CeSbTi retains strong Lewis acid sites after K poisoning, which accounts for its better alkali resistance. Meanwhile, the weakening of Brønsted acid site strength results in the lower low-temperature catalytic activity of K/CeSbTi.

To further investigate the NO adsorption behavior, in situ DRIFTS experiments of NO adsorption were carried out on CeSbTi and K/CeSbTi at elevated temperatures (Figure 4c,d). For the CeSbTi catalyst, after exposure to NO + O₂, peaks attributed to bridging nitrate (1625 and 1243 cm⁻¹), bidentate nitrate (1582 and 1283 cm⁻¹), and monodentate nitrate (1538 and 1396 cm⁻¹) were observed [33,34]. As the temperature increased, the intensity of these peaks gradually decreased. Notably, the monodentate nitrate peaks (1538 and 1396 cm⁻¹) preferentially disappeared or transformed at lower temperatures, indicating their poor thermal stability. The bridging nitrate (1625/1243 cm⁻¹) and bidentate nitrate (1582/1283 cm⁻¹) species exhibited relatively higher stability. After K poisoning (K/CeSbTi, Figure 4d), the types of nitrate species remained unchanged, but the peak intensities decreased significantly. Moreover, the remaining species showed little change with increasing temperature, suggesting that K poisoning promotes the formation of nitrate species with enhanced thermal stability. After K poisoning, more stable adsorbed NO_x species are formed on the catalyst surface, which is unfavorable for the low-temperature SCR reaction.

2.4.2. Reaction Between NH₃ and NO_x Species

To explore the reaction between adsorbed NH₃ and NO_x species, an in situ DRIFTS experiment of pre-adsorbed NH₃ followed by NO + O₂ was carried out at 250 °C. As shown in Figure 5a, after pre-adsorption of NH₃ on the CeSbTi catalyst, NH₃ species adsorbed on Lewis acid sites (1603 and 1181 cm⁻¹) and Brønsted acid sites (1419 and 1360 cm⁻¹) were observed on the catalyst surface [35,36]. After the introduction of NO + O₂, all peaks gradually weakened over time and disappeared after 20 min. Simultaneously, peaks attributed to bidentate nitrate (1581 and 1339 cm⁻¹), monodentate nitrate (1538 and 1373 cm⁻¹), and bridging nitrate (1613 cm⁻¹) species emerged and grew over time [37]. For the K/CeSbTi catalyst (Figure 5b), obvious NH₃ species adsorbed on Lewis and Brønsted acid sites were also observed, indicating that CeSbTi remains highly acidic after potassium poisoning. After the introduction of NO + O₂, all peaks gradually weakened and disappeared after 30 min. Bridging and bidentate nitrate species were eventually observed on K/CeSbTi after 60 min. The fact that adsorbed NH₃ species on both catalysts can react with gaseous NO is consistent with an Eley–Rideal (E-R)-type pathway.

To explore the reaction between adsorbed NO_x and NH₃ species, an in situ DRIFTS experiment of pre-adsorbed NO + O₂ followed by NH₃ was carried out at 250 °C. As shown in Figure 5c, after pre-adsorption of NO + O₂ on the CeSbTi catalyst, peaks attributed to bridging nitrate (1608 and 1232 cm⁻¹), monodentate nitrate (1548 and 1388 cm⁻¹), and bidentate nitrate (1580 cm⁻¹) species were observed on the catalyst surface [38]. After the introduction of NH₃, the intensity of these nitrate peaks decreased only slightly over time and did not disappear by the end of the experiment. In addition, after 10 min of NH₃ flow, characteristic peaks of NH₃ species adsorbed on Lewis acid sites (1562 and 1206 cm⁻¹) and Brønsted acid sites (1418 cm⁻¹) appeared on the CeSbTi catalyst. For the K/CeSbTi catalyst (Figure 5d), after pre-adsorption of NO + O₂, peaks attributed to bridging nitrate, monodentate nitrate, and bidentate nitrate species were observed on the catalyst surface [39]. After the introduction of NH₃, the intensity of these nitrate peaks decreased only slightly over time and did not disappear by the end of the experiment. Moreover, after 5 min of NH₃ flow, characteristic peaks of NH₃ species adsorbed on Lewis acid sites (1206 cm⁻¹) appeared on the K/CeSbTi catalyst. Because the NO_x species pre-adsorbed on CeSbTi and K/CeSbTi hardly reacted with NH₃, these results further support that the catalytic mechanism over both catalysts is dominated by an Eley–Rideal (E-R) pathway.

2.5. Reaction Mechanism

As shown in Figure 6. First, NH₃ gas is adsorbed on the catalyst surface, and the adsorbed NH₃ is oxidized to NH₂. The generated NH₂ reacts with gaseous NO to produce N₂ and H₂O, which follows the Eley–Rideal (E-R) reaction mechanism. However, overoxidation of NH₂ generates NH, which further reacts with NO in the non-selective catalytic reduction (NSCR) reaction to form N₂O and H₂O. This side reaction is the main cause of reduced N₂ selectivity of the catalyst.

For the K-poisoned CeTi catalyst, K species reduce the number of acid sites on the catalyst surface, which is unfavorable for NH₃ adsorption. In addition, K species shield the Ce active sites and lower the Ce³⁺/Ce⁴⁺ ratio, thereby decreasing the concentration of defect oxygen. This further inhibits oxygen adsorption, reduces the redox ability of the catalyst, and suppresses the activation of adsorbed NH₃ species.

Compared with the CeTi catalyst, the CeSbTi catalyst exhibits much better retention of catalytic activity after K poisoning. Characterization results show no significant differences in specific surface area or crystal structure between the fresh and K-poisoned catalysts, indicating that these factors are not the primary cause of the observed activity difference. Notably, XPS results further reveal that antimony oxide species show a clear tendency to bind with K after poisoning, and the decrease in the Ce³⁺/Ce⁴⁺ ratio on CeSbTi is significantly smaller than that on CeTi. These results indirectly support the hypothesis that antimony oxide species act as “alkali capture sites”, preferentially binding to K and thereby protecting a portion of Ce active sites from K attack.

3. Materials and Methods

3.1. Catalyst Preparation

CeTi catalyst was synthesized by a co-precipitation method. Typically, 4.34 g of Ce(NO₃)₃·6H₂O (10 mmol, AR, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) was dissolved in 100 mL of ethanol (AR, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) solution and stirred for 10 min. Then, 6.81 mL of tetrabutyl titanate (20 mmol, AR, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) was added and the mixture was stirred for 30 min. Ammonia solution (25 wt%, 98%, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) was added dropwise to adjust the pH to 10, and the mixture was allowed to settle for 24 h. The precipitate was filtered by suction, rinsed with deionized water and ethanol until the pH reached 7, dried in an oven at 110 °C for 12 h, and finally calcined in a muffle furnace (KSL-1100X, Hefei Kejing Materials Technology Co., Ltd., Hefei, China) at 500 °C for 5 h with a heating rate of 5 °C/min. The molar ratio of Ce:Ti was 1:20.

CeSbTi catalysts with different Sb loadings were also prepared by the co-precipitation method. Different amounts of SbCl₃ (AR, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) and 0.434 g (1 mmol) of Ce(NO₃)₃·6H₂O were dissolved in 100 mL of ethanol solution, and the subsequent procedure was the same as that for CeTi. The molar ratios of Ce:Sb:Ti were 1:0.5:20, 1:1:20, 1:1.5:20, and 1:2:20, corresponding to SbCl₃ amounts of 0.114 g, 0.228 g, 0.342 g, and 0.456 g, respectively. Based on preliminary activity screening (Figure 1a), the catalyst with a molar ratio of 1:1:20 (denoted as CeSbTi) was selected for further characterization.

K-poisoned catalysts (1 wt% K) were prepared as follows: 1 g of CeTi or CeSbTi was dispersed in deionized water, and 0.0212 g of KNO₃ (0.21 mmol, AR, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) was added. The suspension was stirred for 20 min and then dried at 70 °C using a rotary evaporator (RE-52, Shanghai Yarong Biochemical Instrument Factory, Shanghai, China). The resulting powder was calcined in air at 500 °C for 5 h. The obtained catalysts were denoted as K/CeTi and K/CeSbTi, respectively. The same method was used to prepare catalysts poisoned with 1 wt% Pb, Na, Ca, and Mg (results shown in Figure 1d).

3.2. Catalyst Characterization

X-ray diffraction (XRD) patterns were recorded on a Rigaku SmartLab diffractometer (Rigaku Corporation, Tokyo, Japan) using Cu Kα radiation (λ = 1.5406 Å). The measurements were performed in the 2θ range of 10–80° with a scanning speed of 6°/min. The crystallite sizes of anatase TiO₂ were calculated using the Scherrer equation:

D = Kλ/(β cos θ),

(1)

where K = 0.89, λ = 1.5406 Å, β is the full width at half maximum (FWHM) of the (101) diffraction peak, and θ is the Bragg angle.

UV-vis-NIR diffuse reflectance spectra (DRS) were recorded on an Agilent Cary 5000 UV-vis-NIR spectrophotometer (Agilent Cary, Santa Clara, CA, USA) using BaSO₄ as a reference, with a wavelength range of 200–800 nm. The band gap energy (E_g) was calculated from the absorption edge (λ_edge) using the equation:

E_g = 1240/λ_edge,

(2)

X-ray photoelectron spectroscopy (XPS) was performed on a Thermo Scientific EscaLab 250Xi spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) using monochromatic Al Kα radiation (hν = 1486.6 eV). Binding energies were calibrated using the C 1s peak at 284.6 eV. Spectra were deconvoluted using XPSPEAK (Version: 4.1) software.

Raman spectra were recorded on a Horiba LabRAM HR Evolution confocal Raman microscope (Horiba Scientific, Kyoto, Japan) using a 532 nm laser as the excitation source.

N₂ physisorption was performed on a Micromeritics TriStar II 3020 analyzer (Micromeritics Instrument Corporation, Norcross, GA, USA). Prior to measurement, approximately 0.1 g of sample was degassed at 300 °C for 12 h under vacuum. The specific surface area was calculated using the Brunauer–Emmett–Teller (BET) method. The pore size distribution was determined using the Barrett–Joyner–Halenda (BJH) model applied to the adsorption branch of the isotherm.

NH₃ temperature-programmed desorption (NH₃-TPD) and H₂ temperature-programmed reduction (H₂-TPR) were carried out on a Tianjin Xianquan TP-5076 system (Tianjin Xianquan Instrument Co., Ltd., Tianjin, China). For NH₃-TPD, 0.08 g of catalyst was pretreated in He flow at 300 °C for 1 h, cooled to 50 °C, and then exposed to 1 vol% NH₃/He for 1 h. After purging with N₂ at 50 °C for 1 h to remove weakly adsorbed NH₃, the sample was heated to 500 °C at a ramp rate of 10 °C/min. The desorbed NH₃ was monitored using a thermal conductivity detector (TCD). The integrated peak areas were normalized by sample mass and are reported as relative values. For H₂-TPR, 0.08 g of sample was pretreated in He flow at 300 °C for 60 min and cooled to 50 °C. After baseline stabilization in 10 vol% H₂/Ar flow at 50 °C for 1 h, the temperature was ramped to 900 °C at a rate of 10 °C/min while monitoring H₂ consumption using a TCD.

In situ DRIFTS was conducted on a Thermo Scientific Nicolet iS 50 FTIR spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) with an in situ cell and a high-sensitivity detector. Spectra were acquired in Kubelka–Munk mode.

Steady-state experiments: The catalyst was heated to 300 °C in N₂ flow and held at this temperature for 30 min. After pretreatment, the sample was cooled and background spectra were recorded at various temperatures. Upon reaching the target temperature, NH₃ (or NO + O₂) was introduced for 1 h, followed by purging with N₂ for 15 min to remove weakly adsorbed species. Spectra were then collected at each temperature point (after stabilization for 10 min) under N₂ flow.

Transient experiments: For the reaction between pre-adsorbed NH₃ and NO + O₂, the sample was pretreated in N₂ at 300 °C for 30 min, cooled to 250 °C, and a background spectrum was recorded. NH₃ was then introduced for 1 h, followed by N₂ purging for 15 min. Subsequently, the gas flow was switched to NO + O₂, and spectra were recorded continuously for 1 h using an automatic sampling program. For the reaction between pre-adsorbed NO + O₂ and NH₃, the same pretreatment procedure was applied. After recording the background at 250 °C, NO + O₂ was introduced for 1 h, followed by N₂ purging for 15 min. Finally, NH₃ was introduced, and spectra were recorded at 1 min intervals for 1 h.

3.3. Catalytic Activity Measurements

The catalytic performance was evaluated in a quartz fixed-bed reactor (inner diameter: 8 mm) using 300 mg of catalyst (40–60 mesh) at temperatures ranging from 150 to 500 °C. The total flow rate was 300 mL/min, corresponding to a gas hourly space velocity (GHSV) of approximately 60,000 h⁻¹. The feed gas composition was as follows: 500 ppm NO, 500 ppm NH₃, 5 vol% O₂, balance N₂ (detailed in Table S1). NO, NO₂, and N₂O concentrations were measured online using a portable Fourier transform infrared multi-component gas analyzer (Model DX4000 GASMAT, Gasmet Technologies Oy, Helsinki, Finland). The reaction system (Model VDRT-SMT, WoDe Instrument Co., Ltd., Shanghai, China) was used for catalytic activity testing. The temperature was increased at a rate of 10 °C/min, and data were recorded after stabilizing at each test point for 10 min. All feed gases (NO, NH₃, O₂, and N₂) were supplied by Jinhou Special Gas Co., Ltd. (Jinhou, Haikou, China), Deli Meisel Gas Co., Ltd. (Deli Meisel, Foshan, China), and Date Gas Co., Ltd. (Date, Dalian, China).Detailed gas specifications and sources are provided in Table S1.

The NO conversion (X_NO) and N₂ selectivity were calculated using the following equations:

X_NO = ([NO_x]_in − [NO_x]_out)/[NO_x]_in × 100%,

(3)

$${\mathrm{N}}_2\mathrm{selectivity}(\%)=(1--{\displaystyle \frac{2\left[{\mathrm{N}}_2\mathrm{O}\right]\mathrm{o}\mathrm{u}\mathrm{t}}{\left[\mathrm{N}\mathrm{O}\right]\mathrm{i}\mathrm{n}-\mathrm{N}\mathrm{O}]\mathrm{o}\mathrm{u}\mathrm{t}})\times 100\%,}$$

(4)

where [NO]_in and [NO]_out are the inlet and outlet NO concentrations, respectively, and [N₂O]_out is the outlet N₂O concentration.

4. Conclusions

The modulated CeSbTi mixed oxide catalyst exhibits a wider active temperature window and better alkali resistance than the CeTi catalyst. Physicochemical characterization reveals that the addition of Sb₂O₅ increases the specific surface area and provides more acid sites, which promote NH₃ adsorption and activation. Moreover, the Sb–Ce interaction enhances the Ce³⁺ concentration and oxygen vacancy density, leading to improved redox ability. After K poisoning, a possible interaction between Sb and K may contribute to the protection of Ce active sites; however, other factors—including surface acidity, increased specific surface area, or enhanced redox ability—also contribute to the improved alkali resistance of the CeSbTi catalyst. The catalyst also demonstrates good stability, high N₂ selectivity, and strong anti-toxicity. Collectively, these results suggest that the CeSbTi catalyst is promising for practical applications.