A Comparative Study on the Effect of Surface and Bulk Sulfates on the High-Temperature Selective Catalytic Reduction of NO with NH₃ over CeO₂

Abstract

Herein, two CeO₂ samples dominantly decorated with surface and bulk sulfates were constructed and their distinct effects on high-temperature NH₃-SCR were investigated by strictly controlling the sulfate content at a comparable level. The obtainment of surface and bulk sulfates was revealed using a designed leaching experiment, and further evidenced by the characterization results from XPS and H₂-TPR. In comparison with CeO₂ modified with bulk sulfates (B-CeS), sufficient acid sites with strong intensity were generated on CeO₂ modified with surface sulfates (S-CeS). In addition, due to electron-withdrawing effect from S=O in sulfate species, NH₃ oxidation over S-CeS was greatly suppressed, providing an additional contribution to enhanced performance in high-temperature NH₃-SCR.

1. Introduction

The large and uncontrolled emission of nitrogen oxides (NO_x, x = 1,2) has caused serious environmental problems, and selective catalytic reduction of NO_x with NH₃ (NH₃-SCR) is testified to be one of the most effective methods for NO_x elimination from stationary sources and diesel vehicles [1]. Among the diverse catalysts developed for NH₃-SCR, the vanadium–titanium system, represented by V₂O₅-WO₃(MoO₃)/TiO₂, has been commercialized for SCR denitration for decades. The operation temperature for V₂O₅-WO₃(MoO₃)/TiO₂ catalysts typically lies in the medium temperature zone (300–400 °C), which is compatible with the flue gas temperature at boiler economizer outlets and thus often used for NO_x removal from coal-fired power plants [2]. However, the installation of this catalyst in high-temperature denitration is largely restrained, mainly due to the aggravated oxidation of ammonia by active V₂O₅ and inferior thermal stability of anatase TiO₂ [3,4]. From the viewpoint of practical application, there is an urgent need for treatment of exhaust gases via high-temperature SCR in some workplaces, such as diesel engines, gas turbines and waste incinerators [5,6,7].

As early as the 1990s, Topsøe and coworkers pointed out that acid and redox sites are essential for NH₃-SCR catalysts [8,9]. Whether for the Eley–Rideal (E-R) or the Langmuir–Hinshelwood (L-H) reaction pathway, NH₃ adsorption on a catalyst surface constitutes the first step to initiate a reaction. With an increase in reaction temperature, NH₃ adsorption will be disrupted and desorption tends to occur. As such, it is necessary that the catalysts designed for high-temperature SCR should acquire acid sites with sufficient intensity [10]. On the other hand, their redox properties also need to be carefully regulated, since excessive consumption of NH₃ via heavy oxidation is bound to reduce the available NH₃ for NO reduction. Following these two basic guidelines, a diverse array of catalysts has been explored and such typical combinations like metal ion-exchanged zeolite [11,12] and WO₃/TiO₂ [3,13] are found to exhibit excellent de-NO_x performance even at temperatures higher than 500 °C.

Ceria-based materials, as an important kind of rare-earth oxide, have been extensively studied as efficient V-free NH₃-SCR catalysts due to their abundant reserve, good redox property and tunable acidity [14,15]. Particularly, by taking advantage of the fascinating characteristics of rich oxygen vacancies and an easy Ce³⁺/Ce⁴⁺ redox cycle, substantial attention has been paid to CeO₂ for application in low-temperature NH₃-SCR [16]. Nevertheless, the catalytic behavior at high temperature is less considered. Actually, based on the experimental results of some recent studies, it was shown that by grafting strong acid sites onto the surface of ceria, the NO reduction efficiency at high temperatures (>400 °C) can also be significantly boosted. For example, Yang et al. reported that H₃PO₄-modified CeO₂ exhibited more than 80% NO conversion in the broad temperature range of 250–550 °C [17]. Our recent study revealed that when applying excessive SO₂ treatment (72 h) over CeO₂, the obtained catalysts displayed excellent activity at 500 °C with NO conversion exceeding 90% [18].

It is well known that under actual flue gas conditions, SO₂ is widely present. The interaction of SO₂ with CeO₂ will inevitably lead to the generation of sulfate species, and depending on the extent of the interaction, surface sulfates and bulk sulfates are expected to be formed [19,20]. Many previous studies have proved that irrespective of the specific structures or morphologies, sulfate species modification is conducive to enhance the acid strength of CeO₂ and, simultaneously, disrupt the redox properties to a certain extent [21,22,23,24]. Obviously, this would endow a positive effect on the performance of high-temperature SCR. However, further details are still lacking. For example, what is the difference between surface sulfate and bulk sulfate species on the catalytic performance of CeO₂ for high-temperature SCR, it is still obscure. Herein, two sulfate species-modified CeO₂ catalysts with comparable sulfur content but distinct existing states were fabricated, and the unique function of surface sulfates and bulk sulfates in high-temperature SCR was explored in detail. Results showed that in comparison with bulk sulfates, surface sulfates played a more significant role in promoting NO conversion at high temperatures, which could be reasonably attributed to the much stronger acid properties and worse redox capacity due to close surface coverage of sulfate species on CeO₂.

2. Results and Discussion

2.1. Construction of Target Catalysts

The CeO₂ catalysts modified with sulfate species have been widely investigated in NH₃-SCR, and conventional fabrication is fulfilled according to a bottom-up route, either through vapor sulfation of CeO₂ with SO₂ + O₂ or via wet impregnation of CeO₂ with H₂SO₄/(NH₄)₂SO₄ [24,25,26]. Both methods are facile in operation and one can obtain surface or bulk sulfates by tuning the exposure time or modulating the H₂SO₄ concentration. However, it is worth to mention that to explicitly approach the goal of present study, we needed to control the sulfate content at a similar value. This is challenging for the traditional preparation, since bulk sulfates are usually formed at the expense of surface sulfates. In other words, in order to generate bulk sulfates, the introduced sulfur content should be controlled at a very high value. For this reason, an alternative top-down preparation methodology was used to construct the designated catalyst of CeO₂ modified with bulk sulfates.

It is well reported that when CeO₂ is exposed to a feed stream of SO₂ + O₂ for a short duration, its surface will be covered by an adsorption layer of sulfates [27]. In the present study, this method was utilized to construct the sample of CeO₂ exclusively modified with surface sulfates (S-CeS). However, for the obtainment of CeO₂ modified with bulk sulfates, the traditional method of treating CeO₂ in a SO₂ + O₂ stream for a long duration or at a high temperature was not adopted. Alternatively, a novel but facile top-down strategy was performed. That is, directly calcining the sulfated precursor of Ce₂(SO₄)₃·xH₂O in air at a high temperature (i.e., 900 °C) to convert the majority of metal salts into CeO₂, and the residual is expected to present mainly as bulk sulfates. Obviously, one notable merit of this method is that it can well resolve the dilemma between low sulfur content and bulk state of sulfates.

To confirm the effectiveness of this method, a TG analysis was first performed. It is clearly visible in Figure 1 that two distinct decomposition processes are displayed. The first step is related to a loss of crystalline water. Compared to desorption of crystalline water, a distinct delay is detected for the thermal decomposition of sulfate species. Evidently, decomposition of sulfate groups cannot be initiated until the temperature reaches 700 °C, and a sharp weight loss (29.7%) occurs in the narrow temperature range of 700–850 °C. Notably, by inspecting the TG profile at a much higher temperature (830–1000 °C, inset), it is found that the decomposition is still in progress even when the temperature surpasses 900 °C. This can satisfactorily explain the existence of minor sulfur in the B-CeS sample after the harsh thermal treatment at 900 °C for 1 h.

EDS analysis (

Table 1. Summary of element analysis from EDS and XPS for B-CeS and S-CeS.

Samples	Atomic Concentration (mol.%)				Atomic Ratio (%)
	Ce	O	S	C	Ce3+/(Ce3+ + Ce4+)	Oβ/(Oα + Oβ)	Ce/S *
B-CeS	18.42	58.52	2.13	20.93	22.08	33.06	20.44
S-CeS	16.47	54.32	3.6	25.61	40.97	74.07	20.78

* Ce/S ratio determined from EDS analysis.

) was performed to approach the actual sulfur content in B-CeS and S-CeS. It can be seen from SEM images (Figure 2a,b) that due to the employment of different preparation routes, variations in the macroscopic morphology of the two samples are present, and S-CeS exhibits a more porous appearance with fine particles and tight arrangement. From Table 1, a S signal is detected and the measured content displays a low but comparable value (Ce/S ratio: 20.44 vs. 20.78) for B-CeS and S-CeS. Since a variation in the S content is expected to bring great distinction in acid and redox properties [19,27], the construction of sulfate CeO₂ with comparable sulfate loading provides a reliable comparison of the contribution from different sulfate species on the catalytic performance of ceria in high-temperature SCR.

In order to confirm the existence of different sulfate species in B-CeS and S-CeS, a designed water leaching experiment was carried out. According to previous studies, one notable difference between surface and bulk sulfates comes from their solubility in water [19,28]. Commonly, sulfates are tightly bound to a metal oxide surface and cannot be easily leached away by water. Contrarily, analogue to the salt of cerium sulfate, bulk sulfates interact less with the rigid framework of CeO₂ and can thus be easily dissolved in water. Therefore, in the present work, we innovatively utilize this principle to distinguish bulk sulfates and surface sulfates, and with the aid of Raman characterization, the crucial information related to the existing state of sulfates is revealed.

As shown in Figure 3, a sharp band at 463 cm⁻¹, which can be ascribed to the F_2g vibrational mode of cubic fluorite CeO₂, is dominated in the Raman spectra of B-CeS and S-CeS. In addition, no obvious difference in peak intensity is exhibited, suggesting insignificant disturbance on the bulk structure of CeO₂ with sulfate species. By further inspecting the spectra, a weak and broad peak centered around 1170 cm⁻¹ is detected and it actually represents the signal from sulfate species [19,29]. Unfortunately, no obvious variation in peak shape or position can be found for B-CeS and S-CeS. Nevertheless, we can still take advantage of the sensitivity of Raman characterization in revealing sulfate signals to approach the crucial information of S content changes. It is observed that the peak intensity of sulfate in B-CeS-W (water leached sample) is significantly lower than that of B-CeS. Interestingly, a negligible change in the peak intensity of the sulfate signal in S-CeS-W is detected as compared to the pristine S-CeS. This sharp difference in the solubility of the sulfate species can be attributed to the different existing states of the sulfate species. That is, bulk sulfates are widely present in B-CeS, while surface sulfates are dominant in S-CeS.

2.2. XRD Results

To explore any impact from sulfate introduction on the phase structure of CeO₂, XRD measurements were performed. It is obvious from Figure 4 that the XRD patterns of B-CeS and S-CeS are represented by the characteristic diffractions from fluorite CeO₂ (JCPDS 43-1002), and no peaks from sulfates are detected. Based on these results, we can conclude that in comparison with Raman, XRD shows less sensitivity in revealing the evolution of sulfate species. For S-CeS, the absence of reflection peaks from cerium sulfate is understandable, since the sulfates are well dispersed on CeO₂. However, for B-CeS, we suppose the absence of signal from cerium sulfate is mainly due to the low loading or amorphous structure [29].

When it comes to the signal related to CeO₂, a clear difference is exhibited. In general, the appearance of sharp peaks suggests high crystallinity of CeO₂ in both catalysts. This is basically in agreement with the high-temperature treatment and small surface area shown by the two catalysts (B-CeS: 3.6 m²/g, S-CeS: 10.8 m²/g). Additionally, we can find that although both catalysts are calcined under the same temperature, B-CeS obtains diffraction peaks with a stronger intensity, and the calculated average grain sizes applying the Scherrer equation on the most intense peak at 2θ = 28.6° for B-CeS and S-Ce are 40.2 nm and 31.3 nm, respectively. The results clearly show the sintering behavior of ceria can be influenced by the accompanied anion linked to Ce³⁺ in the precursor.

2.3. XPS Measurement

To obtain a better understanding of the chemical state of elements on the catalyst surface, XPS characterization was performed. Figure 5a presents the S 2p spectra. No obvious difference in the binding energy of S is exhibited, and the appearance of a typical band at 168.7 eV conveys the sulfur element is in the oxidation state of +6, in accordance with the formation of SO₄²⁻ [30]. In addition, a smaller surface atomic concentration of S is observed for B-CeS (2.13%) when compared to S-CeS (3.60%). This can be viewed as convincing evidence for the aggregated feature of sulfate species in B-CeS sample.

The XPS spectra of Ce 3d are shown in Figure 5b. The peaks denoted as u′ and v′ are related to Ce³⁺ species, and u, u″, u‴, v, v″ and v‴ are attributed to Ce⁴⁺ species. From Table 1, we can see that the introduction of surface sulfate species induces a significant augmentation of the surface concentration of Ce³⁺. This is not surprising since Waqif et al. reported that SO₂ can act as a reducing agent and lead to the formation of Ce₂(SO₄)₃ on a CeO₂ surface [31]. Additionally, the obtainment of a much higher surface Ce³⁺ concentration can be attributed to the fact that most S in S-CeS are concentrated on the surface with CeO₂, and therefore more surface Ce³⁺ are generated.

Regarding the O 1s spectra, a significant difference in the peak shape is displayed. For B-CeS, two well resolved peaks are detected, and they can be reasonably attributed to the signals from lattice oxygen O_α (529.0–530.0 eV) and surface oxygen O_β (531.3–531.9 eV). Owing to insufficient contact of the CeO₂ surface with bulk sulfates, O_β shows a much weaker intensity than O_α. However, the case is very different for S-CeS. O_β undergoes an obvious enhancement with the surface coverage of sulfates. In addition, a third peak, O_γ, located at the binding energy of 532.7–533.5 eV, springs out. According to a previous study, this peak can be related to the signal from surface hydroxyl species [32]. Based on the XPS analysis, we know that the different existing states of sulfate species induce a lot of alteration on the surface features of CeO₂, and this is particularly true for the S-CeS sample. For B-CeS, lattice oxygen species from ceria are dominant, while the surface oxygen species from sulfate are enriched in S-CeS.

2.4. H₂-TPR Results

To explore the impact from varied sulfate species on the reduction behavior of CeO₂, H₂-TPR was carried out and is shown in Figure 6. According to previous reports, three species have potential for interaction with H₂ over the catalyst of sulfated ceria. They are the surface and bulk oxygen from CeO₂ and the sulfur in sulfates [27,33]. In B-CeS, all of the three reduction peaks are displayed, with corresponding H₂ consumption at low temperatures (peak α, 300–400 °C,), medium temperatures (peak β, 450–540 °C) and high temperatures (peak γ, 700–800 °C). For S-CeS, peak γ with a comparable intensity is found. This is reasonable since this peak is usually attributed to the reduction of bulk oxygen in the deeper interior of CeO₂. However, for the other two peaks, significant changes are displayed. That is, peak α cannot be observed anymore and peak β shifts to higher temperatures, with its signal strongly intensified.

It is well known that for pure CeO₂ free from any alien element modification, in addition to the reduction of bulk oxygen, the reduction of surface oxygen at a much lower temperature usually accompanies it [33]. This feature is shown by the appearance of a minor peak centered at 360 °C for the B-CeS sample (peak α). However, for S-CeS, the peak α disappears. In view of the fact that surface sulfate species are abundant, we suppose that the disappearance of the reduction peak from oxygen species on the ceria surface actually reflects that most of them reacted with SO₂ in generating surface sulfate species. This can also explain the presence of the large β peak shown for the S-CeS sample.

Based on the comprehensive characterizations using Raman, XPS and H₂-TPR, it is concluded that two kinds of sulfate species are established for CeO₂. That is, through SO₂ + O₂ treatment, the surface of CeO₂ is exclusively covered with a layer of sulfate. As a sharp contrast, the sulfate species are present mainly as bulk sulfates through directly calcining Ce₂(SO₄)₃ at an elevated temperature.

2.5. NH₃-SCR Performance

The NO conversion and N₂ selectivity as a function of the reaction temperature for pure CeO₂ and sulfated CeO₂ at 200–500 °C is shown in Figure 7a,b. As expected, a much poorer activity was exhibited by pure CeO₂, with a NO conversion of less than 50% across the entire test range. In addition, the appearance of a negative NO conversion at 500 °C reflects the occurrence of serious NH₃ oxidation [32], which is also compatible with the very poor N₂ selectivity (Figure 7b). Clearly, the introduction of sulfate species distinctly promotes the catalytic performance. S-CeS shows a more pronounced improvement in reaction efficiency than B-CeS, which acquires more than 90% NO conversion in the high-temperature range of 350–450 °C. When the reaction temperature is further increased, a distinct deactivation was observed for both samples. However, the negative NO conversion was no longer present, demonstrating efficient suppression of NH₃ oxidation via sulfate modification. From the results of the activity test, we can tentatively propose that surface sulfates contribute more to promoting high-temperature SCR activity than bulk sulfates. To further confirm this, a direct comparison of the activity from B-CeS and B-CeS-W was achieved by taking advantage of the complete removal of bulk sulfates in the B-CeS-W sample. From Figure 7c, we can see that the catalytic performance at 400–500 °C shows negligible declining after removal of bulk sulfates, although obvious differences in low temperature range are displayed. This can be viewed as solid evidence that in comparison with surface sulfates, the contribution from bulk sulfates in boosting the catalytic performance of CeO₂ at high temperatures is insignificant.

To further explore the reaction stability of catalysts operated at high temperatures, the results of NO conversion/N₂ selection as a function of reaction time are presented in Figure 7d. Both catalysts exhibit no obvious activity and selectivity change after successive operation at 450 °C for 240 min, revealing a good reaction stability of the sulfated samples.

2.6. The Acidity of Prepared Catalysts (NH₃-TPD)

To disclose possible reasons for the different catalytic behaviors of surface and bulk sulfates in promoting high-temperature NH₃-SCR performance, the acid and redox properties were explored. Figure 8 presents the NH₃-TPD results. In line with previous studies, both catalysts showed a wide range of NH₃ desorption [22,24]. Through further examination, two desorption peaks can be differentiated. The low-temperature desorption with a peak centered at 150 °C can be attributed to NH₃ adsorbed on the weak acid sites from CeO₂. It is well reported that due to the nature of basicity, pure CeO₂ possesses weak acidity. With introduction of sulfate species, the acidity is strongly enhanced, as evidenced by the appearance of another desorption peak at a temperature of 263 °C.

Interestingly, despite their comparable sulfate content, the distribution of weak and strong acid sites over the two catalysts varies sharply. For B-CeS, weak acidity is dominant, as can be deduced from the appearance of a maximum NH₃ desorption signal at 150 °C (peak α). In view of the fact that insufficient contact is present between sulfate species and CeO₂, more surface sites from ceria can be provided for NH₃ adsorption. On the other hand, due to the aggregation state of sulfate in B-CeS, fewer sites from sulfates are available for NH₃ adsorption. It is reasonable that this changes when bulk sulfates are substituted by surface sulfates. From the XPS and H₂-TPR results, we know that abundant sulfates are bound to the CeO₂ surface. With this significant difference in sulfate species distribution, we can reasonably conclude that the amount of strong acid sites (peak β) increases at the expense of the weak sites provided by CeO₂.

2.7. The Oxidation Properties with Respect to NO and NH₃ (NO Oxidation and NH₃ Oxidation)

The oxidation properties of catalysts with respect to reactants are investigated in terms of NO oxidation and NH₃ oxidation. The NO oxidation performance is shown in Figure 9a. In general, both catalysts show poor activity in converting NO to NO₂, which is evidenced by the high temperature required to initiate the reaction (250 °C and the obtained poor maximum conversion efficiency (<20%). It has been previously reported that although ceria has excellent redox properties, it shows poor NO oxidation performance and, due to the strong electron-withdrawing effect from S=O in sulfate group, the modification of sulfate species can induce a further deterioration in NO oxidation performance [30]. From this point of view, it is reasonable that the B-CeS displayed a better NO oxidation performance than the S-CeS, since the main body of ceria is less disrupted by bulk sulfates in B-CeS. By comparing the NO oxidation performance with the high-temperature NH₃-SCR results, we can conclude that the ability to oxidize NO is not responsible for the different SCR behaviors at high temperatures.

The results of the NH₃ + O₂ test are shown in Figure 9b. In line with the NO oxidation performance, the S-CeS catalyst presents a worse oxidation capacity for NH₃. It is possible that due to the strong coverage of sulfate species on CeO₂, the inherent surface oxygen on ceria is not available for S-CeS. As a result, a poor oxidation ability is exhibited, which can suppress ammonia oxidation to some extent at high temperatures. Based on the above analysis, we can conclude that in addition to the change in acid sites, the reduced oxidation capacity of reductant NH₃ constitutes another reason for the better performance of the S-CeS sample in high-temperature NH₃-SCR.

3. Materials and Methods

3.1. Catalyst Preparation

CeO₂ modified with bulk sulfates was constructed through a novel top-down strategy. Ce₂(SO₄)₃·xH₂O was employed directly as the precursor, and with help of a harsh thermal treatment (static air, 900 °C, 1 h), most of the metal salt was converted into CeO₂. Notably, a minute amount of precursor was not decomposed and still existed as bulk phase in the mixture. For simplicity, the obtained sample is denoted as B-CeS.

The CeO₂ modified with surface sulfates was obtained via a vapor sulfation method. Pure CeO₂ derived from thermal treatment of cerium nitrate hydrate (900 °C, 1 h) was exposed to a stream of 2000 ppm SO₂ + O₂ to obtain the designated sample of CeO₂ modified with surface sulfate (S-CeS). By controlling the treatment time, a similar value in sulfate content as B-CeS is guaranteed.

A designed experiment of water leaching was performed to discriminate bulk sulfates and surface sulfates, by taking advantage of their different behaviors in water solubility. The B-CeS and S-CeS catalysts were dropped into distilled water, washed to remove the soluble sulfates, sonicated, filtered and dried to obtain the water-washed samples, which are named B-CeS-W and S-CeS-W, respectively.

3.2. Catalyst Characterization

The specific surface area was characterized using the N₂ sorption technique at −196 °C on an Micromeritics ASAP-2020 adsorption apparatus (Norcross, GA, USA). The samples were vacuum-pretreated at 300 °C for 3 h before measurement.

TG was performed on a Netzsch thermal analyzer STA 449C (Erich NETZSCH GmbH & Co.Holding KG, Selb, Germany) with N₂ as carrier gas and a test temperature range of 0–1000 °C with a temperature rise rate of 10 °C/min.

Surface morphologies of the catalyst samples were investigated using a FEI NANO SEM430 (Fei company, Lausanne, Switzerland) scanning electron microscope. All SEM images were obtained from secondary electrons with an accelerating voltage of 15 kV. The elemental composition and distribution were determined using an Oxford X-Max20 EDS equipment (Oxford Instruments Private Limited, Oxford, UK).

Raman spectra of samples were collected on a LabRAM Aramis (Japan Hobiba, Kyoto, Japan) Laser Raman spectroscopy with a 532 nm Ar⁺ laser beam.

XRD was performed on a Philips X’pert Pro diffractometer (Philips Analytical, Overijssel, The Netherlands) with a Ni filter and a Cu target K_α radiation source (wavelength 0.15418 nm). The operating voltage and current were set to 40 kV and 40 mA, and data were collected in the range 2θ = 10-80° with a scanning speed of 10 °C/min.

XPS was performed on a PHI 5000 VersaProbe X-ray photoelectron spectrometer (PHI Corporation, Kanagawa, Japan) with a monochromatic Al K_α radiation source (1486.6 eV) and an acceleration power of 15 kW. Prior to testing, the sample was degassed to below 5 × 10⁻⁷ Pa at room temperature in an ultra-high vacuum chamber. Sample charge effects were compensated for by correcting the C 1s spectrum to 284.6 eV. The accuracy of the binding energy was 0.1 eV.

Temperature-programmed reduction by H₂ (H₂-TPR) was carried out in a quartz U-tube reactor connected to a thermal conduct detector (TCD) with an H₂/Ar mixture (7% H₂ by volume) as a reductant. A total of 10 mg sample was used for each measurement. Before introducing it to the H₂-Ar stream, the sample was pretreated in the N₂ stream at 200 °C for 1 h. The H₂ consumption profile was collected from room temperature to 900 °C at a rate of 10 °C/min.

The acid properties were determined from temperature-programmed desorption (TPD) of ammonia using a Nicolet IS10 FT-IR spectrometer (Thermo Scientific, Waltham, MA, USA) equipped with a 2 m path-length gas cell (2 L volume). Catalysts were evaluated in a fixed-bed quartz reactor (5 mm internal diameter) operating with a steady-state flow. About 100 mg sample was primarily pretreated using argon (100 mL/min) at 200 °C for 1 h, then it was cooled to room temperature and exposed to 500 ppm NH₃/Ar (100 mL/min) for 1 h. Thereafter, excessive physical absorbed ammonia was removed by purging with argon at room temperature for 1 h. All NH₃-TPD tests were carried out by increasing the temperature from 30 to 600 °C at a rate of 10 °C/min and an argon flow rate of 100 mL/min.

The NH₃/NO oxidation experiments were carried out on a fixed-bed reactor with a gas composition of 500 ppm NH₃ (or NO), 5 vol.% O₂, and N₂ as equilibrium gas at a gas flow rate of 100 mL/min. A certain amount (200 mg) of the sample was loaded into the quartz tube reactor, purged with N₂ at 200 °C for 1 h and then cooled to room temperature. After the reaction gas was adsorbed to saturation, the reactions were carried out at different temperatures. The concentrations of NO and NH₃ were recorded online using a Thermofisher IS10 FTIR spectrometer, and the catalytic performance data as a function of temperature were collected after being kept stable for 30 min at each temperature point.

3.3. Activity Measurement

The catalytic performance was evaluated in a fixed-bed quartz reactor with 5 mm internal diameter under steady-state flow. The feed gas was 500 ppm NO, 500 ppm NH₃, 5% O₂, and total gas flow was controlled at 100 mL/min. The WHSV was 60,000 mL/(gh). A total of 100 mg catalyst was sieved through a 40–60 mesh. Before the tests, the catalyst was pretreated in a purified Ar stream at 200 °C for 30 min to avoid surface impurities. Then, the mixed gases were switched on and activity data were collected at every target temperature after stabilizing for 30 min in the range of 200–500 °C. The concentration of effluent gases was analyzed on a Nicolet IS10 FT-IR spectrometer equipped with a 2 m path-length gas cell (2 L volume). The NO conversion was calculated based on the following equations:

$$\mathrm{NO} \mathrm{conversion} (\%)=\frac{{\left[\mathrm{NO}\right]}_{\mathrm{in}}-{\left[\mathrm{NO}\right]}_{\mathrm{out}}}{{\left[\mathrm{NO}\right]}_{\mathrm{in}}}\times 100\%$$

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

4. Conclusions

By explicitly controlling the existing states of sulfate species, the effect of surface and bulk sulfates on high-temperature catalytic performance of ceria in NH₃-SCR was investigated in the present study. It was found that introduction of sulfate species induced little modification on the bulk structure of CeO₂. However, both the adsorption and reaction behaviors were significantly altered. In comparison with pure CeO₂, introduction of sulfates notably enhanced the high-temperature NH₃-SCR performance. In addition, an apparent difference in reaction efficiency between surface and bulk sulfates was exhibited. Due to close contact between sulfate species and ceria, the active surface oxygen species from ceria were largely eliminated, resulting in alleviated NH₃ oxidation for S-CeS. On the other hand, the strong interaction between SO₂ and surface oxygen species from ceria created sufficient sulfate species. As a result, more strong acid sites were generated in S-CeS. With the obtainment of these unique properties (more strong acid sites and reduced NH₃ oxidation capacity), the catalyst of ceria modified with surface sulfates displayed enhanced catalytic performance in high-temperature NH₃-SCR.