Co-Deactivation of Cu-SSZ-13 Catalyst by K₂SO₄ Solid-State Diffusion and Hydrothermal Aging

Abstract

Cu-SSZ-13, the most widely used catalyst in diesel selective catalytic reduction (SCR) systems, often suffers severe deactivation, including hydrothermal aging and ash poisoning. In comparison with traditional impregnation in laboratory work, a more realistic solid-state diffusion method was employed to simulate K₂SO₄ poisoning on a commercial Cu-SSZ-13 catalyst with high aluminum and copper contents. Hydrothermal aging at 650 °C alone induces severe framework dealumination and transformation of isolated Cu²⁺ ions to copper aluminate (CuAlO_x) species. K₂SO₄ poisoning alone is more prone to detached Cu²⁺ ions and aluminum terminal hydroxyl group to form CuSO₄ and Al₂(SO₄)₃. The presence of water vapor during K₂SO₄ poisoning dramatically reduces SCR activity by accelerating the ion-exchange between K⁺ and Cu²⁺ and zeolite dealumination. These synergistic effects promote extensive detachment of active Cu species, resulting in the formation of predominating inert sulfates, along with a small amount of CuO_x clusters. These findings are expected to provide a theoretical basis for designing catalysts with enhanced resistance to both hydrothermal aging and ash poisoning in diesel aftertreatment applications.

1. Introduction

With the progressive tightening and full implementation of emission standards such as China VI and Euro VI, efficient removal of nitrogen oxides (NO_x) in diesel vehicle exhaust faces unprecedented challenges [1,2,3]. Ammonia selective catalytic reduction (NH₃-SCR) has been widely regarded as the primary NO_x mitigation technology in the past decades [4]. Among various NH₃-SCR catalysts, copper-based zeolite, particularly Cu-SSZ-13 with a chabazite (CHA) structure, has become the mainstream choice for meeting China VI requirements due to its excellent NO_x conversion over a wide temperature range (especially at low temperatures), high N₂ selectivity, and superior hydrothermal stability [5,6].

However, the stability and poison resistance of these catalysts remain severe challenges in practical applications. During diesel operation, sulfur-containing components in the fuel are oxidized to sulfur oxides [7]. These sulfur oxides can further react with impurities such as alkali metals (e.g., K) originating from lubricating oil and residual urea, to form sulfates (e.g., K₂SO₄). In some diesel aftertreatment systems, SCR catalyst is coated onto a wall-flow diesel particulate filter (DPF) to form a selective catalytic reduction filter (SCRF), promoting heating to reduce NO_x emissions during cold-start and low-load conditions, and minimizing system volume [8]. Consequently, thermally stable sulfates gradually accumulate as ash on the SCRF sidewalls, leading to both physical blockage and chemical deactivation of the catalyst [8]. Additionally, biodiesel, as a renewable alternative to crude oil-refined diesel, is gradually being promoted worldwide. Its production introduces alkaline catalysts and herein contains more K and Na than regular diesel, resulting in an increase in alkali content in the ash from 0.2% to 6% [9]. Previous studies generally attribute alkali-induced deactivation to several mechanisms [8,10,11]. First, alkali species promote the transformation of isolated Cu²⁺ into aggregated CuO_x or copper aluminate species (i.e., CuAl₂O₄ and CuAlO₂), which not only reduce the number of active sites but also enhance side reactions of NH₃ oxidation. Second, alkali species or copper-derived compounds may physically block zeolite micropores. Third, alkali metal ions can undergo ion exchange with Brønsted acid sites, weakening NH₃ storage capacity. Fourth, severe alkali poisoning can directly damage the zeolite framework and eventually cause complete catalyst deactivation.

At present, wet impregnation is the most commonly employed method for simulating ash poisoning in laboratory work because it is operationally simple and allows precise control of loading [12,13,14]. However, this method causes the dissociation of soluble salt cations and anions in the water and completes the liquid-state ion exchange to a large extent in the suspension, greatly accelerating the poisoning of the catalyst. This scenario differs substantially from real operating conditions, where alkaline impurities in the form of aerosols in exhaust gradually accumulate on the catalyst coating and progressively diffuse into the zeolite channels in the solid state under hydrothermal environments. Although aerosol deposition and solid-state diffusion have been used in some studies, most of them focus on vanadium-based catalyst systems. For example, Lei et al. [15] compared the deactivation of V₂O₅-WO₃/TiO₂ exposed to KCl introduced by wet impregnation, vapor deposition, and solid-state diffusion. They found that wet impregnation induced deactivation primarily through the formation of V₂O₃ caused by the interaction of potassium with oxygen in vanadium oxides, while vapor deposition and solid-state diffusion mainly produced a K₂S₂O₇-V₂O₅ eutectic phase that led to deactivation. Larsson et al. [16] reported that wet impregnation had more severe poisoning effects than the aerosol method, whereas Ali et al. [17] concluded that V₂O₅-WO₃/TiO₂ catalysts subjected to prolonged KCl solid-state diffusion showed a greater decrease in denitration activity compared to the impregnation method based on activity tests and DFT calculations, demonstrating the importance of poisoning treatment parameters. Few systematic investigations about ash poisoning methods have been performed on zeolite-based catalysts. By using the aerosol method to simulate potassium poisoning, Shwan et al. [18] found that the deactivation of Fe-BEA zeolites was also caused by the exchange of potassium with isolated iron ions. Kern et al. [19] explored the effects of various alkali elements (K, Na, Mg, and Ca) on Fe-BEA and Fe-MFI using wet impregnation and aerosol treatments, and obtained that these zeolites had better resistance to alkali poisoning than vanadium-based catalysts. Particularly, no similar solid-state diffusion methods were applied to Cu-SSZ-13 to simulate ash poisoning.

Diesel vehicle exhaust typically contains approximately 10 vol.% water vapor. In particular, Cu-SSZ-13 with a low Si/Al ratio contains a high proportion of hydrolysis-susceptible bonds with great hydrophilicity, which accelerates framework dealumination and ultimately leads to structural collapse [20]. Although the influence of water vapor on hydrothermal aging has been widely investigated, it is often neglected as an essential environmental factor in simulating ash poisoning. Tarot et al. [21,22] reported that impregnation of Cu/FER catalysts in water caused more severe deactivation than impregnation in ethanol, attributing the effect to water-facilitated migration of Cu species and enhanced formation of extra-framework CuO_x species. Wang et al. [23] compared the effects of different alkali metal elements (Na and K) impregnated on the hydrothermal stability of Cu-SSZ-13 and found that alkali poisoning significantly decreased hydrothermal stability and resulted in severe deactivation. Zhang et al. [24] reported the combined effects of hydrothermal aging and SO₂ poisoning on Cu/SSZ-13. Nevertheless, current research has not yet provided an investigation into the ash poisoning under wet conditions or using solid-state diffusion method on Cu-SSZ-13 catalysts.

Using potassium sulfate as a typical ash component, this work employs a solid-state diffusion approach to simulate ash poisoning of a commercial Cu-SSZ-13, aiming to elucidate the synergistic deactivation effects of hydrothermal aging and ash poisoning. By comparing diffusion behavior under wet and dry conditions, the influence of water vapor on potassium migration and Cu-species evolution along a diffusion-dominated poisoning pathway was clarified.

2. Results

2.1. Catalytic Activity

Figure 1a shows the NH₃-SCR activities of Cu-SSZ-13 catalyst after different deactivation treatments. After aging in dry air at 650 °C for 10 h, CuSSZ-DA exhibits even better activity than the pristine catalyst (Figure S1a), with a broad temperature window of 175–550 °C. It has been reported that Cu-CHA zeolites with high Cu and Al content, particularly those prepared by the one-pot method, inevitably contain residual CuO_x that can be redispersed into isolated Cu²⁺ ions under mild thermal treatments, leading to an enhancement in catalytic activity [25,26]. Such a reactivation is confirmed by the UV-vis spectra (Figure S1b) in which CuO_x signals at ~260 nm and 400–500 nm are weakened to some extent and, correspondingly, the Cu²⁺ signal at 205 nm is strengthened over CuSSZ-DA. However, the NO_x conversion of CuSSZ-HA decreases across the whole temperature range, and the temperature window is narrowed to 250–550 °C, demonstrating the importance of water vapor in thermal deactivation. Notably, CuSSZ-DA-KS, chemical poisoned under dry conditions, displays deactivation characteristics similar to CuSSZ-HA, although its low-temperature (<200 °C) NO_x conversion is slightly higher and the loss of high-temperature (>300 °C) activity is more pronounced. In contrast, CuSSZ-HA-KS subjected to K₂SO₄ solid-state diffusion poisoning in hydrated conditions shows the most severe performance deterioration, with a maximum NO_x conversion of only 65% at 250 °C. All the catalysts maintain high N₂ selectivity above 95% (Figure 1b).

In the SCR reaction, NH₃ oxidation serves as a major competitive reaction, particularly at high temperatures, where it accelerates NH₃ consumption and consequently reduces reductant availability. As shown in Figure 2a, CuSSZ-DA exhibits a seagull-shaped profile due to various dominating active sites in different temperature regions. Low-temperature (<350 °C) NH₃ oxidation is primarily attributed to isolated Cu²⁺ sites, with N₂ as the dominant product. In contrast, the high-temperature (>350 °C) reaction is governed by both Cu²⁺ and CuO_x species, with NO being a characteristic product of CuO_x-catalyzed NH₃ oxidation [27,28]. After hydrothermal aging, CuSSZ-HA exhibits suppressed activity within the whole temperature range, implying a depletion of isolated Cu²⁺ active sites. Meanwhile, the NO formation remains comparable to that of CuSSZ-DA, indicating that the detached Cu²⁺ does not convert into CuO_x but instead forms other relatively inert copper species. Upon the introduction of K₂SO₄, the low-temperature NH₃ conversion declines over CuSSZ-DA-KS due to the loss of isolated Cu²⁺ species. The Cu²⁺ loss would definitely lead to a reduction in high-temperature NH₃ conversion. However, its high-temperature NH₃ oxidation activity increases, and trace NO is produced (Figure 2b), indicating the formation of limited CuO_x species. Consequently, it shows a severe loss in high-temperature SCR activity due to limitations on reductant availability with stronger NH₃ oxidation side reaction. Upon poisoning in the presence of water, the low-temperature NH₃ conversion suffers similar severe suppression to that of CuSSZ-HA, but the conversion above 350 °C markedly increases, accompanied by increased NO formation. This observation reveals that poisoning under wet conditions promotes the transfer of isolated Cu²⁺ to CuO_x species. Nevertheless, the maximum NO concentration is only 11 ppm, indicating that the amount of CuO_x remains quite low and other copper species, such as copper aluminates, may be the predominant products.

2.2. Structural Characterization of the Catalysts

XRD was conducted to explore possible phase changes in the catalyst treated under different conditions. As shown in Figure 3a, all the samples exhibit typical diffraction features of the CHA framework [29,30], indicating that neither K₂SO₄ poisoning nor high-temperature hydrothermal aging causes significant changes in the bulk structure of the SSZ-13 zeolite. However, the diffraction peaks of the other three samples decrease in intensity to varying degrees in comparison with CuSSZ-DA, suggesting that the framework has undergone a certain degree of damage. In addition, no diffraction peaks related to Cu (CuO, CuAl₂O₄, or CuAlO₂) or other elements (K, S, Al, or Si) are observed in any sample, implying that even if these compounds were formed, their quantities are too low or highly dispersed to be detected.

The collapse of the zeolite framework and the formation or introduction of additional species inevitably leads to changes in the textural properties of the catalysts. As shown in Figure 3b, the specific surface area of CuSSZ-HA decreases significantly compared with CuSSZ-DA, while the reduction in pore volume is relatively minor. In contrast, K₂SO₄ poisoning results in an opposite trend for CuSSZ-DA-KS, where the zeolite framework remains largely intact, but the introduced salt blocks the pore channels. Under wet poisoning conditions, both the declines in the textural features become more marked due to simultaneous partial framework collapse and pore blockage.

2.3. Evolution of Copper Species

UV-vis spectroscopy was applied to identify different copper species with characteristic absorption behaviors [31]. As shown in Figure 4a, all the samples display pronounced absorption bands near 205 nm and 800 nm, corresponding to ligand-to-metal charge transfer (LMCT) transitions from framework O²⁻ to coordinated isolated Cu²⁺ and d-d electronic transitions of Cu²⁺, respectively [32], indicating that isolated Cu²⁺ is the dominant copper species in Cu-SSZ-13. As enlarged in Figure 4b, the feature centered at 253 nm is attributed to CuO_x [33], copper aluminate (CuAl₂O₄ and CuAlO₂) [34], and copper sulfate [8]. To further track copper evolution, reference materials including CuSO₄, CuO, CuAl₂O_4, and CuAlO₂ were analyzed (Figure 4d). Compared with CuSSZ-DA, the signal at 253 nm increases in intensity over CuSSZ-HA, due to the relatively poor hydrothermal stability of the zeolite in which partial framework dealumination facilitates the formation of additional copper aluminate species. It shows much stronger adsorption in the 400–500 nm region than CuSSZ-DA, coinciding with the absorption characteristics of CuAl₂O₄ and CuAlO₂. Both the poisoned samples exhibit stronger absorption intensity at 253 nm than their thermally aged counterparts, while the strengthening in the 400–500 nm region is not obvious. are noted correspondingly. These features are consistent with the characteristic fingerprints of CuSO₄ as established in the literature [8,35] instead of copper aluminate, and the results in Figure 4d. It means that the detached Cu²⁺ ions are predominantly captured by extra-framework sulfate species in the poisoned samples. Additionally, the formation of CuO_x cannot be excluded, as mentioned in the NH₃ oxidation measurement (Figure 2b).

Thermogravimetric analysis was employed to monitor the thermal decomposition behavior of the catalysts and some reference sulfates. As shown in Figure 5a, the thermal decomposition process of Cu-SSZ-13 can be divided roughly into two stages. Stage I (50–300 °C) corresponds to the removal of adsorbed water and physically adsorbed species, while Stage II (900–980 °C) in Figure 5b reflects framework collapse accompanied by de-hydroxylation and associated release of O₂ [35]. As shown in Figure 5d, K₂SO₄ remains stable even at 1000 °C due to its exceptionally high thermal stability in air. Al₂(SO₄)₃ and CuSO₄ exhibit mass losses of 68% and 50% within 730–900 °C and 660–830 °C, respectively, corresponding to the thermal decomposition into Al₂O₃ [36] and SO₃ [37] with the release of sulfur oxides and oxygen. The onset temperature for mass loss is defined as the temperature at which the DTG curve begins to decline, while the end temperature corresponds to where the DTG curve levels off. CuSSZ-HA shows similar mass loss features to CuSSZ-DA, while the onset temperature of the two poisoned catalysts shifts towards lower temperatures. Given that the reference materials Al₂(SO₄)₃ and CuSO₄ decompose in lower temperature regions, these observed shifts are likely attributed to the formation of analogous sulfates. The interaction of the sulfates with zeolite may delay their thermal decomposition temperature. The mass gap between the onset and end points in Figure 5b was drawn to represent the total mass loss associated with framework de-hydroxylation and decomposition of sulfate species, and the results are summarized in Figure 5c.

The difference in mass loss between the two aged catalysts reflects the influence of water on framework dealumination. Meanwhile, the increased mass loss of the two poisoned catalysts in comparison with their thermally aged counterparts can be considered the content of thermally decomposable sulfur species in the samples (Figure 5c), formed by solid-state ion-exchange of K⁺ and trapping of sulfate radicals on the zeolite surface. The detailed calculation method was described in the support information. Considering that the total sulfur content in the catalyst-salt mixture is 0.82 wt.%, the decomposable sulfur fraction in CuSSZ-DA-KS and CuSSZ-HA-KS accounts for only approximately 10% and 20%, respectively, indicating the presence of substantial amounts of thermally stable K₂SO₄. Notably, the decomposable sulfur content in CuSSZ-HA-KS is twice that of CuSSZ-DA-KS, which further confirms the synergistic interaction between water vapor and K₂SO₄. The presence of water vapor promotes K⁺ diffusion, while aluminum species released during framework collapse and active Cu²⁺ ions are captured by SO₄²⁻, forming larger quantities of Al₂(SO₄)₃ and CuSO₄.

Hydrogen temperature-programmed reduction (H₂-TPR) was used not only to assess the effects of K₂SO₄ and water vapor on the redox behavior of the catalysts, but also to identify different copper and sulfate species based on their characteristic reduction temperatures. As shown in Figure 6b, the H₂-TPR profiles of CuSSZ-DA and CuSSZ-HA display clearly separated temperature regions: 150–500 °C (region I) corresponds to the reduction in isolated Cu²⁺ species (ZCuOH and Z₂Cu) to Cu⁺ and the reduction in CuO_x species to Cu⁰; 500–760 °C (region II) is associated with the reduction in copper aluminate species (CuAl₂O₄ and CuAlO₂) [28]; and 760–1000 °C (region III) reflects the reduction in Cu⁺ to Cu⁰ [38]. Specifically, the Cu-SSZ-13 contains a relatively high Cu loading and a low Si/Al ratio, which unavoidably leads to the presence of CuO_x and especially copper aluminate species during catalyst synthesis. This feature is consistent with the UV-vis spectra (Figure 4). The H₂ consumption ratio between the low- and high-temperature peaks is approximately 1:1 (Figure 6c), indicating that reduction is dominated by isolated Cu²⁺. With partial severe framework damage, ZCuOH and Z₂Cu can further convert to CuO_x or copper aluminate phases (CuAl₂O₄ and CuAlO₂) [28], resulting in an obvious increase in H₂ consumption in region II for CuSSZ-HA. The high-temperature reduction peak for CuSSZ-HA shifts to lower temperatures, which may be related to decreased copper dispersion and weakened interactions between copper species and damaged zeolite framework.

The high-temperature peaks of the poisoned catalysts become much more complicated, and thus, H₂-TPR was also conducted on some reference sulfates. As shown in Figure 6d, the reduction in K₂SO₄, Al₂(SO₄)₃ and CuSO₄ are above 730 °C, within 500–800 °C, and within 200–500 °C, respectively. The low-temperature peak of CuSSZ-DA-KS and CuSSZ-HA-KS shifts to 295 °C and 310 °C, respectively. These shifts are attributable to the evolution of copper species, likely involving the loss of isolated Cu²⁺ and formation of copper sulfate, whose reduction temperature typically falls within this range [39,40]. The identification of the exact contribution of each species is complex due to their peaks overlapping. However, it is noted that a marked increase in H₂ consumption in region I occurs on CuSSZ-HA-KS (Figure 6c), which is ascribed to the formation of more CuSO₄ that requires two-electron reduction to Cu⁰ at relatively higher temperatures. Meanwhile, K₂SO₄ poisoning also develops a new reduction signal near 740 °C (Figure 6b), which is tentatively assigned to the reduction in aluminum sulfate species based on the results in Figure 6d and Du et al.’s study [41]. The largest H₂ consumption in region II is obtained under wet poisoning, indicating that the framework is more susceptible to dealumination under water vapor attack and formation of more sulfates and aluminates. Correspondingly, the H₂ consumption above 800 °C should decrease substantially for CuSSZ-HA-KS with the reservation of less K₂SO₄. Interestingly, an inverse sequence is observed. As shown in Figure 6d, H₂-TPR measurement was performed on a mixture of K₂SO₄ + CuO at a mass ratio of 1:1. The low- and high-temperature peaks correspond to a complete reduction in CuO and about 88 wt.% of K₂SO₄, respectively. The latter value is larger than the reduction ratio of pure K₂SO₄ (64 wt.%), suggesting a promoted reduction of K₂SO₄ in the presence of copper oxide. Correspondingly, the larger H₂ consumption in region III over CuSSZ-HA-KS could be attributable to the more complete reduction in residual K₂SO₄ catalyzed by CuO_x clusters.

3. Discussion

Although Cu-SSZ-13 retains the characteristic CHA framework after hydrothermal aging and K₂SO₄ poisoning (Figure 3a), these treatments result in catalyst deactivation to different degrees (Figure 1). Detailed deactivation mechanisms are shown in Scheme 1. First, the copper- and aluminum-rich Cu-SSZ-13 prepared by the one-pot method inevitably contains a trace amount of CuO_x (Figure S1b). After mild thermal treatment in dry air at 650 °C for 10 h, these CuO_x clusters can diffuse into the SSZ-13 and be reactivated to form isolated Cu²⁺ ions [25,26].

However, with the introduction of H₂O, framework Si-O(H)-Al bonds undergo hydrolytic cleavage by water, producing Al-OH defect sites and further forming Al(OH)₃, leading to framework dealumination of the hydrothermally aged sample [42]. Meanwhile, a fraction of isolated Cu²⁺ ions detach from ion-exchange sites and are subsequently captured by extra-framework Al species, leading to the formation of copper aluminates (Figure 4a and Figure 6c) rather than CuO_x (Figure 2b) [43]. Compared with Cu-SSZ-13 with high Si/Al (>6) and low Cu loading (<3 wt.%), the commercial catalyst in this study exhibits stronger hydrophilicity and more significant dealumination and copper loss [44,45]. Given that copper aluminate phases are essentially inert for SCR, their formation is considered a primary deactivation factor of CuSSZ-HA.

Compared with H₂O attack, K⁺ can diffuse more readily into the zeolite and exchange with isolated Cu²⁺ and H⁺ at Brønsted acid sites [12], while the comparatively large sulfate anions (with an effective radius ≈ 3.8 Å) are difficult to diffuse into the negatively charged SSZ-13 micropores (with a similar pore diameter ≈ 3.8 Å). UV-vis spectra show a preferential enhancement at ~253 nm compared to the 400–500 nm range (Figure 4), a feature associated with copper sulfate [8,46]. Meanwhile, the poisoned catalysts show more mass loss in the temperatures above 900 °C in the TGA curves (Figure 5), as characteristics of decomposition of CuSO₄ and Al₂(SO₄)₃ [47], and quantified decomposable sulfur contents were estimated. Moreover, H₂-TPR (Figure 6) reveals the emergence of a reduction peak centered at 740 °C, aligning with the reduction in Al₂(SO₄)₃, and an obvious increase in low-temperature H₂ consumption, probably attributed to CuSO₄ reduction [41]. Consequently, free Al and Cu species are likely trapped by these extra-framework sulfate anions to form Al₂(SO₄)₃ and CuSO₄ (Figure 5c and Figure 6d). These inactive sulfate salts block the zeolite channels and further reduce pore volume (Figure 3b). The ion-exchange efficiency depends greatly on the diffusion process parameters [17], and the deactivation of K₂SO₄ is limited in this work.

The chemical poisoning effect of K₂SO₄ is further accelerated in the presence of H₂O. The SCR activity of CuSSZ-HA-KS is approximately 25–50% lower than that of CuSSZ-DA-KS in the whole temperature range (Figure 1). As isolated Cu²⁺ ions serve as the active sites for SCR reactions [48,49], they suffer more significant loss under wet conditions, as implied by the sharply reduced NH₃ oxidation activity at low temperatures (Figure 2a). The absence of stabilizing neighboring Cu²⁺ further facilitates dealumination [28]. In addition to the formation of copper aluminates, these leached Cu and Al species are mainly captured by surface SO₄²⁻ to form CuSO₄ and Al₂(SO₄)₃, respectively (Figure 4, Figure 5c and Figure 6c). This may be ascribed to the enhanced mobility of K⁺ and Cu²⁺ ions in the presence of water and require further investigation.

4. Materials and Methods

4.1. Catalyst Preparation

A commercially available Cu-SSZ-13 (Weifu Environmental Catalysts Co., Ltd., Wuxi, China, synthesized via a one-pot synthesis method [50]), featuring a low Si/Al ratio (Si/Al = 4.9) and high copper loading (3.9 wt.%), was used as the pristine material and denoted as CuSSZ. The catalyst was calcined at 650 °C for 10 h in flowing air and 10 vol.% H₂O/air in a tubular furnace, and the obtained catalysts were denoted as CuSSZ-DA and CuSSZ-HA, respectively. For ash loading, CuSSZ powder and K₂SO₄ (2 wt.%) were manually mixed in a mortar for 10 min using a spatula to ensure sufficient contact and then treated at 650 °C for 10 h in flowing air and 10 vol.% H₂O/air by solid-state diffusion. The dry poisoned and wet poisoned catalysts were denoted as CuSSZ-HA-KS and CuSSZ-DA-KS, respectively.

The aging and poisoning conditions were selected based on testing protocols commonly used to simulate long-term severe exposure. Similar parameters (10 vol.% H₂O/air, 600–850 °C, 0.5–5.7 wt.% alkali loading) have been widely adopted in the literature to evaluate catalyst durability under harsh environments [8,10,23,46,51,52,53,54]. A summary of these literature conditions was presented in Table S1.

4.2. Activity Measurements

NH₃-SCR activity tests were evaluated in a quartz reactor loaded with 250 mg of catalyst (60–80 mesh). Before testing, the catalyst was pretreated in flowing air (250 mL/min) at 550 °C for 20 min. During the measurements, the total gas flow rate was maintained at 1 L/min, with a feeding gas consisting of 500 ppm NO, 500 ppm NH₃, 5 vol.% O₂, 3 vol.% H₂O, and N₂ in balance, corresponding to a gas hourly space velocity (GHSV) of 120,000 h⁻¹. The concentrations of reactants and products in the outlet gas were monitored using a 6030 Fourier transform infrared spectrometer (MKS Instruments, Andover, MA, USA). Activity tests were conducted over the temperature range of 150–550 °C, with intervals of 25 °C below 200 °C and 50 °C above 200 °C. At each temperature, the concentrations of N₂O, NO, NO_2, and NH₃ in the outlet gas were stabilized (with the variation less than 5 ppm) after 20–90 min, depending on reaction temperature, and the steady-state data were recorded. The NO_x (NO + NO₂) conversion and N₂ selectivity were calculated using Equations (1) and (2), respectively. NH₃ oxidation activity was measured using the same procedure, except that the feed gas did not contain any NO. NH₃ conversion was calculated according to Equation (3).

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

(1)

$${\mathrm{N}}_2 \mathrm{selectivity} (\%)=(1-{\displaystyle \frac{{2 \times \left[{\mathrm{N}}_2\mathrm{O}\right]}_{\mathrm{out}}}{{\left[\mathrm{N}{\mathrm{O}}_x\right]}_{\mathrm{in}}+{\left[\mathrm{N}{\mathrm{H}}_3\right]}_{\mathrm{in} }{-\left[\mathrm{N}{\mathrm{O}}_x\right]}_{\mathrm{out}} {-\left[\mathrm{N}{\mathrm{H}}_3\right]}_{\mathrm{out} }}}) \times 100$$

(2)

$${\mathrm{NH}}_3 \mathrm{conversion} (\%)=(1 -{\displaystyle \frac{{\left[{\mathrm{NH}}_3\right]}_{\mathrm{out}}}{{\left[{\mathrm{NH}}_3\right]}_{\mathrm{in}}}}) \times 100$$

(3)

4.3. Characterizations

X-ray diffraction (XRD) patterns were collected on a D8 ADVANCE diffractometer (Bruker, Mannheim, Germany) using Cu Kα radiation (λ = 1.5406 Å) and continuous scanning in the range of 5–45° at a rate of 6°/min. N₂ adsorption-desorption isotherms were measured using a JW-BK200C-02 analyzer (Beijing JWGB, Beijing, China). Before testing, the samples were degassed at 220 °C for 8 h, and measurements were conducted at −196 °C. The specific surface area (S_BET) and total pore volume (V_total) of the catalysts were calculated using the Brunauer-Emmett-Teller (BET) method and the t-plot model, respectively. Ultraviolet-visible (UV-vis) spectra were recorded at room temperature using a UV-vis 2600 spectrophotometer (Shimadzu, Tokyo, Japan) with BaSO₄ powder as the reference, scanning over the wavelength range 200–800 nm. Thermogravimetric analysis (TGA) was carried out on a TGA/DSC 1 system (Mettler Toledo, Greifensee, Switzerland). Approximately 20 mg of sample was heated in air (50 mL/min) at a ramp rate of 10 °C/min, and thermogravimetry (TG) was recorded. H₂ temperature-programmed reduction (H₂-TPR) experiments were conducted on an AutoChem II 2920 chemisorption analyzer (Micromeritics, Norcross, GA, USA) equipped with a thermal conductivity detector (TCD). Typically, 100 mg of catalyst was loaded into a U-shaped quartz reactor and pretreated in 10% O₂/He (50 mL/min) at 300 °C for 30 min. After cooling to room temperature, the gas was switched to 3% H₂/Ar (50 mL/min), and the temperature was increased to 1000 °C at a rate of 10 °C/min. An isopropyl alcohol cold trap was installed between the quartz reactor and the TCD to condense volatile components during reduction.

5. Conclusions

In this work, the deactivation mechanisms of a Cu-SSZ-13 catalyst during hydrothermal aging and/or K₂SO₄ poisoning simulated by a solid-state diffusion method were investigated. Some findings are drawn as follows:

(1) The studied one-pot-synthesized Cu-SSZ-13 with a high copper loading (3.9 wt.%) and a low Si/Al ratio (Si/Al = 4.9) undergoes severe dealumination and framework collapse after hydrothermal aging at 650 °C for 10 h, leading to transformation of isolated Cu²⁺ ion to copper aluminates, while the thermal aging in dry air even reactivates residual CuO_x in the pristine zeolite to Cu²⁺.

(2) Using the solid-state diffusion method, K₂SO₄ poisoning under dry conditions forces some Cu²⁺ to detach from ion-exchange sites. Different from thermal aging, the detached Cu and Al species are mostly captured by SO₄²⁻ anions on the external surface to form CuSO₄ and Al₂(SO₄)₃.

(3) The presence of water vapor during the solid-state diffusion process results in further deterioration of SCR performance, because it significantly accelerates the diffusion of K⁺ and Cu²⁺ as well as zeolite structural damage. Consequently, more CuO_x clusters and especially inert sulfates are produced.

It should be noted that the severe dealumination and copper loss observed here are mostly attributable to the relatively low hydrothermal stability of the one-pot synthesized catalyst with a low Si/Al ratio and high Cu loading. Future studies on hydrothermally stable Cu-SSZ-13 catalysts should be conducted to generalize these findings and explore possible roles of water on the solid-state diffusion process of ash salts.