Selective catalytic reduction (SCR) of NO_x with NH₃ as the reductant has been shown to be a well-established NO_x abatement technology in diesel exhaust aftertreatment systems [1-3]. Extensive investigations have been conducted on copper-ion-exchanged zeolite catalysts as potential NH₃-SCR catalysts [4-6]. Recently, small-pore-sized Cu-chabazite (CHA) molecular sieves, e.g., Cu-SSZ-13 and Cu-SAPO-34, have received extensive attention due to their extraordinary deNO_x activities, broad activity temperature window, and high hydrothermal stability [5, 7]. In addition, both Cu-SAPO-18 and Cu-SSZ-39 molecular sieves, which possess an AEI construction similar to CHA, exhibit exceptional NO conversion and hydrothermally stable properties [8-12]. In general, the deNO_x performance and hydrothermal resistance over copper-zeolite catalyst are related to active Cu²⁺ sites and acidic sites. Meanwhile, the amount of acidic sites and Cu²⁺ substances in Cu-SAPO-18 can be tuned by altering the quantity of Si and Cu [13-15], which could facilitate the selection of more effective and hydrothermally stable Cu-SAPO-18 catalysts. These realities enable Cu-SAPO-18 as a potential alternative in diesel exhaust aftertreatment systems.

In practical aftertreatment applications, the resistance of SCR catalysts to thermal and chemical damage is an especially significant issue [16]. The total mileage of diesel vehicles is considerably high, and SCR catalysts are repeatedly exposed to high temperatures during the operation of the engine, which could cause the breakdown of catalyst texture and the aggregation of metal active species [17]. Furthermore, some inorganic compositions (e.g., K, Na, S, P, Ca, and Mg), which are derived from the contaminants of biodiesel, and the lubricant and fuel additives, can gradually aggregate on the catalyst surface, resulting in gradual pore blocking and the loss of active sites of SCR catalysts [16-18]. Finally, SCR catalysts will undergo a decline in catalytic activity. The influence of alkali and alkaline earth metals on SCR catalysts has attracted more interest due to their significant content in diesel derivatives [18].

Recently, the deactivation of V₂O₅/WO₃-TiO₂, caused by K, Na, Ca, and Mg, has been investigated, concluding that the introduction of these metals brings about the decrease in acidity and redox ability of the catalyst [19-21]. There are some reports on the effect of alkali and alkaline earth metals on the structure, acidity, and active species of different Cu-based zeolites. For instance, Fehrmann et al. [22] announced that Cu-zeolite catalytic agents, which had large surface area and strong acidity, showed favorable K resistance. However, Lezcano-Gonzalez et al. [17] reported that the contamination of Cu-SSZ-13 by Ca²⁺ led to pore blocking, resulting in its deactivation. Gao et al. [23] introduced Li⁺, K⁺, Cs⁺, Mg²⁺, and Ca²⁺ into SSZ-13, resulting in the decrease of Brønsted acidic sites (H⁺). Meanwhile, the remaining H⁺ content is consistent with these hydrated metal ions diameters, which is presented in the order of Mg²⁺ > Ca²⁺ ≈ Li⁺ > Na⁺ > K⁺ > Cs⁺. Additionally, when K⁺ was introduced into Cu-SAPO-34, the Cu²⁺ content and the number of acidic sites is reduced, leading to the decrease in NH₃-SCR activity [24]. However, to our knowledge, there is no comprehensive study regarding the effect of contaminants (K, Na, Ca, and Mg) on the framework, the content of copper species, and the availability of acidic sites on Cu-SAPO-18.

Consequently, Cu-SAPO-18 catalyst, doped with diverse contaminants (K, Na, Ca, and Mg), was prepared, with the aim of elucidating the impact of these metals on the construction, availability of acidic sites and content of copper species on Cu-SAPO-18. NO conversion is discussed based on the variation in these physicochemical properties of Cu-SAPO-18. Additionally, the influence of metals on the NH₃-SCR reaction was also investigated through kinetic examination of NH₃-SCR.

The Cu-SAPO-18 catalyst (SiO₂:Al₂O₃:P₂O₅:Cu-TETA:DIPEA: H₂O = 0.6:1:0.9:0.1:1.65:45) was directly synthesized, and the elaborate synthesis procedure was explained in other studies [11, 13]. The final Cu-SAPO-18 powder was doped with K, Na, Ca, and Mg, containing 0.5 and 1.0 mmol/g_catal, by instilling the desired concentration of KNO₃, NaNO₃, Ca(NO₃)₂, and Mg(NO₃)₂ solution into the Cu-SAPO-18 catalyst. Then, the slurries were dried under air overnight at 90 ℃. Next, the powders were calcined at 600 ℃ for 5 h. The resulting powders were labeled Cu/M-SAPO-18(x), where "M" denotes the impregnated metals (M = K, Na, Ca, and Mg) and "x" stands for the metal amount. Finally, a certain amount of Cu/M-SAPO-18(x) was coated on the cordierite, respectively. The obtained monolithic catalysts were applied to evaluate NO conversion. More details are displayed in the Supporting Information.

N₂ adsorption-desorption was conducted on the V-Sorb 2800 P instrument to obtain the pore volume and BET specific surface area of the catalysts, and the samples were degassed at 250 ℃ for 5 h under vacuum before the measurement. ²⁷Al MAS nuclear magnetic resonance (NMR) spectra were obtained on an Agilent 600 DD2 spectrometer with a Larmor frequency of 156.25 MHz. The X-ray diffraction (XRD) charts were acquired from an X'Pert Pro X-ray diffractometer with Cu-K_α radiation, at 40 mA and 40 kV. H₂ temperature-programmed reduction (H₂-TPR) was recorded on an automatic chemisorption instrument (Auto Chem 2920II). The catalyst (50 mg) was preprocessed under Ar gas flow at 550 ℃ for 1.5 h and, subsequently, the catalyst was heated from ambient temperature to 1000 ℃ under a 10% H₂/Ar gas mixture. As for NH₃ temperature-programmed desorption (NH₃-TPD), the catalyst pretreatment process was consistent with that of TPR. The NH₃ adsorption of catalyst took place at 100 ℃, and the process was retained for 45 min. Ultimately, the catalyst temperature was elevated from 100 to 600 ℃ under Ar gas flow. The electron paramagnetic resonance (EPR) spectrum was obtained with a JEOL-FA200 apparatus. The catalyst (10 mg) placed in a tube was subjected to EPR measurements at –170 ℃ with a modulated frequency of 100 kHz and an amplitude of 5 G.

The catalytic performance of monolithic Cu-SAPO-18 was carried out in a stainless-steel tubular reactor under a fixed bed. A total of 250 g/L of Cu-SAPO-18 was inserted into the constant temperature zone of the reactor. Then, a thermocouple was stuck into the uniform temperature region to detect the bed temperature. The simulated reactant gases consisted of NO, NH₃, O₂, and H₂O, which accounted for 0.1%, 0.11%, 5%, and 10%, respectively. Meanwhile, N₂ was used as a carrier gas to purge these reactant gases into the reactor. The H₂O was produced by an injection pump (TYD01-01) containing deionized water. The overall gas flow velocity was 1045 mL/min. After the finish of the SCR reaction inside the catalyst bed, the remaining content of NO was investigated in a waste gas emission analyzer (FGA-4100-5G). Moreover, an NH₃ trap was installed in front of the analyzer to avert errors arising from excessive NH₃. The NO conversion was computed as ([NO]_inlet–[NO]_outlet)/[NO]_inlet × 100%.

The NH₃-SCR dynamic tests over fresh and contaminated Cu-SAPO-18 were executed in the reactor described above, using 60 g/L monolithic catalyst. The concentration of the simulated reaction gas and the corresponding overall gas flow velocity are the same as depicted above. Dynamic steady-state evaluations were conducted, ranging from 140 to 190 ℃. The exhaust fumes discharged from the fixed bed were monitored by a flume emission analyzer. The calculation formula for the NH₃-SCR reaction rates is exhibited in the Supporting Information.

Table 1 shows the elemental compositions of the fresh and contaminated Cu-SAPO-18 samples, and the compositions were determined by ICP-OES technology. The molar ratio of Si:Al:P in all samples is 0.27:0.68:1.0. The Cu content in fresh Cu-SAPO-18 is 2.10 wt%, and other contaminated samples present similar Cu content. Furthermore, the contents of K, Na, Ca, and Mg in Cu-SAPO-18 are analogous to the theoretical results (0.5 mmol/g_catal or 1.0 mmol/g_catal).

To determine the influence of contaminants on the Cu-SAPO-18 structure, XRD and BET tests were conducted. Fig. 1a illustrates that both fresh Cu-SAPO-18 and Cu/M-SAPO-18(0.5) catalysts exhibit the typical AEI structure [11, 25] and retain similar intensities in diffraction peaks, suggesting that the loading of 0.50 mmol/g_catal contaminants has little effect on AEI structure or the crystallinity of Cu-SAPO-18. Moreover, there are no CuO nor K, Na, Ca, or Mg peaks observed, which indicates that they are highly dispersed on Cu-SAPO-18 or that there is a detection limitation because of their small sizes [3]. When the metal content increases to 1.0 mmol/g_catal, the Cu/K-, Cu/Ca-, and Cu/Mg-SAPO-18(1.0) catalysts all preserve the typical AEI framework but is accompanied by a decrease in their crystallinity. This reveals that their zeolite frameworks suffer some damage. For the Cu/Na-SAPO-18(1.0) catalyst, its crystallinity decreases noticeably (as shown in Fig. S2). Meanwhile, a rather strong tridymite diffraction peak is detected [26, 27], illustrating that the introduction of excessive Na into the Cu-SAPO-18 catalyst causes partial decomposition of its structure. In addition, significant diffraction peaks of CuO species over Cu/Na-SAPO-18(1.0) catalyst at 35.5° and 38.6° are detected [13, 28], causing speculation that a large amount of Na facilitates the aggregation of copper species as CuO clusters over the external surface of the Cu/Na-SAPO-18(1.0) catalyst, resulting in partial structure collapse.

As displayed in Table 1, the introduction of contaminants into Cu-SAPO-18 affects the BET surface area and pore volume of the catalyst. It is worth noting that the specific surface area and pore volume decrease for all catalysts with 0.5 mmol/g_catal contaminants, in contrast to the fresh catalyst, to which the partial pore blocking by copper species is ascribed [29, 30]. With the increase in contaminant content to 1.0 mmol/g_catal, the Cu/M-SAPO-18(1.0) exhibits a greater decrease in the BET surface area and pore volume, which is due to the partial deterioration of the zeolite construction. In addition, such a decrease in the BET surface area and pore volume varies with distinct metals. For instance, the BET surface area for Cu/K-, Cu/Na-, Cu/Ca-, and Cu/Mg-SAPO-18(1.0) is reduced by 40.1%, 56.0%, 39.4%, and 38.0%, respectively, and the pore volume decreases by 40.7%, 58.1%, 40.7%, and 39.4%, respectively. This phenomenon demonstrates that the destruction on the Cu-SAPO-18 structures caused by Na is more severe than that from K, Ca, and Mg, which is in accordance with XRD results.

To further affirm the above observations, the ²⁷Al MAS NMR was measured to detect the variation in the Al coordination of Cu-SAPO-18. Fig. 2 presents the ²⁷Al NMR spectra of the fresh and contaminated catalysts. Three peaks appear for all samples at 39, 13, and -13 ppm. The three peaks are attributed to the tetrahedrally coordinated framework of Al atoms (Al^IV), pentahedrally coordinated Al atoms (Al^V), and octahedrally coordinated Al atoms (Al^VI), respectively [31, 32]. The percentage ratio of the diverse Al coordination structures is summarized in Table 2. In terms of the fresh catalyst, the proportion of the framework Al^IV species is 61.8%. After impregnating the catalyst with 0.5 mmol/g_catal contaminants, the percentage of tetrahedral Al atoms in Cu/K-, Cu/Na-, Cu/Ca-, and Cu/Mg-SAPO-18(0.5) catalysts drops to 47.0%, 45.3%, 50.8%, and 55.0%, respectively. This causes the breakage of Si–O–Al bonds, leading to the transformation of framework Al^VI species to pentahedral and octahedral Al atoms [33]. With the increase in contaminants to 1.0 mmol/g_catal, the percentage of Al^IV species for Cu/M-SAPO-18(1.0) catalysts decreases more than that for Cu/M-SAPO-18(0.5) catalysts. It is remarkable that the loss of framework Al^IV species in Cu/Na-SAPO-18(1.0) is greater than that for the other Cu/M-SAPO-18(1.0) catalysts, implying a more serious destruction in the Cu/Na-SAPO-18(1.0) framework. These data are consistent with BET results.

The H₂-TPR curves were analyzed to identify the distribution of different copper species in all samples. As seen from Fig. 3, five reduction peaks (A, B, C, E, and F) in the fresh sample are monitored. Peak A at 244 ℃ and peak C at 331 ℃ are the representative reduction peaks in the isolated Cu²⁺ substances, residing at the ellipsoidal cavity near the six-membered ring and in the center of the hexagonal prism, respectively [5, 34, 35]. Peak B at 289 ℃ corresponds to the reduction of CuO to Cu⁰ via a single-step procedure [35]. The unstable and high steady Cu⁺ substances were reduced to Cu⁰, resulting in the generation of peak E at 570 ℃ and peak F above 900 ℃. The Cu⁺ substances are derived from two aspects, the incompletely reduced Cu²⁺ and the inherent existence in the sample [36, 37]. Nevertheless, when K, Na, Ca, and Mg are introduced into Cu-SAPO-18, a new peak (D) is detected. According to the ²⁷Al-NMR results, the contaminated catalysts possess a greater extra-framework Al percentage than the fresh catalyst, and the extra-structure Al can react with isolated Cu²⁺ substances to generate CuAl₂O₄-like species [23, 38, 39]. Thus, the new peak (D) is assigned to the reduction of CuAl₂O₄-like species.

Based on Fig. 3a, after impregnating the fresh catalyst with 0.5 mmol/g_catal contaminants, the peak intensity of isolated Cu²⁺ substances in the ellipsoidal cavity (peak A) decreases significantly, and the isolated Cu²⁺ substances in the hexagonal prism (peak C) are also slightly reduced. Such a decline is interpreted by the replacement of partially isolated Cu²⁺ by K⁺, Na⁺, Ca²⁺, and Mg²⁺, separately. Concomitantly, the peak intensity of CuO and CuAl₂O₄-like substances increases, owing to the fact that the replaced Cu²⁺ ions, free from the ion exchange sites, will aggregate into CuO species or react with extra-framework Al, forming CuAl₂O₄ during calcination [23, 24]. This will block the micropore of the Cu-SAPO-18 catalyst. After increasing the amount of contaminants to 1.0 mmol/g_catal, the fall in the peak intensity of isolated Cu²⁺ substances at low reduction temperatures is dramatic, indicating that isolated Cu²⁺ species in the ellipsoidal cavity are so unstable that they are prone to be replaced. Simultaneously, more CuO and CuAl₂O₄ species appear, arising from the transformation of many Cu²⁺ substances during calcination. Strikingly, notwithstanding the great many CuO species generated in Cu/Na-SAPO-18(1.0) during calcination, the formation of CuO species would cause severe damage to its framework [39], which could be demonstrated by XRD and BET results.

In addition, EPR analysis was also measured to further investigate the coordination environment of isolated Cu²⁺ and semi-quantitatively confirm the isolated Cu²⁺ content [27, 40]. As Fig. 4 shows, the Cu²⁺ substances in the fresh and contaminated hydrated Cu-SAPO-18 catalysts all show one peak, which is an axially symmetric signal, implying that there is only one type of isolated Cu²⁺ substance in the hydrated samples. The EPR signal values of g_∥= 2.393 mT, g_⊥= 2.06 mT, and A_∥= 13.11 mT correspond to the isolated Cu²⁺, which is octahedrally coordinated to the three oxygens of the zeolite structure and the three water molecules, residing in the ellipsoidal cavity near the six-membered ring [41, 42]. However, two kinds of isolated Cu²⁺ species are observed in H₂-TPR profiles, which might be due to the migration of isolated Cu²⁺ from one location to another during thermal treatment [5, 41].

Concurrently, the amount of total isolated Cu²⁺ is semi-quantitatively determined via the double integration of the EPR spectra and the outcomes are plotted in Fig. 4c. After K, Na, Ca, and Mg impregnation, the quantity of total isolated Cu²⁺ in contaminated Cu/M-SAPO-18(x) decreases compared to the fresh catalyst, and higher contaminant content corresponds to a greater decrease in the sum of Cu²⁺ substances, revealing the H₂-TPR outcome. Moreover, the extent of the K, Na, Ca, and Mg contaminated effect on the Cu²⁺ substances of Cu-SAPO-18 reveals a disparity. For instance, the total isolated Cu²⁺ contents in Cu/K-, Cu/Na-, Cu/Ca-, and Cu/Mg-SAPO-18(0.5) are reduced by 11.3%, 20.9%, 7.7%, and 8.4%, respectively. Increasing the contaminants to 1.0 mmol/g_catal, the impregnation of K, Na, Ca, and Mg decreases the sum of the total Cu²⁺ substances by 30.4%, 48.9%, 21.2%, and 24.3%, respectively. Consequently, it can be concluded that the impregnation of Cu-SAPO-18 catalyst with K and Na gives rise to greater substitution of Cu²⁺ species, especially with Na. Such a substitution will lead to the reduction in Lewis acid sites generated by isolated Cu²⁺.

The variation in acidic sites of fresh and contaminated Cu-SAPO-18 was investigated by measuring NH₃-TPD. Fig. 5 displays the NH₃-TPD curves. Three NH₃ desorption peaks appear, which are attributed to NH₃ molecule desorption from weak Brønsted acidic sites (peak A, Si–OH and P–OH), Lewis acidic sites (peak B), and strong Brønsted acidic sites (Si–OH–Al) and strong Lewis acidic sites related to Cu²⁺ species (peak C) [14, 33, 43, 44]. Additionally, the K-, Na-, Ca-, and Mg-contaminated Cu-SAPO-18 catalysts also exhibit three similar desorption peaks, and the number of acidic sites is calculated from the NH₃-TPD curves and summarized in Table 3.

As shown in Table 3, the number of overall acidic sites within Cu/M-SAPO-18(0.5) is less than that of the fresh Cu-SAPO-18 catalyst. This phenomenon is due to the Brønsted acidic sites (H⁺) and Lewis acidic sites (Cu²⁺ species) being replaced by K, Na, Ca, and Mg, and the pore blocking of Cu/M-SAPO-18(0.5), which could cause partial coverage of acidic sites [17, 23, 45]. Moreover, Cu/K-SAPO-18(0.5) catalyst exhibits weak desorption over the entire temperature range, suggesting that the K⁺ located at ion exchange sites of Cu-SAPO-18 shows low affinity to NH₃ [23, 24]. However, it seems that Na⁺ could adsorb NH₃ molecules, as demonstrated by the increased number of weak and moderate acidic sites (peak A and peak B) in Cu/Na-SAPO-18(0.5) below 300 ℃ [23]. Meanwhile, the adsorption of NH₃ molecules onto Ca²⁺ and Mg²⁺ at moderate temperatures corresponds to the increase in the content of moderate acidic sites (peak B) of Cu/Ca- and Cu/Mg-SAPO-18(0.5) samples [17, 23]. With the increase in contaminants to 1.0 mmol/g_catal, the overall acidic sites in Cu/M-SAPO-18(1.0) decreases strikingly, which is explained by a greater loss of H⁺ and Cu²⁺, and the covering of acidic sites, induced by local structure collapse. The impregnation of K, Na, Ca, and Mg on the raw sample took place at the expense of Brønsted acidic sites, Cu²⁺ substances, and the surface area, leading to the decrease in the overall acidic sites of Cu/M-SAPO-18(x).

Furthermore, the decrease in the overall acidic sites within Cu/M-SAPO-18(x) follows the order Cu/K- > Cu/Na- > Cu/Ca- > Cu/Mg-SAPO-18(x). This is opposite to the effective hydrated diameter of the contaminants, which is presented in the order K⁺ < Na⁺ < Ca²⁺ < Mg²⁺ [46]. Although two K⁺ or Na⁺ can substitute for two H⁺ or one Cu²⁺, and one Ca²⁺ or Mg²⁺ can replace two H⁺ or one Cu²⁺, K⁺ and Na⁺ with smaller hydrated diameters are more prone to enter the pores of Cu-SAPO-18 catalyst than Ca²⁺ and Mg²⁺ with larger hydrated diameters. Consequently, a larger quantity of Brønsted acidic sites and isolated Cu²⁺ is easier to exchange with K⁺ and Na⁺, leading to a greater reduction in total acidic sites in Cu/K- and Cu/Na-SAPO-18(x).

During the deNO_x reaction, NH₃ molecules could adsorb on both isolated Cu²⁺ and Brønsted acidic sites, while the deNO_x reaction solely happens on isolated Cu²⁺, identified as the active sites [5, 33]. At low temperatures, only the NH₃ molecules adsorbed on isolated Cu²⁺ take part in the deNO_x reaction. Thus, the isolated Cu²⁺ species at low reduction temperatures affect the low-temperature NO conversion. At high temperatures, the adsorption of NH₃ on the isolated Cu²⁺ is unstable. Therefore, the NH₃-SCR reaction occurs by supplying the NH₃ from the Brønsted acidic sites to the Cu²⁺ substances [47, 48]. This indicates that the NO conversion is involved in the acidic sites and Cu²⁺ substances.

Fig. 6 depicts the NH₃-SCR performance over the fresh and contaminated Cu-SAPO-18. The fresh Cu-SAPO-18 catalyst reveals outstanding catalytic activity during the entire temperature range, and the NO conversion exceeds 90% in the temperature region of 200–525 ℃ because of the existence of a great many acidic sites and Cu²⁺ substances. After the doping of 0.5 mmol/g_catal or 1.0 mmol/g_catal of contaminants to Cu-SAPO-18, all contaminated catalysts suffer deactivation, and their catalytic activities decrease in the entire temperature range. The higher the contaminant content, the more pronounced the decrease in NO conversion. Such a decrease is due to the reduction in acidic sites and Cu²⁺ content, and the structural damage. Meanwhile, the decline in isolated Cu²⁺ in the ellipsoidal cavity is responsible for the decrease in the low-temperature deNO_x activity [15]. The NH₃ oxidation induced by the large quantity of CuO species leads to the decline in high-temperature NH₃-SCR activities [7]. Simultaneously, the deactivation effect of contaminants on Cu-SAPO-18 is different, which is presented by the order K > Na > Ca > Mg. For instance, the catalytic performance of Cu/Mg-SAPO-18(0.5) is approximately 87% from 275 to 450 ℃. In contrast, the highest NO conversion within Cu/K-SAPO-18(0.5) is only 78% at 350–475 ℃. Furthermore, comparing Cu/Ca- to Cu/Na-SAPO-18(0.5), the former exhibits slightly higher catalytic performance (around 82%) than the latter from 250 to 475 ℃. For the 1.0 mmol/g_catal contaminants, the highest NO conversion over Cu/K-, Cu/Na-, and Cu/Ca-SAPO-18(1.0) catalysts is 24% at 525 ℃, 40% at 525 ℃, and 60% at 475 ℃, respectively, corresponding to a decline of 76%, 60%, and 40% compared with the fresh catalyst, respectively. However, the highest catalytic activity of Cu/Mg-SAPO-18(1.0) is 85% at 425–450 ℃. The deactivation in the fresh catalyst is related to the decreased level in isolated Cu²⁺ and acidic sites, and the framework damage.

Fig. 7 shows the relative number of acidic sites (according to the NH₃-TPD results) and isolated Cu²⁺ (according to the EPR results) as a function of the actual metal content. The introduction of K, Na, and Ca into the Cu-SAPO-18 results in a reduced number of acidic sites more significant than that of isolated Cu²⁺. This shows that the K⁺, Na⁺, and Ca²⁺ ions are more apt to exchange with H⁺ than Cu²⁺. However, Mg²⁺ seems to easily substitute for the Cu²⁺ instead of H⁺, as demonstrated by the more distinct diminished level of Cu²⁺ substances in Cu/Mg-SAPO-18(x). Simultaneously, the Cu²⁺ ions, free from the Cu-SPAO-18 catalyst, can transform into the CuO species during calcination, which can cause the catalyst structural damage and decline in specific areas. Furthermore, the Cu²⁺ content of Cu/M-SAPO-18(x) follows the order: Cu/Na- > Cu/K- > Cu/Mg- > Cu/Ca-SAPO-18(x). The decrease in the BET specific area for Cu/M-SAPO-18(x) follows the order: Cu/Na- < Cu/K- < Cu/Ca- < Cu/Mg-SAPO-18(x). The reduction in the number of acidic sites of Cu/M-SAPO-18(x) is presented in the order: Cu/K- > Cu/Na- > Cu/Ca- > Cu/Mg-SAPO-18(x), which corresponds to the extent of catalyst deactivation. A conclusion that can be drawn is that deNO_x activity can be jointly affected by the isolated Cu²⁺, catalyst structure, and acidic sites. However, the acidic sites might affect the NH₃-SCR activity more significantly than isolated Cu²⁺ and catalyst structure. Additionally, the kinetics studies were carried out at low temperature to investigate the effect of the contaminants on NH₃-SCR mechanism. As shown in Fig. 8, all catalysts possess similar apparent activation energies, indicating that the existence of K, Na, Ca, and Mg has little influence on deNO_x mechanism over Cu-SAPO-18 [23]. Moreover, the NO conversion speeds correspond to the NH₃-SCR activities. This suggests that a larger number of acidic sites and Cu²⁺ substances in the catalyst correspond to better catalytic activity.

Cu-SAPO-18 contaminated by K, Na, Ca, and Mg suffers deactivation in the NH₃-SCR reaction. Such deactivation becomes more significant by increasing the contaminant content from 0.5 to 1.0 mmol/g_catal. The deactivation of Cu-SAPO-18 is connected with the decline in the content of acidic sites and Cu²⁺ substances, and the framework destruction of the catalyst. When the K, Na, Ca, and Mg was introduced into Cu-SAPO-18, these metal ions, with different hydrated ion diameters, substitute for the H⁺ or Cu²⁺ to different extents. This leads to different degrees of decrease in Cu²⁺ substances and acidic sites. Cu²⁺, unable to interact with the ion exchange sites, transform into the CuO species and CuAl₂O₄-like species during high-temperature calcination. This causes Cu-SAPO-18 to experience micropore blockage and further covers the acidic sites of the catalyst. Finally, the deactivation of Cu-SAPO-18 follows the order of Cu/K- > Cu/Na- > Cu/Ca- > Cu/Mg-SAPO-18(x), which corresponds to the effect of the contaminants on the acidic sites. Thus, to some extent, the influence of acidic sites on NO conversion seems to be more conspicuous. Finally, the evaluation of NH₃-SCR dynamics was conducted. It can be concluded that the introduction of contaminants into Cu-SAPO-18 shows little influence on the deNO_x mechanism.

The authors are grateful to the tests offered by the Analysis and Testing Center, Huazhong University of Science and Technology, Wuhan, China.