English

• 0.   Introduction

  With the popularization of diesel vehicles in people's life, the issues of nitrogen oxides (NO_x) emission is increasingly serious. NO_x has been an important source of air pollution which could cause significantly harm to both the environment (e. g. acid rain, ozone depletion, photochemical smog, and greenhouse effect) and human health^[1-3], the selective catalytic reduction of NO_x with NH₃ (NH₃-SCR) is an efficient NO_x elimination technology among all NO_x emission control tech-nologies for the diesel exhausts^[4-6].

  Zeolites (e.g. SAPO-34, BEA, and ZSM-5) as supports with a large specific surface area, channel complex and orderly advantages, can promote the dispersion of active metal species, and abundant surface acid sites from the skeleton of zeolites are beneficial to the adsorption of NH₃^[7-9]. Among the reported materials, Cu/BEA catalysts have shown relatively excellent catalytic performance and lower cost, which have attracted wide attention^[10-12]. However, poor hydrothermal stability of Cu/BEA limited its commercial application due to dealumination of the zeolite and subsequent formation of the inactive species after the hydrothermal treatment^[13]. To meet the increasingly stringent regulations on emitted NO_x, it is essential to improve the hydrothermal stability of Cu/BEA. Zhao et al. ^[14] found that Ce and Nb additives could not only improve the NO_x conversion of Cu/BEA, but also improve its hydrothermal stability. The hydrothermal stability of Cu-exchanged BEA was obviously improved by the introduction of Fe as the co-active component^[15]. Our previous work has reported that the hydrothermal stability of Cu/BEA was effectively enhanced by the alkaline metal Ba due to the formation of more isolated Cu²⁺ species from the enhanced interactions between Cu and other surrounding atoms by the addition of Ba^[16]. The above works mainly focused on stabilizing the active Cu species, but there is little literature to further stabilize the skeleton structure of BEA to our best knowledge. Considering that the hydrothermal stability of the BEA catalyst has an important relationship with the hydrolysis of Si-O-Al bonds in the skeleton of BEA^[17]. Therefore, the additional silanizing reagent (e. g. ethyl orthosilicate, TEOS) deposited on the surface of BEA would be hydrolyzed firstly to protect against the attack of H₂O to Si-O-Al bonds and finally stabilize the structural skeleton of BEA^[18].

  Herein, the introduction of extraneous Si into CuCe/BEA attempts to improve the hydrothermal stability. ²⁷Al nuclear magnetic resonance (²⁷Al NMR), X-ray diffraction (XRD), NH₃-temperature programmed desorption (NH₃-TPD), in situ diffuse reflectance infra-red transform spectroscopy (in situ DRIFTS), H₂-temperature programmed reduction (H₂-TPR), and X-ray photoelectron spectrum (XPS) are employed to characterize and compare the effects of Si on the structure and active species of CuCe/BEA catalyst before and af-ter the hydrothermal treatment, with a focus on investi-gating the enhanced influences of Si on the hydrothermal stability of CuCe/BEA catalyst.

  1.   Experimental

  1.1   Catalyst preparation

  The Cu/BEA catalyst was prepared by an incipient wetness impregnation method. Firstly, 12.32 g commercial H/BEA (n_si/n_Al =25, Nankai University Catalyst Co., Ltd., China) was dropped into 9.24 mL solution contained 1.13 g Cu(NO₃)₂·3H₂O (AR, Kelong, Chengdu, China), and then, the mixture was stirred to ensure that the copper ions were dispersed well. Next, the mixture was dried in a water bath to a solid powder and calcined at 550 ℃ for 3 h in the static air in a tubular furnace. Finally, the calcined Cu/BEA powders were coated on the cordierite monolith (2.5 cm³, 62 cell per cm², Corning Ltd., USA), and the amount of catalyst deposited on the monolith substrate was about 160 g·L^-1. The fresh catalyst was denoted as Cu, and the mass fraction of copper was 3% (measured by inductively coupled plasma atomic emission spectrometry, ICP-AES).

  CuCe/BEA catalyst was prepared by a two-step method. Firstly, 10.00 g commercial H/BEA was ion-exchanged with 100.00 mL solution contained 4.34 g Ce(NO₃)₃ (AR, Kelong, Chengdu, China), which was then filtered, washed with distilled water and dried to obtain Ce/BEA material. Secondly, 9.24 mL solution contained 1.13 g Cu(NO₃)₃·3H₂ O was impregnated on 12.32 g Ce/BEA powder by an incipient wetness impregnation method as mentioned above. The fresh catalyst was denoted as CuCe.

  CuCeSi/BEA catalyst was obtained by modifying the CuCe/BEA with TEOS (AR, Kelong, Chengdu, China). Firstly, 2.25 mL TEOS was added to 20.00 mL solution of equal proportions of water and ethanol. Then, 10.00 g CuCe/BEA powder was added, stirred magnetically for 12 h, dried in a water bath, and calcined at 550 ℃ for 3 h in the static air in a tubular furnace. Finally, the coated fresh catalyst was denoted as CuCeSi.

  The monolithic Cu, CuCe and CuCeSi catalysts were hydrothermally treated at 700 ℃ for 10 h in a fixed-bed quartz flow reactor using air, containing 10% (volume fraction) water vapor flowing at a space velocity of 30 000 h^-1, The hydrothermally treated catalysts were denoted as Cu-HT, CuCe-HT and CuCeSi-HT, respectively.

  1.2   Catalyst characterization

  ²⁷Al NMR experiments were carried out on JEOL-ECZ600 instrument in order to detect dealuminization in BEA structure before and after the hydrothermal treatment.

  The XRD measurements were achieved by the Rigaku D/MAX ray diffractometer using Cu Kα radiation (λ =0.154 18 nm). The X-ray tube was operated at 45 kV and 25 mA. The samples were investigated in the 2θ range of 5°~60° at a scanning speed of 0.02 (°)· min^-1.

  SEM images of the catalyst were observed by field emission scanning electron microscopy (FESEM) on a Hitachi SU8220 instrument operating at 10 kV.

  NH₃-TPD was also performed on the thermal conductivity detector (Xianquan, TP5076). And approximately 100 mg sample was pretreated under the He gas (30 mL·min^-1) at 550 ℃ for 30 min, and then cooled to 120 ℃ in the above atmosphere. Next, the gas was switched into the NH₃ (30 mL·min^-1) until saturation. After that, the catalyst was flushed by He (30 mL· min^-1) to remove physical adsorbed NH₃. Finally, the NH₃-TPD desorption process was started at a linear heating rate of 10 ℃ ·min^-1 from 120 to 800 ℃ under the He flow (30 mL·min^-1).

  The Nicolet 6700 spectrometer with a high-temperature environmental cell which contained a KBr window and a DTGS detector was used to measure the in situ DRIFTS of catalysts. Before each test, the sample was pretreated in the pure N₂ (300 mL·min^-1) flow from 50 to 350 ℃ at a rate of 10 ℃ min^-1, and pretreated at 350 ℃ for 30 min. Then, the background spectra were acquired from 350 to 50 ℃ per 50 ℃. Pure N₂ was switched to the adsorption gas (volume fraction of 0.1% NH₃, N₂ as the balance gas) at 50 ℃. Afterwards, the temperature was raised at the rate of 10 ℃ ·min^-1 from 50 to 350 ℃, and the adsorption gas spectra were determined at per 50 ℃. Finally, spectra of catalysts were measured by the means of subtracting the background diffuse reflection spectra.

  The H₂-TPR was carried on the thermal conductivity detector (Xianquan, TP5076). And approximately 100 mg sample was pretreated under volume fraction of 10% O₂/N₂ (30 mL·min^-1) at 550 ℃ for 30 min. Then, the sample was cooled to 50 ℃ in above atmosphere, and the H₂- TPR desorption process was started at a linear heating rate of 10 ℃·min^-1 from 50 to 900 ℃ under a volume fraction of 10% H₂/N₂ (30 mL·min^-1) flow.

  The XPS was carried out on a spectrometer (XSAM-800, KRATOS Co.) with Al Kα radiation under ultra-high vacuum (UHV). The C1s peak (284.6 eV) was used for the calibration of binding energy value.

  1.3   NH₃-SCR activity measurements

  Catalyst activity measurements for NH₃ -SCR were tested in a fixed bed quartz flow reactor. The monolith substrate loaded into a horizontal quartz microreactor and the reaction temperature was measured by two K-type thermocouples. The feed gases were regulated by mass-flow controllers and composed of volume fraction of 0.02% NO (c_{NOx, in}), 0.02% NH₃, 5% H₂O and 10% O₂, and N₂ was used as the balance gas with a gas hourly space velocity of 40 000 h^-1. Before each test, the catalyst was pretreated at 550 ℃ in the reaction atmosphere. The concentration of outlet gases (c_{NOx, out}) were monitored by an FTIR spectrometer (Antaris IGS, Thermo Fisher Scientific), and the test temperature was varied between 550~150 ℃. The NO_x conversion was calculated by the followed equation:

  $ {\rm{N}}{{\rm{O}}_x}\;{\rm{conversion}} = \left( {1-\frac{{{c_{{\rm{N}}{{\rm{O}}_{x, {\rm{ out}}}}}}}}{{{c_{{\rm{N}}{{\rm{O}}_{x, {\rm{ in}}}}}}}}} \right) \times 100\% $

  “c_{NOx, out}”and“c_{NOx, in}”represent the outlet and inlet reactant concentrations of the reactor, respectively.

  2.   Results and discussion

  2.1   Structural and morphological properties

  ²⁷Al NMR spectra are employed to characterize the dealuminization of CuCe and CuCeSi before and after the hydrothermal treatment, as shown in Fig. 1. All four samples display three peaks: the peak of chemical shift at 54 is ascribed to the tetracoordinated framework aluminum atoms (Al^Ⅳ), the peak of chemical shift at 12 is attributed to the pentacoordinated aluminum atom (Al^Ⅴ), and the peak of chemical shift at-13 is assigned to the hesacoordinated aluminum atom (Al^Ⅵ). BEA molecular sieve is mainly composed of tetrahedral SiO₄ and AlO₄, Al^Ⅳ is the main Al species in the structure of BEA molecular sieve, showing the largest NMR peak. Al^Ⅴ may be due to the partial incomplete structure existing in the synthesis of BEA molecular sieve, or the hydrolytic destruction of some BEA zeolite supports occurred during the preparation of the catalyst. AlVI is due to the hydrolysis of Si-O-Al bonds in the hydrothermal process, which results in the recoordination of Al^Ⅳ and Al^Ⅴ with H₂O molecules to form Al atoms with hesacoordinated^[19-21]. The relative content of Al^Ⅳ over CuCeSi was slightly higher than that in CuCe, as listed in Table 1. Al^Ⅳ species are significantly decreased due to the dealuminization for both catalysts after the hydrothermal treatment, although CuCeSi-HT catalyst still possesses higher relative content of Al^Ⅳ after the hydrothermal treatment, indicating that the introduction of Si can inhibit the attack of H₂O molecules on the Si-O-Al bonds, inhibit the occurrence of dealuminization and protect the structure of BEA during the process of the hydrothermal treatment.

  Figure 1

  

  Figure 1.  ²⁷Al NMR spectra of CuCe and CuCeSi catalysts before and after the hydrothermal treatment

  下载: 全尺寸图片 幻灯片

  Table 1

  

  Table 1.  Integrated area (A) of Al^Ⅳ atoms in CuCe and CuCeSi catalysts

  下载: 导出CSV

  Sample	AAlⅣ	AAlⅤ	AAlⅥ	AAlⅣ/ (AAlⅣ+AAlⅤ+AAlⅥ) / %
  CuCe	922.7	94.2	32.0	88.0
  CuCeSi	945.6	5.1	70.5	92.6
  CuCe-HT	530.1	103.2	78.1	74.5
  CuCeSi-HT	648.9	35.3	52.3	88.1

  To investigate the effect of Si on the structure of BEA, XRD patterns of CuCe, CuCeSi, and their hydrothermal samples are shown in Fig. 2. Only diffraction peaks ascribed to the typical structure of BEA were detected over all samples and no obvious other diffraction peaks ascribed to Cu_xO_y and SiO₂ species^[22-24] (Fig. 2a). No peaks assigned to metallic or metal oxides species can be observed in Fig. 2a, which could be that they are highly dispersed, or the concentration is too low to meet the detection limitation of XRD, or they do present as isolated metal species in the BEA molecular skeleton^[25]. The main diffraction peak at 2 θ=22.5° was not shifted and no significant difference between the crystallinity of CuCe and CuCe Si indicates little effect of Si on the structures of BEA. Compared with the fresh catalysts, the intensity of the characteristic diffraction peaks of BEA decreased to a certain extent for the HT catalysts, indicating that the structure of catalysts is destroyed during the process of the hydrothermal treatment, which is consistent with the ²⁷Al NMR results.

  Figure 2

  

  Figure 2.  XRD patterns of CuCe and CuCeSi catalysts before and after the hydrothermal treatment (a) and crystallinity loss ratio (b)

  下载: 全尺寸图片 幻灯片

  In order to further evaluate the sample crystallinity changes before and after the hydrothermal treatment more accurately, it is assumed that the crystallinity of fresh Cu catalyst was 100%, and the relative content changes were calculated through the main diffraction peak of 2θ =22.5°, the results are shown in Fig. 2b. Although the crystallinity of both catalysts was affected by the hydrothermal treatment, the crystallinity of CuCe was declined by 15.4%, while that of CuCeSi was only 7.1%. The above phenomena suggest that the modification of Si can prevent the destruction of the skeleton structure of BEA zeolite during the process of the hydrothermal treatment, protect the integrity of the skeleton structure.

  The SEM images displayed in Fig. 3 show the effect of Si on the morphology of CuCe before and after the hydrothermal treatment. CuCe and CuCeSi have the characteristic regular morphology of BEA zeolite, but due to the low amount of Si, there was no obvious difference between them. After hydrothermally treated at 700 ℃ for 10 h, the surface morphology of the catalysts are severely damaged, but CuCeSi-HT could retain more regular morphology than CuCe- HT^[26]. Combined with ²⁷Al NMR and XRD, the results show that Si modification can effectively inhibit the hydrolysis of Si-O-Al bonds, protect the skeleton and morphological of BEA and improve the hydrothermal stability of CuCe.

  Figure 3

  

  Figure 3.  SEM images of CuCe and CuCeSi catalysts before and after the hydrothermal treatment

  下载: 全尺寸图片 幻灯片

  2.2   NH₃ adsorption

  The hydrolysis of BEA skeleton will lead to the loss of surface acidity of zeolite during the process of the hydrothermal treatment, and it is well known that the surface acidity of the catalyst plays a crucial role in NH₃-SCR reaction. So NH₃- TPD was employed to investigate the influence of Si on the surface acidity of CuCe before and after the hydrothermal treatment (Fig. 4). Three desorption peaks were detected in the NH₃-TPD curves of all samples in Fig. 4. The low-temperature desorption peak (α) at 150~250 ℃ can be classified as NH₃ adsorbed on the weak silanol (Si-OH) group, which is the Lewis acid sites generated by the structural defect. The medium-temperature desorption peak (β) at 250~400 ℃ can be attributed to the NH₃ species adsorbed on the strongly acidic hydroxyl group, that is, the Brønsted acid sites^[27]. The high-temperature desorption peak (γ) above 400 ℃ can be attributed to the strong Lewis acid sites generated by the introduced copper ions^[28]. The surface acidity of CuCe is slightly depressed after introducing Si, which may be caused by the hydrolysis of TEOS at the acid sites of BEA, part of the acid sites of BEA was covered by SiO₂ after calcined, preventing it from adsorbing NH₃. Moreover, the amount of acid sites of the fresh catalysts was decreased significantly after the hydrothermal treatment, which may be caused by H₂O molecules attack the Si-O-Al bonds during the hydrothermal treatment process leading to structural collapse of BEA zeolite, resulting in less acidic sites. Interestingly, Si modification can inhibit the loss of surface acidity of CuCe during the hydrothermal treatment, and the amount of acidic sites of CuCeSi-HT was even higher than CuCe- HT. NH₃-TPD results show that the modification of Si can stabilize the surface acidity of the catalyst and inhibit the loss of surface acidity of catalyst during the hydrothermal treatment.

  Figure 4

  

  Figure 4.  NH₃-TPD profiles of catalysts before (a) and after (b) the hydrothermal treatment

  下载: 全尺寸图片 幻灯片

  As a probe molecule, NH₃ can further explore the changes of acidity and acid sites of catalysts before and after the hydrothermal treatment. The in situ DRIFTS results of NH₃ adsorption on the catalysts after the hydrothermal treatment are shown in Fig. 5. Brønsted and Lewis acid sites can be both detected on the two HT catalysts via NH₃ adsorption. The Brønsted acid sites (3 361, 3 282 and 3 178 cm^-1) are mainly provided by zeolite supports, while the Lewis acid sites (1 627, 1 251 and 1 159 cm^-1) are mainly derived from Si-OH in the structure and the introduced active copper species^[29-31]. It can be observed that the Brønsted acid sites are the dominant acid sites for both catalysts. In order to more intuitively compare the adsorption of NH₃ between two catalysts and their acidity at the same temperature, the comparison results of NH₃ adsorption spectra at 350 ℃ are listed in Fig. 5c. It can be clearly seen that, CuCeSi-HT has more Brønsted acid sites than CuCe-HT at 350 ℃, which is from the Si-O-Al bonds on the zeolite skeleton structure, and this result is consistent with the NH₃-TPD, indicating that Si modification can protect the surface acid sites. Combined with XRD and ²⁷Al NMR results, CuCeSi-HT catalyst with a more complete skeleton structure due to the introduction of Si provides more Brønsted acid sites for adsorption and activation of NH₃ than CuCe catalyst, which may improve NH₃-SCR activity.

  Figure 5

  

  Figure 5.  In situ DRIFTS results of NH₃ adsorption over CuCe-HT (a) and CuCeSi-HT (b); Comparison of NH₃ adsorption of catalysts under the hydrothermal treatment at 350 ℃ (c)

  下载: 全尺寸图片 幻灯片

  2.3   Redox properties

  Isolated Cu²⁺ species are the main active species of CuCe for NH₃-SCR reaction. As shown in Fig. 6, H₂-TPR profiles of all samples can be divided into four reduction peaks^[32-34]: the first peak (α) centered at 262 ℃ and the third peak (δ) located at 447 ℃ can be assigned to the two-step reduction of isolated Cu²⁺ to Cu⁰; the second peak (β) at 357 ℃ is related the reduction of CuO to Cu⁰, and the last peak (γ) at 554 ℃ is related to Cu⁺ self-reduction (SCu⁺) to Cu⁰. H₂ consumptions were quantitatively calculated according to the CuO reference sample, and the results are summed up in Table 2. The decrease of isolated Cu²⁺ species was clearly observed in CuCe modified by Si, possibly due to the blocking and covering of the structures and active components by SiO₂. H₂ consumptions of all peaks were decreased after the hydrothermal treatment, but it can be seen from the quantitative results that the amount of isolated Cu²⁺ species in CuCeSi-HT was unexpectedly higher than that in CuCe-HT, and its amount of CuO was lower than that of CuCe-HT.

  Figure 6

  

  Figure 6.  H₂-TPR spectra of fresh (a) and hydrothermal treated (b) CuCe and CuCeSi catalysts

  下载: 全尺寸图片 幻灯片

  Table 2

  

  Table 2.  Quantitative analysis of the H₂-TPR profiles of the fresh and hydrothermal treated CuCe and CuCeSi catalysts

  下载: 导出CSV

  Sample	H2 consumption / (μmol·g-1)				Total H2 consumption / (μmol·g-1)
  	α (Cu2+ to Cu+)	β (CuO to Cu0)	δ (Cu+ to Cu0)	γ (SCu+ to Cu0)
  CuCe	77.0	62.7	173.2	70.3	383.2
  CuCeSi	61.5	66.5	56.3	153.2	337.5
  CuCe-HT	41.3	64.2	109.5	9.6	224.6
  CuCeSi-HT	50.4	54.2	36.4	67.4	208.5

  XPS has been widely used in the semiquantitative evaluation of catalyst surface species. In order to further characterized the chemical state of Cu species, the Cu2p XPS spectra of the catalysts before and after the hydrothermal treatment are shown in Fig. 7. All samples show characteristic peaks of Cu²⁺ species, including the main peaks of the Cu2p spin orbit split^[35] (Cu2p_3/2 and Cu2p_1/2 at 933~937 eV and 953~956 eV, respectively) and one satellite peak near 944.6 eV which is assigned to CuO^[36]. Cu2p_3/2 peaks are deconvoluted to the band at 933.8 and 936.7 eV, which are assigned to CuO and isolated Cu²⁺ species in ionexchanged positions, respectively^[37]. In order to better analyze the differences in the distribution of surface elements before and after the hydrothermal treatment of catalysts, the quantitative calculation results are shown in Table 3. There are significant differences in the distribution of copper species among different catalysts. For fresh catalysts, the content of isolated Cu²⁺ species on the surface of CuCeSi is lower than that of CuCe due to Si coverage, which is consistent with the TPR results (Table 3). After the hydrothermal treatment, the content of isolated Cu²⁺ is increased for both catalysts, possibly because the copper species in the bulk phase migrated to the surface and mainly existed in isolated Cu²⁺ and CuO forms. Isolated Cu²⁺ of 0.82% was detected on CuCeSi-HT, higher than that of 0.64% in CuCe- HT, which is consistent with H₂-TPR. In addition, the ratio of Cu²⁺/CuO increased from 0.28 to 0.42, indicating that the addition of Si can inhibit the aggre-gation of isolated Cu²⁺. Combined with the results of H₂-TPR and XPS, it can be concluded that the addition of Si can inhibit the aggregation of isolated Cu²⁺ to form CuO during the process of the hydrothermal treatment, so more isolated Cu²⁺ species could improve the low-temperature SCR activity, while less CuO species can weaken the oxidation of high-temperature NH₃. Therefore, the addition of Si may be beneficial to maintain the NH₃-SCR activity of CuCeSi-HT in the entire reaction temperature range after hydrothermal treatment.

  Figure 7

  

  Figure 7.  Cu2p XPS spectra for the fresh (a) and hydrothermal treated (b) CuCe and CuCeSi catalysts

  下载: 全尺寸图片 幻灯片

  Table 3

  

  Table 3.  Surface chemical composition and contents over catalysts before and after the hydrothermal treatment

  下载: 导出CSV

  Sample	Mass fraction of surface Cu / %		Mass ratio of isolated Cu2+ to CuO
  	Isolated Cu2+	CuO
  CuCe	0.29	2.03	0.14
  CuCeSi	0.26	1.86	0.14
  CuCe-HT	0.64	2.32	0.28
  CuCeSi-HT	0.82	1.97	0.42

  2.4   NH₃-SCR activity

  The curves of NO_x conversion as a function of reaction temperatures in NH₃-SCR reaction over Cu, CuCe and CuCeSi catalysts are shown in Fig. 8. It can be seen that both CuCe and CuCeSi catalysts show similar NO_x conversions except higher NH₃-SCR activity of CuCeSi catalyst than that of CuCe catalyst above 500 ℃ in Fig. 8a. Their low-temperature NO_x conversion both higher than that of Cu catalyst, and their values of T₉₀ (the reaction temperature when the NO_x conversion rate reaches 90%) are shifted to 186 ℃ from 218 ℃ of Cu catalyst, decreased by about 30 ℃. After the hydrothermal treatment at 700 ℃ for 10 h, the catalytic activities have a significant loss over all catalysts at the temperature below 500 ℃ (as shown in Fig. 8b). However, the maximum NO_x conversion over CuCeSi-HT was 85% in the temperature range of 175 to 400 ℃, which is higher than CuCe-HT and Cu-HT (80% and 78%, respectively). Furthermore, CuCeSi-HT can retain over 80% NO_x conversion between 200~550 ℃, while CuCe-HT and Cu-HT have over 80% NO_x conversion only in the high temperature range, indicating that Si modification can improve the hydro-thermal stability of the CuCe catalyst to some extent.

  Figure 8

  

  Figure 8.  NO_x conversion as a function of the reaction temperature over Cu, CuCe and CuCeSi during NH₃-SCR

  Before (a) and after (b) the hydrothermal treatment, reaction conditions: 0.02% NO, 0.02% NH₃, 5% H₂O, 10% O₂ (volume fraction), GHSV=40 000 h^-1

  下载: 全尺寸图片 幻灯片

  Combined with the characterization results to analyze, for fresh catalysts, the amount of isolated Cu²⁺ and the surface acid sites are slightly reduced in CuCeSi compared with CuCe due to the covering effect of SiO₂ generated by TEOS in the calcined process on the catalyst surface. But their NH₃-SCR activity are almost the same, probably because CuCeSi possesses enough isolated Cu²⁺ and acid sites to maintain the optimal activity. After the hydrothermal treatment, the attack of H₂O molecules on the Si-O-Al bonds causes the collapse of zeolite skeleton structure, resulting in the decrease of isolated Cu²⁺ and surface acid sites content, which leads to the decrease of catalysts activity. However, after the addition of Si, TEOS hydrolyze on the acid sites and cover the surface to play a certain protective role, thus preventing H₂O molecules from attacking the Si-O-Al bonds in the process of hydrothermal treatment, which is conducive to stabilizing the structure of zeolite skeleton. The more complete skeleton structure helps to stabilize the acid sites on the catalyst surface (particularly Brønsted acid sites), and inhibit the aggregation of isolated Cu²⁺. Therefore, compared with CuCe-HT, CuCeSi-HT has more isolated Cu²⁺ and more surface acid sites, and finally, exhibits better NH₃-SCR activity than CuCe-HT.

  3.   Conclusions

  In summary, the hydrothermal stability of CuCe catalyst is effectively improved through the modification of Si. The presence of Si stabilizes the skeleton structure of the zeolite, which improves the amount of acid sites on the catalyst surface and the dispersion of copper ions, and finally leads to the improvement of the hydrothermal stability of the catalyst. The results of XRD, SEM and ²⁷Al NMR reveal that the addition of Si can protect the Si-O-Al bonds, inhibit the attack of H₂O molecules in the process of hydrothermal treatment, so as to protect the skeleton structure of zeolite and improve the stability. The NH₃-TPD and in situ DRIFTS results show that more complete skeleton structure is beneficial to protect the structural acid sites on the surface of zeolite and reduce the loss of the amount of acid sites during the process of the hydrothermal treatment. The results of XPS and H₂-TPR show that although the modification of Si covered some active sites on the catalyst surface, a more complete skeleton structure is beneficial to improve the dispersion of copper ions and inhibit the migration and aggregation of copper ions to form CuO. Therefore, the modification of Si can provide more acid sites and highly dispersed Cu species which finally improve the hydro-thermal stability of CuCe catalyst for NH₃-SCR. This study proves the feasibility of adding Si to stabilize zeolite skeleton, which is helpful to improve the hydrothermal stability of catalysts and develop catalysts with excellent activity for NH₃-SCR.

  
  
• 1. [1]

     Park E, Chin S, Jeong J, et al. Microporous and Mesoporous Mater., 2012, 163:96-101 doi: 10.1016/j.micromeso.2012.07.009

  2. [2]

     陈佳炜, 赵茹, 周仁贤. (无机化学学报), 2018, 34(12):2135-2142 http://www.wjhxxb.cn/wjhxxbcn/ch/reader/view_abstract.aspx?flag=1&file_no=20181202&journal_id=wjhxxbcnCHEN Jia-Wei, ZHAO Ru, ZHOU Ren-Xian. Chinese J. Inorg. Chem., 2018, 34(12):2135-2142 http://www.wjhxxb.cn/wjhxxbcn/ch/reader/view_abstract.aspx?flag=1&file_no=20181202&journal_id=wjhxxbcn

  3. [3]

     Sun M T, Huang B C, Ma J W, et al. Acta Phys.-Chim. Sin., 2016, 32:1501-1510 doi: 10.3866/PKU.WHXB201603171

  4. [4]

     Qu L, Li C T, Zeng G M, et al. Chem. Eng. J., 2014, 242(15):76-85

  5. [5]

     Schmieg S J, Oh S H, Kim C H, et al. Catal. Today, 2012, 184(1):252-261 doi: 10.1016/j.cattod.2011.10.034

  6. [6]

     Liu Y, Gu T T, Weng X L, et al. J. Phys. Chem. C, 2012, 116(31):16582-16592 doi: 10.1021/jp304390e

  7. [7]

     Kosinov N, Liu C, Hensen E J M, et al. Chem. Mater., 2018, 30(10):3177-3198 doi: 10.1021/acs.chemmater.8b01311

  8. [8]

     Shi L, Yu T, Wang X Q, et al. Acta Phys. Chim. Sin., 2013, 29(7):1550-1557 doi: 10.3866/PKU.WHXB201304283

  9. [9]

     Yu R, Zhao Z C, Huang S J, et al. Appl. Catal. B, 2020, 269:8

  10. [10]

      Mihai O, Widyastuti C R, Andonova S, et al. J. Catal., 2014, 311:170-181 doi: 10.1016/j.jcat.2013.11.016

  11. [11]

      Urrutxua M, Pereda-Ayo B, Gonzalez-Velasco J R, et al. Catal. Today, 2016, 273:72-82 doi: 10.1016/j.cattod.2016.02.054

  12. [12]

      Mohan S, Dinesha P, Kumar S. Chem. Eng. J., 2019, 384:10 https://www.sciencedirect.com/science/article/abs/pii/S1385894719326658

  13. [13]

      Wilken N, Wijayanti K, Kamasamudram K, et al. Appl. Catal. B, 2012, 111:58-66 https://www.sciencedirect.com/science/article/abs/pii/S0926337311004450

  14. [14]

      Zhao Y, Choi B, Kim D. Chem. Eng. Sci., 2017, 164:258-269 doi: 10.1016/j.ces.2017.02.009

  15. [15]

      Wang A Y, Wang Y L, Walter E D, et al. Catal. Today, 2017, 320:91-99

  16. [16]

      Lin Q J, Liu J Y, Liu S, et al. Dalton Trans., 2018, 47(42):15038-15048 doi: 10.1039/C8DT03156H

  17. [17]

      Gao F, Washton N M, Wang Y L, et al. J. Catal., 2015, 331:25-38 doi: 10.1016/j.jcat.2015.08.004

  18. [18]

      Bonaccorsi L, Calabrese L, Alioto S, et al. Renewable Energy, 2017, 110:79-86 doi: 10.1016/j.renene.2016.09.030

  19. [19]

      Buchholz A, Wang W, Xu M, et al. Microporous Mesoporous Mater., 2002, 56(3):267-278 doi: 10.1016/S1387-1811(02)00491-2

  20. [20]

      Chen Z, Fan C, Pang L, et al. Appl. Catal. B, 2018, 237:116-127 doi: 10.1016/j.apcatb.2018.05.075

  21. [21]

      Zhao Z C, Yu R, Zhao R R, et al. Appl. Catal. B, 2017, 217:421-426 doi: 10.1016/j.apcatb.2017.06.013

  22. [22]

      Baran R, Averseng F, Wierzbicki D, et al. Appl. Catal. A, 2016, 523:332-342 doi: 10.1016/j.apcata.2016.06.008

  23. [23]

      Tang J J, Yu T, Xu M H, et al. Chem. Eng. Sci., 2017, 168:414-422 doi: 10.1016/j.ces.2017.04.053

  24. [24]

      Lin Q J, Feng X, Zhang H L, et al. J. Ind. Eng. Chem., 2018, 65:40-50 doi: 10.1016/j.jiec.2018.04.009

  25. [25]

      Zhang D, Yang R T. Appl. Catal. A, 2017, 543:247-256 doi: 10.1016/j.apcata.2017.06.021

  26. [26]

      Sazama P, Sadovska B, Dedecek J, et al. J. Catal., 2014, 318:22-33 doi: 10.1016/j.jcat.2014.06.024

  27. [27]

      Beale A M, Gao F, Lezcano-Gonzalez I, et al. Chem. Soc. Rev., 2016, 46(48):7371-7382

  28. [28]

      Li Y H, Song W Y, Liu J, et al. Chem. Eng. J., 2017, 330:926-935 doi: 10.1016/j.cej.2017.08.025

  29. [29]

      Liu J X, Liu J, Zhao Z, et al. Ind. Eng. Chem. Res., 2017, 56(20):5832-5841

  30. [30]

      Shan Y L, Sun Y, Du J P, et al. Appl. Catal. B, 2020, 275:118-129

  31. [31]

      Gao F, Kwak J H, Szanyi J, et al. Top. Catal., 2013, 56(15):1441-1459 doi: 10.1007/s11244-013-0145-8

  32. [32]

      Gao F, Walter E D, Washton N M, et al. Appl. Catal. B, 2015, 162:501-514 doi: 10.1016/j.apcatb.2014.07.029

  33. [33]

      Jouini H, Mejri I, Petitto C, et al. Microporous Mesoporous Mater., 2018, 260:217-226 doi: 10.1016/j.micromeso.2017.10.051

  34. [34]

      叶青, 闫立娜, 霍飞飞, et al. (无机化学学报), 2012, 28(1):103-112 http://www.wjhxxb.cn/wjhxxbcn/ch/reader/view_abstract.aspx?file_no=20120118&flag=1YE Qing, YAN Li-Na, HUO Fei-Fei, et al. Chinese J. Inorg. Chem., 2012, 28(1):103-112 http://www.wjhxxb.cn/wjhxxbcn/ch/reader/view_abstract.aspx?file_no=20120118&flag=1

  35. [35]

      Chen B H, Xu R N, Zhang R D, et al. Environ. Sci. Technol., 2014, 48(23):13909-13916 doi: 10.1021/es503707c

  36. [36]

      Wilken N, Nedyalkova R, Kamasamudram K, et al. Top. Catal., 2013, 56(1):317-322 doi: 10.1007/s11244-013-9973-9

  37. [37]

      Imyen T, Yigit N, Dittanet P, et al. Ind. Eng. Chem. Res., 2016, 55(51):13050-13061 doi: 10.1021/acs.iecr.6b03990

• 

  Figure 1  ²⁷Al NMR spectra of CuCe and CuCeSi catalysts before and after the hydrothermal treatment

  下载: 全尺寸图片 幻灯片
  

  Figure 2  XRD patterns of CuCe and CuCeSi catalysts before and after the hydrothermal treatment (a) and crystallinity loss ratio (b)

  下载: 全尺寸图片 幻灯片
  

  Figure 3  SEM images of CuCe and CuCeSi catalysts before and after the hydrothermal treatment

  下载: 全尺寸图片 幻灯片
  

  Figure 4  NH₃-TPD profiles of catalysts before (a) and after (b) the hydrothermal treatment

  下载: 全尺寸图片 幻灯片
  

  Figure 5  In situ DRIFTS results of NH₃ adsorption over CuCe-HT (a) and CuCeSi-HT (b); Comparison of NH₃ adsorption of catalysts under the hydrothermal treatment at 350 ℃ (c)

  下载: 全尺寸图片 幻灯片
  

  Figure 6  H₂-TPR spectra of fresh (a) and hydrothermal treated (b) CuCe and CuCeSi catalysts

  下载: 全尺寸图片 幻灯片
  

  Figure 7  Cu2p XPS spectra for the fresh (a) and hydrothermal treated (b) CuCe and CuCeSi catalysts

  下载: 全尺寸图片 幻灯片
  

  Figure 8  NO_x conversion as a function of the reaction temperature over Cu, CuCe and CuCeSi during NH₃-SCR

  Before (a) and after (b) the hydrothermal treatment, reaction conditions: 0.02% NO, 0.02% NH₃, 5% H₂O, 10% O₂ (volume fraction), GHSV=40 000 h^-1

  下载: 全尺寸图片 幻灯片

  Table 1.  Integrated area (A) of Al^Ⅳ atoms in CuCe and CuCeSi catalysts

  Sample	AAlⅣ	AAlⅤ	AAlⅥ	AAlⅣ/ (AAlⅣ+AAlⅤ+AAlⅥ) / %
  CuCe	922.7	94.2	32.0	88.0
  CuCeSi	945.6	5.1	70.5	92.6
  CuCe-HT	530.1	103.2	78.1	74.5
  CuCeSi-HT	648.9	35.3	52.3	88.1

  下载: 导出CSV

  Table 2.  Quantitative analysis of the H₂-TPR profiles of the fresh and hydrothermal treated CuCe and CuCeSi catalysts

  Sample	H2 consumption / (μmol·g-1)				Total H2 consumption / (μmol·g-1)
  	α (Cu2+ to Cu+)	β (CuO to Cu0)	δ (Cu+ to Cu0)	γ (SCu+ to Cu0)
  CuCe	77.0	62.7	173.2	70.3	383.2
  CuCeSi	61.5	66.5	56.3	153.2	337.5
  CuCe-HT	41.3	64.2	109.5	9.6	224.6
  CuCeSi-HT	50.4	54.2	36.4	67.4	208.5

  下载: 导出CSV

  Table 3.  Surface chemical composition and contents over catalysts before and after the hydrothermal treatment

  Sample	Mass fraction of surface Cu / %		Mass ratio of isolated Cu2+ to CuO
  	Isolated Cu2+	CuO
  CuCe	0.29	2.03	0.14
  CuCeSi	0.26	1.86	0.14
  CuCe-HT	0.64	2.32	0.28
  CuCeSi-HT	0.82	1.97	0.42

  下载: 导出CSV
•