Automobiles and industrial combustion of fossil fuels produce harmful nitrogen oxides (NOx) [1]. The abatement of NOx released from diesel engines remains a challenge in the field of environmental catalysis. One of the most promising technologies for NOx emission control is the selective catalytic reduction (SCR) of NOx with NH3 [2].
Conventional three-way catalysts used in the exhaust after-treatment devices of internal combustion engines proved ineffective when the engine was operated under highly oxidizing conditions [3]. Copper-based zeolite catalysts, such as Cu-Beta [4, 5, 6], Cu-ZSM-5 [7, 8], Cu-SSZ-13 [9, 10], have excellent catalytic activity and are drawing interest from the auto industry as potential catalysts for the urea-SCR system. Phosphorus-containing silicoaluminophosphate crystalline chabazite (CHA) molecular sieves supporting a copper catalyst (Cu/SAPO-34) are another promising candidate for NH3-SCR of NOx. Post-synthesis cationic exchange-impregnation methods have traditionally been used to prepare Cu/SAPO-34 catalysts, which involve introducing copper species into small-pore zeolites. However, the small pores (~0.38 nm) can limit the distribution of Cu within the zeolitic crystals, resulting in the metal preferentially locating close to the external surface [11]. Martínez-Franco et al. [12] have reported a direct preparation of Cu-SAPO-34 zeolite using a low-cost copper-amine complex (Cu2+ with tetraethylenepentamine, Cu-TEPA) as the template and diethanolamine as a co-template. This methodology allowed the introduction of extra-framework copper species in the CHA cages, and Cu-SAPO-34 is promising for the SCR of NOx.
Peden’ group [13] reported that Cu-SAPO-34 catalysts are kinetically limited in the NH3-SCR of NOx at low reaction temperatures because their small pore sizes dramatically restricting reactant accessibility to the microporous channels of the catalytically active sites. To improve their catalytic performance, it is imperative to develop meso-microporous Cu-SAPO-34 catalysts that have less diffusion limitation for the NH3-SCR of NOx. Internal mesopores can be created among the microporous particles during the synthesis of meso-microporous Cu-SAPO-34 catalysts. Chemical reactions still take place in the micropores of meso-microporous catalysts, but the accessibility to the catalytic active sites is improved.
In this work, a series of meso-microporous Cu-SAPO-34 catalysts are prepared via a one-pot hydrothermal crystallization synthesis method. Cu-TEPA is used as the template and N,N′-diisopropylethylamine (DIPEA) as a cooperative organic structure-directing agent. The molar ratio of DIPEA to Cu-TEPA is varied and the Si/Al ratio is fixed. The one pot synthesis of Cu-SAPO-34 is used to prepare catalysts with high solid yields and controlled locations of the Si species and extra-framework copper ions. The prepared Cu-SAPO-34 catalysts possess excellent low-temperature NH3-SCR performance.
1.3 g tetraethylenepentamine (TEPA, 98 wt%, Aldrich, Beijing, China) was added to an aqueous solution containing 1.56 g copper (II) sulfate (98 wt.%, Alfa, Shanghai, China) and 6.24 g distilled water, and the mixture was stirred for 1 h to ensure complete dissolution. 60.36 g distilled water and 6.53 g phosphoric acid (85 wt%, Aldrich, Beijing, China) were added to the solution successively and stirred for 5 min to give a gel. A certain amount of DIPEA (99 wt%, Aldrich, Beijing, China) was introduced to the gel mixture and stirred for 4 h. 4.59 g pseudo-boehmite (75 wt%, Condea, Beijing, China) and 1.50 g silica gel (Ludox AS40 40 wt%, Aladdin, Beijing, China) were added to the gel and stirred for another 12 h. The final molar composition of the zeolite gels were: 1 Al2O3:0.85 P2O5:(5-40) DIPEA: 0.185 Cu-TEPA:1 SiO2:111 H2O.
The gel was reacted in a stainless-steel autoclave at 175 °C and under autogenous pressure for 6 days. The crystalline product was filtered, washed with deionized water and dried at 110 °C overnight. The organic template was removed by heating in air at 550 °C for 6 h. Four hierarchical Cu-SAPO-34 samples were prepared and are as H-Cu-SAPO-34-x, where, x represents the molar ratio of DIPEA/Cu-TEPA used in the synthesis of the gel. For comparison, conventional Cu/SAPO-34 catalyst was synthesized according to a previously published method [14].
The morphology of samples was observed by scanning electron microscopy (SEM) on an FEI Quanta 200F microscope (Oregon, USA) operating at 20 kV with a working distance of 11 mm. Transmission electron microscopy (TEM) images were obtained with an FEI Tecnai 20 (Oregon, USA) at 200 kV. The Brunauer-Emmett-Teller surface area and pore volume of the samples were measured by N2 adsorption-desorption on an ASAP 2010 (Georgia, USA) analyzer (Micromeritics). Powder X-ray diffraction (XRD) measurements were undertaken using Cu Kα radiation (λ = 0.15406 nm) on a Bruker D8 Advance diffractometer (Germany) in the scanning range of 5°-50° (2θ) to identify the phase and determine the relative crystallinity. 27Al magic angle spinning nuclear magnetic resonance (MAS NMR) was recorded using a standard Bruker MAS probehead at a spin rate of 20 kHz (Georgia, USA). Diffuse reflectance ultraviolet-visible spectra (UV-Vis DRS) were recorded with a Hitachi U-4100UV-Vis spectrophotometer (Tokyo, Japan) between 200-800 nm using dehydrated BaSO4 as the internal reference standard. Elemental analysis of the materials prepared was determined by a Varian 715-ES inductively coupled plasma-atomic emission spectrometer (ICP-AES) (California, USA). X-ray photoelectron spectroscopy (XPS) experiments were conducted on a commercial VG Multilab 2000 system (UK). Temperature-programmed reduction (TPR) experiments were performed in a custom-made apparatus (China University of Petroleum, Beijing). Prior to reduction, the samples (50 mg) were heated from room temperature to 600 °C under N2 flow rate of 30 mL/min and then cooled down to 100 °C with a purging N2 flow. The TPR process was carried out with 10% H2 -90% N2 at a flow rate of 30 mL/min and a temperature ramp rate of 10 °C/min up to 700 °C. Electron paramagnetic resonance (EPR) spectra were obtained using a Bruker EMX EPR spectrometer (Germany) in the microwave region of l-3 cm using a high-frequency modulation of magnetic field of 100 kHz, and magnetic field of up to 5000 G at -173 and 22 °C in a quartz glass ampoule with an internal diameter of 3 mm.
SCR activity tests of the catalysts were performed using a fixed-bed quartz flow reactor at atmospheric pressure. The gas composition was as follows: 0.1% NO, 0.1% NH3, 3% O2, with N2 balance. The gas flow rate was 500 mL/min. Different gas hourly space velocities (GHSV) were obtained by changing the amount of catalyst (20-400 mg, 40-60 mesh). The effluent gases, including NO, O2, NH3 and N2, were analyzed by a NEXUS 670-FTIR spectrometer (Wisconsin, US) equipped with a multiple-path gas cell (2 m). NO conversion was calculated according to the following equation:
Turnover frequencies (TOF) were evaluated by dividing the moles of NO molecules converted per second by the molars of Cu atoms in the catalysts.
A series of H-Cu-SAPO-34 catalysts were prepared using varying amounts of DIPEA in the gel synthesis. Conventional Cu/SAPO-34 catalyst was synthesized for comparison. SEM was used to observe the morphology of the Cu/SAPO-34 and H-Cu-SAPO-34 samples (Fig. 1). The H-Cu-SAPO-34-5 sample consists of irregular particles with different diameters (Fig. 1(b)). Increasing the molar ratio of DIPEA to Cu-TEPA resulted in large spherical particles with diameters between 20 and 30 µm for H-Cu-SAPO-34-10 (Fig. 1(c)). Higher magnification images (not shown here) revealed that the large particles are made from the intergrowth of small cubic-like zeolite crystals with sizes in the range of 3-6 µm. Further increasing the molar ratio of DIPEA to Cu-TEPA from 10 to 40 during gel synthesis decreased the sizes of the agglomerated crystals to about 10 µm with well-defined morphology (Fig. 1(d) and (e)). This demonstrates an improvement in the level of crystallinity. The morphology of the H-Cu-SAPO-34 catalysts is notably different from the conventional microporous Cu/SAPO-34 particles, which have a smooth crystal surface and a size of approximately 5 µm (Fig. 1(a)).
A selection of representative TEM images of the H-Cu-SAPO-34-20 catalyst is given in Fig. 2. The mesopores with different sizes are distributed homogeneously on the surface of the H-Cu-SAPO-34-20 crystals [15, 16], and CuOx nanoparticles are present on the surface of the catalyst. Furthermore, Cu ions do not migrate and aggregate inside the exchange sites to form CuOx particles.
Fig. 3 displays the N2 adsorption-desorption isotherms and the BJH pore size distributions of the Cu/SAPO-34 and H-Cu-SAPO-34 catalysts. Conventional microporous Cu/SAPO-34 (Fig. 3(a-1)) shows type-I N2 adsorption- desorption isotherms. The H-Cu-SAPO-34 catalysts showed the representative characteristics of type-IV adsorption-desorption, with a sharp inflection at a relative pressure of 0.85< p/p0 < 1.0 and a well-defined hysteresis loop, indicating well-formed mesoporous structural frameworks [17]. Furthermore, the isotherms of H-Cu-SAPO-34 samples exhibited a clear, steep increase in N2 adsorption volume in the region of p/p0 < 0.05. This increase suggests the presence of microporosity in the four H-Cu-SAPO-34 samples. The pore size distribution of H-Cu-SAPO-34 samples (Fig. 3(b)) illustrates the existence of a mesopore structure with pore sizes of approximately 25 nm. Combined with the SEM and TEM analyses, the N2 adsorption-desorption results strongly suggest that a full meso-microporous structure exists in the four H-Cu-SAPO-34 samples (Table 1 and Fig. 3(b)).
Importantly, changes in the preparation conditions leads to a systematic change of the textural properties of H-Cu-SAPO-34 samples (Table 1). Increasing the molar ratio of DIPEA to Cu-TEPA for the gel synthesis increased the micropore surface area markedly from 180 m2/g for H-Cu-SAPO-34-5 to 370 m2/g for H-Cu-SAPO-34-20 and then resulted in a slight decrease to 361 m2/g for H-Cu-SAPO-34-40.
The XRD patterns of Cu/SAPO-34 and H-Cu-SAPO-34 catalysts are displayed in Fig. 4. Characteristic diffraction peaks belonging to a CHA-type structure, which are in good agreement with those of SAPO-34, were present for all of the crystalline samples [18]. An additional characteristic peak attributed to SiO2 was present in the pattern obtained for H-Cu-SAPO-34-5. The intensity of the peaks ascribed to the CHA-type structure increased slightly relative to the molar ratio of DIPEA to Cu-TEPA. There were no peaks related to CuOx or CuAl2O4 species in the catalysts [19, 20]. Combined with the TEM results, this demonstrates that most of the Cu species are well dispersed on theH-Cu-SAPO-34 catalysts.
27Al MAS-NMR spectroscopy was used to study the coordination state of the aluminum in the Cu/SAPO-34 and H-Cu-SAPO-34 catalysts. The three main peaks centered at δ =38, 10.3 and −11.5 are assigned to tetrahedrally, pentahedrally and octahedrally coordinated Al, respectively (Fig. 5) [21, 22, 23]. A weak shoulder peak at δ = ~67 is likely the result of Al with distorted tetrahedral coordination [13]. The formation of pentahedrally and octahedrally coordinated Al features is caused by reversible SAPO-34 hydrolysis and these features are lost upon dehydration [8, 24].
UV-Vis DRS was performed to better understand the local environment and nuclearity of the Cu species in the Cu/SAPO-34 and H-Cu-SAPO-34 catalysts. There are two main characteristic absorption bands for all of the samples (Fig. 6); the absorption band centered at ~230 nm is attributed to charge transfer lattice O2−→ Cu2+ [6, 25, 26], and the other centered at 750 nm is ascribed to d-d Cu2+ transition in disperse CuO particles [27]. The shoulder peak adsorbed at 650 nm is related to bulk CuO [28], which was also detected by TEM. Liu et al. [26] attributed a peak at 450 nm to dimeric copper ((Cu-O-Cu)2+) species, which result from the dehydration of hydroxyl-bridged dimers. Gang et al. [29] suggested that CuAl2O4 species display absorption bands at 740 nm. These bands were not detected in our results, suggesting that CuAl2O4 species are not present, which is consistent with the XRD results. UV-Vis DRS shows the presence of highly dispersed CuO clusters and bulk CuO, which were not detected by XRD because of their small size.
XPS and ICP experiments were carried out to characterize the atomic distribution and composition of Cu/SAPO-34 and H-Cu-SAPO-34 catalysts (Table 2). All of the Cu-SAPO-34 samples possess similar element contents (P, Si, Al and Cu). Importantly, the surface Cu/Al ratio is much lower than that in the bulk, indicating that the Cu ions are predominantly at the exchanged sites inside the SAPO-34 pores. Increasing the molar ratio of DIPEA to Cu-TEPA used in the gel synthesis gradually decreases, then dramatically increases, the Si/Al molar ratio on the surface of the H-Cu-SAPO-34 catalysts. Thus, the ratio of DIPEA to Cu-TEPA can be used to effectively control the location of Si species and the performance of H-Cu-SAPO-34 catalysts.
The binding energy of the Cu 2p transition peaks provides information on the state of Cu species. The binding energies of the catalysts determined by XPS are shown in Fig. 7. Cu species from CuO result in two shake-up satellite peaks centered at 942.2 and 962.5 eV. The main Cu 2p3/2 transition peak is located at 932.6 (with a shoulder peak at 935) eV and the Cu 2p1/2 transition peak is at 952.8 eV [30]. These satellites are not observed for Cu+ and Cu0 species and this feature can be used to distinguish between Cu2+ and Cu+ or Cu0 species [31]. The shakeup satellites observed for the Cu/SAPO-34 and H-Cu-SAPO-34 samples are attributed to the existence of CuO on the surface of the catalysts. The peak attributed to Cu2+ for the H-Cu-SAPO-34-20 catalyst is bigger than that for the other H-Cu-SAPO-34 samples, which may contribute to the improved NH3-SCR performance of this catalyst (see below).
H2-TPR was used to investigate the reducibility of Cu species and to identify their locations and chemical states. The H2-TPR curves of the catalysts are displayed in Fig. 8. H-Cu-SAPO-34-5 has three H2 consumption peaks in the range of 100-700 °C. The peak centered at 230 °C is assigned to the reduction of isolated Cu2+ ions (Cu2+→Cu+) [13, 32]. The peak located at 303 °C is ascribed to the reduction of highly dispersed nanosized CuO (Cu2+→Cu+ and Cu+→Cu0) [33, 34]. Finally, the peak at ~520 °C is a result of the reduction of Cu+ to Cu0 [32]. Increasing the molar ratio of DIPEA to Cu-TEPA shifted the reduction peaks assigned to isolated Cu species from 230 to 223 °C and then to 228 °C. Increasing the DIPEA/Cu-TEPA ratio from 5 to 20 increased H2 consumption by isolated Cu species, with a further increase to 40 resulting in decreased H2 consumption (Table 3). The maximum H2 consumption of H-Cu-SAPO-34-20 can be explained by this sample having the greatest proportion of Cu2+, as shown by the XPS results. The differences in the reduction ability of the H-Cu-SAPO-34 catalysts may affect their NH3-SCR performance. The H2/Cu values of H-Cu-SAPO-34-5 and H-Cu-SAPO-34-10 are much lower than those of H-Cu-SAPO-34-20 and H-Cu-SAPO-34-40 (Table 3), indicating that the former samples have a higher content of Cu+ than the latter ones. There are three types of Cu species in all the samples: highly dispersed CuO clusters both internally and on the external surface, and isolated Cu2+ and Cu+ ions located at the exchange sites. The isolated Cu2+ and Cu+ ions are the active centers for NH3-SCR. The lower temperature required for reduction and greater number of active centers of H-Cu-SAPO-34-20 improved its NH3-SCR performance. However, the high content of CuO in this catalyst may result in poor high-temperature reduction performance.
Only isolated Cu2+ ions (3d9, 2d5/2) are EPR active; all other Cu species ([Cu(II)-O-Cu(II)]2+ or Cu+ ions) are EPR silent [35]. Therefore, EPR is an excellent technique to qualitatively and quantitatively identify the amount of isolated Cu2+ ions and probe the structure and coordination environment of paramagnetic isolated Cu2+ ions in the catalysts. The framework of chabazite is formed from units of two six-membered rings of aluminum-oxygen and silicon-oxygen tetrahedra, bound face to face to form a hexagonal prism. Eight-membered rings form at the pore openings on the particle surface. There are four types of cationic sites in the chabazite framework: Site I is displaced from a six-membered-ring into the ellipsoidal cavity; Site II is located near the face of the ellipsoidal cavity; Site III is located in the center of the hexagonal prism; and Site IV is located near the eight-membered-ring at the pore [36]. Fig. 9 shows the EPR spectra of Cu/SAPO-34 and the H-Cu-SAPO-34 samples. Xiao’s group [19] reported an EPR signal of g// = 2.378 with A = 140 G for Cu2+ in site I for a hydrated sample. Similar EPR signals, such as g// = 2.39 with A = 111 G [13] and g// = 2.394 with A = 131 G [36], for Cu-CHA catalysts with site I Cu2+ species have been reported. The Cu2+ species of the SAPO-34 samples prepared show the same coordination environment, with g// = 2.384 and A = 132 G, indicating an identical location for the Cu2+ irrespective of the DIPEA/Cu-TEPA ratio. This result, combined with the results of UV-Vis DRS and H2-TPR, indicates that the isolated Cu2+ ions are located in site I. The total isolated Cu2+ content was quantified by the signal area integrations and the H-Cu-SAPO-34-20 sample possesses the highest content of isolated Cu2+ ions (Table 3). These results are consistent with the findings of the H2-TPR and XPS analyses and NH3-SCR performance.
NO conversion as a function of reaction temperature between 100 and 550 °C was measured for conventional Cu/SAPO-34 and H-Cu-SAPO-34 catalysts (Fig. 10). The low-temperature (100-250 °C) NO conversion increased from H-Cu-SAPO-34-5 to a maximum at H-Cu-SAPO-34-20 and then decreased for H-Cu-SAPO-34-40. The H-Cu-SAPO-34-20 catalyst exhibited the best performance because this sample has the highest isolated Cu2+ content (XPS and H2-TPR) and more readily reduced active Cu species. The conventional Cu/SAPO-34 catalyst exhibited the lowest low-temperature performance. Consequently, we conclude that the mesopores in the H-Cu-SAPO-34 catalysts facilitate physical transport of NO and improve accessibility of reactants to catalytic active sites.
NH3-SCR reactions were carried out with various GHSVs over the H-Cu-SAPO-34-20 sample (Fig. 11(a)). Increasing the GHSV from 50000 to 400000 h−1 resulted in a decrease in NH3-SCR performance at low temperatures. High-temperature SCR performance was unaffected by varying the GHSV. SCR reactions were conducted to establish kinetically benign (i.e., mass-transfer limitation free) regimes. Fig. 11(b) presents NO conversion ratios compared with that at GHSV = 800000 h−1 as a function of reaction temperature using data from the GHSV experiments with H-Cu-SAPO-34-20. NO conversion is expected to increase n-fold as the space velocity decreases n-fold in the absence of interparticle mass transfer limitations. We found that interparticle mass transfer was more than 99% free of limitation using conditions of GHSV ≥ 400000 h−1 and temperature ≤ 200 °C. (NO conversions ≤ 15%; data points below the dashed horizontal line in Fig. 12).
The apparent activation energy for NO oxidation over H-Cu-SAPO-34 catalysts (Fig. 13) was calculated using reaction data (Fig. 12). The linearity indicates the NO oxidation at 100-200 °C is first order with respect to NO over Cu-SAPO-34; like other SCR catalysts. The rate constants estimated from these plots show apparent activation energy of approximately 98 kJ/mol and similar Ea values for NO conversion for all four H-Cu-SAPO-34 samples, suggesting an identical reaction rate-limiting mechanism. This rate is somewhat higher than those reported previously for conventional CHA-based SCR catalysts (e.g., 62-89 kJ/mol) [13, 32, 37], indicating that the formation of mesopores in H-Cu-SAPO-34 catalysts lowers the pore diffusion limitations and improves the accessibility of reactants to the catalytically active sites. The formation of mesopores greatly improves NH3-SCR performance [38].
A series of H-Cu-SAPO-34 catalysts with full mesopores have been synthesized with varying ratios of DIPEA to organometallic Cu complex in the synthesis gel. The mesopores in H-Cu-SAPO-34 catalysts remarkably lower the diffusion limitation of reactant molecules. These catalysts exhibit excellent catalytic NH3-SCR reaction performance at low temperature. The best catalytic performance was obtained over the H-Cu-SAPO-34-20 sample, for which the ratio of DIPEA to Cu-TEPA was 20, and NO conversion reached 94% at 150 °C. The main active sites for the NH3-SCR reaction are the isolated Cu2+ ions in site I, displaced from the six-membered ring into the ellipsoidal cavity.
2016, Vol. 37 Issue (5): 750-759
PDF (1207 KB)
2016, Vol. 37 Issue (5): 750-759
PDF (1207 KB)