Due to its small pore, moderate acidity and excellent thermal and hydrothermal stability, the chabazite-type SAPO-34 is the most promising catalyst for the conversion of methanol to light olefins (MTO), which is a flexible alternative way to get olefins from biomass, natural gas, shale gas and coal. Much progress have been made in the industrial MTO processes based on this catalyst in the past few years [1, 2].
However, rapid deactivation of the catalyst makes increasing its lifetime the biggest challenge. This has been attributed to the formation of bulky polycyclic aromatic compounds (coke precursor) [3, 4], which block the internal channels and hinder the further diffusion of reactants and products [1, 4, 5, 6, 7, 8]. The deactivation process is so fast that once some cages near the external surface of crystals have been reached by the coke precursors, the whole catalyst is deactivated instantly [9, 10, 11, 12, 13]. Nanosized crystals of SAPO-34 have been developed to increase the accessible cages near external surface [6, 11, 14, 15, 16, 17, 18, 19]. However, the total coke is increased with the crystal size decrease although the diffusion limitation is reduced [20, 21, 22, 23]. This means that the intrinsic acidity has to be considered as the coking of SAPO-34 cannot be suppressed just by diminishing the crystal size. On the other hand, AlPO4-34 is also deactivated quickly during the MTO process by losing the ability to form intermediates [24, 25]. Clearly, a suitable criterion for the acid sites and strength should be ascertained that will make SAPO-34 into an efficient catalyst.
In our experiments, phosphoric acid, and 35 wt% TEAOH, TEA, DEA, and MOR (Alfa Aesar) were all analytical grade. Pseudoboehmite (water loss at 600 °C being 34.75 wt%) and silica sol (4.5 mol/L) were obtained from Shandong Aluminum Plant. Typically, in the synthesis of SAPO-34 (TEA), 1.74 g aluminum isopropoxide was first dissolved in 10 mL distilled water, then 1.51 mL H3PO4 (85%) was added followed by 0.25 mL silica sol and 2 mL TEA respectively. The mixture was stirred until a uniform gel was obtained. The homogeneous gel with the molar ratio Al2O3:P2O5:0.1SiO2:3TEA:50H2O was kept stirred for 2 hours, and then the mixture was sealed in Teflon-lined stainless steel autoclaves and heated at 170 °C for 3 d.
Suitable single crystals of DEA-SAPO-34 and PIPZ-SAPO-34 (dimensions 0.20×0.18×0.12 and 0.24×0.18×0.16 mm3) were carefully selected and glued to a thin glass fiber with superglue adhesive. The intensity data by X-ray diffraction were collected on a Bruker SMART 1000 CCD diffractometer equipped with normal focus, 2.4-KW sealed tube X-ray source using monochromatic Mo Kα (λ = 0.71073 Å) radiation.
The structure was solved by direct methods and refined on F2 by full matrix least squares using the SHELXL97 program. Positional parameters for the Al, P, and F or O atoms were located by direct methods. Remaining non-hydrogen atoms were routinely located from the Fourier difference maps during the course of the refinement.
The base framework is AlPO4 with P and Al at the T positions being strictly alternating. An isolated Si atom in SAPO-34 can only occupy the P site. For the 5-Si island, only one Si atom occupies an Al site and the other four Si atoms can only occupy the four P sites. So in the structure of SAPO-34, the location of Si atoms can be distinguished.
29Si MAS NMR spectra were recorded at room temperature on a Varian Infinityplus-400 spectrometer fitted with a T3 probe. Spin speed was 5.5 kHz. The resonance frequencies observed were 79.5 MHz. Chemical shifts were recorded with respect to TMS for 29Si.
Protonic acid sites are formed from the Si-O-Al linkage. An Si(OAl)4 entity (named isolated Si) is the weakest acid site in SAPO-34. It results from a negatively charged lattice that is balanced initially by the trapped template ions since the organic template is the sole available cation source during the crystallization [26, 27, 28, 29, 30]. More than 30 templates have been used to direct the formation of SAPO-34. They can be classified into three types according to the charges of the cations trapped in the chabazite cage (CHA cage), namely, as R+ (tetraethyl-ammonium hydroxide, triethylamine), 2R+ (diethylamine, morpholine) [31, 32, 33, 34, 35, 36, 37, 38] and 2R2+ (piperazine), (Fig. 1, Table S1) [39, 40, 41, 42], which determine the number of isolated Si atoms in the lattice [4, 7, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52]. Insufficient Si means a deficiency of isolated Si atoms which lead to the partial crystallization of SAPO-34. When this is in excess, Si island will be formed in the lattice by Si-O-Si linkages with Si-O-Al at the edge of the islands to maintain the charge balance. These are the stronger acid sites in SAPO-34 [53, 54, 55, 56]. The density of acid sites depends on the positive charge of trapped template ions, while the strength of acid sites depends on the amount of added SiO2.
It is known that each unit cell of the CHA structure comprises three CHA cages with 36 T atoms, which means that there are 12 atoms per CHA cage (T denotes non-oxygen atoms) [30]. Each T atom in a CHA cage is shared by three adjacent cages. For the R+ type ion, when TEAOH or TEA was used as the template, there is only one isolated Si atom in the lattice of a CHA cage, and this gives Si/(Si+Al+P) = (1÷3)/12 = 1/36, Si/Al = 1/(3×6) (0.056). This means that a minimum ratio of SiO2/Al2O3 = 0.11 is needed to synthesize SAPO-34 with one isolated acid site in each CHA cage. In the case of DEA, MOR (2R+), and PIPZ (2R2+), the ratio of SiO2/Al2O3 is 0.22 and 0.44, respectively (Table S1).
When the Si atoms in the lattice exceed these limits, Si islands appear. A 5-Si island spans six adjacent CHA cages, which are the periodic unit that forms a whole 5-Si island and named a 6-cage group (Fig. 2(a)). The three negative charges of the 5-Si island are balanced by three of six caged R+ ions. Meanwhile, three isolated Si atoms are formed to match with the other three trapped R+ ions (Fig. 2(a)). This gives the lowest SiO2/Al2O3 ratio required to form periodic Si island in SAPO-34-R+ (0.45). In the range of the SiO2/Al2O3 ratio between 0.11 for forming isolated Si atoms and 0.45 for periodic Si islands, non-periodic Si islands appear sporadically. With the increase of lattice Si, two 5-Si islands are formed in an 8-cage group (Fig. 2(b)). Two types of 8-Si island are formed in the 7-cage group and 8-cage group, respectively (Fig. 2(c) and (d)). Si islands with more than eight Si atoms can also been found in 7, 9 and 10-cage groups (Fig. 2(e)-(g)).
Periodic Si islands in SAPO-34-2R+ and SAPO-34-2R2+ have a similar distribution of Si as SAPO-34-R+. However, more isolated Si atoms are formed in each CHA cage to match the increase of positive charge of the trapped 2R+ or 2R2+ ions. For example, in SAPO-34-2R+, trapped 2R+ ions of each CHA cage give two positive charges. A 5-Si island which spans six adjacent CHA cages has only three negative charges. So nine isolated Si atoms are formed to balance the charges of the trapped ions (Fig. 2(h)). Thus, the lowest SiO2/Al2O3 ratio to form periodic Si island in SAPO-34-2R+ and SAPO-34-2R2+ are 0.8 and 1.4, respectively (Table S1).
This means that isolated Si atoms are the basic form of the existence of Si in the SAPO-34 framework. Its number increases with the positive charges of trapped ions, regardless of the presence of Si islands or not. This result enables us to understand the puzzling 29Si MAS NMR spectra of SAPO-34 samples with a higher Si content. Whether in SAPO-34-TEAOH or SAPO-34-TEA in Refs. [38, 57, 58, 59, 60], in SAPO-34-MOR or SAPO-34-DEA in Refs. [4, 26, 39], or SAPO-34-PIPZ in the present work (Fig. S1), there is always a strong peak (δ = 92.6 ) ascribed to isolated Si (Si(4OAl)) accompanied by weak peaks at 96, 102 and 111, which are assigned to the chemical shift of Si(nAl) (n = 3-0) in islands. Therefore, in the formation of SAPO-34, with more positive charges on the template ions, there is a higher density of acid sites; with more Si content in the framework, there are stronger acid sites. In order to suppress coking of the MTO catalyst, SAPO-34 possessing the weakest strength and lowest density of acid sites would be a solution.
However, Si islands are more suitable than isolated Si atoms for stabilizing the structure [61, 62, 63, 64]. It is challenging to prepare SAPO-34 with a low SiO2/Al2O3 ratio in the presence of TEAOH or TEA, because their conformations are flexible and can easily direct the AFI structure, which is independent of SiO2 and would result in the quick nucleation of SAPO-5 or AlPO4-5 at very low SiO2 concentration [56]. It is well acknowledged that a faster reaction impacted largely by charging conditions. Thus, strategies for reducing the rate of the initial reactions (lowering the crystallization temperature from 200 to 165-170 °C or increasing the H2O/Al2O3 ratio from 50 to 100 or using two steps for the crystallization, first at 120 °C for 12 h, then at 200 °C for 36 h) have been used. They have shown the effects of suppressing the nucleation of AFI zeolites and produced SAPO-34 as the pure product with high crystallinity and crystal yield. SAPO-34 samples that only have isolated Si atoms (Fig. S2) have been prepared from the gel with Al2O3:P2O5: (0.10-0.45)SiO2:2R:(50-100)H2O at 165-200 °C for 48-120 h. The 29Si MAS NMR spectra of the samples confirmed the existence of only isolated Si in the SAPO-34 framework (Fig. S2, Table S5).
In summary, the charge of the trapped template ions in the CHA cages of SAPO-34 is the key that determines the distribution of isolated Si atoms and Si islands. The category of the trapped template ions (R+, 2R+ and 2R2+) determine the lowest concentration of SiO2 needed for forming isolated Si atoms or Si islands, namely the density of acid sites, while the amount of SiO2 determines the strength of the acid sites. In order to suppress the coking of MTO catalysts, SAPO-34 with weak strength and low density of the acid sites is a promising solution.
2016, Vol. 37 Issue (2): 227-233
PDF (792 KB)
2016, Vol. 37 Issue (2): 227-233
PDF (792 KB)