Meng Yuan; Wang Haiyan; Sun Na; Yang Zhanxu; Wang Yujia
(School of Petrochemical Engineering, Liaoning Shihua University, Fushun, 113001)
Abstract: Silicoaluminophosphates SAPO-11 molecular sieves with small particle size and hierarchical pores were synthesized using the directing agent method. The effect of crystallization time on the particle structure, morphology, pore structure properties, and acid properties of SAPO-11 molecular sieves were investigated. Unlike the SAPO-11 molecular sieves synthesized with the conventional method, the results of XRD, SEM, BET and NH3-TPD analyses showed that the SAPO-11 molecular sieves synthesized by the directing agent method in a shorter crystallization time exhibited fine and uniform morphology. By increasing the crystallization time, the particle size of SAPO-11 molecular sieve was significantly reduced, and the mesoporous structure (intercrystalline pores) was formed. Furthermore, the external speci fic surface area and the total speci fic surface area reached 81.7 m2/g and 192.0 m2/g, respectively, which effectively reduced the pore mass transfer resistance and signi ficantly increased the number of acid sites. The results of n-dodecane hydroisomerization revealed that the Pt/SAPO-11 prepared with the novel method exhibited higher catalytic activity and better hydroisomerization selectivity than that synthesized by the conventional hydrothermal method. Thus, the small particle molecular sieve showed a promising industrial application prospect to be used as catalyst support.
Key words: directing agent method; small particle; SAPO-11; mesoporous structure; hydroisomerization
At present, as a silicoaluminophosphate (SAPO)molecular sieve, SAPO-11 with one dimensional tenmembered ring channel and AEL pore structure has been widely used in various re fining and chemical industries[1-4].SAPO-11 molecular sieve has obvious advantages and wide application prospects in the isomerization of alkanes,owing to its exchangeable cations, suitable acidity and channel structure, as well as excellent thermal and hydrothermal stability[5-7]. However, SAPO-11 molecular sieves synthesized by the traditional hydrothermal method ganerally have microporous structure and large particle size[8-9], which are unfavorable to the mass transfer of large organic molecules. For long chainn-alkanes, it is generally believed that the isomerization process occurs in the pore mouths of SAPO-11, which is known as the pore mouth and key-lock (PMKL) shape selectivity. This implies that only the outer layer of a SAPO-11 particle contributes to the catalytic activity[10-12]. Therefore, it is of signi ficant importance to overcome the inherent diffusion limitations and expose more available pore mouths[2,13-14].Recently, zeolites containing mesoporosity have been extensively studied for various acid-catalyzed reactions.Mesoporous molecular sieves are usually created by adding templates and surfactants[15], but the general disadvantages of the above method show that adding expensive template or active agent would significantly increase the preparation cost. Zhang, et al.[16]investigated the synthesis of cavitary SAPO-11 molecular sieves by using polystyrene (PS) micro-spheres as templates,which could greatly increase the speci fic surface area of the molecular sieves, leading to a decreased crystallinity.The preparation of mesoporous molecular sieves needs not only to create mesoporous structures inside single zeolite particles, but also to form intercrystalline pores by the accumulation of multiple particles[17]. Consequently,the study of the synthesis and crystallization of SAPO-11 with small particle size, large specific surface area,and more active centers has become a key research topic. Zhang, et al.[18]also synthesized small particle SAPO-11 molecular sieve with high crystallinity by means of a HF-containing system, but found that the proportion of strong acids increased, which could lead to cracking reactions. Zhu, et al.[19]studied the preparation of small particle SAPO-34 molecular sieves via seed crystallization method and colloidal seed crystallization method, respectively. The results showed that the size of the particles obtained by the seed method depends on the seed particles added, which has large limitations. Ren,et al.[20]synthesized the nanorod-assembled mordenite(MOR) by using directing agent, which is generally called the directing agent method. By addition of colloidal seed crystals (guiding liquid gel), the particle size could be reduced. The addition of the guiding gel can promote the formation of crystal nuclei and shorten the crystallization process, which can effectively avoid the formation of heterocrystals, and modify the particle size and acid properties of the molecular sieve[21-22].
By adding the guiding liquid (gel obtained by short-term crystallization at low temperature with the same raw material ratio) in the secondary synthesis process, this study adopted the directing agent method to prepare small particle molecular sieves with more intercrystalline pores.The external speci fic surface area and speci fic surface area of SAPO-11 are expected to be improved, so that more surface active sites could be exposed in catalytic reaction.The structure and morphology of the SAPO-11 molecular sieves were analyzed by XRD, SEM, BET, and NH3-TPD techniques. In addition, the catalytic performance of zeolites in the n-dodecane hydroisomerization process was investigated.
Phosphoric acid (85%), pseudo-boehmite (PB, 75%) and silica sol (30%) were purchased from the Sinopharm Chemicl Reagent Company, Ltd., di-n-propylamine(DPA, 98%), and diisopropylamine (DIPA, 98%) were obtained from Alladin, China. All chemicals were used as received without further purification. The SAPO-11 zeolites were synthesized by using the starting gel with a molar composition of 0.1 SiO2:1.0 P2O5:1.0 Al2O3:0.6 DPA:0.6 DIPA:50 H2O. Phosphoric acid and PB were mixed with deionized water and stirred vigorously for 2 h. Subsequently, silica sol was introduced into the mixture and was then stirred for 2 h, followed by addition of DPA and DIPA. The mixture was then transferred to an autoclave and heated successively at 190 °C for 4 h,8 h, 16 h, 24 h, and 48 h. After crystallization, the SAPO-11 composite zeolite products were washed, dried at 120 °C overnight, and then calcined at 600 °C for 12 h to completely remove the template. Finally, the products were named S4, S8, S16, S24, and S48 according to different crystallization duration.
The gel obtained in the above-mentioned conventional synthesis method was transferred into an autoclave and was subjected to ageing at 120 °C for 12 h to generate the guiding liquid. The steps of the conventional synthesis method were repeated, and after an equal amount of template agent was added into the aluminum phosphate aqueous solution, silica sol was introduced into the mixture and stirred vigorously, followed by adding dropwise a certain amount of guiding liquid.After being stirred for a period of time, the reactant gel was transferred to the autoclave and heated at 190 °C for 4 h, 8 h, 16 h, 24 h, and 48 h, respectively. After crystallization, the SAPO-11 composite zeolite products were washed, dried at 120 °C overnight, and then calcined at 600 °C for 12 h to completely remove the template.Finally, the products were named D4, D8, D16, D24, and D48 according to different crystallization duration.
The Pt/SAPO-11 catalysts were prepared by the incipient wet impregnation method with H2PtCl6solution serving as the metal precursor. The Pt content was kept at 0.5% for all samples[23]. The resulting catalysts were dried at 373 K and calcined at 823 K for 4 h. These catalysts called Pt/SAPO-11 were labeled as Pt/S4, Pt/S8, Pt/S16, Pt/S24, Pt/S48, Pt/D4, Pt/D8, Pt/D16, Pt/D24, and Pt/D48.
The phase composition was analyzed by a Bruker D8 advance X-ray powder diffractometer. The experimental conditions covered: a Cu Kα radiation (λ = 0.15418 nm)operated at 40 kV and 40 mA, a wide-angle scanning range of 2θ = 5°—40°, and a small-angle scanning range of 2θ = 0.5°—5°. The morphology of the samples was measured using a Hitachi high-tech SU8010 series ultra-high-resolution field emission scanning electron microscope SU8010. The pore structure of the samples was examined by the N2isothermal adsorptiondesorption on an Autosorb IQ2-MP gas sorption analyzer(Quantachrome Instruments). The total specific surface area (SBET) was calculated according to the multi-point Brunauer, Emmett and Teller (BET) method. The surface area (Sm) was obtained from the t-plot method, while the pore volume (Vpore) was measured by t-plot analysis of the adsorption isotherm. The temperature-programmed desorption of ammonia (NH3-TPD) experiments were performed on a Micromeritics ASAP2920 automated chemisorption instrument.
The hydroisomerization performance evaluation was performed on a continuous-flow fixed-bed microreactor. Before reaction, 10 mL (about 6 g) of the solid catalyst with a particle size of 20—40 mesh were put into a reaction tube, and were then subjected to reduction in a hydrogen atmosphere at 673 K and 2.0 MPa for 4 h. The reaction was completed after reduction. The reaction conditions covered: a LHSV of 1.0 h-1, a n(H2)/n(n-C12) ratio of 200, a reaction pressure of 2.0 MPa,and a reaction temperature of 613 K. The products were analyzed online using a 7820A gas chromatograph(Agilent) equipped with a HP-PONA capillary column(50 m×0.2 mm×0.5 μm) and a flame ionization detector.
3.1.1 XRD analysis
Figure 1(a) shows the XRD patterns of S4, S8, S16, S24,and S48, while Figure 1(b) shows the XRD patterns of D4,D8, D16, D24, and D48. Obviously, all diffraction peaks for SAPO-11 matched the simulated powder diffraction pattern for typical SAPO with AEL topology (2θ=8.06°,9.44°, 20.36°, 21.09°, 22.10°, 22.48°, 22.74°)[24-26].
The SAPO-11 molecular sieves synthesized by conventional hydrothermal method within a short crystallization time showed a distinct characteristic diffraction peak of SAPO-5 crystal at 2θ = 7.40°[27](S4 and S8 samples). However, the SAPO-11 molecular sieves synthesized by the directing agent method did not show peaks of heterocrystalline phase which was intrinsic to SAPO-5, even though the crystallization time was short. It is a great advantage of the directing agent method to change the crystallization direction of the original gel into a certain zeolite, which can inhibit the formation of heterocrystalline phases, indicating that the synthesized products are pure SAPO-11[28]. By comparing Figure 1(a)and Figure 1(b), it is found that the peak of sample D48 is sharp and high, indicating that the crystallinity of SAPO-11 molecular sieve synthesized by the directing agent method is significantly higher than that of SAPO-11 molecular sieve synthesized by the conventional hydrothermal method. Interestingly, we also found that broader Bragg peaks are presented for SAPO-11 samples synthesized by the directing agent method, which has further proved that the directing agent method can synthesize small particle SAPO-11 molecular sieves[10].
Figure 1 XRD patterns of S4, S8, S16, S24, S48,D4, D8,D16, D24, and D48 samples
3.1.2 SEM analysis
The SEM images in Figure 2 and Figure 3 show signi ficant differences in the morphology of the samples.It can be seen that with the extension of the crystallization time, the SAPO-11 samples prepared by the directing agent method showed rugby-like particles closely packed with thin rod-shaped prisms, which might be formed due to the preferential growth in the longitudinal direction during the synthesis of SAPO-11 molecular sieves. Figure 2 shows that the particle size of sample S4 is 5—6 μm,and that of sample D4 is about 2 μm. In Figure 3, the particle size of sample S48 is 8—10 μm, while that of sample D48 is 2—3 μm. Apparently, the particle size of SAPO-11 molecular sieves mixed by using the directing agent method is significantly reduced, which is mainly because there is a highly dispersed microcrystalline nucleus in the guiding gel which can be entered into the crystal growth period quickly, while a large number of silicate and aluminate ions are polymerized into the microcrystalline nuclei, so that the crystal would grow continuously without increase in the number, and therefore the SAPO-11 molecular sieve synthesized by the directing agent method shows a fine and uniform morphology[28]. By comparing the SEM images in Figure 2,it can be observed that the distance between single particles of the small particle SAPO-11 molecular sieves synthesized by the directing agent method was greatly shortened, but no cluster occurred, thereby forming lots of intercrystalline pore structures. In addition, the sample D24 was grown into denser small particles when the crystallization time reached 24 h, while the morphology of the sample D48 presented more dense spherical particles when the crystallization time reached 48 h, indicating that the addition of directing agent not only generated samples with smaller particle size, but also could greatly reduce the crystallization time of SAPO-11 molecular sieves.
Figure 2 SEM images of SAPO-11 zeolites synthesized by hydrothermal and directing agent methods at different crystallization times(A: D4; B: D8; C:D16; D:D24; a: S4; b: S8; c: S16; d: S24)
Figure 3 SEM images at low magni fications of SAPO-11 zeolites synthesized by hydrothermal and directing agent methods(E:D48; e:S48)
Figure 4 N2 adsorption-desorption isotherms (a) and pore size distribution (b) of S24, S48, D24, and D48
3.1.3 N2adsorption-desorption analysis
The N2adsorption-desorption isotherms and pore size distribution of S24, S48, D24, and D48 samples are presented in Figure 4 (a) and (b), respectively. It can be seen from Figure 4 that the type IV adsorption-desorption isotherms for typical hierarchical structure of material were observed for samples D24 and D48, while the type I adsorption-desorption isotherms for typical microporous structure of material were observed for samples S24 and S48. According to the pore size distribution of the SAPO-11 molecular sieves, the pore size of samples D24 and D48 was concentrated in the range of 2 nm-5 nm. A large number of intercrystalline pores was formed via the accumulation of small particles, resulting in hierarchical SAPO-11 molecular sieves.
The porosity and speci fic surface area of different SAPO-11 molecular sieve samples were characterized by the nitrogen adsorption measurements, as shown in Table 1. It can be seen that the total specific surface area and external specific surface area of the samples prepared by the directing agent method were significantly higher than those prepared by the conventional hydrothermal method. In particular, the external speci fic surface area of sample D24 (27.7 m2/g) was about 1.5 times greater than that of sample S24 (17.5 m2/g). Moreover, the external specific surface area of sample D48 (81.7 m2/g) was three times bigger than that of sample S48 (26.4 m2/g).This is because the particle size of the molecular sieves synthesized by the directing agent method is reduced due to the addition of the guiding fluid and the formation of lots of intercrystalline pore structure, which increased the speci fic surface area of the molecular sieves. Meanwhile,the number of unit cells on the small particle SAPO-11 molecular sieves was large, and many pores were exposed, which greatly improved the isomerization activity of the SAPO-11 molecular sieves.
Table 1 Comparison of pore structure parameters of different SAPO-11 molecular sieves
3.1.4 Acidity characterization
The acid properties of the SAPO-11 samples were investigated via the NH3-TPD technique, with the results displayed in Figure 5. The test results revealed that two desorption peaks occurred, viz.: a sharp peak at ca.200 ℃ and a shoulder peak at ca. 300 ℃, corresponding to the weak and medium acid sites, respectively. It can be observed that in the same range of composition, the amount and strength of total acids gradually increased with the extension of crystallization time, while the NH3-TPD desorption peak at the center of medium-strong acid showed a tendency of moving to the high temperature region.Whereas, within the same crystallization time, the acid content of the SAPO-11 sample synthesized by the directing agent method decreased, while the proportion of medium acids increased slightly. This is due to the addition of the guiding fluid in the system, which can cause a large number of microcrystalline nuclei in the guiding liquid growing on the surface of the particles and promoting Si into the framework of the molecular sieves, and therefore the Si/Al ratio is relatively high and the acid content is low, which can be confirmed by the skeletal Si/Al ratio of the four samples shown in Table 2. Different synthesis conditions would cause Si to enter the molecular sieve framework in different ways, which would result in acidic changes.
Table 2 Comparison of framework Si/Al ratio of SAPO-11 samples
SAPO-11 has the same AEL topology as AlPO4-11, which is formed by replacing the P and Al atoms in the AlPO4-11 skeleton with the Si atom isomorph. Generally, the Si atoms are substituted by two mechanisms: one mechanism is substitution of a Si atom for a P atom (SM2), which will generate a protonic acid center (Si(4Al)) and give rise to weak acid sites; the second mechanism is to replace two Si atoms for an Al atom and a P atom simultaneously (SM3),which often occurs simultaneously with SM2 substitution.SM2 substitution produces a weak acid site with a small amount of Si, while SM2 and SM3 mechanisms occur simultaneously with the increase of Si amount to form a Si island in the SAPO-11 skeleton and produce Si (nAl) (0 <n<4) environments, thus generating acid sites with a medium strength at the border of silicon domains. In the case of SAPO-11 synthesized by the directing agent method, the molecular sieves are epitaxially grown on the surface due to the existence of a large number of microcrystalline nuclei in the guiding liquid. Therefore, Si atoms are more likely to enter the framework, and SM2 and SM3 mechanisms can occur simultaneously, forming more medium strength acid centers. In comparison, when SAPO-11 is synthesized by conventional hydrothermal method, the silica sol particles move to the vicinity of the nucleus of SAPO-11 slowly, and the Si atoms enter the skeleton tardily, so SM2 mechanism occurs with more weak acid centers being generated.Based on the above-mentioned reasons, the addition of the guiding fluid will lead to a reduced acid amount of the SAPO-11 molecular sieve and an increased proportion of medium strength acid.
Figure 5 NH3-TPD pro files of various SAPO-11 samples—D24;—S24;—S48;—D48
In order to compare the effects of different crystallization duration on the catalytic performance of molecular sieves prepared by using the two different methods,hydroisomerization reaction was performed over the four catalysts Pt/S24, Pt/S48, Pt/D24, and Pt/D48 withn-dodecane used as a model compound under the conditions covering a reaction temperature of 613 K, a reaction pressure of 2 MPa, a space velocity of 1.0 h-1,and a hydrogen/oil volume ratio of 200:1. The results are shown in Figure 6 and Table 3.
Figure 6 Isomerization yield plotted against n-dodecane conversion■—Pt/S24; ●—Pt/S48; ▲—Pt/D24; ▼—Pt/D48
The isomer yields of the four catalysts are shown in Figure 6, indicating that Pt/D48 has the highest isomer yield, while Pt/D24 has an intermediate isomer yield as compared to those of the traditional zeolite catalysts.
As for samples Pt/D24 and Pt/D48, the conversion rate of then-dodecane during hydroisomerization recation was higher than that of samples Pt/S24 and Pt/S48. The selectivity of the isomerized products of samples Pt/D24 and Pt/D48 increased by about 20% as compared to that of samples Pt/S24 and Pt/S48. The selectivity of isomeric hydrocarbons also increased by about 10%, the product yield was significantly improved, and the selectivity of the cracked products was greatly reduced. These results suggested that the small particle Pt/SAPO-11 catalyst with intercrystalline pore structure has a reduced reactant mass transfer resistance and product residence time, which can prevent the ole fin intermediates and alkyln-carbon ions from experiencing excessively long diffusion through the pores of acidic sites during the reaction[29], leading to pore blockage coupled with the occurrence of secondary cracking reactions and the reduction of the isomerization products yield.The small particle Pt/SAPO-11 catalyst is favorable to the isomerization of alkanes,which can inhibit the occurrence of side reactions of cracking to a certain extent. On the other hand, changes in the acidity of the molecular sieve may have an effect on the selectivity of isomerization. However,SAPO-11 molecular sieves are relatively mild in acidity,since they mainly consist of weak acids and medium strong acids, and do not contain strong acid centers which are similar to the case of ZSM-5 zeolite. Therefore, the effect of acidity on its isomerization selectivity is not obvious.
Table 3 Comparison of catalytic performance of four SAPO-11 molecular sieves for hydroisomerization of n-C12
In summary, a novel synthesis strategy has been developed to prepare small particle SAPO-11 molecular sieve with intercrystalline pore structure. The SAPO-11 molecular sieve synthesized by the directing agent method has smaller particle size, higher crystallinity, larger external specific surface area and total speci fic surface area, stronger acidity,and more active sites than that synthesized by traditional hydrothermal method. Merits of such a preparation method include simplicity, low cost, versatility, and tunability.Importantly, the particle size and the external speci fic surface area, which are key structural factors affecting diffusion properties, can be controlled. Under the same reaction conditions as those of SAPO-11 molecular sieve synthesized by traditional hydrothermal method, the conversion ofn-dodecane during hydroisomerization over the small particle Pt/SAPO-11 catalyst could reach up to 95.42%, and the yield of isomeric hydrocarbons coud be equal to 65.13%,among which, the yield of multi-chain isomers was 29.36%.
Acknowledgements: We thank the National Natural Science Fund of China (2016-Z0030) and Natural Science Foundation of Liaoning Province (L2017 LQN008, L2019014).