3.1 The crystal structure of synthesized SSZ-13
Primarily, a series of processing parameters, including the silica/alumina ratio, alkali/silicon ratio, water/silicon ratio, amount of seed, aging time, and crystallization temperature, were systematically optimized (as shown in supporting information Fig. S1). The key factor, namely the ratio of n(CC)/n(SiO2), impacting the catalyst feature was principally focused. The XRD patterns of synthesized materials with different ratio of m(CC)/m(SiO2) were presented in Fig. 2. Without the addition of CC template, the product was ANA-zeolite in pure phase (Fig. 2a). With the addition of CC template agent, the characteristic peak of ANA at 2θ = 15.9° gradually disappeared. The secondary structural unit of SSZ-13 was gradually formed in the mixed gel. As the m(CC)/m(SiO2) = 0.14, pure SSZ-13 phase generated with a relative crystallinity of 101.2% (Fig. 2d, Table S1). It could obviously prove that the CC template greatly accelerated the formation of SSZ-13 structure. In addition, the excessively increasing the amount of CC template agent did not show apparent influence on the crystal structure of SSZ-13. Therefore, the m(CC)/m(SiO2) ratio of 0.14 was the optimal condition to obtain a fully crystallized SSZ-13 molecular sieve with low cost (taking the commercially available SSZ-13 as a reference). The optimized condition details for the synthesis of SSZ-13 using CC template were described in supporting information and Fig.S1.
3.2 Morphology of catalysts
Figure 3 shows the SEM images of prepared SSZ-13 catalysts with different ratios of the m(CC)/m(SiO2) = 0.07, 0.14 and 0.32, respectively. The sample prepared by m(CC)/m(SiO2) = 0.07 was irregular particles dominated by cubic shape and mixed with many different size crystals (Fig. 3a). It was thought that with lower addition of template agent, it was hard to achieve completely crystallization of amorphous gel into regular SSZ-13 molecular sieves. The particles were severely aggregated rather than homogeneously dispersive. The serious aggregation was ascribed to the lack of specification template agent and slow nucleation rate, as well as the insufficient crystal nucleus formation during the hydrothermal synthesis. With the m(CC)/m(SiO2) ratio increased to 0.14, the irregular particles gradually disappeared, as shown in Fig. 3b. Even though the particles demonstrated slightly aggregation, the cubic crystals were relatively homogeneously distributed. The appropriate amount of template could meet the condition for rapid and ordered growth of catalyst crystals. Compared with synthesized sample with m(CC)/m(SiO2) = 0.14, it could be clearly seen that the purchased commercial SSZ-13 sample was still in serious agglomeration form (Fig. 3d). It may indicate more excellent performance of as-prepared SSZ-13 using choline chloride template. With the m(CC)/m(SiO2) ratio increased to 0.32, the particles became more terrible agglomeration (Fig. 3c), and the crystal morphology changed to be ambiguous. It was considered that the excessive amount of choline chloride template might promote the growth of larger size crystal, and consequently induce the locally sudden agglomeration. It was consistency with the XRD analysis that the m(CC)/m(SiO2) = 01.4 was the optimal condition to obtain relative uniform and cubic particles without undesired agglomeration. And besides, the synthesis time was shortened to within 4 h, highly improved the preparing efficiency.
3.3 Variation of microstructure
The FTIR spectra of the SSZ-13 prepared with different m(CC)/m(SiO2) ratios are shown in Fig. 4. Roughly, the peak vibrations of the synthesized samples were similar. The typical absorption peaks at 460, 526, 647, 777 and 1069 cm− 1 were assigned to the SSZ-13 molecular sieve. The asymmetric stretching of Si-O-Si or Al-O-Si bond was a wide and strong peak near 1069 cm− 1, which was classified to be T-O-T vibration. Likewise, the peak at 777 cm− 1 was the symmetrical vibration of Al-O. The peak vibration of a single six-member ring was at around 526 cm− 1. And at ca. 647 cm− 1 was the peak vibration of double six-member ring. The strong absorption peak near 460 cm− 1 was attributed to the bending vibration of Si-O in silicon-oxygen tetrahedron. The peaks at 1631, 3450 and 3615 cm− 1 were the stretching of -OH. Two weak peaks at 2924 and 2849 cm− 1 were ascribed to the stretching of -CH2- derived from the residual organic template.
From Fig. 4, it can be clearly seen that the absorption peak at 1069 cm− 1 became narrower with the m(CC)/m(SiO2) ratio increase, indicating the gradually ordered and regular of the SSZ-13 crystal. The FTIR is extremely sensitive to the change of silica-alumina microstructure. With the addition of chorine chloride, absorption peaks at 647 and 526 cm− 1 gradually appeared revealing the formation of bi-hexagonal ring unit in SSZ-13 framework. The formation of bi-hexagonal ring structure proved the template facilitating the crystallization process; likely, the increase of the absorption peak intensity suggested the crystallization getting gradually completed. Obviously, the catalyst prepared with m(CC)/m(SiO2) = 0.14 showed the sharpest peak at 1069 cm− 1 (Fig. 4c) implying the relative ordered Si-O-Si or Al-O-Si structure without extra distortion. It is consistent with the results discussed in XRD and SEM analyses.
3.4 the pore structure of as-prepared SSZ-13
The reactions on molecular sieve are very sensitive to the framework structure. The adsorption-desorption isotherms of synthesized samples with different m(CC)/m(SiO2) are shown in Fig. 5. And their surface area and pore volume are listed in Table 1. The results suggested that the synthesized SSZ-13 samples possessed Langmuir adsorption-desorption isothermal curves. The SSZ-13 samples had homogenously distributed microporous channels. In the low-pressure region (10− 6 < P/P0 < 0.01), an obvious spike was observed, which was induced by the adsorption of nitrogen molecules in the micropores. The adsorption plateau appeared at P/P0 > 0.1, approaching to adsorption saturation. At P/P0 > 0.95, the adsorption isotherm gradually increased and arrived the highest point due to capillary condensation. At the relative pressure P/P0 = 0.45, desorption curve is closed to form the hysteretic ring, indicating the synthesized zeolite contained stacking mesopores formed by the accumulation of small crystal particles. Such small particles could be also found in the SEM images (Fig. 3). Particularly, the catalyst prepared with m(CC)/m(SiO2) = 0.14 showed more significant mesopore structure with an obvious hysteretic ring (Fig. 5b). As can be seen from Table 1, the total specific surface area of samples prepared at m(CC)/m(SiO2) = 0.07, 0.14 and 0.32 were 419.177 m2.g− 1, 641.706 m2.g− 1, and 583.623 m2.g− 1, respectively; and the mesopore volume were 0.0162, 0.0377, and 0.0250 cm3.g− 1, respectively. Undoubtedly, the proper amount of template agent with m(CC)/m(SiO2) = 0.14 achieved the largest specific surface area and highest meso-pore volume.
Table 1
Pore structure characteristics of as-prepared zeolites
Sample | SBET(m2.g− 1) | Smic(m2.g− 1) | Sext(m2.g− 1) | Vmic(cm3.g− 1) | Vmeso(cm3.g− 1) |
m(CC)/m(SiO2) = 0.07 | 419.177 | 410.642 | 8.535 | 0.15699 | 0.0162 |
m(CC)/m(SiO2) = 0.14 | 641.706 | 636.058 | 5.648 | 0.23980 | 0.0377 |
m(CC)/m(SiO2) = 0.32 | 583.623 | 569.036 | 14.586 | 0.20884 | 0.0250 |
3.5 the acidity of the SSZ-13
The ammonia temperature-programmed desorption method (NH3-TPD) could reveal the acid properties of the as-prepared SSZ-13. The NH3-TPD of the SSZ-13 catalysts prepared with different m(CC)/m(SiO2) and purchased SSZ-13 were presented in Fig. 6. There were two typical desorption peaks found in the samples. The peak near 200 oC was the low temperature desorption region, which was attributed to the interaction between acid level and ammonia gas. NH3 could be desorbed from these weakly acidic surface hydroxyl groups, forming the weak acid potential of SSZ-13. The peak in the range of 490 ~ 540 oC was the high temperature desorption peak from the bridging hydroxyl group (≡ Si-OH-Al≡), corresponding to the strong acid center. The desorption peaks in the low temperature region were stronger, suggesting the weak acid sites were dominant. With the increase of m(CC)/m(SiO2), the strong acid potential tended to shift to high temperature region. The corresponding peak area increased greatly at m(CC)/m(SiO2) = 0.14. From Table 2, the weak and strong acidity of m(CC)/m(SiO2) = 0.14 were 0.672 and 0.553 mmol/g, respectively. In accordance with the SEM images, at m(CC)/m(SiO2) = 0.14, the particle size distribution became more uniform and the specific surface area increased significantly, indicating a larger amount of acid sites on the surface. The optimized strong and weak acid sites was obtained by regulated the m(CC)/m(SiO2), which would be conducive to the catalytic performance of such catalyst under relatively low temperatures.
Table 2
Acidity of as-prepared SSZ-13 measured by NH3-TPD
Samples | Weak acid site | Strong acid site |
Temperature (oC) | Acidity (mmol/g) | Temperature (oC) | Acidity (mmol/g) |
m(CC)/m(SiO2) = 0.07 | 200 | 0.424 | 505 | 0.507 |
m(CC)/m(SiO2) = 0.14 | 210 | 0.672 | 536 | 0.553 |
m(CC)/m(SiO2) = 0.32 | 201 | 0.450 | 511 | 0.801 |
SSZ-13 | 203 | 0.354 | 510 | 0.558 |
3.6 Catalytic performance for MTO
Based on the optimal conditions described, the catalytic activity of the synthesized SSZ-13 material for MTO reaction was evaluated. As a typical acid catalytic reaction, the SSZ-13 would demonstrate the role of bi-functional catalyst. The catalytic performance of purchased (using the TMAdaOH as template) and synthesized (using the CC as template) SSZ-13 for MTO was shown in Figs. 7 and 8. The products were mainly C1 ~ C3 alkanes and alkenes. The ethylene and propylene were the dominant compounds, and their yield reached the highest at 60 ~ 90 min. The yields were maintained over 35% during the test. The main by-products were CH4, C2H6, and C3H8. In the early stage (before 40 min) of the MTO reaction, the methanol was firstly decarbonylated for the formation of intermediates, such as acetate, formate, methyl acetate, and dimethoxymethane; while the zeolite was carbonylated during the induction period. The yield of by-products was continuously decreased during the reaction period indicating the improvement of reaction selectivity and deactivation of catalyst.
Both the SSZ-13 catalysts maintained a high methanol conversion rate over 95% within a reaction time of 120 min, and it would drastically drop down to ca. 80% in 80 min. Eventually, the conversion rate would maintain at ca 75% in a long reaction period (Fig. 7). The synthesized SSZ-13 showed a better conversion rate than that of purchased one. The catalytic lifetime of the molecular sieve was defined as duration that the methanol conversion was kept above 90%, i.e. the time from the beginning of the reaction to the methanol conversion dropping to 90%. From the NH3-TPD analyses, the weak acidity of synthesized and purchased SSZ-13 were 0.354 and 0.450 mmol/g, respectively; and the strong acidity were 0.558 and 0.721 mmol/g, respectively. Consequently, the as-prepared and purchased samples exhibited obvious differences in the catalytic lifetime. Their catalyst lifetimes were 172 and 160 min, respectively. The reaction intermediates, such as methyl benzene, would be gradually converted into polycyclic aromatic hydrocarbons, which thus might accumulate on the internal surface and outside the cage of SSZ-13, leading to congestion of channel and lower the mass transfer rate. Eventually, it would become difficult for the reactants to contact with the active center resulting in catalyst deactivation. It was considered that the appropriate acidity and well-developed pore structure of synthesized SSZ-13 could slow down the carbon deposition rate and significantly increase the lifetime of the catalyst. Primarily, the formation of dimethyl ether initiated on the weak acid center, and sequentially the conversion of dimethyl ether to light olefins would occur on the strong acid center (Cheng 2016).
Both the two SSZ-13 catalysts showed induction period where the selectivity was quite low in the beginning 40 min, as shown in Fig. 8. The light olefin selectivity reached the highest at the reaction time of ca. 60 min. The methanol adsorbed at the acid site was the first step of MTO conversion; and the deprotonation of carbenium ions and shift of methyl groups would be the followed reaction. The framework structure of SSZ-13 could accelerate the deprotonation of carbenium ions and convert the methyl group into the alkene (Li 2020). Such induction reaction could be occurred under low reaction temperature (i.e., < 450 oC). The active hydrocarbon species would form inside the zeolite at the initial stage. Theoretically, the synthesized SSZ-13 with higher acidity would possess strong adsorption capacity. The light olefine selectivity of as-prepared SSZ-13 was slightly lower than that of purchased SSZ-13. It was thus thought that the excessive micropore hindered the timely desorption and diffusion of intermediate products, and consequently reduced the light olefin selectivity. Although the light olefin selectivity of synthesized SSZ-13 was slightly lower than that of purchased SSZ-13, the initial selectivity could maintain above 50% within 160 min with the conversion rate over 90%. The larger amount of weak acid sites and relative smaller amount of strong acid sites on the as-prepared SSZ-13 using choline chloride as template were beneficial for the low temperature catalytic performance and longer lifetime.