Crystallization Mechanism of Cage-Based, Small-Pore Molecular

Article pubs.acs.org/cm

Crystallization Mechanism of Cage-Based, Small-Pore Molecular Sieves: A Case Study of CHA and LEV Structures Sang Hyun Ahn, Hwajun Lee, and Suk Bong Hong* Center for Ordered Nanoporous Materials Synthesis, Division of Environmental Science and Engineering, POSTECH, Pohang 37673, Korea S Supporting Information *

ABSTRACT: Here we have investigated the crystallization mechanisms of SSZ-13 and SAPO-34 with the CHA topology and NU-3 and SAPO35 with the LEV topology using 1H−13C CP MAS NMR and IR spectroscopies. The nucleation of these cage-based, small-pore molecular sieves begins with the formation of large 20-hedral cha or 17-hedral lev cages, with incorporation of organic structure-directing agents (SDAs) alone or together with inorganic cations, in both aluminosilicate and silicoaluminophosphate compositions. The next two steps are the construction of multiple-cha or multiple-lev cages in an appropriate arrangement by sharing 8-rings and their subsequent coupling to form smaller 8-hedral double 6-ring units, leading to viable nuclei of CHA or LEV molecular sieves. The initial formation of the large cages, especially in the presence of large SDA molecules, can in our view be generalized to other cage-based zeolite systems.

■

INTRODUCTION Zeolites and related microporous materials with crystallographically well-defined channels and cavities have found increasing use as ion-exchangers, adsorbents, and catalysts.1,2 Today there are more than 230 distinct zeolite structures approved by the Structure Commission of the International Zeolite Association.3 From both scientific and technological points of view, a comprehensive understanding of the mechanisms of zeolite crystallization is of fundamental importance in that such information is essential for the a priori design of zeolites with desired structural features for particular applications. However, the formation mechanisms of these metastable crystals are still far from well understood.4−7 We have recently investigated the crystallization mechanism of UZM-9,8 a silica-rich (3.5 ≤ Si/Al ≤ 6) LTA-type zeolite, synthesized via the so-called charge density mismatch (CDM) approach,9−13 in which tetraethylammonium, tetramethylammonium, and Na+ ions were simultaneously used as structuredirecting agents (SDAs). The LTA structure is made up of 14hedral ([4668]) sod cages that are linked via 6-hedral ([46]) double 4-ring (d4r) units, creating a 26-hedral ([4126886]) lta cage at the unit cell center.3 The combined use of elemental analysis and 13C MAS NMR and IR spectroscopies allowed us to propose that UZM-9 nucleation begins with the construction of large lta cages with 48 tetrahedral atoms (T atoms) followed by that of, in turn, sod cages and d4r units with 24 and 8 T atoms, respectively.8 From the crystallization mechanism study of UZM-5 (UFI),14 a new zeolite structure discovered via the CDM approach,9 we were also able to draw essentially the same conclusion: the formation of cages large enough to © 2017 American Chemical Society

accommodate organic SDA molecules, like lta cages, takes precedence over that of much smaller d4r units. While UZM-5 can crystallize from the same synthesis mixture as that used in UZM-9 crystallization but at a higher temperature (150 vs 100 °C),13 its structure is built from 4-hedral ([45546481]) wbc and 8-hedral ([4454]) rth cages, in addition to sod and d4r cages, as building units.3 The purpose of this study is to check whether the trend found in the nucleation and crystal growth of UZM-9 and UZM-5 could be generalized to other cage-based systems, such as the CHA topology, regardless of their framework composition, even in zeolite synthesis using organic SDAs only. SSZ-13 and SAPO-34 are two molecular sieves with the same framework topology (CHA) but different compositions (aluminosilicate vs silicoaluminophosphate (SAPO)) that have been successfully commercialized as an active catalyst itself for the selective catalytic reduction of nitrogen oxides (NOx) with NH3 and as an active catalyst itself for the methanol-to-olefins reaction, respectively.15,16 As shown in Figure 1, the CHA structure comprises only two types of cages: large 20-hedral ([4126286]) cha cages with 36 T atoms and much smaller 8hedral ([4662]) double 6-ring (d6r) units within which any organic SDA cannot be encapsulated.3 Hence, CHA molecular sieves should be more suitable as a model system for proving whether large cages are formed before smaller ones than the materials made up of three cage types or more. If such is the Received: March 9, 2017 Revised: June 8, 2017 Published: June 15, 2017 5583

DOI: 10.1021/acs.chemmater.7b00980 Chem. Mater. 2017, 29, 5583−5590

Article

Chemistry of Materials

crystallized SSZ-13 or SAPO-34 and NU-3 or SAPO-35. The yield of each product was calculated by dividing the mass of the product obtained after crystallization for a given time by the total mass of the oxide forms of all the components in the synthesis mixture except water. Crystal morphology and average size were determined by a JEOL JSM-6510 scanning electron microscope (SEM). Thermogravimetric analyses (TGA) were performed on an SII EXSTAR 6000 thermal analyzer, where the weight losses related to the combustion of organic SDAs were further confirmed by differential thermal analyses (DTA) using the same analyzer. Elemental analysis for Si, Al, P, and Na was carried out by a Jarrell-Ash Polyscan 61E inductively coupled plasma spectrometer in combination with a PerkinElmer 5000 atomic absorption spectrophotometer. The C, H, and N contents of the samples were analyzed by using a Vario EL III elemental organic analyzer. The 27Al, 29Si, 31P, and 13C solution NMR spectra were recorded in 5 mm quartz tubes on a Bruker Avance III 600 spectrometer. The 27Al, 29 Si, 31P, and 1H−13C CP MAS NMR spectra were collected on a Bruker DRX500 spectrometer at a spinning rate of 21.0 kHz. Detailed measurement conditions for the NMR experiments performed here are given in Table S1. The IR spectra in the structural region were recorded on a Nicolet 6700 FT-IR spectrometer using the KBr pellet technique. The concentration of solid sample in the KBr pellets was kept constant at 0.02 g sample per g KBr, and 256 scans were accumulated to obtain the IR spectra. The Raman spectra were measured on a Bruker RFA 106/S FT-Raman spectrometer equipped with an Nd:YAG laser operating at 1064 nm. The samples were exposed to a laser power of 100−250 mW at the spectral resolution of 4 cm−1. Typically, 400−1600 scans were accumulated for obtaining the Raman spectra.

Figure 1. (a) CHA and (b) LEV structures and their two different building units.

case, another good system would then be the LEV structure consisting of 17-hedral ([496583]) lev cages with 30 T atoms and d6r units (Figure 1). LEV structure can also be synthesized as both aluminosilicate (NU-3) and SAPO (SAPO-35) compositions. Here we use powder X-ray diffraction, elemental and thermal analyses, and multinuclear solution and solid-state NMR and IR spectroscopies to show that no matter which framework composition the CHA and LEV molecular sieves have, their nucleation is initiated by the construction of larger cages (i.e., cha and lev cages, respectively) accompanying the incorporation of organic SDA species.

■

EXPERIMENTAL SECTION

Molecular Sieve Syntheses. The reagents employed in this study included NaOH (50% aqueous solution, Aldrich), N,N,N-trimethyl-1adamantylammonium hydroxide (TMAdAOH, 25%, Sachem), tetraethylammonium hydroxide (TEAOH, 35% aqueous solution, Aldrich), 1-adamantanamine (AA, 97%, Aldrich), hexamethyleneimine (HMI, 99%, Aldrich; Caution! flammable and toxic), aluminum hydroxide (Al(OH)3·0.5H2O, Aldrich), aluminum isopropoxide (≥98%, Aldrich), orthophosphoric acid (85%, Aldrich), and colloidal silica (Ludox AS40, Aldrich). The chemical compositions of the synthesis mixtures used in the crystallization of SSZ-13, SAPO-34, NU-3, and SAPO-35 and their crystallization temperatures are listed in Table 1, and further synthesis details can be found elsewhere.17−19,28 After heating each synthesis gel as a function of time at a given temperature, the solid products and mother liquors were separated by centrifugation (16 000 rpm, 10 min). The recovered solids were redispersed in deionized water using an ultrasonic bath (100 W, 42 kHz) for 60 min with subsequent centrifugation, which was repeated three times. Finally, the resulting solids were dried overnight at room temperature. If required, some solid products were refluxed three times in 1.0 M NaNO3 solutions (1 g of solid/100 mL of solution) for 6 h. Analytical Methods. Powder X-ray diffraction (XRD) patterns were recorded on a PANalytical X’Pert diffractometer (Cu Kα radiation) with an X′Celerator detector. The relative crystallinities of a series of solid products recovered at different times were determined by comparing the area of the most intense X-ray peaks around 2θ = 20.4 and 22.0°, corresponding to the (201)̅ and (122) reflections of the CHA and LEV structures, respectively,9 with those of fully

■

RESULTS AND DISCUSSION Figure 2 shows the powder XRD patterns of a series of solid products obtained after heating the SSZ-13, SAPO-34, NU-3, and SAPO-35 synthesis mixtures as a function of time at 160, 200, 180, and 200 °C, respectively,17−19 and Tables S2−S5 list their chemical composition data. Three X-ray peaks around 2θ = 9.4, 15.9, and 20.4° due to the (100), (111̅), and (201̅) reflections of the CHA structure3 became detectable when the SSZ-13 and SAPO-34 gels were heated for 3 d and 4 h, respectively. The same trend was also observed for the synthesis of LEV molecular sieves (i.e., NU-3 and SAPO-35). The much slower crystallization rates of SSZ-13 and NU-3, compared to their SAPO versions, can be partly attributed to their lower crystallization temperature (160 or 180 vs 200 °C). The SEM images in Figure S1 reveal that molecular sieve crystals studied here have a rhombohedral morphology, although their crystal sizes are different from one another. To investigate the formation pathway for SSZ-13, SAPO-34, NU-3, and SAPO-35, the mother liquors and solid products separated after each crystallization for different times were characterized by solution and solid-state 29Si, 31P, and 27Al NMR spectroscopies, respectively. Despite considerable differ-

Table 1. Chemical Compositions of the Gels Used in the Synthesis of Molecular Sieves Studied Here and Their Crystallization Temperatures materiala

IZA code

gel compositionb

crystallization temperatureb (°C)

ref

SSZ-13 SAPO-34 NU-3 SAPO-35

CHA CHA LEV LEV

10(TMAdA)2O·10Na2O·2.5Al2O3·100SiO2·4400H2O 1.0(TEA)2O·1.0Al2O3·1.0P2O5·0.3SiO2·100H2O 20AA·3.21Na2O·2.5Al2O3·60SiO2·2400H2O 1.5HMI·1.0Al2O3·1.0P2O5·0.3SiO2·100H2O

160 200 180 200

17 19 18 28

a

All materials except NU-3 crystallized under static conditions. NU-3 was synthesized with rotation (60 rpm). bTMAdA, N,N,N-trimethyl-1adamantylammonium hydroxide; TEA, tetraethylammonium hydroxide; AA, 1-adamantanamine; HMI, hexamethylenimine. 5584


Article


−90 and −98 ppm that can be attributed to the Q2 and Q3 units, respectively, or their combinations with the prismatic hexamer (d3r) and with the cubic octamer (d4r), respectively.20−22 Also, the concentration of silicate monomer in both mother liquors was found to be significantly larger than that of any of the other (alumino)silicate species. These results strongly suggest that the crystallization of both SSZ-13 and NU-3 does not require the presence of any particular type of their structural building unit (i.e., d6r) in the solution phase. On the other hand, 1 d of heating of the SSZ-13 and NU-3 gels at 160 and 180 °C, respectively, when both of the resulting solid products are amorphous (Figure 2), led to notable differences in the 29Si NMR spectra of mother liquors. Only the three low-field 29Si lines are observed from the spectrum of the SSZ-13 synthesis mother liquor, implying that, during the crystallization, the higher-molecular-weight (alumino)silicate species were consumed faster in the solution phase than the low-molecular-weight ones. As shown in Figure 3, however, there are no detectable lines in the 29Si NMR spectrum of the NU-3 synthesis mother liquor. We also note that although the Al concentrations (Si/Al = 20 and 12, respectively) in both starting synthesis gels are much smaller than the Si concentrations, the 27Al NMR spectra of the mother liquors separated after SSZ-13 and NU-3 fully crystallize are still characterized by one broad line around 65 ppm attributed to the qn(3Si) species23 with little intensity change (Figure S2). Therefore, the migration of Al from the solution phase to the solid phase during the synthesis of these two zeolites appears to be insignificant in comparison to that of Si. This assumption correlates well with the observation that all X-ray amorphous solids obtained during their synthesis gave one very broad 29Si line around −107 ppm or higher (Figure 3), indicating their high-silica nature (Tables S2 and S4). Figure 4 shows the 31P NMR spectra of some mother liquors separated from SAPO-34 and SAPO-35 syntheses at 200 °C for different times. As long reported,24−26 the spectra of the

Figure 2. Powder XRD patterns for a series of solid products obtained after (a) SSZ-13, (b) SAPO-34, (c) NU-3, and (d) SAPO-35 syntheses at 160, 200, 180, and 200 °C for different times, respectively.

ences in the concentration of organic and inorganic SDAs, as well as of Al, in the starting gels used for SSZ-13 and NU-3 syntheses, the 29Si NMR spectra of the mother liquors of their starting synthesis gels were found to be almost identical to each other. As shown in Figure 3, both spectra are characterized not only by three lines around −72, −81, and −82 ppm corresponding to the silicate monomer, dimer, and cyclic timer, respectively, but also by two broader 29Si lines around

Figure 4. Solution 31P NMR spectra (left) of some mother liquors and 31 P MAS NMR spectra (right) of the corresponding solid products recovered after (a) SAPO-34 and (b) SAPO-35 syntheses at 200 °C for different times, respectively.

Figure 3. Solution 29Si NMR spectra (left) of several mother liquors and 29Si MAS NMR spectra (right) of the corresponding solid products separated after (a) SSZ-13 and (b) NU-3 syntheses at 160 and 180 °C for different times, respectively. 5585


Article


35 syntheses at 160, 200, 180, and 200 °C, respectively, which are all X-ray amorphous, and of the fully crystallized phases of these four molecular sieves. For comparison, the solution 13C NMR spectra in D2O of TMAdA+, TEA+, AA, and HMI, which were used as organic SDAs in their syntheses, respectively, are also given. The 1H−13C CP MAS NMR results clearly show that each of organic species in these solids remains intact. We should note here that while all four organic SDAs are quite smaller in size than the dimensions (8.4 × 8.4 × 8.2 and 8.1 × 8.1 × 7.0 Å3, respectively) of cha and lev-cages in CHA and LEV molecular sieves, respectively, they are considerably larger than their 8-ring window sizes (3.8 × 3.8 and 3.6 × 4.8 Å2, respectively).3 When encapsulated within the cha or lev cages, therefore, they cannot be removed without cage breaking. Both elemental and thermal analyses reveal that, despite the amorphous nature, the four solids recovered at the very beginning of SSZ-13, SAPO-34, NU-3, and SAPO-35 syntheses have considerable organic contents (according to TGA/DTA results, for example, 10.1, 12.8, 3.1, and 8.6 wt %, respectively; Figure S4). More importantly, although these amorphous solids were refluxed three times in 1.0 M NaNO3 solutions for 6 h, they were found to still retain at least about 60% of their original organic contents (Figure S4). This suggests that most of their organics are positioned in the nonexchangeable sites, most likely within cha and lev cages, respectively, or within analogous cages that can be transformed into each of the cha or lev cages. It is also remarkable that the organic contents in the first solids separated during the crystallization of SSZ-13 and SAPO-34 are at least 70% of those in the fully crystallized products (Tables S2 and S3). However, the organic contents in similar amorphous solids obtained during the crystallization of NU-3 and SAPO-35 are only 40% or smaller of those in the corresponding, fully crystallized materials (Tables S4 and S5). Therefore, we speculate that the role of organic SDAs in the nucleation stage is more important for CHA molecular sieves than for LEV ones, regardless of the framework composition or the presence of any inorganic SDAs in synthesis mixtures. In our previous Raman study,28 we have shown that HMI in asmade SAPO-35 exists as the protonated form but not as the neutral one. To check whether this is also the case for AA used in NU-3 synthesis, we performed Raman measurements on asmade NU-3, as well as on AA and AA:HCl. As can be found in Figure S5, neutral AA exhibits a weak band around 3260 cm−1 due to the N−H stretching mode.29 However, neither protonated AA nor as-made NU-3 prepared here shows this band. Therefore, it is clear that AA is located as the protonated form inside the NU-3 pores. Figure 6 shows the IR spectra in the 500−900 cm−1 region of a series of solid products isolated after crystallization of SSZ-13 and SAPO-34 as a function of time at 160 and 200 °C, respectively. An IR band appearing at 637 and 648 cm−1 in the spectra of fully crystallized SSZ-13 and SAPO-34, which is due to the d6r unit in the CHA structure,30,31 begins to appear in the IR spectra of the products obtained after at least 4 d and 4 h of heating in their syntheses, respectively. Notice that these crystallization times are considerably longer than those (1 d and 1 h, respectively; Figure S4) required to obtain amorphous solids containing non-negligible amounts of organic SDAs. The intensity of the d6r band increases until SSZ-13 and SAPO-34 fully crystallize. Given that the detection limit of IR spectroscopy for the atomic ordering of zeolites is at least comparable to that of NMR spectroscopy,30 we conclude that

mother liquors of both starting SAPO gels are characterized by a number of very sharp 31P lines: among them, the lowest-field line around 3 ppm can be assigned to free H3PO4 molecules and H2PO4− and (H3PO4)nm+ ions, and all the other higherfield lines to those species that are bonded to Al monomers or dimers. The spectra of the mother liquors separated after 1 d of SAPO-34 and SAPO-35 syntheses, when both molecular sieves fully crystallize, reveal that the relative intensity of the lowestfield 31P line becomes considerably stronger than that of any of the other lines. This implies that the water-soluble aluminophosphate (AlPO4) species identified by 31P NMR spectroscopy are the nutrients for the growth if AlPO4-based molecular sieves. Figure 4 also shows the 31P MAS NMR spectra of the solids isolated after crystallization of SAPO-34 and SAPO-35 at 200 °C for 1 h, which are still amorphous (Figure 2). Their spectra are dominated by a very broad 31P line around −16 ppm or slightly higher. This suggests that most of the P atoms in these solids with Al/(P + Si) = 1.0 ± 0.1 (Tables S3 and S5) are coordinated with at least one water molecule and/or are located at connectivity defects.26−28 We note here that their 27Al MAS NMR spectra are characterized not only by a small line around 40 ppm, typical of tetrahedral Al, but also by a much stronger line around 0 ppm (Figure S3). This line shape is substantially different from that seen in the 27Al MAS NMR spectra of the amorphous aluminosilicate solids, obtained during SSZ-13 and NU-3 syntheses, consisting of the tetrahedral 27Al line only (Figure S2). Figure 5 shows the 1H−13C CP MAS NMR spectra of the first solids recovered after SSZ-13, SAPO-34, NU-3, and SAPO-

Figure 5. 1H−13C CP MAS NMR spectra of the solid products separated after (a) SSZ-13, (b) SAPO-34, (c) NU-3, and (d) SAPO-35 syntheses at 160, 200, 180, and 200 °C for different times, respectively. The solution 13C NMR spectra in D2O of their organic SDAs (i.e., TMAdAOH, TEABr, AA, and HMI), showing the assignment of each resonance, are given as the bottom traces, respectively. 5586


Article


3 and SAPO-35 give the d6r band at 665 and 649 cm−1 that is observable from the IR spectra of the products obtained after 7 d and 3 h of heating in the crystallization of these two materials having the same LEV topology, respectively. Thus, compared with the 1H−13C CP MAS NMR results, the formation of larger lev cages, with complete incorporation of organic SDAs, should precede that of small d6r units, regardless of their framework composition and presence of inorganic cations in the synthesis mixture. It is worth noting that relatively strong bands appearing around 540, 560, and 530 cm−1 in the IR spectra of fully crystallized SSZ-13, NU-3, and SAPO-34, respectively, are assignable to the single 6-ring (s6r) unit in the CHA or LEV structure.32 Figure 7 compares changes in the organic SDA/(Al or Si), Na+/Al, (organic SDA + Na+)/Al, and Si/Al or Si/(Al + P + Si) ratios of the solid products, calculated from the chemical composition data in Tables S2−S5, as a function of relative crystallinity during the synthesis of CHA and LEV molecular sieves with different framework compositions. While the early solids obtained during the crystallization of SSZ-13, which are still X-ray amorphous (Figure 2), exhibit a notable decrease in Na+/Al ratio with increasing the crystallization time, the Na+/Al ratio of similar solids isolated in the synthesis of NU-3 remains almost unchanged. However, because the organic SDA/Al ratio is nearly the same for both series of solid products, the structure-directing effects of organic SDAs at nucleation appear to be stronger in the cationic SDA-assisted SSZ-13 synthesis than in the neutral SDA-mediated NU-3 synthesis. Figure 7 also shows that during the nucleation of SAPO-34 and SAPO-35 (Figure 2), the organic SDA/Si ratio of the solid products isolated increases with increasing crystallization time. It is interesting to note here that the extent of increase in organic SDA/Si ratio is considerably lower in SAPO-34 nucleation than in SAPO-35 nucleation. This suggests that the nucleation of the former SAPO material is less organic SDA-dependent.

Figure 6. IR spectra in the structural region of the solid products recovered after (a) SSZ-13, (b) SAPO-34, (c) NU-3, and (d) SAPO35 syntheses at 160, 200, 180, and 200 °C for different times, respectively.

the nucleation of CHA molecular sieves may begin with the formation first of large cha cages and then of smaller d6r units at both chemical compositions. A similar conclusion can also be derived from the IR data of the solid products recovered after NU-3 and SAPO-35 syntheses at 180 and 200 °C for different times, respectively. As shown in Figure 6, fully crystallized NU-

Figure 7. Organic SDA/(Al or Si) (■), Na+/Al (●), (organic SDA + Na+)/Al (▲), and Si/Al (▽) or Si/(Al + P + Si) (left-pointing triangle) ratios in the solid products recovered after (a) SSZ-13, (b) SAPO-34, (c) NU-3, and (d) SAPO-35 crystallization at 160, 200, 180, and 200 °C for different times, respectively, where M+ is both an organic SDA (i.e., TMAdA+, TEA+, AA+, or HMI+) and/or an inorganic SDA (Na+). 5587


Article


Figure 8. Schematic illustrations of possible formation pathways for CHA (left) and LEV (right) molecular sieves in the presence of organic SDAs alone or with inorganic cations.

resulting rhombohedral crystals, therefore, the formation of cha or lev cage dimers and/or analogous cages, the shortest in length, can be logically hypothesized (step b). The next step is the proper arrangement of such cage dimers (step c), which may be more important in the nucleation of LEV molecular sieves. This is because the lev cages are connected to one another via shared s6rs, as well as via 8-ring windows. Finally, coupling among the cage dimers results in d6r formation (step d). A similar self-assembly process in which organic and/or inorganic SDAs are involved would allow the nuclei of CHA or LEV molecular sieves to further grow in the solid phase and/or at the interface with the aluminosilicate or SAPO solution (step e). The formation pathways in Figure 8 are essentially identical with the trend reported in the crystallization mechanism studies on UZM-5 and UZM-9 synthesized using the CDM approach.8,14 Therefore, it is most likely that these formation pathways could be applied to explain the crystallization mechanism of other cage-based zeolites. They also provide some insight into the organic SDA-mediated synthesis of zeolites, especially of cage-based materials: if the formation sequence of zeolite cages is strongly dependent on their size, this could then be one reason the phase selectivity of the crystallization can differ according to the type of organic additives used in zeolite synthesis.2,6

Figure 8 shows the formation pathways for CHA and LEV molecular sieve crystals drawn based on the characterization results presented thus far. Small aluminosilicate and SAPO species are arranged in a structured manner around the organic SDAs (and/or the hydrated Na+ ions) to make cha and lev cages at the nucleation stage in the solid phase and/or at the solid−liquid interface (step a), respectively, no matter whether they are structurally incomplete or complete. The incorporation of organic SDAs into this cage is consistent with the 1H−13C CP MAS NMR results in Figure 5. The negligible increase in Si/Al ratio observed for the X-ray amorphous solids isolated at early times of SSZ-13 and NU-3 syntheses (Figure 8) implies that the migration of both Si and Al from the solution phase to the solid phase during the nucleation of these two zeolites may not be significant, in good agreement with the solution and solid-state 29Si and 27Al NMR results shown in Figures 3 and S2. However, this is not the case of their SAPO version, i.e., SAPO-34 and SAPO-35, because a notable decrease in the Si/ (Al + P + Si) ratio of SAPO solids, which is a clear indication of the Si release from the initially formed solids to the solution phase, is observed at early crystallization times. Once such a single cha or lev cage and/or something similar is formed, the cations compensating the negative framework charges created by Al substitution during the synthesis of SSZ13 and NU-3 or by Si substitution during the synthesis of SAPO-34 and SAPO-35 can be located not only within this large cage but also in the proximity of its outer surface. Here we assume the formation of multiple cha or multiple lev cages by sharing their 8-rings which can be more facile in the presence of large organic SDAs rather than of smaller inorganic cations. This is because the formation of many single cha or lev cages in random orientations may make it very difficult to allow further nucleation. Furthermore, the growth of rhombohedral crystals (Figure S1) cannot be favorable for LEV materials, when their lev cages share both single 6-rings (s6rs) and 8-ring windows prior to d6r formation. In fact, the formation of particular, survivable nuclei during zeolite crystallization is a challenging task that requires repeated dissolution/recrystallization.14 To improve this situation while maintaining the morphology of the

■

CONCLUSIONS

A plausible formation pathway for molecular sieve crystals with CHA and LEV topologies has been proposed based on the results from powder X-ray diffraction, elemental and thermal analyses, multinuclear solution and solid-state NMR, and IR spectroscopic measurements of a series of solid products and mother liquors recovered after SSZ-13, SAPO-34, NU-3, and SAPO-35 syntheses at 160, 200, 180, and 200 °C for different times, respectively. It was found that the construction of large cha or lev cages, with encapsulation of organic SDAs alone or along with inorganic SDAs, precedes that of much smaller d6r units in both aluminosilicate and SAPO compositions. Combined with the trend found in the formation of UZM-5 and UZM-9 zeolite crystals, therefore, the nucleation of other 5588


Article


Zeolites UZM-4 and UZM-5. Angew. Chem., Int. Ed. 2003, 42, 1737− 1740. (10) Lewis, G. J.; Miller, M. A.; Moscoso, J. G.; Wilson, B. A.; Knight, L. M.; Wilson, S. T. Experimental Charge Density Matching Approach to Zeolite Synthesis. Stud. Surf. Sci. Catal. 2004, 154, 364−372. (11) Miller, M. A.; Moscoso, J. G.; Koster, S. C.; Gatter, M. G.; Lewis, G. J. Synthesis and Characterization of the 12-Ring Zeolites UZM-4 (BPH) and UZM-22 (MEI) via the Charge Density Mismatch Approach in the Choline-Li2O-SrO-Al2O3-SiO2 System. Stud. Surf. Sci. Catal. 2007, 170, 347−354. (12) Kim, S. H.; Park, M. B.; Min, H.-K.; Hong, S. B. Zeolite Synthesis in the Tetraethylammonium−Tetramethylammonium Mixed-Organic Additive System. Microporous Mesoporous Mater. 2009, 123, 160−168. (13) Park, M. B.; Jo, D.; Jeon, H. C.; Nicholas, C. P.; Lewis, G. J.; Hong, S. B. Zeolite Synthesis from a Charge Density Perspective: The Charge Density Mismatch Synthesis of UZM-5 and UZM-9. Chem. Mater. 2014, 26, 6684−6694. (14) Park, M. B.; Ahn, N. H.; Broach, R. W.; Nicholas, C. P.; Lewis, G. J.; Hong, S. B. Crystallization Mechanism of Zeolite UZM-5. Chem. Mater. 2015, 27, 1574−1582. (15) Bull, I.; Xue, W.-M.; Burk, P.; Boorse, R. S.; Jaglowski, W. M.; Koermer, G. S.; Moini, A.; Patchett, J. A.; Dettling, J. C.; Caudle, M. T. Process for Reducing Nitrogen Oxides Using Copper CHA Zeolite Catalysts. U.S. Patent 8404203, 2013. (16) Tian, P.; Wei, Y.; Ye, M.; Liu, Z. Methanol to Olefins (MTO): From Fundamentals to Commercialization. ACS Catal. 2015, 5, 1922− 1938. (17) Robson, H.; Lillerud, K. P. Verified Syntheses of Zeolitic Materials, 2nd ed.; Elsevier: Amsterdam, Netherlands, 2001. (18) Glyn, D. S.; Thomas, V. W. Zeolite NU-3. U.S. Patent 4372930, 1983. (19) Wilson, S.; Barger, P. The Characteristics of SAPO-34 which Influence the Conversion of Methanol to Light Olefins. Microporous Mesoporous Mater. 1999, 29, 117−126. (20) Engelhardt, G.; Michel, D. High-Resolution Solid-State NMR of Silicates and Zeolites; Wiley: Chichester, U.K., 1987. (21) Thangaraj, A.; Kumar, R. NMR Study of Soluble Aluminosilicate Precursors in Zeolite Y Synthesis. Zeolites 1990, 10, 117−120. (22) Cheng, C.-H.; Shantz, D. F. 29Si NMR Studies of Zeolite Precursor Solutions. J. Phys. Chem. B 2006, 110, 313−318. (23) Harris, R. K.; Samadi-Maybodi, A.; Smith, W. The Incorporation of Aluminum into Silicate Ions in Alkaline Aqueous Solutions, Studied by Aluminum-27 n.m.r. Zeolites 1997, 19, 147−155. (24) Mortlock, R. F.; Bell, A. T.; Radke, C. J. 31P and 27Al NMR Investigations of Highly Acidic, Aqueous Solutions Containing Aluminum and Phosphorous. J. Phys. Chem. 1993, 97, 767−774. (25) Mortlock, R. F.; Bell, A. T.; Radke, C. J. 31P and 27Al NMR Investigations of the Effects of pH on Aqueous Solutions Containing Aluminum and Phosphorous. J. Phys. Chem. 1993, 97, 775−782. (26) Briend, M.; Peltre, M. J.; Lamy, A.; Man, P. P.; Barthomeuf, D. Dependence of the Acidic Properties of SAPO-37 Molecular Sieve-On Si Content and Heat Trearment. J. Catal. 1992, 138, 90. (27) Buchholz, A.; Wang, W.; Arnold, A.; Xu, M.; Hunger, M. Successive Steps of Hydration and Dehydration of silicoaluminophosphates H-SAPO-34 and H-SAPO-37 Investigated by In Situ CF MAS NMR Spectroscopy. Microporous Mesoporous Mater. 2003, 57, 157. (28) Jung, H. J.; Shin, C.−H.; Hong, S. B. Si Distribution in Silicoaluminophosphate Molecular Sieves with the LEV Topology: A Solid-State NMR Study. J. Phys. Chem. B 2005, 109, 20847−20853. (29) Lin-Vien, D.; Colthup, N. B.; Fateley, W. G.; Grasselli, J. G. Handbook of Infrared and Raman Characteristic Frequencies of Organic Molecules; Academic: San Diego, 1991. (30) Mozgawa, W.; Jastrzębski, W.; Handke, M. Vibrational Spectra of D4R and D6R Structural Units. J. Mol. Struct. 2005, 744−747, 663− 670. (31) Król, M.; Mozgawa, W.; Jastrzębski, W.; Barczyk, K. Application of IR Spectra in the Studies of Zeolites from D4R and D6R Structural Groups. Microporous Mesoporous Mater. 2012, 156, 181−188.

cage-based molecular sieves may also begin with the formation of larger cages rather than of small ones among their different structural building units. This may be particularly prone to the crystallization process of such zeolites which includes the use of larger organic SDAs.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.chemmater.7b00980. NMR measurement conditions and chemical composition, 27Al NMR, and Raman data of a series of mother liquors and/or solid products recovered after SSZ-13, SAPO-34, NU-3, and SAPO-35 syntheses at 160, 200, 180, and 200 °C for different times, respectively, and changes in the organic content of some solid products by Na+ ion exchange at 80 °C 6 h (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*(S.B.H.) E-mail: [email protected]. ORCID

Suk Bong Hong: 0000-0002-2855-1600 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the National Creative Research Initiative Program (2012R1A3A2048833) through the National Research Foundation of Korea funded by the Korea government (MSIP).

■

REFERENCES

(1) Camblor, M. A.; Hong, S. B. Synthetic Silicate zeolites: Diverse Materials Accessible through Geoinspiration. In Porous Materials; Bruce, D. W., O’Hare, D., Walton, R. I., Eds.; Wiley: Chichester, U.K., 2011; pp 265−325. (2) Moliner, M.; Rey, F.; Corma, A. Towards the Rational Design of Efficient Organic Structure-Directing Agents for Zeolite Synthesis. Angew. Chem. Int. Ed. 2013, 52, 13880−13889. (3) Baerlocher, Ch.; McCusker, L. B. Database of Zeolite Structures. http://www.iza-structure.org/databases/. (4) Chang, C. D.; Bell, A. T. Studies on the Mechanism of ZSM-5 Formation. Catal. Lett. 1991, 8, 305−316. (5) Burkett, S. L.; Davis, M. E. Mechanisms of Structure Direction in the Synthesis of Pure-Silica Zeolites. 1. Synthesis of TPA/Si-ZSM-5. Chem. Mater. 1995, 7, 920−928. (6) Davis, M. E.; Zones, S. I. A Perspective on Zeolite Synthesis: How Do You Know What You’ll Get? In Synthesis of Porous Materials; Occelli, M. L., Kessler, H., Eds.; Marcel Dekker: New York, 1997; pp 1−34. (7) Cundy, C. S.; Cox, P. A. The Hydrothermal Synthesis of Zeolites: Precursors, Intermediates and Reaction Mechanism. Microporous Mesoporous Mater. 2005, 82, 1−78. (8) Park, M. B.; Lee, Y.; Zheng, A.; Xiao, F.-S.; Nicholas, C. P.; Lewis, G. J.; Hong, S. B. Formation Pathway for LTA zeolite Crystals Synthesized via a Charge Density Mismatch Approach. J. Am. Chem. Soc. 2013, 135, 2248−2255. (9) Blackwell, C. S.; Broach, R.; Gatter, M. G.; Holmgren, J. S.; Jan, D. Y.; Lewis, G. J.; Mezza, B. J.; Mezza, T. M.; Miller, M. A.; Moscoso, J. G.; Patton, R. L.; Rohde, L. M.; Schoonover, M. W.; Sinkler, W.; Wilson, B. A.; Wilson, S. T. Open-Framework Materials Synthesized in the TMA+/TEA+ Mixed-Template System: The New Low Si/Al Ratio 5589


Article

Chemistry of Materials (32) Mozgawa, W.; Król, M.; Barczyk, K. FT-IR Studies of Zeolites from Different Structural Groups. Chemik 2011, 65, 671−674.

5590


Crystallization Mechanism of Cage-Based, Small-Pore Molecular

Recommend Documents