A phase-transfer crystallization pathway to synthesize ultrasmall silicoaluminophosphate for enhanced catalytic conversion of dimethylether-to-olefin

Hongxin Ding a, Jiajia Ding b, Wei Liu b, Xiaoling Zhao a, Qijin Chi c, Kake Zhu *a, Xinggui Zhou a and Weimin Yang *b
aState Key Laboratory of Chemical Engineering, East China University of Science and Technology, Shanghai 200237, P. R. China. E-mail: kakezhu@ecust.edu.cn
bShanghai Research Institute of Petrochemical Technology, Sinopec, Shanghai 201208, P. R. China. E-mail: yangwm.sshy@sinopec.com
cDepartment of Chemistry, Technical University of Denmark, Kemitorvet Building 207, DK-2800 Kongens Lyngby, Denmark

Received 14th October 2018 , Accepted 30th November 2018

First published on 30th November 2018

We present a phase-transfer crystallization approach to the synthesis of ultrasmall (<100 nm) SAPO-34 crystallites. This method features porogen-free and tumbling crystallization in biphasic toluene–water media. The nanosized SAPO-34 catalyst has demonstrated prolonged catalysis lifetime (approximately 5 fold) and higher ethylene/propylene selectivity (by ca. 5%) for dimethylether-to-olefin conversion.

Zeolites and zeotype silicoaluminophosphates (SAPOs) are a family of microporous crystalline solids that are used as acid catalysts in chemical transformation, owing to their unique shape selectivity.1–4 Among them, SAPO-34 with a CHA topology is an industrial catalyst for the production of light olefins (i.e. ethylene and propylene) from methanol (MTO) or dimethylether (DTO), an important step to derive olefin feedstocks from non-petroleum sources such as coal or natural gas.5 The outstanding catalytic selectivity is attributed to its intermediate acidity, relatively large cavity (0.94 × 1.27 nm) and narrow pore aperture (0.38 × 0.38 nm) that allows for the egression of only small molecules.6,7 As the catalytic transformation proceeds through intricately combined aromatic and olefin cycles, co-produced bulky aromatics could deactivate the catalyst and shorten the catalyst lifespan.8,9 Indeed, one of the major efforts for the MTO or DTO catalyst design is to find ways to prolong the catalyst lifetime without sacrificing the selectivity to light olefins.10–15 It has been determined that the deactivation behaviour is closely related to the diffusion properties of SAPO-34 crystals, i.e., reduction of diffusion length (Thiele theorem) or/and generation of auxiliary transport macro- or mesoporosity can significantly increase the catalyst lifetime.16–22

Towards this end, a persistent quest for ways to fabricate nanocrystalline or hierarchically porous SAPO-34 was sought in the past decade.23–29 Synthetic strategies including the soft-template method,30 hard-template method,31 dry gel conversion (DGC),32 microwave-assisted crystallization,33–35 post-treatment with alkali solution or HF,36 and so forth have been reported. Hitherto, only the microwave-assisted crystallization,33,34 DGC,32,37 and time-release of Si (ref. 38) could produce sub 100 nm SAPO-34 crystals. These protocols, however, suffer from difficulties in upscaling, low yields, deterioration of acidity or the use of expensive additives. The practical challenge in downsizing SAPO-34 crystals is to identify simple, inexpensive, scalable and efficient ways to produce the material while simultaneously preserving the acidity, a goal that is yet to be achieved.

Herein, we show the preparation of ultra-small SAPO-34 crystallites from toluene–water medium by a phase-transfer crystallization pathway. The starting mixture, consisting of an aluminium isopropoxide (AIP) dissolved in toluene and an immiscible aqueous mixture of H3PO4, tetraethyl ammonium hydroxide (TEAOH) and tetraethyl orthosilicate, was crystallized at 473 K under tumbling for 48 h to produce SAPO-34-N (with the suffix indicating nano), as elaborated in the ESI. Toluene was chosen due to its excellent ability to solubilize AIP, chemical inertness, low-toxicity and vapour pressure under hydrothermal conditions. An estimated yield of 87% based on an inorganic basis is achieved, which is comparable to the yield (89%) of SAPO-34-C (ESI, C stands for conventional) derived from a standard hydrothermal synthesis.

The high phase purity of SAPO-34-N was confirmed by the powder X-ray diffraction (XRD) patterns of the as-synthesized sample (Fig. 1a), showing typical diffraction peaks of the CHA structure free of unconverted raw materials or impurities. The scanning electron micrograph (SEM) image (Fig. 1d) reveals that SAPO-34-N consists of uniform aggregated cubes, a typical shape for trigonal CHA crystals synthesized using TEAOH as a structure directing agent (SDA). Particle sizes based on the distribution statistics vary from 50 to 100 nm, with a weighted average value of 65 nm. The aggregated particulates consist of even smaller primary crystallites of 20–40 nm (Fig. S1, ESI). To the best of our knowledge, this is the smallest average crystallite size ever reported. In contrast, the standard SAPO-34-C is made up of 200 nm crystals as shown in Fig. 1e. TEM gives the similar information for the aggregated crystallites (Fig. 1f), and the primary crystallites are found to have aligned along the same crystallographic orientation, thus giving rise to a single crystal like diffraction pattern (inset of Fig. 1f), i.e., a typical feature of mesocrystals.39

image file: c8ce01752b-f1.tif
Fig. 1 XRD patterns (a), adsorption rates of propylene by fitting eqn 2 in the ESI (b), N2 physisorption isotherms (c), FE-SEM images of SAPO-34-N (d) and SAPO-34-C (e), TEM image (f) and SAED pattern (the inset in (f)) of SAPO-34-N.

As the visual method often fails to detect the twinning of domain intergrowths in zeolites, which hurdles intracrystalline diffusion, direct measurements of transient uptake curves are conducted using propylene as a probe molecule to manifest the influence of crystal size over diffusion properties. The yielded characteristic diffusion times (L2/Deff, units: s) using Fick's second law (ESI), where L is the effective radius of crystallites and Deff is the diffusivity, were used to estimate the extent of diffusion constraint mitigation. As shown in Fig. 1b and Table 1, by plotting the adsorbate concentration as a function of time, the L2/Deff value decreases from 1.03 × 104 s for SAPO-34-C to 2.01 × 103 s for SAPO-34-N, as a result of crystal size reduction.

Table 1 Compositions and textural properties of SAPO-34-C and SAPO-34-N
Sample Molar composition S BET (m2 g−1) S micro (m2 g−1) S ext (m2 g−1) V micro (cm3 g−1) V total (cm3 g−1) Yield (%) L 2/Deff (s)
SAPO-34-N Si0.12Al0.50P0.38O2 633 392 241 0.28 0.59 83 2.01 × 103
SAPO-34-C Si0.13Al0.49P0.38O2 450 395 55 0.26 0.33 89 1.03 × 104

N2 physisorption was employed to reveal the textural properties of SAPO-34-N (Fig. 1c). The isotherms can be categorized as a type I isotherm, with a corresponding micropore volume of 0.28 cm3 g−1. In addition, a jump in the N2 uptake due to the capillary condensation is also found at a relative pressure above 0.85, which is ascribed to the presence of interstices resulting from the packing of particulates. A micropore volume of 0.26 cm3 g−1 for SAPO-34-C is deduced. The comparable micropore volumes between the two samples suggest that the inherent crystallinity and framework integrity were well preserved by the biphasic synthesis. With the reduction of crystal sizes, an increase in the BET surface area from 450 m2 g−1 for SAPO-34-C to 633 m2 g−1 for SAPO-34-N is seen, which is largely ascribed to the increment of external surface area (Table 1).

31P, 27Al, and 29Si MAS NMR spectra for SAPO-34-C and SAPO-34-N were collected and compared to disclose the effect of synthesis over the chemical environment of the constructing elements. The 27Al spectra (Fig. 2a and b) show one intense peak at 36 ppm, which originates from tetra-coordinated Al in the Al(OP)4 environment, and a weak signal at −9 ppm attributable to a small amount of penta-coordinated aluminium atoms is also visible, as a result of coordination with one molecule H2O.40 The sole peak at −29 ppm in the 31P spectra (Fig. 2c and d) indicates that all P atoms are tetra-coordinated into the framework in the form of P(OAl)4 species. Both samples exhibit a broadened peak in the 29Si MAS NMR spectra (Fig. 2e and f), which can be regarded as a convolution of six peaks centred at −85, −91, −96, −101, −106 and −111 ppm, assignable to Si atoms in various local environments of Si(OT)n(OH)4−n, Si(4Al), Si(3Al,1Si), Si(2Al,2Si), Si(1Al,3Si) and Si(0Al,4Si), respectively.41 For SAPO-34-N, quantitative deconvolution (Table 2) shows that the proportion of Si atoms located at Si single sites or small patches is predominant, as evidenced by the 51.9% concentration of Si(4Al) and a rather small amount (6.0%) of Si(0Al, 4Si). In contrast, the fractions of Si(4Al) and Si(0Al, 4Si) are, respectively, only 29.5% and 10.6% in SAPO-34-C. Meanwhile, the comparable fractions of Si(OT)n(OH)4−n species (Table 2) between the samples suggest that the defective sites are not increased for SAPO-34-N. The results support that the present synthesis method enables the formation of highly dispersed Si species that are well incorporated into the CHA framework.

image file: c8ce01752b-f2.tif
Fig. 2 27Al (a and b), 31P (c and d), and 29Si (e and f) MAS NMR spectra, DRIFTS (g) and NH3-TPD (h) of SAPO-34-N and SAPO-34-C.
Table 2 Distribution of silicon atoms by deconvolution of different 29Si MAS NMR spectra
Sample Chemical environments (%)
Si(OT)n(OH)4−n Si(4Al) Si(3Al) Si(2Al) Si(1Al) Si(0Al)
SAPO-34-N 24.1 51.9 5.9 4.4 7.6 6.0
SAPO-34-C 27.2 29.5 14.7 11.2 7.7 10.6

In order to probe acidity, the diffuse reflectance infrared Fourier-transform spectra (DRIFTS) of the two samples are compared in Fig. 2g. The weak bands at 3743 and 3677 cm−1 are assigned to weak acid sites such as Al–OH/P–OH and Si–OH species located on the external surface of the crystals,42 respectively, whereas the doublet bands at 3625 and 3598 cm−1 are associated with stretching modes of bridged strong Brønsted OH group Al–(OH)–Si sites. The complementary NH3-TPD profiles for the two samples show that both samples possess comparable amounts of strong acid sites, as shown by the presence of high temperature desorption profiles at 690 and 715 K, respectively (Fig. 2h). These results, together with the NMR data, consistently indicate that the acid strength pertaining to the CHA topology was largely preserved for SAPO-34-N.

Controlled experiments (see ESI Fig. S2 and S3 and Table S1) demonstrate that Al dissolved in toluene assisted the formation of small size crystals compared to insoluble pseudoboehmite (Fig. S2 and S3), dynamic crystallization suppressed severe aggregation of primary crystallites that afforded bulkier crystals (Fig. S2 and S3), and toluene extracted isopropanol and minimized its effect on crystallization (Fig. S2 and S3), which are all indispensable for a successful synthesis.

To shed light on the underlying formation mechanism, the crystallization process was monitored by combined time-dependent characterization techniques, as shown in Fig. S4–S9 (ESI). During the hydrothermal synthesis of SAPO-34-C, a layered precursor of ca. 405 nm formed at the early stage was gradually converted to CHA crystals, as shown by the digital photographs, dynamic light scattering (DLS) measurements of size, XRD, FT-IR and SEM variations displayed in Fig. S4–S7 (ESI). Temporal overlap of precursor depolymerisation, nucleation and crystal growth was observed. In striking contrast, in biphasic medium, it was the formation of a colloid in the aqueous phase at the early stage (0.5–3 h) that preceded crystallization, showing that AIP was transferred from the toluene phase to an aqueous phase upon heating. DLS analyses of the colloid size show that the size increased from 45.5 to 59.9 and 60.1 nm from 0.5 to 2 and 3 h during crystallization (Fig. S5, ESI). The temporal XRD patterns and the corresponding crystallization curves (Fig. S6a–c, ESI) demonstrate a fast nucleation rate for biphasic synthesis, and relative crystallinity reaches 80% in 3 h whereas 24 h is required for SAPO-34-C to reach the same stage. In the meantime, the evolution of order tracked by FT-IR shows that the characteristic vibration out of double-6 rings (d6r) that construct the CHA structure at 640 cm−1 (Fig. S6d, ESI) becomes detectable as early as 1 h. The intensity of the characteristic peaks of SAPO-34 in the range of 1500–400 cm−1 becomes sharp with elapsed time.41 The crystal size and shape variations were also monitored by SEM/TEM imaging (Fig. S7–S9, ESI). Irregularly shaped amorphous precursors quickly transformed into uniform spherical crystallites in 2 h and grew into rhombohedral crystals of ca. 100 nm after 24 h (Fig. S7, ESI). Thereafter, the crystal growth via oriented attachment of primary crystallites led to the formation of aligned small crystallites into larger ones, as also corroborated by the TEM image (Fig. S9, ESI) showing aggregated particulates ranging from 50 to 80 nm. The chemical composition analyses depicted in Fig. S10 (ESI) show that the content of P, Si and Al in the solid only varied in the initial 3 h and remained nearly constant for the rest of the crystallization time.

On the basis of the above-mentioned experiments and considerations, we propose a possible formation mechanism as illustrated in Scheme 1. In the phase-transfer synthesis of the precursor mixture, AIP dissolved in toluene underwent hydrolysis and released AlOx moieties into the aqueous phase upon heating and tumbling, whereby they quickly combined with other components (P, Si, TEAOH) to form an amorphous colloid. This is why AlOx was not captured by the temporal studies. As the colloid already contained all the necessary components for SAPO-34 growth, it was further transformed into crystallites of comparable size. The rapid formation of colloid precursors results in the formation of numerous size focused crystallites.3,35,43 It is speculated that the supersaturation ratio and consequently the nucleation rate are high in the phase-transfer synthesis, as all precursors have participated in the nucleation process and no precursors have been left after the first 3 h of crystallization. This reasoning is plausible as high supersaturation is also responsible for the formation of zeolitic nanocrystals in DGC or microwave-assisted synthesis.32,33 In the late phase of crystallization, as most nutrients have been consumed in the nucleation stage, further growth is restricted by the lack of nutrients.44 Fast nucleation rates and short timespan often produce size focused nanocrystallites;43 the presence of toluene and tumbling crystallization have prevented otherwise growth by attachment.45 In contrast, in standard hydrothermal synthesis, the formation of a poorly ordered precipitate precedes crystallization, (Scheme 1) whose dissolution is a necessary step for nucleation.46 The dissolution and release of precursors overlap with the nucleation and crystal growth processes, and some precursors are spectators during the nucleation process, which may lead to a relatively low supersaturation ratio in the nucleation stage and result in a low nucleation rate contributing to low numbers of crystals. In the sequential growth step, further growth into bulky ones is readily fed by the dissolution of precursors.

image file: c8ce01752b-s1.tif
Scheme 1 Schematic illustration of the possible process for the formation of SAPO-34-N and SAPO-34-C zeolites.

SAPO-34-N and SAPO-34-C were tested as catalysts for the DTO reaction, which was performed on a fixed-bed reactor at 723 K with a DME WHSV of 2 h−1 g-DME g-cat.−1. The lifetime and selectivity of the samples are compared in Fig. 3, S11(ESI) and Table S2 (ESI). The nanosized SAPO-34-N catalyst exhibited a significantly prolonged catalysis lifetime and improved selectivity to light olefins compared with SAPO-34-C. Notably, the lifetime of sample SAPO-34-N with full conversion of methanol can reach up to 102 min, which is about five-fold improvement compared to that of SAPO-34-C (22 min). Furthermore, the SAPO-34-N catalyst shows a high selectivity to ethylene and propylene up to 86.8%, giving about 5% higher selectivity than the SAPO-34-C catalyst (81.9%). The enhanced propylene olefin selectivity is ascribed to the presence of more intermediate acid sites associated with isolated Si(4Al) sites in SAPO-34-N.6,16,33,47

image file: c8ce01752b-f3.tif
Fig. 3 Dimethyl ether conversion (a), total product selectivity (b), and selectivity over SAPO-34-C (c) and SAPO-34-N (d). Reaction conditions: 723 K, WHSV = 2 h−1 g-DME g-cat.−1.

The thermogravimetric analysis (TGA) shows that there is a larger amount of coke found on the spent SAPO-34-C catalyst than that from SAPO-34-N (Fig. S12, ESI), and the corresponding GC-MS data (Fig. S13, ESI) for extracts demonstrate that there are more polymethylbenzenes and polyaromatic compounds, such as hexamethylbenzene, polymethylnaphthalene, anthracene and pyrene, entrapped in spent SAPO-34-C than in the spent SAPO-34-N catalyst. These results, together with the enhanced diffusion properties of SAPO-34-N and the corresponding coke formation mechanism,8,9 suggest that the diffusion property improvement and the presence of more isolated Si(4Al) sites in SAPO-34-N have alleviated heavy coke precursor formation with respect to the SAPO-34-C catalyst.

In summary, we have shown that nanosized SAPO-34-N crystallites can be facilely prepared via a colloid-mediated crystallization pathway enabled by biphasic synthesis, with the merits of being porogen-free and having high throughput. Thanks to the significantly enhanced transport efficiency, retained structural integrity and more dispersed Si incorporation, the nanosized SAPO-34-N exhibits excellent performance in catalysing DTO reactions. This synthesis method is expected to be expanded for synthesis of other high-performance hierarchical SAPO zeolites. The influence of other parameters, such as the Si content, concentration, nature of the Si source, etc., on the catalytic properties of SAPOs will be further explored and reported later.

KZ is grateful for the financial support from the National Natural Science Foundation of China (21576082).

Conflicts of interest

There are no conflicts to declare.

Notes and references

  1. A. Corma, Chem. Rev., 1995, 95, 559–614 CrossRef CAS.
  2. V. Valtchev, G. Majano, S. Mintova and J. Pérez-Ramírez, Chem. Soc. Rev., 2013, 42, 263–290 RSC.
  3. S. Mintova, J.-P. Gilson and V. Valtchev, Nanoscale, 2013, 5, 6693–6703 RSC.
  4. J. Li, A. Corma and J. Yu, Chem. Soc. Rev., 2015, 44, 7112–7127 RSC.
  5. P. Tian, Y. Wei, M. Ye and Z. Liu, ACS Catal., 2015, 5, 1922–1938 CrossRef CAS.
  6. Y. Li, M. Zhang, D. Wang, F. Wei and Y. Wang, J. Catal., 2014, 311, 281–287 CrossRef CAS.
  7. C. Li, C. Paris, J. Martínez-Triguero, M. Boronat, M. Moliner and A. Corma, Nat. Catal., 2018, 1, 547–554 CrossRef.
  8. U. Olsbye, S. Svelle, M. Bjorgen, P. Beato, T. V. W. Janssens, F. Joensen, S. Bordiga and K. P. Lillerud, Angew. Chem., Int. Ed., 2012, 51, 5810–5831 CrossRef CAS PubMed.
  9. B. P. C. Hereijgers, F. Bleken, M. H. Nilsen, S. Svelle, K.-P. Lillerud, M. Bjørgen, B. M. Weckhuysen and U. Olsbye, J. Catal., 2009, 264, 77–87 CrossRef CAS.
  10. Z. Xie, Z. Liu, Y. Wang and Z. Jin, Natl. Sci. Rev., 2015, 2, 167–182 CrossRef CAS.
  11. J. Zhong, J. Han, Y. Wei, P. Tian, X. Guo, C. Song and Z. Liu, Catal. Sci. Technol., 2017, 7, 4905–4923 RSC.
  12. Q. Sun, Z. Xie and J. Yu, Natl. Sci. Rev., 2018, 5, 542–552 CrossRef.
  13. I. Yarulina, A. D. Chowdhury, F. Meirer, B. M. Weckhuysen and J. Gascon, Nat. Catal., 2018, 1, 398–411 CrossRef.
  14. M. Dusselier and M. E. Davis, Chem. Rev., 2018, 118, 5265–5329 CrossRef CAS PubMed.
  15. J. D. Rimer, A. Chawla and T. T. Le, Annu. Rev. Chem. Biomol. Eng., 2018, 9, 283–309 CrossRef PubMed.
  16. D. Chen, K. Moljord and A. Holmen, Microporous Mesoporous Mater., 2012, 164, 239–250 CrossRef CAS.
  17. M. Milina, S. Mitchell, P. Crivelli, D. Cooke and J. Pérez-Ramírez, Nat. Commun., 2014, 5, 3922 CrossRef.
  18. Q. Y. Qian, C. Vogt, M. Mokhtar, A. M. Asiri, S. A. Al-Thabaiti, S. N. Basahel, J. Ruiz-Martinez and B. M. Weckhuysen, ChemCatChem, 2014, 6, 3396–3408 CrossRef CAS.
  19. Q. Sun, N. Wang, G. Guo, X. Chen and J. Yu, J. Mater. Chem. A, 2015, 3, 19783–19789 RSC.
  20. L. Zhang and Y. Huang, J. Mater. Chem. A, 2015, 3, 4522–4529 RSC.
  21. A. D. Chowdhury, K. Houben, G. T. Whiting, M. Mokhtar, A. M. Asiri, S. A. Al-Thabaiti, S. N. Basahel, M. Baldus and B. M. Weckhuysen, Angew. Chem., Int. Ed., 2016, 55, 15840–15845 CrossRef CAS PubMed.
  22. T. Yue, W. Liu, L. Li, X. Zhao, K. Zhu, X. Zhou and W. Yang, J. Catal., 2018, 364, 308–327 CrossRef CAS.
  23. P. Wu, M. Yang, W. Zhang, S. Xu, P. Guo, P. Tian and Z. Liu, Chem. Commun., 2017, 53, 4985–4988 RSC.
  24. C. Wang, M. Yang, M. Li, S. Xu, Y. Yang, P. Tian and Z. Liu, Chem. Commun., 2016, 52, 6463–6466 RSC.
  25. Q. Sun, N. Wang, R. Bai, X. Chen and J. Yu, J. Mater. Chem. A, 2016, 4, 14978–14982 RSC.
  26. G. Guo, Q. Sun, N. Wang, R. Bai and J. Yu, Chem. Commun., 2018, 54, 3697–3700 RSC.
  27. Y. Qiao, M. Yang, B. Gao, L. Wang, P. Tian, S. Xu and Z. Liu, Chem. Commun., 2016, 52, 5718–5721 RSC.
  28. C. Wang, M. Yang, P. Tian, S. Xu, Y. Yang, D. Wang, Y. Yuan and Z. Liu, J. Mater. Chem. A, 2015, 3, 5608–5616 RSC.
  29. Q. Sun, N. Wang, R. Bai, G. Chen, Z. Shi, Y. Zou and J. Yu, ChemSusChem, 2018, 11, 3812–3820 CrossRef CAS PubMed.
  30. D. Jin, G. Ye, J. Zheng, W. Yang, K. Zhu, M.-O. Coppens and X. Zhou, ACS Catal., 2017, 7, 5887–5902 CrossRef CAS.
  31. A. Z. Varzaneh, J. Towfighi and S. Sahebdelfar, Microporous Mesoporous Mater., 2016, 236, 1–12 CrossRef CAS.
  32. Y. Hirota, K. Murata, S. Tanaka, N. Nishiyama, Y. Egashira and K. Ueyama, Mater. Chem. Phys., 2010, 123, 507–509 CrossRef CAS.
  33. Z. Li, J. Martinez-Triguero, J. Yu and A. Corma, J. Catal., 2015, 329, 379–388 CrossRef CAS.
  34. S. Lin, J. Li, R. Sharma, J. Yu and R. Xu, Top. Catal., 2010, 53, 1304–1310 CrossRef CAS.
  35. H. van Heyden, S. Mintova and T. Bein, Chem. Mater., 2008, 20, 2956–2963 CrossRef CAS.
  36. D. Verboekend, M. Milina and J. Perez-Ramirez, Chem. Mater., 2014, 26, 4552–4562 CrossRef CAS.
  37. J. Zheng, D. Jin, Z. Liu, K. Zhu, X. Zhou and W. Yuan, Ind. Eng. Chem. Res., 2018, 57, 548–558 CrossRef CAS.
  38. J. Zheng, J. Ding, D. Jin, G. Ye, K. Zhu, X. Zhou, W. Yang and W. Yuan, Chem. Commun., 2017, 53, 6132–6135 RSC.
  39. E. V. Sturm and H. Coelfen, Crystals, 2017, 7, 207 CrossRef.
  40. C. S. Blackwell and R. L. Patton, J. Phys. Chem., 1984, 88, 6135–6139 CrossRef CAS.
  41. G. Liu, P. Tian, Y. Zhang, J. Li, L. Xu, S. Meng and Z. Liu, Microporous Mesoporous Mater., 2008, 114, 416–423 CrossRef CAS.
  42. I. Halasz, B. Moden, A. Petushkov, J. Liang and M. Agarwal, J. Phys. Chem. C, 2015, 119, 24046–24055 CrossRef CAS.
  43. F. Wang, V. N. Richards, S. P. Shields and W. E. Buhro, Chem. Mater., 2014, 26, 5–21 CrossRef CAS.
  44. S. Mintova, N. H. Olson, V. Valtchev and T. Bein, Science, 1999, 283, 958–960 CrossRef CAS PubMed.
  45. M. Kumar, M. K. Choudhary and J. D. Rimer, Nat. Commun., 2018, 9, 2129–2138 CrossRef PubMed.
  46. J. Zheng, W. Zhang, Z. Liu, Q. Huo, K. Zhu, X. Zhou and W. Yuan, Microporous Mesoporous Mater., 2016, 225, 74–87 CrossRef CAS.
  47. Z. Li, J. Martinez-Triguero, P. Concepcion, J. Yu and A. Corma, Phys. Chem. Chem. Phys., 2013, 15, 14670–14680 RSC.


Electronic supplementary information (ESI) available: Experimental details and supporting data. See DOI: 10.1039/c8ce01752b

This journal is © The Royal Society of Chemistry 2019