DOI: 10.1002/chem.201402887
Full Paper
& Zeolites
Atomic Force Microscopy of Novel Zeolitic Materials Prepared by Top-Down Synthesis and ADOR Mechanism** Rachel L. Smith,[a] Pavla Elisˇov,[b] Michal Mazur,[b] Martin P. Attfield,[a] Jirˇ Cˇejka,*[b] and Michael W. Anderson*[a]
Abstract: Top-down synthesis of 2D materials from a parent 3D zeolite with subsequent post-synthetic modification is an interesting method for synthesis of new materials. Assembly, disassembly, organisation, reassembly (ADOR) processes towards novel materials based on the zeolite UTL are now established. Herein, we present the first study of these materials by atomic force microscopy (AFM). AFM was used to monitor the ADOR process through observation of the changes in crystal surface and step height of the products. UTL surfaces were generally complex and contained grain boundaries and low-angle intergrowths, in addition to regular terraces. Hydrolysis of UTL to IPC-1P did not have adverse
Introduction Traditional syntheses of zeolites are based on trial-and-error approaches using bottom-up protocols.[1] They start from a reaction mixture (gel) and after hydrothermal treatment provide three-dimensional (3D) crystalline zeolites.[2, 3] In some cases, the reactions proceed via a two-dimensional (2D) zeolite intermediate, for example MCM-22P, which is subsequently transformed to regular 3D structures by calcination.[4–6] Numerous layered zeolitic materials have been reported in recent years, many based on swelling and pillaring methods[7–9] or from post-synthetic treatment.[10] It is now becoming increasingly attractive to study the transformation of 3D zeolites into 2D ones.[11]
effects on the surfaces as compared to UTL. The layers remained intact after intercalation and calcination forming novel materials IPC-2 and IPC-4. Measured step heights gave good correlation with the X-ray diffraction determined d200spacing in these materials. However, swelling gave rise to significant changes to the surface topography, with significantly less regular terrace shapes. The pillared material yielded the roughest surface with ill-defined surface features. The results support a mechanism for the majority of these materials in which the UTL layers remain intact during the ADOR process as opposed to dissolving and recrystallising during each step.
Recently, we have proposed and demonstrated another synthetic protocol starting from three-dimensional zeolite UTL (germanosilicate).[10, 12, 13] UTL zeolite was successfully hydrolysed to individual layers (IPC-1P) by removal of the bridging D4R units, thus preserving the structure of the UTL layers. Subsequently, the interlayer spaces were swollen and used as extended building units to prepare the pillared material IPC-1PI (Scheme 1; colour image available in Supporting Information as Figure S1). More interestingly, IPC-1P layers were stabilised with diethoxydimethylsilane and condensed to new zeolite IPC-2 (OKO topology) or condensed without stabilisation to another new zeolite IPC-4 (PCR topology, Scheme 1). All these materials form a family of materials having the same structure
[a] R. L. Smith, Dr. M. P. Attfield, Prof. M. W. Anderson School of Chemistry University of Manchester Oxford Road, Manchester, M13 9PL (UK) Fax: (+ 44) 0161-275-4598 E-mail: [email protected] [b] P. Elisˇov, M. Mazur, Prof. J. Cˇejka Department of Synthesis and Catalysis J. Heyrovsky´ Institute of Physical Chemistry Academy of Sciences of the Czech Republic, v.v.i. Dolejsˇkova 3, 182 23 Prague 8 (Czech Republic) Fax: (+ 420) 286582307 E-mail: [email protected] [**] ADOR = assembly, disassembly, organisation, reassembly Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/chem.201402887. Chem. Eur. J. 2014, 20, 10446 – 10450
Scheme 1. ADOR processes to novel materials from hydrolysis of zeolite UTL.
10446
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Paper of individual layers differing in their connectivity. This new approach for the synthesis of zeolites was denoted the ADOR strategy: assembly, disassembly, organisation, reassembly.[13] X-ray diffraction (XRD) studies have been invaluable in monitoring the changes in the interlayer spacing throughout these processes. However, XRD gives no information towards the surface structure of the layers after each modification. Atomic force microscopy (AFM) is a method that allows measurement of surface topography and height of surface features to extremely high resolution (0.1 nm vertical resolution). AFM has been used to study the surfaces of zeolites[14–21] and related materials[22, 23] for a number of years to elucidate information relating to their growth mechanisms. Herein, we present the first AFM observations of this family of materials in order to establish how each subsequent modification affects the structural integrity of the layers. We also aim to provide mechanistic insight into the ADOR mechanism including the important question as to whether the UTL layers remain intact or undergo a dissolution/recrystallisation process during the ADOR process.[11, 13]
Results and Discussion Germanosilicate UTL is the first example of a material used in a 3D to 2D zeolite transformation, a so-called top-down synthesis, with the preservation of the original character of the layers. The changes upon hydrolysis of UTL (IPC-1P) and the subsequent intercalation of the layers were monitored by Xray powder diffraction. The XRD patterns are shown in Figure 1
Figure 1. XRD pattern from 1-408 2q for UTL, lamellar IPC-1P, swollen IPC1SW, pillared IPC-1PI, IPC-2 and IPC-4.
and the d200-spacing values are listed in Table 1. The dominant UTL reflection corresponding to the (200) reflection (half of the unit cell along the a axis) was found at 6.158 2q. Upon hydrolysis to IPC-1P, this reflection shifted to 8.358 2q. It corresponds to a reduced d200-spacing of 1.06 nm as a result of the removal of the double-four-ring (D4R) bridges between layers. The individual layers in IPC-1P were recondensed to form new zeolite structures either by direct condensation or stabilisation.[11, 13] Depending on the chosen linkage, two novel zeolites were Chem. Eur. J. 2014, 20, 10446 – 10450
www.chemeurj.org
Table 1. Calculated d-spacing from the (200) reflection from XRD and the average step height found by AFM. Sample name
d-spacing [nm]
Step height [nm], 0.1 nm
UTL IPC-1P IPC-2 IPC-4 IPC-1SW IPC-1PI
1.44 1.06 1.14 0.90 3.47 3.72
1.4 1.0 1.2 1.0 3.9 not measurable
prepared: IPC-2 and IPC-4.[13] The original D4R units in UTL were replaced by new single-four-ring (S4R) units in IPC-2 and by single Si O Si bridges in IPC-4. The three zeolites are structurally related since they have the same UTL-like layers and differ only in the layer connectivities. The size of their channel systems decreases with the size of the linkage in order: UTL (14-12-ring) > IPC-2 (12-10-ring) > IPC-4 (10-8-ring). The changed size of the channel system in these novel zeolites is reflected in the decreased d200-spacing in the case of IPC-2 (1.14 nm) and IPC-4 (0.90 nm) as compared with UTL (1.44 nm). The interlayer space of IPC-1P can be modified by introduction of a long-organic chain surfactant during the swelling treatment.[10, 12] In the swollen IPC-1SW, the dominant interlayer peak shifted to a lower 2q value (2.548 2q) indicating an increase in the d200-spacing. The d200-spacing value increased from 1.06 nm (in IPC-1P) up to 3.47 nm. To keep the interlayer space permanently expanded, silica amorphous pillars were subsequently introduced.[12] In the pillared IPC-1PI the d200spacing increased to 3.72 nm indicating additional expansion of the interlayer region that is likely to result from the hightemperature soaking in tetraethylorthosilicate (TEOS) during the pillaring process.[12] Scanning electron microscopy (SEM) showed that the UTL crystals had a rectangular morphology with crystal sizes ranging from 20 to 60 mm (Figure 2 a). The effect of the modifications on the surfaces of the materials and the changes in interlayer distance were monitored by AFM. A summary of the observed step heights compared to the observed d200-spacing for all materials is given in Table 1. The UTL surfaces before hydrolysis were complex with a high degree of faulting (Figure 3 a, b). Square and rectangular terraces were both observed. Lowangle intergrowths and grain boundaries were evident on the surface and were common to all crystals investigated. The angles of the intergrowths were very shallow with typical measurements by AFM of 0.798, 1.848 and 2.778 (see Supporting Information for data, Figure S2). Grain boundaries were visible in a number of different directions and did not necessarily align with the principal crystallographic directions in the crystal. There was no evidence of growth across the grain boundaries. The average step height of the terraces was found to be 1.4 nm ( 0.1 nm). This corresponds to a UTL monolayer, that is, half the unit cell (2.9 nm) or the d200-spacing found by XRD (1.44 nm). Hydrolysis of UTL resulted in the layered IPC-1P material. SEM (Figure 2 b) showed the overall crystal morphology of UTL (Figure 2 a) was unaffected by the hydrolysis and there was no
10447
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Paper ure 3 a, b). The AFM images showed no signs of damage or cracks on the surface. The measured step height of the terraces reduced from 1.4 nm in UTL to 1.0 nm ( 0.1 nm) in IPC-1P. This agreed with the reduced d200-spacing from XRD after hydrolysis (1.06 nm). The D4Rs were successfully hydrolysed to give layers of FER profile that retained the morphology and surface topography of the parent UTL material. The measured step heights were uniform which indicated homogeneous hydrolysis throughout the crystal. Intercalation of diethoxydimethylsilane or octylamine into IPC-1P, followed by calcination, resulted in novel zeolites IPC-2 and IPC-4, respectively. This process did not further change the topography of the zeolite surface as, compared to IPC-1P and UTL, the characteristic square and rectangular terrace patterns, low angle defects and intergrowths were all still present (Figure 3 e–h). Reconnection of the layers by both methods did not affect the surface structure other than the respective step heights. The step height of IPC-4 (1.0 0.1 nm) was comparable to the d200-spacing value found by XRD (0.90 nm). In the case of IPC-2, the average step height was measured as 1.2 nm ( 0.1 nm). Again, this correlated well with the observed d200spacing (1.14 nm). SEM also showed that the morphology of the crystals were unchanged and undamaged by the intercalation step (Figure 2 c, d). The swollen material IPC-1SW had an entirely different surface topography (Figure 3 i, j). The terraces that were visible Figure 2. Scanning electron microscopy images of a) UTL; b) IPC-1P; c) IPCwere irregularly shaped as compared to the UTL terraces. This 2; d) IPC-4; e,f) IPC-1SW; g,h) IPC-1PI. could be explained by the severe conditions used for swelling, which was performed under high pH in the presence of hydroxide ions (25 % solution C16TMA-OH). These conditions can lead to partial dissolution of the crystals and thus to irregularity in the crystal shapes and surface features. Nevertheless, the basic conditions and presence of C16TMA + cation did not result in the formation of mesoporous phases (M41S type), which can sometimes be observed.[24] This was confirmed by the XRD pattern and sorption properties obtained from a calcined sample of the swollen IPC-1SW. Significant changes were also apparent in the SEM images of IPC-1SW (Figure 2 e, f). The crystal surfaces were observed to be uneven at 12 000x magnification (Figure 2 f). Many of the crystals were broken and larger crystals Figure 3. Vertical deflection atomic force microscopy images of a, b) UTL; c, d) IPC-1P; e, f) IPC-2; g, h) IPC-4; i, j) (circa 60 mm) were less common. IPC-1SW; k, l) IPC-1PI. The AFM images (Figure 3 i, j) showed that the surfaces were uneven as a result of the swelling. This is visible on the terramacroscopic damage to the (100) face. The surfaces of these ces in Figure 3 j and led to a change in height of the terrace of crystals (Figure 3 c, d) were comparable to the UTL surface (FigChem. Eur. J. 2014, 20, 10446 – 10450
www.chemeurj.org
10448
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Paper 11 nm across the white line and 23 nm across the blue line (see Supporting Information for data, Figure S3). This is an interesting effect and not one we have seen before. The surface displays contiguous nanoscopic features across the unevenness suggesting that some small parts within the structure are not swollen, circa 5–10 layers. We can conjecture either inhomogeneous swelling in a small region or, perhaps more likely, that incomplete hydrolysis of the D4Rs in UTL has prevented layer swelling in some isolated areas. This effect is on such a small scale that it would not affect the diffraction patterns or the other AFM images (Figure 3 a–h), in which the relative changes in height are much lower. This effect is magnified in IPC-1SW in which the increase in height is more significant. Step heights were measured with an average height of 3.9 nm ( 0.1 nm). This is a 0.4 nm deviation from the XRD d200-spacing (3.47 nm). However, it is worth noting that this material has a wider range of height values than the other materials, with measured heights ranging from 3.5 to 4 nm. This is possibly due to inhomogeneous distribution of the surfactant molecules giving a range of d-values. The large, irregular features across many of the crystals were found to have heights in the region of 200 nm. This could be a result of redeposition of amorphous silica on the surface after dissolution. Upon the subsequent pillaring treatment with TEOS, the surface of IPC-1PI became very rough and ill-defined (Figure 3 k, l), which made it impossible to measure individual terrace step heights. The surfaces also appeared rough by SEM (Figure 2 g, h). The rough surfaces developed due to silica deposition on the surface from the excess of TEOS used in the pillaring step.
Conclusion AFM was successfully utilised to observe the changes to the surfaces of the different IPC materials and the transformations were monitored by measurement of the step heights. In all cases, the average step heights of the materials found by AFM were closely linked to the d200-spacing found by XRD. AFM revealed that the layers remain intact and undamaged after hydrolysis, and subsequent intercalation, leading to IPC-2 and IPC-4. XRD showed an increased interlayer distance upon swelling of IPC-1P and the step height found by AFM was in the order of the calculated increase. However, it was evident from the SEM and AFM images that the swelling step had serious consequences on both the overall morphology and the surface topography of the crystals. The swelling step clearly induces some form of mild dissolution on the surfaces, as the terrace shapes for IPC-1SW are noticeably different. Hydroxide ions are known to dissolve zeolites[20, 26] and in this case some form of surface dissolution is likely to have occurred. The use of TEOS in the pillaring step resulted in silica formation on the surfaces of the IPC-1SW crystals, which further altered the crystal surface of IPC-1PI. Questions have previously been raised about the mechanism of the ADOR process[11, 13] concerning the possible dissolution and recrystallisation of the UTL layers during each step. In some materials, the 3D to 2D transformation is thought to occur through a solution-mediated recrystallisation Chem. Eur. J. 2014, 20, 10446 – 10450
www.chemeurj.org
mechanism.[11] The AFM evidence suggests that this is not the case for materials IPC-1P, IPC-2 and IPC-4 as the terrace shapes, low-angle intergrowths and grain-boundary features were very similar across these materials and to those observed on the surface of UTL. Although the morphological similarities cannot in themselves categorically rule out a dissolution/recrystallisation mechanism, taken in conjunction with chemical considerations, the weight of evidence is strongly in favour of ADOR. The key chemical considerations are that the zeolites are highly unlikely to dissolve and recrystallize under the lower temperature solution treatments used to form IPC-1P and IPC2. Further, materials IPC-2 and IPC-4 are very unlikely to recrystallise in the absence of a structure-directing agent.[1] Combination of these points supports the theory that in most of the cases the UTL layers remain intact during the ADOR process and simply reassemble in different ways depending on the conditions.
Experimental Section Synthesis of UTL Zeolite UTL was prepared following the procedures reported previously.[25] The reaction gel was prepared as followed: template (6R,10S)-6,10-dimethyl-5-azoniaspiro[4.5]decane (67.83 g) was dissolved in water (500 g) and the template was ion-exchanged into HO form using AG 1-X8 resin (Bio-Rad). Then, GeO2 (38.76 g) was dissolved in the solution of template and SiO2 (Cab-O-Sil M5, 44.5 g) was added. The mixture was stirred at room temperature for 30 min. The gel was charged into a 1000 mL Teflon-lined autoclave and heated at 175 8C for 7 days. The solid was filtered, washed with water and dried at 60 8C. Calcination was carried out in the flow of air at 550 8C for 8 h.
Preparation of layered IPC-1P by hydrolysis of calcined UTL Calcined UTL (1 g) was hydrolysed in CH3COOH (1 m, 150 mL) at 85 8C overnight. The product was then isolated by filtration, washed with water and dried at 60 8C. Hydrolysed product was denoted IPC-1P.
Intercalation of silylating agent into IPC-1P leading to IPC-2 IPC-1P (1 g) was intercalated with diethoxydimethylsilane (0.1 g) in HNO3 solution (1 m, 10 g) at 170 8C for 16 h. The product was isolated by filtration, washed with water, dried at 60 8C and calcined (750 8C, 8 h, 1 8C min 1). The calcined zeolite was denoted IPC-2.
Intercalation of octylamine into IPC-1P leading to IPC-4 IPC-1P (1 g) was intercalated with octylamine (65 g) at 75 8C for 16 h. The product was isolated by centrifugation and dried at room temperature. The solid was calcined at 750 8C for 6 h (2 8C min 1). The calcined zeolite was denoted IPC-4.
Swelling of IPC-1P IPC-1P (1 g) was added to a 25 % solution of C16TMA-OH (20 mL, prepared by ion-exchanging from chloride form). The slurry was stirred for 16 h at ambient temperature. The product, IPC-1SW, was separated by centrifugation, washed with water and dried at 60 8C.
10449
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Paper Pillaring treatment Pillaring was carried out with IPC-1SW (1 g) in tetraethylorthosilicate (TEOS, 50 mL). The mixture was stirred and heated at 85 8C under reflux for 16 h. The solid was isolated by centrifugation and dried at ambient temperature. Then, water was added to the dried powder (100 mL/1 g) and stirred overnight. The product was centrifuged again and dried at 60 8C. Final calcination was carried out in air at 650 8C for 10 h (2 8C min 1). The structure and crystallinity of zeolites were determined by X-ray powder diffraction using Bruker AXS D8 Advance diffractometer equipped with a graphite monochromator and a position-sensitive detector Vntec-1 using CuKa radiation in Bragg–Brentano geometry. Samples for scanning electron microscopy were dusted onto carbon tabs and gold-coated. Imaging was performed on an FEI Quanta 200 ESEM in secondary electron mode. A working distance of 11 mm and voltage 20 kV were used. Atomic force microscopy was performed in contact mode in air on a JPK Nanowizard II BioAFM mounted on an inverted Axiovert 200 MAT optical microscope. Silicon nitrite tips (Bruker probes NP-10, spring constant 0.58 N m 1) were used with a scan rate of 1–2 Hz. Data processing was performed on the JPK Data Processing software. Line-by-line fitting followed by terrace plane fitting were used to level the height images for cross section analysis. Example cross section measurements used for these materials are given in the Supporting Information (Figure S4).
Acknowledgements R.L.S., M.P.A., and M.W.A. would like to thank the EPSRC for funding. P.E., M.M. and J.Cˇ thank the Czech Science Foundation for financial support (Centre of Excellence - P106/12/G015). Keywords: assembly · atomic force microscopy · top-down synthesis · UTL · zeolites [1] [2] [3] [4] [5]
M. E. Davis, R. F. Lobo, Chem. Mater. 1992, 4, 756 – 768. C. S. Cundy, P. A. Cox, Chem. Rev. 2003, 103, 663 – 701. C. S. Cundy, P. A. Cox, Microporous Mesoporous Mater. 2005, 82, 1 – 78. W. J. Roth, J. Cˇejka, Catal. Sci. Technol. 2011, 1, 43 – 53. G. Bellussi, A. Carati, C. Rizzo, R. Millini, Catal. Sci. Technol. 2013, 3, 833 – 857.
Chem. Eur. J. 2014, 20, 10446 – 10450
www.chemeurj.org
[6] F. Solnea, O. Ramos, M. K. de Pietre, H. O. Pastore, RSC Adv. 2013, 3, 2084 – 2111. [7] A. Corma, V. Fornes, S. B. Pergher, T. L. M. Maesen, J. G. Buglass, Nature 1998, 396, 353 – 356. [8] W. J. Roth, C. T. Kresge, J. C. Vartuli, M. E. Leonowicz, A. S. Fung, S. B. McCullen, Stud. Surf. Sci. Catal. 1995, 94, 301 – 308. [9] A. Corma, U. Diaz, M. E. Domine, V. Forns, Angew. Chem. 2000, 112, 1559 – 1561; Angew. Chem. Int. Ed. 2000, 39, 1499 – 1501. [10] W. J. Roth, O. V. Shvets, M. Shamzhy, P. Chlubn, M. Kubu˚, P. Nachtigall, J. Cˇejka, J. Am. Chem. Soc. 2011, 133, 6130 – 6133. [11] W. J. Roth, P. Nachtigall, R. E. Morris, J. Cˇejka, Chem. Rev. 2014, 114, 4807 – 4837. [12] P. Chlubn, W. J. Roth, H. F. Greer, W. Zhou, O. Shvets, A. Zukal, J. Cˇejka, R. E. Morris, Chem. Mater. 2013, 25, 542 – 547. [13] W. J. Roth, P. Nachtigall, R. E. Morris, P. S. Wheatley, V. R. Seymour, A. E. Ashbrook, P. Chlubn, L. Grajciar, M. Polozˇij, A. Zukal, O. Shvets, J. Cˇejka, Nat. Chem. 2013, 5, 628 – 633. [14] M. W. Anderson, J. R. Agger, J. T. Thornton, N. Forsyth, Angew. Chem. 1996, 108, 1301 – 1304; Angew. Chem. Int. Ed. Engl. 1996, 35, 1210 – 1213. [15] J. R. Agger, N. Pervaiz, A. K. Cheetham, M. W. Anderson, J. Am. Chem. Soc. 1998, 120, 10754 – 10759. [16] J. R. Agger, N. Hanif, C. S. Cundy, A. P. Wade, S. Dennison, P. A. Rawlinson, M. W. Anderson, J. Am. Chem. Soc. 2003, 125, 830 – 839. [17] L. I. Meza, M. W. Anderson, J. R. Agger, Chem. Commun. 2007, 2473 – 2475. [18] L. I. Meza, M. W. Anderson, J. R. Agger, C. S. Cundy, C. B. Chong, R. J. Plaisted, J. Am. Chem. Soc. 2007, 129, 15192 – 15201. [19] R. Brent, M. W. Anderson, Angew. Chem. 2008, 120, 5407 – 5410; Angew. Chem. Int. Ed. 2008, 47, 5327 – 5330. [20] L. I. Meza, M. W. Anderson, B. Slater, J. R. Agger, PCCP 2008, 10, 5066 – 5076. [21] R. Brent, P. Cubillas, S. M. Stevens, K. E. Jelfs, A. Umemura, J. T. Gebbie, B. Slater, O. Terasaki, M. A. Holden, M. W. Anderson, J. Am. Chem. Soc. 2010, 132, 13858 – 13868. [22] P. Y. Moh, P. Cubillas, M. W. Anderson, M. P. Attfield, J. Am. Chem. Soc. 2011, 133, 13304 – 13307. [23] M. A. Holden, P. Cubillas, M. P. Attfield, J. T. Gebbie, M. W. Anderson, J. Am. Chem. Soc. 2012, 134, 13066 – 13073. [24] C. T. Kresge, W. J. Roth, Chem. Soc. Rev. 2013, 42, 3663 – 3670. [25] O. V. Shvets, N. Kasian, A. Zukal, J. Pinkas, J. Cˇejka, Chem. Mater. 2010, 22, 3482 – 3495. [26] A. Cˇizˇmek, L. Komunjer, B. Subotic´, M. Sˇiroki, S. Roncˇevic´, Zeolites 1991, 11, 258 – 264.
Received: April 1, 2014 Published online on July 7, 2014
10450
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Paper
& Zeolites
Atomic Force Microscopy of Novel Zeolitic Materials Prepared by Top-Down Synthesis and ADOR Mechanism** Rachel L. Smith,[a] Pavla Elisˇov,[b] Michal Mazur,[b] Martin P. Attfield,[a] Jirˇ Cˇejka,*[b] and Michael W. Anderson*[a]
Abstract: Top-down synthesis of 2D materials from a parent 3D zeolite with subsequent post-synthetic modification is an interesting method for synthesis of new materials. Assembly, disassembly, organisation, reassembly (ADOR) processes towards novel materials based on the zeolite UTL are now established. Herein, we present the first study of these materials by atomic force microscopy (AFM). AFM was used to monitor the ADOR process through observation of the changes in crystal surface and step height of the products. UTL surfaces were generally complex and contained grain boundaries and low-angle intergrowths, in addition to regular terraces. Hydrolysis of UTL to IPC-1P did not have adverse
Introduction Traditional syntheses of zeolites are based on trial-and-error approaches using bottom-up protocols.[1] They start from a reaction mixture (gel) and after hydrothermal treatment provide three-dimensional (3D) crystalline zeolites.[2, 3] In some cases, the reactions proceed via a two-dimensional (2D) zeolite intermediate, for example MCM-22P, which is subsequently transformed to regular 3D structures by calcination.[4–6] Numerous layered zeolitic materials have been reported in recent years, many based on swelling and pillaring methods[7–9] or from post-synthetic treatment.[10] It is now becoming increasingly attractive to study the transformation of 3D zeolites into 2D ones.[11]
effects on the surfaces as compared to UTL. The layers remained intact after intercalation and calcination forming novel materials IPC-2 and IPC-4. Measured step heights gave good correlation with the X-ray diffraction determined d200spacing in these materials. However, swelling gave rise to significant changes to the surface topography, with significantly less regular terrace shapes. The pillared material yielded the roughest surface with ill-defined surface features. The results support a mechanism for the majority of these materials in which the UTL layers remain intact during the ADOR process as opposed to dissolving and recrystallising during each step.
Recently, we have proposed and demonstrated another synthetic protocol starting from three-dimensional zeolite UTL (germanosilicate).[10, 12, 13] UTL zeolite was successfully hydrolysed to individual layers (IPC-1P) by removal of the bridging D4R units, thus preserving the structure of the UTL layers. Subsequently, the interlayer spaces were swollen and used as extended building units to prepare the pillared material IPC-1PI (Scheme 1; colour image available in Supporting Information as Figure S1). More interestingly, IPC-1P layers were stabilised with diethoxydimethylsilane and condensed to new zeolite IPC-2 (OKO topology) or condensed without stabilisation to another new zeolite IPC-4 (PCR topology, Scheme 1). All these materials form a family of materials having the same structure
[a] R. L. Smith, Dr. M. P. Attfield, Prof. M. W. Anderson School of Chemistry University of Manchester Oxford Road, Manchester, M13 9PL (UK) Fax: (+ 44) 0161-275-4598 E-mail: [email protected] [b] P. Elisˇov, M. Mazur, Prof. J. Cˇejka Department of Synthesis and Catalysis J. Heyrovsky´ Institute of Physical Chemistry Academy of Sciences of the Czech Republic, v.v.i. Dolejsˇkova 3, 182 23 Prague 8 (Czech Republic) Fax: (+ 420) 286582307 E-mail: [email protected] [**] ADOR = assembly, disassembly, organisation, reassembly Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/chem.201402887. Chem. Eur. J. 2014, 20, 10446 – 10450
Scheme 1. ADOR processes to novel materials from hydrolysis of zeolite UTL.
10446
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Paper of individual layers differing in their connectivity. This new approach for the synthesis of zeolites was denoted the ADOR strategy: assembly, disassembly, organisation, reassembly.[13] X-ray diffraction (XRD) studies have been invaluable in monitoring the changes in the interlayer spacing throughout these processes. However, XRD gives no information towards the surface structure of the layers after each modification. Atomic force microscopy (AFM) is a method that allows measurement of surface topography and height of surface features to extremely high resolution (0.1 nm vertical resolution). AFM has been used to study the surfaces of zeolites[14–21] and related materials[22, 23] for a number of years to elucidate information relating to their growth mechanisms. Herein, we present the first AFM observations of this family of materials in order to establish how each subsequent modification affects the structural integrity of the layers. We also aim to provide mechanistic insight into the ADOR mechanism including the important question as to whether the UTL layers remain intact or undergo a dissolution/recrystallisation process during the ADOR process.[11, 13]
Results and Discussion Germanosilicate UTL is the first example of a material used in a 3D to 2D zeolite transformation, a so-called top-down synthesis, with the preservation of the original character of the layers. The changes upon hydrolysis of UTL (IPC-1P) and the subsequent intercalation of the layers were monitored by Xray powder diffraction. The XRD patterns are shown in Figure 1
Figure 1. XRD pattern from 1-408 2q for UTL, lamellar IPC-1P, swollen IPC1SW, pillared IPC-1PI, IPC-2 and IPC-4.
and the d200-spacing values are listed in Table 1. The dominant UTL reflection corresponding to the (200) reflection (half of the unit cell along the a axis) was found at 6.158 2q. Upon hydrolysis to IPC-1P, this reflection shifted to 8.358 2q. It corresponds to a reduced d200-spacing of 1.06 nm as a result of the removal of the double-four-ring (D4R) bridges between layers. The individual layers in IPC-1P were recondensed to form new zeolite structures either by direct condensation or stabilisation.[11, 13] Depending on the chosen linkage, two novel zeolites were Chem. Eur. J. 2014, 20, 10446 – 10450
www.chemeurj.org
Table 1. Calculated d-spacing from the (200) reflection from XRD and the average step height found by AFM. Sample name
d-spacing [nm]
Step height [nm], 0.1 nm
UTL IPC-1P IPC-2 IPC-4 IPC-1SW IPC-1PI
1.44 1.06 1.14 0.90 3.47 3.72
1.4 1.0 1.2 1.0 3.9 not measurable
prepared: IPC-2 and IPC-4.[13] The original D4R units in UTL were replaced by new single-four-ring (S4R) units in IPC-2 and by single Si O Si bridges in IPC-4. The three zeolites are structurally related since they have the same UTL-like layers and differ only in the layer connectivities. The size of their channel systems decreases with the size of the linkage in order: UTL (14-12-ring) > IPC-2 (12-10-ring) > IPC-4 (10-8-ring). The changed size of the channel system in these novel zeolites is reflected in the decreased d200-spacing in the case of IPC-2 (1.14 nm) and IPC-4 (0.90 nm) as compared with UTL (1.44 nm). The interlayer space of IPC-1P can be modified by introduction of a long-organic chain surfactant during the swelling treatment.[10, 12] In the swollen IPC-1SW, the dominant interlayer peak shifted to a lower 2q value (2.548 2q) indicating an increase in the d200-spacing. The d200-spacing value increased from 1.06 nm (in IPC-1P) up to 3.47 nm. To keep the interlayer space permanently expanded, silica amorphous pillars were subsequently introduced.[12] In the pillared IPC-1PI the d200spacing increased to 3.72 nm indicating additional expansion of the interlayer region that is likely to result from the hightemperature soaking in tetraethylorthosilicate (TEOS) during the pillaring process.[12] Scanning electron microscopy (SEM) showed that the UTL crystals had a rectangular morphology with crystal sizes ranging from 20 to 60 mm (Figure 2 a). The effect of the modifications on the surfaces of the materials and the changes in interlayer distance were monitored by AFM. A summary of the observed step heights compared to the observed d200-spacing for all materials is given in Table 1. The UTL surfaces before hydrolysis were complex with a high degree of faulting (Figure 3 a, b). Square and rectangular terraces were both observed. Lowangle intergrowths and grain boundaries were evident on the surface and were common to all crystals investigated. The angles of the intergrowths were very shallow with typical measurements by AFM of 0.798, 1.848 and 2.778 (see Supporting Information for data, Figure S2). Grain boundaries were visible in a number of different directions and did not necessarily align with the principal crystallographic directions in the crystal. There was no evidence of growth across the grain boundaries. The average step height of the terraces was found to be 1.4 nm ( 0.1 nm). This corresponds to a UTL monolayer, that is, half the unit cell (2.9 nm) or the d200-spacing found by XRD (1.44 nm). Hydrolysis of UTL resulted in the layered IPC-1P material. SEM (Figure 2 b) showed the overall crystal morphology of UTL (Figure 2 a) was unaffected by the hydrolysis and there was no
10447
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Paper ure 3 a, b). The AFM images showed no signs of damage or cracks on the surface. The measured step height of the terraces reduced from 1.4 nm in UTL to 1.0 nm ( 0.1 nm) in IPC-1P. This agreed with the reduced d200-spacing from XRD after hydrolysis (1.06 nm). The D4Rs were successfully hydrolysed to give layers of FER profile that retained the morphology and surface topography of the parent UTL material. The measured step heights were uniform which indicated homogeneous hydrolysis throughout the crystal. Intercalation of diethoxydimethylsilane or octylamine into IPC-1P, followed by calcination, resulted in novel zeolites IPC-2 and IPC-4, respectively. This process did not further change the topography of the zeolite surface as, compared to IPC-1P and UTL, the characteristic square and rectangular terrace patterns, low angle defects and intergrowths were all still present (Figure 3 e–h). Reconnection of the layers by both methods did not affect the surface structure other than the respective step heights. The step height of IPC-4 (1.0 0.1 nm) was comparable to the d200-spacing value found by XRD (0.90 nm). In the case of IPC-2, the average step height was measured as 1.2 nm ( 0.1 nm). Again, this correlated well with the observed d200spacing (1.14 nm). SEM also showed that the morphology of the crystals were unchanged and undamaged by the intercalation step (Figure 2 c, d). The swollen material IPC-1SW had an entirely different surface topography (Figure 3 i, j). The terraces that were visible Figure 2. Scanning electron microscopy images of a) UTL; b) IPC-1P; c) IPCwere irregularly shaped as compared to the UTL terraces. This 2; d) IPC-4; e,f) IPC-1SW; g,h) IPC-1PI. could be explained by the severe conditions used for swelling, which was performed under high pH in the presence of hydroxide ions (25 % solution C16TMA-OH). These conditions can lead to partial dissolution of the crystals and thus to irregularity in the crystal shapes and surface features. Nevertheless, the basic conditions and presence of C16TMA + cation did not result in the formation of mesoporous phases (M41S type), which can sometimes be observed.[24] This was confirmed by the XRD pattern and sorption properties obtained from a calcined sample of the swollen IPC-1SW. Significant changes were also apparent in the SEM images of IPC-1SW (Figure 2 e, f). The crystal surfaces were observed to be uneven at 12 000x magnification (Figure 2 f). Many of the crystals were broken and larger crystals Figure 3. Vertical deflection atomic force microscopy images of a, b) UTL; c, d) IPC-1P; e, f) IPC-2; g, h) IPC-4; i, j) (circa 60 mm) were less common. IPC-1SW; k, l) IPC-1PI. The AFM images (Figure 3 i, j) showed that the surfaces were uneven as a result of the swelling. This is visible on the terramacroscopic damage to the (100) face. The surfaces of these ces in Figure 3 j and led to a change in height of the terrace of crystals (Figure 3 c, d) were comparable to the UTL surface (FigChem. Eur. J. 2014, 20, 10446 – 10450
www.chemeurj.org
10448
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Paper 11 nm across the white line and 23 nm across the blue line (see Supporting Information for data, Figure S3). This is an interesting effect and not one we have seen before. The surface displays contiguous nanoscopic features across the unevenness suggesting that some small parts within the structure are not swollen, circa 5–10 layers. We can conjecture either inhomogeneous swelling in a small region or, perhaps more likely, that incomplete hydrolysis of the D4Rs in UTL has prevented layer swelling in some isolated areas. This effect is on such a small scale that it would not affect the diffraction patterns or the other AFM images (Figure 3 a–h), in which the relative changes in height are much lower. This effect is magnified in IPC-1SW in which the increase in height is more significant. Step heights were measured with an average height of 3.9 nm ( 0.1 nm). This is a 0.4 nm deviation from the XRD d200-spacing (3.47 nm). However, it is worth noting that this material has a wider range of height values than the other materials, with measured heights ranging from 3.5 to 4 nm. This is possibly due to inhomogeneous distribution of the surfactant molecules giving a range of d-values. The large, irregular features across many of the crystals were found to have heights in the region of 200 nm. This could be a result of redeposition of amorphous silica on the surface after dissolution. Upon the subsequent pillaring treatment with TEOS, the surface of IPC-1PI became very rough and ill-defined (Figure 3 k, l), which made it impossible to measure individual terrace step heights. The surfaces also appeared rough by SEM (Figure 2 g, h). The rough surfaces developed due to silica deposition on the surface from the excess of TEOS used in the pillaring step.
Conclusion AFM was successfully utilised to observe the changes to the surfaces of the different IPC materials and the transformations were monitored by measurement of the step heights. In all cases, the average step heights of the materials found by AFM were closely linked to the d200-spacing found by XRD. AFM revealed that the layers remain intact and undamaged after hydrolysis, and subsequent intercalation, leading to IPC-2 and IPC-4. XRD showed an increased interlayer distance upon swelling of IPC-1P and the step height found by AFM was in the order of the calculated increase. However, it was evident from the SEM and AFM images that the swelling step had serious consequences on both the overall morphology and the surface topography of the crystals. The swelling step clearly induces some form of mild dissolution on the surfaces, as the terrace shapes for IPC-1SW are noticeably different. Hydroxide ions are known to dissolve zeolites[20, 26] and in this case some form of surface dissolution is likely to have occurred. The use of TEOS in the pillaring step resulted in silica formation on the surfaces of the IPC-1SW crystals, which further altered the crystal surface of IPC-1PI. Questions have previously been raised about the mechanism of the ADOR process[11, 13] concerning the possible dissolution and recrystallisation of the UTL layers during each step. In some materials, the 3D to 2D transformation is thought to occur through a solution-mediated recrystallisation Chem. Eur. J. 2014, 20, 10446 – 10450
www.chemeurj.org
mechanism.[11] The AFM evidence suggests that this is not the case for materials IPC-1P, IPC-2 and IPC-4 as the terrace shapes, low-angle intergrowths and grain-boundary features were very similar across these materials and to those observed on the surface of UTL. Although the morphological similarities cannot in themselves categorically rule out a dissolution/recrystallisation mechanism, taken in conjunction with chemical considerations, the weight of evidence is strongly in favour of ADOR. The key chemical considerations are that the zeolites are highly unlikely to dissolve and recrystallize under the lower temperature solution treatments used to form IPC-1P and IPC2. Further, materials IPC-2 and IPC-4 are very unlikely to recrystallise in the absence of a structure-directing agent.[1] Combination of these points supports the theory that in most of the cases the UTL layers remain intact during the ADOR process and simply reassemble in different ways depending on the conditions.
Experimental Section Synthesis of UTL Zeolite UTL was prepared following the procedures reported previously.[25] The reaction gel was prepared as followed: template (6R,10S)-6,10-dimethyl-5-azoniaspiro[4.5]decane (67.83 g) was dissolved in water (500 g) and the template was ion-exchanged into HO form using AG 1-X8 resin (Bio-Rad). Then, GeO2 (38.76 g) was dissolved in the solution of template and SiO2 (Cab-O-Sil M5, 44.5 g) was added. The mixture was stirred at room temperature for 30 min. The gel was charged into a 1000 mL Teflon-lined autoclave and heated at 175 8C for 7 days. The solid was filtered, washed with water and dried at 60 8C. Calcination was carried out in the flow of air at 550 8C for 8 h.
Preparation of layered IPC-1P by hydrolysis of calcined UTL Calcined UTL (1 g) was hydrolysed in CH3COOH (1 m, 150 mL) at 85 8C overnight. The product was then isolated by filtration, washed with water and dried at 60 8C. Hydrolysed product was denoted IPC-1P.
Intercalation of silylating agent into IPC-1P leading to IPC-2 IPC-1P (1 g) was intercalated with diethoxydimethylsilane (0.1 g) in HNO3 solution (1 m, 10 g) at 170 8C for 16 h. The product was isolated by filtration, washed with water, dried at 60 8C and calcined (750 8C, 8 h, 1 8C min 1). The calcined zeolite was denoted IPC-2.
Intercalation of octylamine into IPC-1P leading to IPC-4 IPC-1P (1 g) was intercalated with octylamine (65 g) at 75 8C for 16 h. The product was isolated by centrifugation and dried at room temperature. The solid was calcined at 750 8C for 6 h (2 8C min 1). The calcined zeolite was denoted IPC-4.
Swelling of IPC-1P IPC-1P (1 g) was added to a 25 % solution of C16TMA-OH (20 mL, prepared by ion-exchanging from chloride form). The slurry was stirred for 16 h at ambient temperature. The product, IPC-1SW, was separated by centrifugation, washed with water and dried at 60 8C.
10449
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Full Paper Pillaring treatment Pillaring was carried out with IPC-1SW (1 g) in tetraethylorthosilicate (TEOS, 50 mL). The mixture was stirred and heated at 85 8C under reflux for 16 h. The solid was isolated by centrifugation and dried at ambient temperature. Then, water was added to the dried powder (100 mL/1 g) and stirred overnight. The product was centrifuged again and dried at 60 8C. Final calcination was carried out in air at 650 8C for 10 h (2 8C min 1). The structure and crystallinity of zeolites were determined by X-ray powder diffraction using Bruker AXS D8 Advance diffractometer equipped with a graphite monochromator and a position-sensitive detector Vntec-1 using CuKa radiation in Bragg–Brentano geometry. Samples for scanning electron microscopy were dusted onto carbon tabs and gold-coated. Imaging was performed on an FEI Quanta 200 ESEM in secondary electron mode. A working distance of 11 mm and voltage 20 kV were used. Atomic force microscopy was performed in contact mode in air on a JPK Nanowizard II BioAFM mounted on an inverted Axiovert 200 MAT optical microscope. Silicon nitrite tips (Bruker probes NP-10, spring constant 0.58 N m 1) were used with a scan rate of 1–2 Hz. Data processing was performed on the JPK Data Processing software. Line-by-line fitting followed by terrace plane fitting were used to level the height images for cross section analysis. Example cross section measurements used for these materials are given in the Supporting Information (Figure S4).
Acknowledgements R.L.S., M.P.A., and M.W.A. would like to thank the EPSRC for funding. P.E., M.M. and J.Cˇ thank the Czech Science Foundation for financial support (Centre of Excellence - P106/12/G015). Keywords: assembly · atomic force microscopy · top-down synthesis · UTL · zeolites [1] [2] [3] [4] [5]
M. E. Davis, R. F. Lobo, Chem. Mater. 1992, 4, 756 – 768. C. S. Cundy, P. A. Cox, Chem. Rev. 2003, 103, 663 – 701. C. S. Cundy, P. A. Cox, Microporous Mesoporous Mater. 2005, 82, 1 – 78. W. J. Roth, J. Cˇejka, Catal. Sci. Technol. 2011, 1, 43 – 53. G. Bellussi, A. Carati, C. Rizzo, R. Millini, Catal. Sci. Technol. 2013, 3, 833 – 857.
Chem. Eur. J. 2014, 20, 10446 – 10450
www.chemeurj.org
[6] F. Solnea, O. Ramos, M. K. de Pietre, H. O. Pastore, RSC Adv. 2013, 3, 2084 – 2111. [7] A. Corma, V. Fornes, S. B. Pergher, T. L. M. Maesen, J. G. Buglass, Nature 1998, 396, 353 – 356. [8] W. J. Roth, C. T. Kresge, J. C. Vartuli, M. E. Leonowicz, A. S. Fung, S. B. McCullen, Stud. Surf. Sci. Catal. 1995, 94, 301 – 308. [9] A. Corma, U. Diaz, M. E. Domine, V. Forns, Angew. Chem. 2000, 112, 1559 – 1561; Angew. Chem. Int. Ed. 2000, 39, 1499 – 1501. [10] W. J. Roth, O. V. Shvets, M. Shamzhy, P. Chlubn, M. Kubu˚, P. Nachtigall, J. Cˇejka, J. Am. Chem. Soc. 2011, 133, 6130 – 6133. [11] W. J. Roth, P. Nachtigall, R. E. Morris, J. Cˇejka, Chem. Rev. 2014, 114, 4807 – 4837. [12] P. Chlubn, W. J. Roth, H. F. Greer, W. Zhou, O. Shvets, A. Zukal, J. Cˇejka, R. E. Morris, Chem. Mater. 2013, 25, 542 – 547. [13] W. J. Roth, P. Nachtigall, R. E. Morris, P. S. Wheatley, V. R. Seymour, A. E. Ashbrook, P. Chlubn, L. Grajciar, M. Polozˇij, A. Zukal, O. Shvets, J. Cˇejka, Nat. Chem. 2013, 5, 628 – 633. [14] M. W. Anderson, J. R. Agger, J. T. Thornton, N. Forsyth, Angew. Chem. 1996, 108, 1301 – 1304; Angew. Chem. Int. Ed. Engl. 1996, 35, 1210 – 1213. [15] J. R. Agger, N. Pervaiz, A. K. Cheetham, M. W. Anderson, J. Am. Chem. Soc. 1998, 120, 10754 – 10759. [16] J. R. Agger, N. Hanif, C. S. Cundy, A. P. Wade, S. Dennison, P. A. Rawlinson, M. W. Anderson, J. Am. Chem. Soc. 2003, 125, 830 – 839. [17] L. I. Meza, M. W. Anderson, J. R. Agger, Chem. Commun. 2007, 2473 – 2475. [18] L. I. Meza, M. W. Anderson, J. R. Agger, C. S. Cundy, C. B. Chong, R. J. Plaisted, J. Am. Chem. Soc. 2007, 129, 15192 – 15201. [19] R. Brent, M. W. Anderson, Angew. Chem. 2008, 120, 5407 – 5410; Angew. Chem. Int. Ed. 2008, 47, 5327 – 5330. [20] L. I. Meza, M. W. Anderson, B. Slater, J. R. Agger, PCCP 2008, 10, 5066 – 5076. [21] R. Brent, P. Cubillas, S. M. Stevens, K. E. Jelfs, A. Umemura, J. T. Gebbie, B. Slater, O. Terasaki, M. A. Holden, M. W. Anderson, J. Am. Chem. Soc. 2010, 132, 13858 – 13868. [22] P. Y. Moh, P. Cubillas, M. W. Anderson, M. P. Attfield, J. Am. Chem. Soc. 2011, 133, 13304 – 13307. [23] M. A. Holden, P. Cubillas, M. P. Attfield, J. T. Gebbie, M. W. Anderson, J. Am. Chem. Soc. 2012, 134, 13066 – 13073. [24] C. T. Kresge, W. J. Roth, Chem. Soc. Rev. 2013, 42, 3663 – 3670. [25] O. V. Shvets, N. Kasian, A. Zukal, J. Pinkas, J. Cˇejka, Chem. Mater. 2010, 22, 3482 – 3495. [26] A. Cˇizˇmek, L. Komunjer, B. Subotic´, M. Sˇiroki, S. Roncˇevic´, Zeolites 1991, 11, 258 – 264.
Received: April 1, 2014 Published online on July 7, 2014
10450
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
