Read fyfe.PDF text version

THE RIGAKU JOURNAL VOL. 12 / NO.1 / 1995

ZEOLITE STRUCTURE ANALYSIS WITH POWDER X-RAY DIFFRACTION AND SOLID-STATE NMR TECHNIQUES

G. T. KOKOTAILO AND C. A. FYFE

Department of Chemistry, University of British Columbia, Vancouver, BC, CANADA Large advances in the development of catalytic materials were made with the synthesis of zeolites, which are porous crystalline materials that can be used as catalysts, catalyst supports, sorbents and ion exchangers. Zeolite structures consist of T-atoms tetrahedrally coordinated to oxygen atoms, where T can be AI, Si, or any other element capable of isomorphous substitution for Si. The unavailability of good quality, large single crystals makes the powder diffraction technique essential. Originally the powder diffraction patterns of these materials were used as fingerprints, but on careful analysis and with improved diffraction techniques a wealth of information may be derived from these data. The availability of rotating anode and synchrotron X-ray and neutron sources and the application of the Rietveld refinement method results in considerable improvement in the quality of structural information that can be derived. Further, the application of solid-state NMR techniques in conjunction with the X-ray diffraction method greatly increases the short-range-order information obtainable from these systems.

Introduction The importance of zeolites in the petroleum and chemical industries can scarcely be overestimated. The first application was the use of rareearth exchanged synthetic faujasite as cracking catalysts [1-3]. ZSM-5 catalysts selectively convert methanol to gasoline, and are used in distillate dewaxing, ethylbenzene synthesis, xylene isomerization, toluene disproportionation, as well as a host of other processes. The size and shape of the channel system and the nature and location of cations determine the absorption and rate of diffusion into the zeolite. The stability and catalytic properties are also functions of the structural characteristics. The understanding of the physical and catalytic properties of these materials is dependent on our knowledge of the structural features of the frameworks. These are also affected by crystallite size, faulting, twinning, and the nature of the cation sites. Solid-state NMR is a very sensitive probe of the local environment of a particular atom in the structure, while X-ray diffraction is sensitive to long-range order or the periodic structure of the framework. The two techniques complement each other and together provide more detailed information on the structure. There have been some very novel and exciting applications in these two areas and in the development of the solid-state NMR method. Vol. 12 No. 1 1995

Factors Affecting Characterization Amorphous and less stable components can be removed from zeolite samples by NaOH treatment [4]. In zeolite systems, perturbations in the framework structure, crystal morphology, extraframework material, phase purity, crystallite size, and the setting and occupation of cation sites can produce differences in the X-ray diffraction patterns. The first requirement is good, clean crystalline material that will yield very high-resolution patterns. This can be accomplished by the judicious choice of pH, temperature, recipe components, and their mixing and synthesis times. This results in a considerable improvement of the Xray diffraction pattern, intensity and resolution, which also increases the ability to characterize the zeolite. The pH used is dependent on the caustic stability of the sample for varying periods. This reaction is time-temperature related. The improvement in crystallinity of high-silica zeolite A sample treated for five minutes with 12.5% NaOH solution at room temperature is noted in Fig. 1. Similarly the X-ray diffraction pattern of a mordenite sample indicated the presence of zeolite Beta. A 15-minute treatment with a 12.5% NaOH solution at room temperature removed the Beta component completely as shown in Fig. 2. Some templates are difficult to remove from zeolites requiring high-temperature treatments that may lead to disorganization of the framework. The organic template can be decomposed at a low tem3

Fig. 1 X-ray diffraction patterns of high silica zeiolite A. (A) As synthesized. (B) Treated for five minutes with 12.5% NaOH solution at room temperature.

Fig. 2. X-ray diffraction patterns of synthetic mordenite (A) As synthesized with beta zeolite impurity. (B) Treated for 15 minutes with 12.5% NaOH solution at room temperature.

perature (-300°C) and the decomposition products expelled by high-temperature hydrothermal treatment. Cation exchangers may be added to the water. The zeolite is then calcined at a high temperature, preferably 500°C, to burn off any residue that has not been previously expelled [5]. Thus the organic material is removed without disrupting the framework with a corresponding improvement in the adsorptive and thermal properties. The X-ray diffraction pattern of a synthetic ferrierite shows considerable improvement in crystallinity from

Fig. 3. X-ray diffraction patterns of ferrierite, as synthesized with template (A) Calcined four hours at 500°C in air. (B) Calcined two hours at 350°C in air. Treated with water for two hours at 100°C Calcined four hours at 500°C in air.

this treatment (Fig. 3) as well as improved adsorptive properties [5]. Cation Effects Erionite and offretite are natural zeolites that can be converted to active catalysts by ion exchange. Most of the Na can be removed, but the K content cannot be readily decreased. The structures of erionite [6] and offretite [7] are related by

4

The Rigaku Journal

a variation in stacking order, which also blocks off the twelve-membered ring (12MR) channel in offretite as seen in Fig. 4. Singlecrystal analysis showed that the K+ ions were in the cancrinite cages which have only 6MR exit openings. In order to remove the K from these cages, a driving force was necessary that was provided by calcining the calcium-exchanged from at a temperature of at least 400°C. Although the calcination is time/temperature related, it is possible to cause the required migration of calcium into the cancrinite cages at a lower temperature. The calcium ions migrated into the cancrinite cages by forcing the K out into the large cages and were then removed by subsequent exchange [8]. This migration of ions may be tracked with X-

ray diffraction. The diffraction pattern of a naturally occurring erionite zeolite is shown in Fig. 5a. This same sample was ion-exchanged with calcium nitrate and calcined at 500°C. This process was repeated, resulting in a decrease in K content from 5.2% to 0.44% while the calcium content increased from 3.3% to 6.06%. The K content can be reduced to below 0.23% [8]. The normal hexane absorption was not affected, indicating that the structure had not been adversely affected. The X-ray diffraction pattern showed a shift in the lines (Fig. 5b) indicating a reduction of 0.28 Å, in the C parameter which can be accounted for by the difference in ionic radius of calcium, 1.97 Å compared to 2.31 Å for K. If the calcined calcium exchanged erionite is

Fig. 4 Views of erionite and offretite frameworks (A) Erionite. (B) Offretite.

Vol. 12 No. 1 1995

5

heated in water at 100°C for two hours, the X-ray pattern reverts to essentially its original form, as seen in Fig. 5c. X-ray diffraction can, thus, be used to track the interchange of K and Ca cations in the cancrinite cages. Similar cation migration to stable positions was noted on calcination of La exchanged zeolites X and Y (synthetic faujasite) [9]. In this case the La migrates into the sodalite cages, forming a stable complex, and contributes greatly to the hydrothermal stability of the catalyst. Simulation of Powder Patterns An old and useful approach to determining the framework topology of a zeolite is that of model building. Trial model building is not a random process: All available data, lattice parameters and symmetry information from diffraction studies, size of channels from diffusion studies, density, ring ellipticity from absorption and diffusion rates, nature of the channel system from sorption and diffusion, transmission electron microscopy lattice imaging, and MAS NMR are used together to build a model consistent with this information. Good crystalline samples that yield highresolution diffraction patterns and high quality MAS NMR are primary requirements. The coordinates of the individual atoms are adjusted so that the interatomic distances correspond as closely as possible to the predicted distances. A least-squares procedure that minimizes these distances was described by Meier and Villager [10]. This Distance Least Square (DLS) refinement gives idealized model structures using prescribed interatomic distances and unit-cell constants for a given space group, and establishes the positional parameters of the atoms. A Smith plot [11] of this data provides the simulated pattern that can now be compared with the experimental one. An "Rfactor", that is an estimate of the "goodness of fit," can be determined. Rietveld [12] developed a method of minimizing the difference between the observed powder-diffraction pattern and the one calculated from a model, which greatly increased the scope of powder diffraction in structure determination. By refining the fit of the powder pattern he was able to circumvent the problem of overlapping diffraction lines. Two groups have subsequently modified the method to accommodate more complex peak shapes [12, 13] and fur-

Fig. 5. X-ray diffraction patterns of erionite. (A) Natural erionite (B) Calcium exchanged and calcined. (C) Treated with water at 100°CC

ther facilitate refinement of powder patterns. Effect of Cations on Powder Patterns The structure of ferrierite was determined by Vaughn [15]. Differences in the powder patterns of samples from different locations and synthetic ferrierite suggested a cation effect. In order to resolve these discrepancies, simulated patterns were obtained for the ferrierite framework with the cations [16], Fig. 6a, the Mg complex in the structure of Fig. 6b, and the ferrierite framework with the Mg complex removed, Fig. 6c. The effect of the Mg complex Mg(H2O)6++ situated in the cavities with 8MR is obvious. Obtaining plots of structures with partial occupancy of cation sites may be used with powder data to resolve the occupancy of the sites. Combined Use of High-Resolution MAS NMR and X-Ray Powder Diffraction, ZSM-12 The catalytic and sorptive properties of a zeolite are determined by its topology or framework. Therefore precise information on both the shortand long-range order of the zeolite is essential.

6

The Rigaku Journal

Fig. 6 Simulated plots of ferrierite. (A) Framework and cations. (B) Mg complex. (C) Ferrierite framework with Mg complex removed.

The structure of ZSM-12, a high silica zeolite [17], was refined by the use of high-resolution solid state NMR and powder diffraction. These two techniques are sensitive to local and long-range order and therefore are complementary, and they resulted in the resolution of the structure. Highly crystalline ZSM-12 was hydrothermally synthesized by Kokotailo, Fyfe and coworkers [18] and dealuminated by steaming at a higher temperature [19]. A Rietveld refinement of the synchrotron powder data resulted in positional parameters that yielded the cell dimensions a=24.863, b=5.02, c=24.328 Å, and = 107.7° [20]. There were extra reflections in the synchrotron powder pattern (Fig. 7) that could not be indexed in the model proposed by Lapierre et al. [21] but could be with a double C parameter. SysVol. 12 No. 1 1995

tematic extinctions led to two possible space groups, Cc or C2/c, with 14 and 7 inequivalent sites resp. However, the 29Si MAS NMR spectrum of ZSM-12, Fig. 8 has 7 resonances of equal intensity, indicating 7 independent T-sites, and restricting the choice to the space group C2/c. In this case, the application of NMR and X-ray powder techniques resulted in the confirmation of the proposed structure. The high quality crystalline sample and the high-resolution NMR and Xray diffraction data revealed the pseudosymmetry, making refinements and the assignment of the space group possible. ZSM-11 The method of synthesis of ZSM-11, an end member of the pentasil family, developed by Chu [22] was modified by Kokotailo, Fyfe and co-

7

Fig. 7 Observed, calculated and difference X-ray powder diffraction patterns of ZSM-12.

Fig. 8

29

Si MAS NMR spectrum of ZSM-12.

workers [23] to yield a highly crystalline siliceous sample from which very highly resolved X-ray powder patterns and 1D and 2D 29Si MAS NMR were obtained. The structure originally proposed, based on model building and X-ray diffraction [24], had a tetragonal framework with space group 14m2. There are seven independent T atoms in the structure, five with an occupancy of 2, and two with an occupancy of 1. Further studies [25, 26] indicated

that the structure was temperature-dependent. Variable-temperature NMR, Fig. 9, shows that ZSM-11 undergoes a displacive phase transition in the temperature range 298-342 K. The hightemperature phase is tetragonal with space group I 4 m2. A full pattern Rietveld refinement of the powder data at 373 K, Fig. 10, confirmed the model [27]. The results of a 2D INADEQUATE experiment carried out at 340 K shows the complete assignment of connectivities. The room-temperature powder diffraction data could not be refined to match the 1D NMR data, and between 316 and 329 K the broadened NMR resonances indicate some distortion. The 1D NMR spectrum at 302 K consists of 12 resonances, indicating that the mirror plane is destroyed and the number of independent atoms is increased from 7 to 12. The asymmetric unit remains the same except for the number of independent T-atoms. The 2D-NMR contour plot of ZSM-11 at 303 K is shown in Fig. 12. There are two possible assignments, and subgroup/supergroup relations show that the low-temperature

8

The Rigaku Journal

structure has the I 4 space group. In this case, one- and two-dimensional NMR and powder diffraction have been successfully

used to resolve the temperature-dependent phase transformations of ZSM-11 and to assign their space groups. Summary A wealth of structural information may be derived from careful analysis of X-ray powder data in conjunction with solid state NMR. Good crystalline zeolite and zeolite catalyst samples are necessary and can be prepared to obtain high resolution powder diffraction and NMR spectra. The improvement in synthesis and treatment of zeolite samples, the sensitivity and resolution of powder diffraction and NMR have made this all possible. With the advent of high intensity rotating anode X-ray sources and high resolution diffractometers, an in-house source of powder data is available but with less resolution and sensitivity than a synchrotron source. References [1] K. M. Elliot and S. C. Eastwood, Proc. Am. Petrol.

Fig. 9 29Si MAS NMR spectra of zeolite ZSM-11 recorded at the temperatures indicated without any resolution enhancement.

Just., 43 (11), 272,1962. [2] D. H. Stormont, Oil and Gas J., April 1965. [3] US Patent 3,140,249 and 2,971,903. [4] US Patent 4,703,025.

Fig. 10 Synchrotron X-ray diffraction pattern of ZSM-11 together with the theoretical fit from Rietveld refinement and the difference pattern.

Vol. 12 No. 1 1995

9

Fig. 11 Contour plot of an INADEQUATE experiment on ZSM-11 at 340 K with the 1D MAS NMR spectrum on top.

Fig. 12 Contour plot of an INADEQUATE experiment on ZSM-11 at 303 K with a 1D MAS NMR spectrum on top.

[5] US Patent 4,187,283. [6] L. W. Staples and J. A. Gard, Mineral Mag., 32, 261, 1959. [7] J. M. Bennett and J. A. Gard, Nature, 214, 1005, 1967. [ 8] US Patent 3,640,680. [ 9] D. H. Olson, G. T. Kokotailo, and J. F. Charnell, J. Colloid, and Interface Science, 28 (2), 365, 1968. [10] W. M. Meier and H. Villager, Z. Krist., 129, 411, 1969. [11] D. K. Smith, "A Revised Program for Calculating Powder Diffraction Patterns", UCRL 50264, Lawrence Radiation Laboratory, University of California, Livermore, California, 1967 (Norelco Reporter 15,57,1968). [12] H. M. Rietveld, J. Appl. Cryst., 2, 65, 1969. [13] G. Malmros and J. O. Thomas, J. Appl. Cryst., 10, 7, 1977. [14] R. A. Young, P. E. Mackee, and R. B. van Greele, J. Appl. Cryst., 10, 262,1977. [15] P. A. Vaughn, Acta Cryst., 21, 983,1966. [16] G. T. Kokotailo and J. L. Schlenker, Advances in X-Ray Analysis; Editors, D. K. Smith, C. S. Barrett, D. E. Leyden, and P. K. Predicki, Plenum Press, N.Y., 24, 49, 1981.

[17] US Patent 3,832,449, 1974. [18] C. A. Fyfe, H. Strobl, G. T. Kokotailo, C. T. Pasztor, G. E. Barlow, and S. Bradley, Zeolites, 8, 132, 1988. [19] C. A. Fyfe, G. C. Gobbi, and G. J. Kennedy, J. Phys. Chem., 88, 3248,.1984. [20] C. A. Fyfe, H. Gies, G. T. Kokotailo, B. Marler, and D. E. Cox, J. Phys. Chem., 94, 3718, 1990. [21] R. B. Lapierre, A. C. Rohrman, J. L. Schlenker, J. D. Wood, M. K. Ruben, and W. J. Rohrbaugh, Zeolites, 5, 346, 1985. [22] US Patent 3,709,979, 1973. [23] C. A. Fyfe, Y. Feng, H. Grondy, G. T. Kokotailo, and A. Mar, J. Phys. Chem., 95, 3747, 1991. [24] G. T. Kokotailo, P. Chu, S. L. Lawton, and W. M. Meier, Nature, 275, 119, 1978. [25] B. H. Toby, M. H. Eddy, C. A. Fyfe, G. T. Kokotailo, H. Strobl, and D. E. Cox, J. Mater. Res., 3, 360, 1988. [26] C. A. Fyfe, H. Gies, G. T. Kokotailo, C. Pasztor, H. Strobl, and D. E. Cox, J. Am. Chem. Soc., 111, 2470, 1989. [27] H. Gies, B. Marler, C. A. Fyfe, G. T. Kokotailo, Y. Feng, and D. E. Cox, J. Phys. Chem. Solids, 52, 1235,1991.

10

The Rigaku Journal

Information

fyfe.PDF

8 pages

Find more like this

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

647289


You might also be interested in

BETA
fyfe.PDF
Microsoft Word - 1322Nafees.doc