1906
Chem. Mater. 2007, 19, 1906-1917
In situ Small-Angle and Wide-Angle X-ray Scattering Investigation on Nucleation and Crystal Growth of Nanosized Zeolite A Wei Fan,† Masaru Ogura,†,‡ Gopinathan Sankar,*,§ and Tatsuya Okubo*,† Department of Chemical System Engineering, The UniVersity of Tokyo, Tokyo 113-8656, Japan, and DaVy-Faraday Research Laboratory, The Royal Institution of Great Britain, London W1S 4BS, United Kingdom ReceiVed NoVember 28, 2006. ReVised Manuscript ReceiVed February 4, 2007
The entire sequence of crystallization processes, starting with the formation of precursor particles, proceeding through the nucleation stage, and finishing with complete transformation into Linde type A (LTA) zeolite nanocrystals and crystal growth, from a clear solution using tetramethylammonium cation with the composition 11.25:1.8:13.4:x:700 SiO2:Al2O3:(TMA)2O:NaOH:H2O (x ) 0.6, 0.9, 1.2, or 1.5) has been monitored by simultaneous in situ small-angle X-ray scattering (SAXS) and wide-angle X-ray scattering (WAXS). Primary units with a size of ca. 0.5 nm and precursor particles with a size of ca. 4.5 nm are formed during the crystallization. The precursor particles with a size of 4.5 nm play an important role during the nucleation process. The influences of Na+ on the crystallization are studied by varying the concentration of Na+ cations in the synthesis solution. It is found that the Na+ cations affect not only the nucleation process but also crystal growth process. When x < 1.5, the number density of the precursor particles increases with the concentration of Na+, which is due to the structure-making influence and charge-compensating role of Na+. However, when the concentration of Na+ is relatively high (x ) 1.5), the 6 nm sized particles in the synthesis solution are observed at the beginning of the synthesis, which is due to the salt-outing effect of Na+. On the other hand, the studies on the structure of the precursor particles suggest that they do not contain long-range order, but some medium-range order related to the crystalline LTA structure, which become ordered during the course of hydrothermal treatment.
Introduction Synthesis of novel porous crystals, in particular zeolites, has received considerable interest because of its wide applications in various fields, such as molecular-sieving separation, shape-selective catalysis, and the ion-exchange process.1,2 Although several attempts have been made so far to understand the formations of zeolites from basic aluminosilicate hydrogel under hydrothermal conditions, the crystallization mechanisms of zeolites have not yet been fully elucidated. A molecular description of zeolite nucleation and crystal growth is highly desired, because of the important implications for both the synthesis of new zeolites by rational approaches and the modification of zeolites with particular properties.3,4 However, because of the highly inhomogeneous nature of the gel formed during the synthesis and the interdependence of numerous synthesis parameters, the elucidation of the crystallization mechanisms of zeolites is still one of the most difficult challenges in this research field. * Corresponding author. E-mail:
[email protected] (T.O.);
[email protected] (G.S.). Tel: 81-3-5841-7348 (T.O.); 44-20-7409-2992 (G.S.). Fax: 81-3-5800-3806 (T.O.); 44-20-7629-3569 (G.S.). † The University of Tokyo. ‡ Present Address: Institute of Industrial Science, The University of Tokyo, Tokyo 153-8505, Japan. § The Royal Institution of Great Britain.
(1) Barrer, R. M. Hydrothermal Chemistry of Zeolites; Academic Press: London, 1982. (2) Breck, D. W. Zeolite Molecular SieVes; Wiley: London, 1982. (3) Cundy, C. S.; Cox, P. A. Microporous Mesoporous Mater. 2005, 82, 1-78. (4) Cundy, C. S.; Cox, P. A. Chem. ReV. 2003, 103, 663-701.
Recently, the crystallization of nanosized zeolites synthesized from a clear solution has been widely studied5-11 as a model system to understand the crystallization mechanisms of zeolites, because the presence of a limited number of welldefined amorphous precursor particles simplifies the interpretation of the results and decreases the probabilities for ambiguous conclusions. Several attempts have been made over the years, using a suite of techniques including dynamic light scattering, NMR and combined in situ small-angle X-ray scattering (SAXS) and wide-angle X-ray scattering (WAXS) to understand the nucleation and crystal growth processes associated with the formation of Si-ZSM-5.7,12,13 de Moor et al.7 have proposed an assembly scheme regarding the nucleation and crystal (5) Cheng, C. H.; Shantz, D. F. J. Phys. Chem. B 2005, 109, 1911619125. (6) Cheng, C. H.; Shantz, D. F. J. Phys. Chem. B 2005, 109, 1391213920. (7) de Moor, P.; Beelen, T. P. M.; Komanschek, B. U.; Beck, L. W.; Wagner, P.; Davis, M. E.; van Santen, R. A. Chem.sEur. J. 1999, 5, 2083-2088. (8) Mintova, S.; Olson, N. H.; Valtchev, V.; Bein, T. Science 1999, 283, 958-960. (9) Yang, S. Y.; Navrotsky, A.; Wesolowski, D. J.; Pople, J. A. Chem. Mater. 2004, 16, 210-219. (10) Yang, S. Y.; Navrotsky, A. Chem. Mater. 2004, 16, 3682-3687. (11) Watson, J. N.; Iton, L. E.; Keir, R. I.; Thomas, J. C.; Dowling, T. L.; White, J. W. J. Phys. Chem. B 1997, 101, 10094-10104. (12) Dokter, W. H.; Vangarderen, H. F.; Beelen, T. P. M.; Vansanten, R. A.; Bras, W. Angew. Chem., Int. Ed. 1995, 34, 73-75. (13) Davis, T. M.; Drews, T. O.; Ramanan, H.; He, C.; Dong, J. S.; Schnablegger, H.; Katsoulakis, M. A.; Kokkoli, E.; McCormick, A. V.; Penn, R. L.; Tsapatsis, M. Nat. Mater. 2006, 5, 400-408.
10.1021/cm062827j CCC: $37.00 © 2007 American Chemical Society Published on Web 03/23/2007
Nucleation and Crystal Growth of Nanosized Zeolite A
growth by investigating the crystallization of Si-ZSM-5 using in situ SAXS/WAXS. More recently, a study13 on the crystallization of Si-ZSM-5 at room temperature using SAXS and high-resolution transmission electron microscopy (HRTEM) has suggested that precursor particles evolve to zeolite crystals through several intermediate states that can contribute to an aggregative growth. In comparison with pure silicate zeolites, the crystallization mechanisms of aluminosilicate zeolites are more complex because of several equilibrium reactions among the various aluminosilicate species and the different physical and chemical properties of the nanosized precursor particles. Therefore, only limited studies have dealt with the crystallization mechanisms of aluminosilicate zeolites. Mintova et al.8 have reported the synthesis of nanosized aluminosilicate zeolite Linde type A (LTA) from a colloidal solution in the presence of tetramethylammonium cation (TMA+) and demonstrated the steps of generation and densification of precursor particles during the crystallization employing HRTEM. However, because the silicate and aluminosilicate intermediates formed during the crystallization are fragile and difficult to separate from the synthesis solution without changing their properties, an investigation into the complete sequence of crystallization events using in situ methods, starting from the formation of the precursor particles, proceeding through the nucleation process, and ending by the growth of aluminosilicate zeolites, is greatly needed. Furthermore, the structure of precursor particles formed during the crystallization is believed to be important for the determination of final product and influence on the properties of synthesized zeolites. Although the studies of Si-ZSM514-16 have suggested that the precursor particles formed during the crystallization contain building units of MFI, it is still difficult to clearly elucidate the structure of the precursor particles, especially in the case of aluminosilicate zeolites, because there are few techniques to characterize the disordered structures of precursor particles.3 In the present work, we monitor the formations and consumptions of precursor and crystal particles during the crystallization of nanosized LTA in the presence of TMA+ employing simultaneous in situ SAXS/WAXS. The influences of Na+ on the formation of precursor particles and the following crystallization are studied by varying the concentration of Na+ in the initial solutions. Furthermore, the structure of the precursor particles is investigated by the high-energy X-ray diffraction (HEXRD) technique, a method for obtaining the atomic arrangement in disordered materials that has been employed to study the structures of zeolite precursor particles by the authors.17 Experimental Section Synthesis. The LTA synthesis solutions with the composition 11.25:1.8:13.4:x:700 SiO2:Al2O3:(TMA)2O:NaOH:H2O were pre(14) (15) (16) (17)
Chang, C. D.; Bell, A. T. Catal. Lett. 1991, 8, 305-316. Schoeman, B. J.; Regev, O. Zeolites 1996, 17, 447-456. Burkett, S. L.; Davis, M. E. J. Phys. Chem. 1994, 98, 4647-4653. Wakihara, T.; Kohara, S.; Sankar, G.; Saito, S.; Sanchez-Sanchez, M.; Overweg, A. R.; Fan, W.; Ogura, M.; Okubo, T. Phys. Chem. Chem. Phys. 2006, 8, 224-227.
Chem. Mater., Vol. 19, No. 8, 2007 1907 pared according to previous studies,8,18 where x ) 0.6, 0.9, 1.2, or 1.5 by varying the concentration of NaOH in the initial synthesis solution. Instead of colloidal silica used in the previous paper,8,18 tetramethylammonium silicate solution was selected as a silicon source,19 because our preliminary investigations show that colloidal silica interferes with SAXS measurements. All in situ measurements were performed in a specially designed sample cell with an electric heater. For ex situ measurements, the clear solution was transferred to a sealed polypropylene bottle and heated at 100 °C. The crystalline nanoparticles were purified by repeated centrifugation (20 000 rpm, 30 min), followed by redispersion in distilled water using an ultrasonic bath several times until the pH value of the colloidal solution was ca. 8.0. The intermediate products, such as aluminosilicate precursor particles and zeolite crystals, were recovered according to the method developed by Kremer et al.20,21 A surfactant solution of 2.02 g of cetyltrimethylammonium bromide (CTMABr, Aldrich) dissolved in 108.00 g of ethanol was slowly added into the synthesis solutions after heating for different periods with stirring. The products were washed with ethanol and dried in a vacuum oven at room temperature for 24 h. Characterization. In situ SAXS/WAXS measurements were started as soon as the sample cell containing the synthesis solution was introduced into an electric heater. The in situ measurements were carried out at station 6.2 of the Synchrotron Radiation Source at Daresbury Laboratory, U.K., which operates at 2 GeV with a typical current between 150 and 220 mA. Both SAXS and WAXS data were collected at a wavelength of 1.4000 Å. For SAXS measurements, a long-distance camera with a sample-to-detector distance of 3.5 m and a short-distance camera with a sample-todetector distance of 1.3 m were used. The corresponding Q ranges are 2.30-0.09 and 7.0-0.2 nm-1 for the long- and short-distance cameras, respectively. SAXS and WAXS patterns were collected using RAPID detectors.22 The scattering from water was used as a background. The scattering of a standard sample (wet rat tail collagen) and the diffraction pattern of zeolite Na-A (Aldrich) were used to calibrate the patterns of SAXS and WAXS, respectively. Ex situ powder XRD patterns were collected using a Bruker AXS M03X-HF equipped with a Cu KR line at 40 kV and 40 mA at a scanning rate of 4°/min with a step of 0.02°. Infrared absorption spectra were measured by JASCO-410 FTIR with a KBr wafer technique and recorded with a resolution of 4 cm-1. TEM experiments were performed on a JEM-2000EX II electron microscope (JEOL, Japan) with an acceleration voltage of 200 kV. Liquid 29Si NMR spectra were recorded on an EX270 (JEOL, Japan) at an 29Si frequency of 53.7 MHz, a recycle time of 4.95 s, and an accumulation time of 2000. Teflon-ined NMR tubes were used to avoid contamination from the leaching of the glass tube due to the alkaline solution. 13C MAS NMR spectra were measured on a CMX-300 (JEOL, Japan) at a spinning rate of 3 kHz with a pulse length of 5 µs, a recycle time of 5 s, and an accumulation time of 1500. The chemical shift was referenced to tetramethylsilane. Thermogravimetric analysis (TGA) and differential thermal analysis (DTA) were performed on PU 4K (Rigaku) equipped with a (18) Mintova, S.; Fieres, B.; Bein, T. Stud. Surf. Sci. Catal. 2002, 142, 223-229. (19) Fan, W.; O’Brien, M.; Ogura, M.; Sanchez-Sanchez, M.; Martin, C.; Meneau, F.; Kurumada, K.; Sankar, G.; Okubo, T. Phys. Chem. Chem. Phys. 2006, 8, 1335-1339. (20) Kremer, S. P. B.; Kirschhock, C. E. A.; Tielen, M.; Collignon, F.; Grobet, P. J.; Jacobs, P. A.; Martens, J. A. AdV. Funct. Mater. 2002, 12, 286-292. (21) Kremer, S. P. B.; Kirschhock, C. E. A.; Jacobs, P. A.; Martens, J. A. C. R. Chim. 2005, 8, 379-390. (22) Helsby, W. I.; Berry, A.; Buksh, P. A.; Hall, C. J.; Lewis, R. A. Nucl. Instrum. Methods Phys. Res., Sect. A 2003, 510, 138-144.
1908 Chem. Mater., Vol. 19, No. 8, 2007
Fan et al.
quadrupole mass spectrometer (MS, Anelva M-QA200TS) using a mixted gas (10% O2/90% He). Inductively coupled plasma mass spectrometry (ICP-MS) was performed after the samples were dissolved in a 6 M KOH solution on an ICP-MS HP4500 (HewlettPackard). HEXRD measurements were carried out on a horizontal twoaxis diffractometer built at the BL04B2 beam station of the Japan Synchrotron Radiation Research Institute (SPring-8), Japan. A bent Si (220) crystal mounted on the monochromator stage fixed at a Bragg angle of 3° in the horizontal plane provides an incident photon energy of 61.63 keV (wavelength ) 0.2012 Å). The samples were pelletized and fixed on the sample stage. Qmax collected in this study was 25 Å-1. The collected data were subjected to wellestablished analysis procedures, including absorption, background, and the Compton scattering corrections, and then normalized to the Faber-Ziman total structure factor, S(Q). The total correlation function, T(r), is obtained from the Fourier transformation of S(Q) using eq 1 T(r) ) 4πFr +
2 π
∫
Qmax
Qmin
{Q[S(Q) - 1]sin(Qr)}dQ
(1)
where F is the total number density and r is the interatomic distance. Analysis of SAXS Data. SAXS is observed when inhomogeneities of electron density exist in samples. The scattering intensity is a function of the scattering angle θ, and can be written in eq 2 I(Q) ) φP(Q)S(Q)
(2)
where φ is the number density of the individual scatters in the sample, and Q is defined by Q ) 4π sin(θ)/λ (θ is the scattering angle and λ is the wavelength of the incident beam). P(Q), the form factor, describes the relation between the geometry of the scatters and their scattering. S(Q), the structure factor, indicates the correlation of the scatters. Using the Debye approximation, we describe the the form factor in eq 3
P(Q) ) IeNe
2
∫
L
0
sin(Qr) 2 r dr Qr
g(r)
∫
L
0
(3)
g(r)r2dr
where Ie is the scattering intensity of an electron, Ne is the number of electrons in one scatter, and g(r) is the electron density correlation function. From the mathematical formula, the form factors of scatters with various shapes can be calculated. The form factors of a homogeneous spherical particle with a radius R and cubic particle with a length a are given by the following formulas in eqs 4 and 5.23 Spherical particles:
[
]
sin(QR) - QRcos(QR)
P(Q) ) 3
(QR)3
2
(4)
Cubic particles:
(
P(Q) )
∫ ∫ 2π
0
2π
0
)
Qa sin θ cos φ Qa sin θ sin φ Qa cos θ 2 sin sin 2 2 2 sin θdθdφ (5) Qa sin θ cosφ Qa sin θ sin φ Qa cos θ 2 2 2 In our study, the size of the crystal particles was analyzed using the methods developed for analyzing spherical and cubic parsin
(23) Pedersen, J. S. AdV. Colloid Interface Sci. 1997, 70, 171-210.
Figure 1. SAXS patterns of a silicate solution (11.25:13.4:1.2:700 SiO2: (TMA)2O:NaOH:H2O) and an aluminate solution (1.8:13.4:1.2:700 Al2O3: (TMA)2O:NaOH:H2O) collected at 100 °C by a short-distance camera.
ticles.19,24 The sizes of the precursor particles were calculated by the Guinier relation.25
Results Nanoparticles in TMA Silicate Solution. The silicate solution before adding an aluminum source with composition 11.25:13.4:1.2:700 SiO2:(TMA)2O:NaOH:H2O was observed by in situ SAXS at 100 °C. Figure 1 shows the in situ SAXS pattern of the solution after heating it for 30 min. For a comparison, the in situ SAXS pattern of aluminate solution without silicon source with the composition 1.8:13.4:1.2: 700 Al2O3:(TMA)2O:NaOH:H2O is also shown in Figure 1. A typical scattering pattern for very small particles is observed from the silicate solution. The Guinier plot shows that the particle size is ca. 0.5 nm in diameter (see the Supporting Information, Figure S1). Large aggregates are not present in the solution. On the other hand, no scattering from any particles can be observed from the aluminate solution. 29Si NMR spectrum (see the Supporting Information, Figure S3) indicates that cubic-like siliceous double-4membered ring (D4R), which has been well-characterized by in both solution and crystalline solid state, for example, as its TMA salt, (TMA)8Si2065H2O26, is the dominant species in this TMA silicate solution.27 Therefore, the structure of the nanoparticles with a size of ca. 0.5 nm is assigned to the siliceous D4R. In the following part, the roles of the siliceous D4R in the crystallization process of LTA are studied after the addition of an aluminum source into the silicate solution. Particles Formed during the Crystallization. The synthesis solutions after adding aluminum source were studied by in situ SAXS/WAXS with a camera length of 1.3 m (Q ) 7.0-0.15 nm-1) at 100 °C. The time-resolved scattering patterns are shown in Figure 2. All of the SAXS data has already been converted from the raw data by carefully (24) de Moor, P.; Beelen, T. P. M.; van Santen, R. A.; Tsuji, K.; Davis, M. E. Chem. Mater. 1999, 11, 36-43. (25) Guinier, A.; Fournet, G. Small-Angle Scattering of X-rays; Wiley: New York, 1955. (26) Harris, R. K.; Knight, C. T. G. J. Mol. Struct. 1982, 78, 273-278. (27) McCormick, A. V.; Bell, A. T. Catal. ReV.-Sci. Eng. 1989, 31, 97127.
Nucleation and Crystal Growth of Nanosized Zeolite A
Chem. Mater., Vol. 19, No. 8, 2007 1909
Figure 2. Time-resolved SAXS patterns for the crystallization from the synthesis solutions with different concentrations of Na+ collected by a shortdistance camera: (a) x ) 1.5, (b) x ) 1.2, (c) x ) 0.9, and (d) x ) 0.6. Scattering types: II ) secondary particles; III ) crystal particles; BR ) Bragg reflection.
subtracting the scattering of water as a background. The Bragg reflection in high Q region (Q ) 5.4 nm-1 corresponding to (200) of LTA, marked as BR in Figure 2) is observed for all synthesis solutions after heating for different periods, indicating the formation of LTA crystals in the synthesis solutions. An increasing scattering marked as III in Figure 3b-d is simultaneously observed in low Q region (Q < 0.2 nm-1) when the Bragg reflection (Q ) 5.4 nm-1) appears. The increasing scattering is attributed to the scattering at the surface of the growing crystal particles. On the other hand, after heating the solution for ca. 15 min, we observe a broad hump marked as II in Figure 2 around Q ) 0.8 nm-1 when x ) 0.6, 0.9, or 1.2. Different from the above cases, when x ) 1.5, a strong scattering in low Q region (Q < 0.6 nm-1) is observed at the beginning of synthesis (Figure 3a), suggesting the existence of larger particles in this stage, whose size will be discussed below. For clarity, snapshots of the SAXS patterns at different reaction periods are shown in Figure 3. For each synthesis solution, the scattering curves are plotted as follows: (1) after heating for 2 min; (2) after heating for 10 min to observe the particles in the initial crystallization stage; (3) at the onset of long-range order formation determined by the appearance of the first signal of Bragg reflection; (4) after heating for 40 min. Independent of Na+ concentration, three different-sized particles are observed during the crystallization. The first one is the crystal particle (III in Figure 3) that corresponds to the increased scattering intensity in the low Q region (Q < 0.2 nm-1). The second one is the particle with a size of ca. 4.5 nm (II in Figure 3b-d) calculated by the Guinier
plot of the broad hump around Q ) 0.8 nm-1 (see the Supporting Information, Figure S2) when x ) 0.6, 0.9, or 1.2. The third is the particle with a size of ca. 0.5 nm (I in Figure 3) corresponding to the small hump around Q ) 4.4 nm-1, which is noted as primary unit. Interestingly, different from the scattering patterns of the solutions with relatively low concentration of Na+ (x ) 0.6, 0.9, or 1.2), a clear hump around Q ) 0.6 nm-1 is observed in the solution with x ) 1.5 at the beginning of the synthesis, suggesting that larger nanoparticles with a size of ca. 6 nm exist in the synthesis solution at this stage. Moreover, the existence of the 6 nm sized particles is observed only when x ) 1.5, indicating that the formation of this particle is dependent on the concentration of Na+. The scattering from the 6 nm sized particles is even stronger than that of the crystals formed later, as shown in Figure 3a, suggesting that the number density of the 6 nm sized particles is quite high. To understand the correlation of the different particle populations, we plotted the evolution of scattering intensity at two fixed Q values (Q ) 4.4 and 0.8 nm-1) corresponding to the 0.5 and 4.5 nm sized particles (6 nm sized particles only in the case of x ) 1.5), respectively, together with the intensities of the Bragg reflection in Figure 4. Independent of Na+ concentration, the intensity of 0.5 nm sized primary units decreases with the reaction time. For the synthesis solutions with relatively low concentration of Na+ (x ) 0.6, 0.9, or 1.2), the 4.5 nm sized particles are formed after being heated for ca. 15 min. The onset time of crystallization determined by the appearance of the Bragg reflection at Q ) 5.4 nm-1 is in agreement with the maximum scattering intensity of the 4.5 nm sized particles as shown in Figure
1910 Chem. Mater., Vol. 19, No. 8, 2007
Fan et al.
Figure 3. Selected SAXS patterns of the crystallization from the synthesis solutions with different concentrations of Na+ after being heated for various reaction periods, collected by a short-distance camera: (a) x ) 1.5, (b) x ) 1.2, (c) x ) 0.9, and (d) x ) 0.6. Scattering types: I ) primary units, II ) secondary particles; III ) crystal particles; BR ) Bragg reflection.
4b-d, indicating that the 4.5 nm sized particles transform to LTA crystals at this stage. For clarity, the 4.5 nm sized particle is noted as a secondary particle. The secondary particles are still amorphous, because no Bragg reflection is observed. For the synthesis solution with a relatively high concentration of Na+ (x ) 1.5), two peaks are observed at 15 and 22 min in the evolution of the scattering intensity at Q ) 0.6 nm-1, as shown in Figure 4a. It is clear that the second peak coincides with the onset time of crystallization seen by the appearance of the Bragg reflection. The second peak may be mainly due to the scattering at the secondary particles formed along with the dissolution of the 6 nm sized particles. Because of the overlap of the scattering patterns of different particles, unambiguous determination of the nature of the secondary particles is prevented. Furthermore, the Guinier plot shows that the size of the secondary particles slightly increases from 4.3 to 4.7 nm during the crystallization. Because both the number density of particles and the size of particle contribute to the resulting scattering intensity, when we study the change in the number density of the particles, the size of the secondary particles must be the same.25,28 It is found that when the secondary particle size is 4.5 nm, the scattering intensity from the secondary particles increases with the concentration of Na+, (28) Orthaber, D.; Bergmann, A.; Glatter, O. J. Appl. Crystallogr. 2000, 33, 218-225.
which indicates that the number density of the secondary particles in the solution increases with the Na+ concentration. Effects of the Na+ Concentration on the Induction Time. The Na+ concentration affects not only the formation of the secondary particles but also the induction time, which is determined by the time of the appearance of the Bragg reflection. The induction times for the crystallization from the synthesis solutions with different concentrations of Na+ at 90 and 100 °C are shown in Figure 6. For all synthesis solutions, the induction time at 100 °C is obviously shorter than that at 90 °C. For the synthesis solutions with relatively low concentration of Na+ (x ) 0.6, 0.9, or 1.2), the induction time slowly increases with the concentration of Na+, although the change is rather small. Contrary to the slow increase in the induction time with the Na+ concentration when x ) 0.6, 0.9, or 1.2, the induction time of the synthesis solution with x ) 1.5 is obviously longer. Especially when the reaction temperature is 90 °C, the induction time of the synthesis solution with x ) 1.5 is ca. 20 min longer than that of the synthesis solution with x ) 1.2. Crystal Growth. To study the influence of Na+ on the crystal growth, we studied the crystallization processes from the synthesis solutions with different concentrations of Na+ by in situ SAXS/WAXS with a camera length of 3.5 m (Q ) 2.3-0.09 nm-1) for SAXS at 100 °C. The time-resolved SAXS patterns and selected SAXS patterns for the crystallization from the synthesis solutions are shown in Figure 7.
Nucleation and Crystal Growth of Nanosized Zeolite A
Chem. Mater., Vol. 19, No. 8, 2007 1911
Figure 4. Time-dependent scattering intensity at fixed angles (Q ) 4.4 and 0.8 nm-1), corresponding to primary units (black dot) and secondary particles (white circle), together with the intensity of the Bragg reflection (Q ) 5.4 nm-1, solid line), for the crystallization from the synthesis solutions with different concentrations of Na+: (a) x ) 1.5, (b) x ) 1.2, (c) x ) 0.9, and (d) x ) 0.6.
Figure 5. Changes in the maximum scattering intensity of the secondary particles formed in the synthesis solutions with different concentrations of Na+.
The upper limit of the Q region (d spacing is ca. 2.73 nm) prevents us from observing the 0.5 nm sized primary units; however, because of the lower limit of the Q region (d spacing is ca. 69.8 nm), it is possible to observe the crystal particles that are larger than 50 nm by fitting the scattering patterns in the low Q region.19 Independent of the Na+ concentration, oscillatory patterns appear in the low Q region. These patterns are attributed to the scattering at the surface of formed crystals, because the increased scattering in the low Q region (Q < 0.2 nm-1) is simultaneously observed with the appearance of the Bragg reflection in WAXS (Figure 8). The presence of oscillatory patterns in SAXS data in the
Figure 6. Induction time of the crystallization from the synthesis solutions with different concentrations of Na+ at 90 and 100 °C.
low Q region is a distinct feature of the monodispersed particles due to the form factor of spherical particle and cubic particle.23,25 Previous studies on the crystallization process of nanosized Si-ZSM-5 crystals using ultra SAXS and SAXS also show the oscillatory patterns in the low Q region.7,29 The calculated SAXS patterns for noninteracting particles using the form factor of spherical particles with different particle size distributions suggest that the maxima and minima of the oscillatory patterns become less pronounced with increasing polydispersity of the particles and that their position shifts to smaller Q region.23,29 The appearance of (29) de Moor, P.; Beelen, T. P. M.; Komanschek, B. U.; Diat, O.; van Santen, R. A. J. Phys. Chem. B 1997, 101, 11077-11086.
1912 Chem. Mater., Vol. 19, No. 8, 2007
Fan et al.
Figure 7. Time-resolved scattering patterns and snapshots of the crystallization from the synthesis solutions with different concentrations of Na+ collected at 100 °C by a long-distance camera: (a, b) x ) 1.5, (c, d) x ) 1.2, and (e, f) x ) 0.9.
Figure 8. Selected WAXS patterns of the crystallization from the synthesis solutions with different concentrations of Na+ collected at 100 °C by a longdistance camera: (a) x ) 1.5, (b) x ) 1.2, and (c) x ) 0.9.
Figure 9. TEM images of the products formed in the synthesis solution with x ) 1.2 after heating for (a) 50 and (b) 120 min at 100 °C.
the oscillatory patterns in our study suggests that particle size distribution of nanosized LTA crystals is relatively narrow.19,30 Figure 9 shows the TEM images of the products recovered from the synthesis solution with x ) 1.2 after heating for 50 and 120 min at 100 °C. It is clear that the crystal particles
after heating for 50 min are spherical. After being heated for 120 min, monodispersed cubic particles with lengths of ca. 120 nm are observed. Therefore, when a crystal particle is smaller than 60 nm, the SAXS oscillatory patterns in the low Q region are fitted with the calculated patterns of spherical particles using eq 4, whereas they are fitted with the patterns of cubic particles using eq 5 when the crystal particle is larger than 60 nm. The determined size of the growing crystals is plotted in Figure 10. It is obvious that the size of crystals formed after heating for 120 min increases with the concentration of Na+. Ex situ measurements also support that the final size of the crystals (after heating for 24 h) increases with the concentration of Na+. Structure of the Secondary Particles. As mentioned above, the secondary particles play an important role in the crystallization of LTA. A method for recovering the nanoparticles from zeolite synthesis solutions using cationic surfactants has been developed by Kremer et al. 20,21. The method was applied here to study the structure of the precursor particles formed during the crystallization of LTA. (30) Fan, W.; Shirato, S.; Gao, F.; Ogura, M.; Okubo, T. Microporous Mesoporous Mater. 2006, 89, 227-234.
Nucleation and Crystal Growth of Nanosized Zeolite A
Figure 10. Changes in the size of the crystal particles formed from the synthesis solutions with different concentrations of Na+ at 100 °C.
Figure 11 shows the powder XRD patterns and FTIR spectra of the samples recovered from the synthesis solution with x ) 1.2 after heating for different reaction periods. Powder XRD patterns in Figure 11 clearly show the appearance of LTA crystals in the sample after heating for 40 min, which is in agreement with previous in situ SAXS/WAXS results shown in Figures 3 and 4, indicating that the crystals were successfully recovered from the solutions. The weak intensities of the diffraction peaks are due to the appearance of the surfactants in the recovered samples. To understand the structure evolution of the secondary particles, we have investigated the FTIR spectra of the recovered samples after heating for different periods. Flanigen and co-workers have shown that IR spectroscopy is sensitive to investigate the structural features of the zeolite framework.31 They identified signals occurring in the region of 500-650 cm-1 as being characteristic of a double ring structure, which makes it possible to observe the vibrational modes of the distinct structural units, such as D4R in LTA structure and 5R in MFI structure. As shown in Figure 11, the absorptions at 1050 cm-1 corresponding to an asymmetric stretching of the Si-O-Si bond and at 550 cm-1 specifically assigned to the D4R structure of the LTA framework are observed in all samples.31 It is of note that the isolated D4R also has a vibration in this region.32 Thus, it is reasonable to observe the vibration at 550 cm-1 even in the sample without heating. It is clear that the vibration at 550 cm-1 shifts to low frequencies and becomes stronger along with heating, which should be attributed to the transformation of isolated D4R to the D4R within LTA framework.33 It should be noted that the recovery procedure could alter the surface species of the secondary particles because of the interaction of surfactant with these species and the change of chemical environments.34 In situ measurements on the structure of the (31) Flanigen, E. M.; Khatami, H.; Szymansk.H. A. AdV. Chem. Ser. 1971, 101, 201-227. (32) Fyfe, C. A.; Fu, G. Y. J. Am. Chem. Soc. 1995, 117, 9709-9714. (33) Kirschhock, C. E. A.; Ravishankar, R.; Verspeurt, F.; Grobet, P. J.; Jacobs, P. A.; Martens, J. A. J. Phys. Chem. B 1999, 103, 49654971. (34) Kragten, D. D.; Fedeyko, J. M.; Sawant, K. R.; Rimer, J. D.; Vlachos, D. G.; Lobo, R. F.; Tsapatsis, M. J. Phys. Chem. B 2003, 107, 1000610016.
Chem. Mater., Vol. 19, No. 8, 2007 1913
secondary particles are highly preferred, whereas it seems to be very difficult to carry out these experiments and obtain reliable results because of the limitation of the current characterization techniques. Taking into account all these points, we used the recovery method to study crystallization mechanisms of zeolites and provided useful informationd for understanding the structure of the precursor particles. ICP measurement suggests that the Si:Al ratio of LTA crystals synthesized from the synthesis solution with x ) 1.2 and recovered by centrifugation is 1.57. The Si:Al ratios of the samples recovered from the synthesis solutions after heating for different periods slightly increase along with heating time. On the other hand, it is found that the asymmetric (ca. 1000 cm-1) and symmetric (ca. 700 cm-1) vibrations of the T-O-T bond, double-ring vibration (ca. 580 cm-1), and TO4 bending vibration (ca. 480 cm-1) shift to the high region along with heating, which also suggests that the Si:Al ratio increases (see the Supporting Information, Figure S4).31 To understand the detailed structure of the secondary particles formed during the crystallization process, we employed HEXRD, a method widely used to investigate the structure of disordered materials,35 to observe the samples recovered after heating for different periods. It is difficult to interpret HEXRD data obtained from in situ measurements, because the O-O scattering from water dominates the total correlation functions, and hence we report only the ex situ result. Figure 12 shows the total correlation functions of the sample recovered from the synthesis solution with x ) 1.2 after heating for 0 and 20 min and LTA crystals together with that of amorphous bulk silica.36 The first peak at 1.7 Å in T(r) is attributed to Si-O and Al-O (the measured Q range does not provide sufficient resolution to discriminate the Si-O and Al-O distances and thus provide an average value). Distinct features are seen at around 2.3, 2.7, and 3.1 Å that are assigned to Na-O, O-O, and Si-Si (Al) correlations, respectively. Comparison of the T(r) of the sample including the secondary particles with that of amorphous bulk silica between 3.5 and 6.0 Å clearly reveals the presence of differences in the medium-range order. It has been known that the silicate ring structure of amorphous bulk silica are smoothly distributed from a three-membered ring (3R) to a ten-membered ring (10R) with a center at a six-membered ring (6R).37 The different medium-range order between the secondary particles and amorphous bulk silica clearly indicates that there is a distinct ring distribution in the secondary particles. The T(r) of the secondary particles is similar to that of LTA crystals, suggesting that the secondary particles contain a similar medium-range order with the crystalline LTA structure. The work on the atomic architecture of amorphous zeolite precursors using HEXRD has demonstrated that the peak from 3.8 to 3.9 Å is mainly assigned to the correlation of second T-O (T ) Si or Al) in 4R, and the peak from 4.2 to 4.6 Å is mainly assigned to (35) Waseda, Y. The Structure of Non-Crystalline Materials; McGrawHill: New York, 1980. (36) Kohara, S.; Suzuya, K. Nucl. Instrum. Methods Phys. Res., Sect. B 2003, 199, 23-28. (37) Kohara, S.; Suzuya, K. J. Phys.: Condens. Matter 2005, 17, S77S86.
1914 Chem. Mater., Vol. 19, No. 8, 2007
Fan et al.
Figure 11. (a) XRD patterns and (b) FTIR spectra of the samples recovered from the synthesis solution with x ) 1.2 after heating for various reaction periods.
Figure 12. Total correlation functions, T(r), of LTA crystals, the samples before heating, after heating for 20 min, and amorphous bulk silica. 4R and 6R structures of LTA framework are shown with various distances between the atoms. Yellow ball, Si; pink ball, Al; red ball, O.
the correlation of second T-O in larger ring structures (6R and/or 8R).17 T(r) of the sample before heating shows a peak at 3.8 Å suggesting the presence of 4R and/or D4R in agreement with the IR spectra of Figure 11. Along with heating, peak shifts at both 3.8 and 4.4 Å are observed, which could be attributed to the ring structures changing from disorder to order (see the Supporting Information, Figure S5). Discussion Primary Units. So far, it has been reported38-40 that silicate in solutions containing tetraalkylammonium (TAA) hydroxide self-assembles to form SiO2 nanoparticles. Different from the silica particles formed in the solution with tetrapropylammonium cation (TPA+), whose size varies from 2 to 5 nm, the nanoparticles in the presence of TMA+ formed in our study are much smaller, ca. 0.5 nm in diameter. Furthermore, our 29Si NMR result suggests that the structure of the nanoparticles in the presence of TMA+ is siliceous D4R. Previous experimental26,27,41,42 and simulation43-45 studies on the TMA silicate solution indicate that TMA+ (38) Rimer, J. D.; Vlachos, D. G.; Lobo, R. F. J. Phys. Chem. B 2005, 109, 12762-12771. (39) Rimer, J. D.; Lobo, R. F.; Vlachos, D. G. Langmuir 2005, 21, 89608971. (40) Fedeyko, J. M.; Rimer, J. D.; Lobo, R. F.; Vlachos, D. G. J. Phys. Chem. B 2004, 108, 12271-12275.
cations directly associate with cagelike polyanions to form a protective shell of hydrophobic hydration that impedes hydrolysis of the central anion, D4R. Six TMA molecules surround the D4R. Each cation occupies a face of the D4R. The reasons for the changes in the size and structure of the nanoparticles are probably due to the difference in the pH between the two systems and different structure-directing effects between TMA+ and TPA+. The pH of the TMA silicate solution is ca. 13.8. However, the pH of the studied silicate solutions in the presence of TPA+ is between 10.0 and 11.0, depending on the compositions.38-40 After the addition of an aluminum source, primary units with a size of ca. 0.5 nm are observed during the whole crystallization process, which was also observed in the previous study.7 The size of the primary units coincides with the size of the siliceous D4R observed in the solution without an aluminum source. Incorporation of one Al atom in siliceous D4R has been reported in aqueous and methanolic (41) Kinrade, S. D.; Knight, C. T. G.; Pole, D. L.; Syvitski, R. T. Inorg. Chem. 1998, 37, 4278-4283. (42) Kinrade, S. D.; Knight, C. T. G.; Pole, D. L.; Syvitski, R. T. Inorg. Chem. 1998, 37, 4272-4277. (43) Caratzoulas, S.; Vlachos, D. G.; Tsapatsis, M. J. Phys. Chem. B 2005, 109, 10429-10434. (44) Caratzoulas, S.; Vlachos, D. G.; Tsapatsis, M. J. Am. Chem. Soc. 2006, 128, 596-606. (45) Caratzoulas, S.; Vlachos, D. G.; Tsapatsis, M. J. Am. Chem. Soc. 2006, 128, 16138-16147.
Nucleation and Crystal Growth of Nanosized Zeolite A
Chem. Mater., Vol. 19, No. 8, 2007 1915
aluminosilicate solution in the presence of TMA+ using 29Si NMR and 27Al NMR.46 The incorporation of Al atom into the silicate anions strongly depends on the compositions and pH of the solutions.46 In our study, the 29Si NMR and 27Al NMR spectra (not shown here) of the synthesis solutions before and after heating for different periods cannot provide reliable information on the formation of Al-containing D4R. In this stage, it is difficult to conclude if the primary units with a size of ca. 0.5 nm contain aluminum. The detailed structure of the primary units is under investigation. The scattering intensity of the primary units decreases along with the crystallization, as shown in Figure 4, indicating that the primary units probably decompose/dissolve during the crystallization and participate in the formation of secondary particles and crystal growth. Secondary Particles. The formation of the secondary particles is observed in the synthesis solutions after heating for ca. 10 min, as shown in Figures 3 and 4. Figure 5 shows that the scattering intensity of the secondary particles is dependent on the concentration of Na+, suggesting that the Na+ cations favor the formation of the secondary particles. The formation of secondary particles during the crystallization of aluminosilicate zeolites is not only observed in our study but has also been observed in the previous studies.8,47 When colloidal silica was used as a silicon source, 40-80 nm sized amorphous particles were formed after adding TMA+ in the colloidal silica solution. The amorphous particles are much larger than the one observed in our study, which suggests that the particles formed in the initial stage of crystallization are critically affected by the silicon source employed. HEXRD in Figure 12 shows more detailed information on the structure of the amorphous secondary particles observed in our study. In the total correlation functions of zeolite frameworks, the distances in the medium-range order between 3.5 and 6.0 Å provide the most important information for identifying the types of ring structures.17 A comparison of the total correlation function, T(r), of LTA with that of the samples including secondary particles between 3.5 and 6.0 Å suggests that the ring structures of the secondary particles become ordered during the course of the hydrothermal treatment. It should be noted that, because the primary units in size of 0.5 nm containing D4R may be recovered from the synthesis solutions together with the secondary particles, the HEXRD results provide the information of not only the secondary particles but also the primary units. Taking into account the similar structure between the secondary particles and the final crystals, as well as the changes in scattering intensity of the secondary particles during the crystallization as shown in Figure 4, it is reasonable to conclude that the secondary particles are precursor particles for the formation of LTA crystals. Roles of the Na+ Cations in the Crystallization. The inorganic and organic cations always play important roles
in zeolite syntheses.1 It has been understood that cations have great influences on the final crystal phase and the crystallization kinetics.27,48,49 Particularly, the cations have a structure-making or structure-breaking influence on water in aqueous solutions.1,3,4,49 It has been reported that inorganic cations are hydrated and have the ability to hold water molecules in charged clusters in aqueous solutions.49 The attractive force between the positively charged centers of inorganic cations and the oxygen atoms of water molecules enhances the tendency of water molecules to form a network. However, large cations do not fit well with the network and induce disorders among water molecules. Therefore, it has been claimed that small alkali metal cations, such as Li+ and Na+, are effective in producing ordered arrays of tetrahedral structures, which can then bond together to form nuclei for zeolite syntheses. At the same time, when the concentration of the inorganic cations is relatively high, gel phases are often formed at the beginning of the zeolite syntheses due to the ability of the cations to hold water molecules, which is called a salting-out effect.49 In our study, the particles with a size of 6 nm are formed at the beginning of the reaction in the synthesis solution with relatively high concentration of Na+ (x ) 1.5), as shown in Figures 2a and 3a. The formation of these particles is concluded as being attributable to the salting-out effect of Na+, because they have the ability to order water molecules around them. If so, these particles are not expected to act as the precursor particles during the crystallization, but probably as a kind of reservoir of amorphous aluminosilicate particles in this oversaturated system. Furthermore, the formation of the particles with a size of 6 nm decreases the supersaturation of the synthesis solution, which impedes the formation of the precursor particles. Therefore, a longer induction time might be observed, as shown in Figure 6. The Na+ cations also play an important role in the formation of precursor particles. Scattering patterns suggest that the number density of the precursor particles increases with Na+ concentration. The roles of Na+ on the formation of precursor particles are explained as structure-making influence and charge-compensating units for Al in the crystalline framework, because a large amount of Al in the framework cannot be compensated only by bulky TMA+ 1,2,30,49. On the other hand, a simulation study on the TMA silicate solution in the presence of Na+ indicates that Na+ quickly settles around the D4R and polarizes the local solvent. Therefore, the local H-bond network of water-water is disrupted by Na+. TMA+ relinquishes its position at the surface of the D4R. As a result, Na+ weakens the clathrate structure hosting the polysilicate anion (D4R) and thus facilitates further polymerization and growth.44 The observation can also be considered to explain the role of Na+ on the formation of precursor particles Furthermore, our results suggest that the main silicate species in the solution before adding aluminum source is D4R (see the Supporting Information, Figure S3). However,
(46) Mortlock, R. F.; Bell, A. T.; Radke, C. J. J. Phys. Chem. 1991, 95, 7847-7851. (47) Singh, P. S.; White, J. W. Phys. Chem. Chem. Phys. 1999, 1, 41314138.
(48) Mintova, S.; Valtchev, V.; Angelova, S.; Konstantinov, L. Zeolites 1997, 18, 269-273. (49) Gabelica, Z.; Blom, N.; Derouane, E. G. Appl. Catal. 1983, 5, 227248.
1916 Chem. Mater., Vol. 19, No. 8, 2007
Fan et al.
Table 1. Si/Al Ratios of LTA Crystals Synthesized with Different Concentrations of Na+ initial composition (SiO2:Al2O3:(TMA)2O:NaOH:H20) 11.25:1.8:13.4:1.5:700 11.25:1.8:13.4:1.2:700 11.25:1.8:13.4:0.9:700 10:2.53:5.30:0.6:570 (colloidal silica30)
Table 2. Number of TMA+ in Different Cages of LTA Crystals Synthesized with Different Concentrations of Na+ no. of TMA+ in cages
Si/Al(ICP) 1.57 1.60 1.63 1.23
concentration of
Na+ (x)
1.5 1.2 0.9 colloidal silica30
R cage
sodalite cage
0.8 0.9 1.0 0.5
1.0 1.0 1.0 1.0
the aluminosilicate species formed in the solution after adding an aluminum source are very complicated and difficult to monitor by SAXS because of the weak contrast (scattering length density, SLD) between the species and solvent, except for the relatively “big” one, D4R. Although the concentration of 4.5 nm particles increases with Na+, a clear decrease in D4R (0.5 nm particles) is not observed in Figure 5, which may be due to the rapid equilibrium reactions among the aluminosilicate species that cannot be observed from our SAXS measurements.50 It should be mentioned that our measurements show that there is no detectable change in pH with varying the molar ratio of NaOH in the initial composition, which could be due to the high molar ratio of TMAOH used in the synthesis solution as shown in the initial composition 11.25:1.8:13.4: x:700 SiO2:Al2O3:(TMA)2O:NaOH:H2O (x ) 0.6, 0.9,1.2, or 1.5). Compared with the effect of TMAOH on pH, that of NaOH is quite weak. Local Environments of TMA+. The previous studies have shown that it is possible to understand the crystallization mechanisms of zeolites by studying the effects of size, geometry, and chemical nature of the SDAs on the crystalline structures of zeolites.51,52 The possible sites of TMA+ in the LTA structure are the sodalite cage and the R cage. The effective diameter of the sodalite cage is 0.7 nm, which suggests that only one TMA+ (molecular size of 0.69 nm) can be occluded in the sodalite cage. Furthermore, because the effective diameter of 6R that is the largest ring of the sodalite cage is 0.28 nm, the TMA+ in the sodalite cage can neither enter nor leave this cage after the crystal formations. Thus, the TMA+ must be incorporated during the formation of the sodalite cage.30 To study the role of TMA+ in the crystallization process, the numbers of TMA+ in the different cages of the LTA crystals synthesized using different concentrations of Na+ are calculated from the weight loss of the TG-DTA experiments, because the location of TMA+ in the different cages can be understood from the their decomposition temperatures.30 The Si:Al ratios of synthesized LTA checked by ICP measurement are shown in Table 1. The number of TMA+ inside each cage is calculated and listed in Table 2. For contrast, the result using a lower concentration of TMA+ is also shown in this table from our previous work.30 Table 2 shows that the number of TMA+ in the sodalite cage is almost constant in all samples with different concentration of Na+ and TMA+. However, the number of TMA+ in the R cage is changed from 0.8 to 1 with decreasing concentration of Na+. In the previous study, the average
number of TMA+ in one R cage, 0.5, is much lower than that in other samples.30 The different local environments of TMA+ in the LTA structure probably suggest that the TMA+ in the sodalite cage plays the SDA role in the formation of LTA, but the TMA+ in the R cage together with Na+ plays a charge-compensating role for a large amount of framework Al. It is well-known that organic cations such as quaternary ammonium compounds can assist the crystallization of highsilica molecular sieves by influencing the rates of product formation as well as product selectivity.30 In the crystallization of Si-ZSM-5 using TPA+ as an SDA, the formation of inorganic-organic composite structures is initiated by overlap of the hydrophobic hydration spheres of the inorganic and organic components, with subsequent release of ordered water to establish favorable intermolecular van der Waals interaction.7,16 Previous study has claimed that TMA+ is hydrophilic, tetraethylammonium cation (TEA+) has mixed character, and TPA+ generates a hydrophobic feature around the propyl hydrocarbon chains.53 The observation suggests that it is difficult for TMA+ to form hydrophobic hydration spheres because of the hydrophilicity of TMA+. However, very recently, the simulation study on TMA silicate solution shows that the TMA+ has partial hydrophobic character, which facilitates the formation of a “continuous” shell of hydrophobic hydration to take on the role of the putative “protective shield”.44,45 Thus, the formation mechanism of the inorganic-organic TMA+ aluminosilicate precursor particles is complicated and debatable. Crystallization Scheme. On the basis of the above results, a crystallization scheme of nanosized LTA in the presence of TMA+ is discussed and shown in Figure 13. Under the hydrothermal conditions, when the concentration of Na+ is relatively low, D4R (pure silicate or aluminosilicate, it is not clear in this stage) with a size of 0.5 nm surrounded by TMA+ and other smaller aluminosilicate species aggregate, which results in the formation of the precursor particles in size of 4.5 nm. When the concentration of Na+ is relatively high, larger particle with a size of 6 nm are formed at the beginning of the synthesis because of the salt-outing effect of Na+. The formation of 6 nm sized particles decreases the supersaturation of the solution, resulting in a prolonged induction time and a slow formation of the precursor particles. The precursor particles with size of 4.5 nm transform into viable nuclei by the reorganization at the interface of the precursor particles with the occluded solution8,54,55or a self-assembly process as observed in the
(50) Swaddle, T. W. Coord. Chem. ReV. 2001, 219, 665-686. (51) Lobo, R. F.; Zones, S. I.; Davis, M. E. J. Inclusion Phenom. 1995, 21, 47-78. (52) Davis, M. E.; Lobo, R. F. Chem. Mater. 1992, 4, 756-768.
(53) Kubota, Y.; Helmkamp, M. M.; Zones, S. I.; Davis, M. E. Microporous Mater. 1996, 6, 213-229. (54) Valtchev, V. P.; Bozhilov, K. N. J. Am. Chem. Soc. 2005, 127, 1617116177.
Nucleation and Crystal Growth of Nanosized Zeolite A
Chem. Mater., Vol. 19, No. 8, 2007 1917
Figure 13. Proposed scheme of nanosized LTA in the presence of TMA+.
previous study on Si-ZSM-5.13 The Na+ cations favor the formation of precursor particles because of its structuremaking influence49 and charge-compensating role,49 as well as its effect on clathrate structure hosting the polysilicate.45 The growth of the viable nuclei to nanosized crystals might occur by the addition of the smaller aluminosilicate species on the crystal surface. Conclusions The results presented here show that the combination of in situ SAXS/WAXS and ex situ HEXRD, XRD, and FTIR measurements is a powerful method for elucidating the whole crystallization process of LTA on a wide range scale covering precursor to product particles. For the crystallization of LTA in the presence of TMA+, three particle populations are observed: (1) primary units with a size of ca. 0.5 nm; (2) precursor particles with a size of ca. 4.5 nm that does not contain long-range order but some medium-range order related to crystalline LTA structure; (3) nanosized LTA crystals. By varying the concentration of Na+ in the synthesis solutions, it is observed that a relatively high concentration of Na+ (x ) 1.5) favors the formation of larger amorphous particles with a size of ca. 6 nm at the beginning of synthesis. The formation of 6 nm sized particles is due to the salt(55) Valtchev, V. P.; Bozhilov, K. N. J. Phys. Chem. B 2004, 108, 1558715598.
outing effect of Na+, which results in a decrease in the supersaturation and prolonged induction time. The formation of precursor particles is affected by the structure-making influence and charge-compensating role of Na+, as well as its effect on clathrate structure hosting the polysilicate. Acknowledgment. T.O., M.O., and G.S. thank the Japan Society for the Promotion of Science for a Grant-in-Aid for Scientific Research. G.S. thanks the Royal Society for a UKJapan visiting fellowship. We thank Dr. C. Martin for the help on the SAXS measurements, Dr S. Kohara for the help on the HEXRD measurements, and Mr. H. Tsunakawa and Mr. K. Ibe for their technical assistance in TEM measurements. Also, we thank Prof. S.P. Elangovan for continuous support for this project. W.F. is also grateful for a grant from the 21st Century COE Program “Human-Friendly Materials based on Chemistry” from the Ministry of Education, Culture, Sports, Science, and Technology of Japan. G.S. thanks the Daresbury Laboratory for the provision of beam time and other in-house facilities. G.S. and T.O. thank SPring-8 for the provision of beam time. Supporting Information Available: SAXS patterns with Guinier plot fittings, 29Si NMR spectrum, changes in wavenumbers for various vibrations, scheme of medium-range-order changes. This material is available free of charge via the Internet at http://pubs.acs.org. CM062827J
