Syntheses of Zeolite Beta Films in Fluoride Media and Investigation of

Feb 18, 2009 - Laboratoire de Matériaux a` Porosité Contrôlée, UMR-7016 CNRS, ENSCMu, UniVersité de Haute Alsace,. 3, rue Alfred Werner, 68093 ...
0 downloads 0 Views 781KB Size
Download PDF
Langmuir 2009, 25, 3549-3555

3549

Syntheses of Zeolite Beta Films in Fluoride Media and Investigation of Their Sorption Properties Alexandra Jakob,† Valentin Valtchev,*,#,† Michel Soulard,† and Delphine Faye‡ Laboratoire de Mate´riaux a` Porosite´ Controˆle´e, UMR-7016 CNRS, ENSCMu, UniVersite´ de Haute Alsace, 3, rue Alfred Werner, 68093 Mulhouse Cedex, France, and Laboratoire de Contamination, Centre National d’Etudes Spatiales, 18, aVenue Edouard Belin, 31401 Toulouse Cedex 9, France ReceiVed October 14, 2008. ReVised Manuscript ReceiVed January 12, 2009 Zeolite beta films were synthesized on glass slides and silicon wafers using the secondary growth method. The hydrothermal synthesis was performed in fluoride media with a wide variation of synthesis parameters. The kinetics of crystal growth under different conditions was studied, and film thicknesses ranging between 0.7 and 40 µm were obtained. A synthesis at 160 °C for 9 days was found to be the best compromise between the duration and the temperature of crystallization, providing films with thickness of 40 µm. Grazing incidence X-ray diffraction (GID) combined with scanning electron microscopy (SEM) analysis revealed the formation of highly crystalline, randomly oriented films. Sorption capacities and kinetics of the films and corresponding powders were studied toward several organic adsorbents (cyclohexane, n-hexane, p-xylene, trimethylbenzene, tetramethoxysilane). It was found that the adsorbed quantity depends both on the kinetic diameter and on the shape of the molecule, while the sorption capacity of the zeolitic films was estimated to be 7.14 mmol g-1 of zeolite by nitrogen sorption tests. Kinetics data showed that the adsorption was faster on zeolite powders compared to corresponding films, revealing the presence of additional diffusion barriers in the intergrown layers.

1. Introduction In space conditions, the outgassing of organic pollutants jeopardizes the performances of certain instruments by contaminating critical surfaces such as thermal radiators and optical systems. Thus, the molecular contamination control appears to be pointed as a key parameter for the optimization of on-orbit satellite functioning. The chemical nature of pollutants has been investigated by the National Aeronautic and Space Administration (NASA), and several types of outgassed molecules were identified, mainly deriving from plasticizers (phthalates, adipates, etc.).1 In addition, hydrocarbons emanating from elastomers, adhesives, and other compounds were also found in notable quantities.2,3 In order to minimize the molecular contamination, techniques involving preflight bakeouts were employed, but rapidly proved to be of limited efficiency, expensive, and time-consuming, as a certain number of outgassing cycles were required to decrease the molecular contamination.4 Therefore, the use of molecular adsorbents has been considered as an appropriate alternative to decrease the concentration of organic contaminants by retaining them into zeolite pores. Amorphous silica and alumina materials were first used, but they did not bring about a significant molecular contamination cutoff due to a lack of affinity toward the organic pollutants.5 Among the molecular adsorbents, the zeolites seem to be particularly appropriate. This family of materials exhibits an ordered microporous network that induces high sorption properties and an outstanding size and shape selectivity. Moreover, * Corresponding author. E-mail: [email protected]. † Universite´ de Haute Alsace. ‡ Centre National d’Etudes Spatiales. # Present address: LCS, University of Caen, France. (1) Chen, P.; Hedgeland, R.; Montoya, A. ; Roman-Velasquez, J.; Dunn, J.; Colony, J.; Petitto, J. 20th Space Simulation Conference: The Changing Testing Paradigm; NASA: Annapolis, MD, 1999. (2) Stein, C.; King, T. R.; Wilson, W. G.; Robertson, R. Proc. SPIE 1998, 3427, 56. (3) Hansen, P. A. Proceedings of the 7th International Symposium on Materials in a Space EnVironment; ESA-SP-399; ESA: Toulouse, France, 1997. (4) Thomson, S.; Hansen, P. A.; Chen, P Proc. SPIE 1996, 2864, 44. (5) Perry, J. L. NASA Tech. Memo. 1995, 108497.

zeolites present great thermal and mechanical stability and are relatively inexpensive. The first trials for on-orbit satellite decontamination were carried out during the development of the Wide Field Planetary Camera 2 (WFPC-2), which was designed to correct the optical aberration of the Hubble Space Telescope.4,6 The employed material was FAU-type, which achieved a 40% decrease in molecular contamination on optics.4 This encouraging result proved the relevancy of zeolite materials as molecular adsorbents. The choice of the zeolite structural type is also of critical influence to achieve a good retaining of organic pollutants. Zeolite beta presents a pore-size comparable to faujasite, but, in contrast to the latter, it is a highly hydrophobic material, hence exhibiting strong physical interactions with organic molecules. Several sorption tests were led on BEA-type powders, showing excellent affinity for linear alkanes and aromatics such as toluene.7-10 Stelzer et al. carried out competitive sorption tests on water/ toluene mixtures as a function of Si/Al ratio.10 It was shown that the quantity of adsorbed water decreased dramatically with the increase of Si/Al ratio in the zeolite beta framework. Another topic of interest deals with the shaping of zeolite materials for their use in space conditions. Unlike powder materials, zeolite films avoid the generation of dust particles in low earth orbit and can be easily embedded in appropriate areas during the mission.4 The on-orbit conditions demand mechanically stable, uniform, and defect-free films in order to reach high efficiency. These requirements may be reached by preparing zeolite films under hydrothermal conditions, as this route provides intergrown films strongly attached to the substrate, in contrast (6) Barengoltz, J.; Moore, S.; Soules, D.; Voecks, G. Jet Propul. Lab. Publ. 1994, 94–001. (7) Denayer, J. F.; Baron, G. V.; Martens, J. A.; Jacobs, P. A. J. Phys. Chem. B 1998, 102, 3077. (8) Reddy, K. S. N.; Eapen, M. J.; Soni, H. S.; Shiralkar, V. P. J. Phys. Chem. 1992, 96, 7923. (9) Lami-Bourgeat, E.; Fajula, F.; Anglerot, D.; Des Courieres, T. Microporous Mater. 1993, 1, 237. (10) Stelzer, J.; Paulus, M.; Hunger, M.; Weitkamp, J. Microporous Mesoporous Mater. 1998, 22, 1.

10.1021/la8033963 CCC: $40.75  2009 American Chemical Society Published on Web 02/18/2009

3550 Langmuir, Vol. 25, No. 6, 2009

to other traditional methods, such as dip- and spin-coating.11-13 Such films were already prepared for various applications, including low-k materials,14-16 sensing devices,17-19 and catalysis.20-22 The greater part of zeolite film syntheses are performed in alkaline media, which induces structural defects such as silanol groups at the crystal surface.23,24 In addition, the film thickness is generally limited to a few microns because of the small size of the crystals synthesized in highly supersaturated systems. Studies concerning the synthesis of BEA-type films in alkaline media showed that the film growth strongly depends on the hydrophobicity of the substrate, and a mechanism of the formation was proposed.25,26 In contrast to alkaline media, the fluoride route is expected to provide films with an enhanced control of their characteristics in term of thickness and defects. BEA-type crystals prepared from a near-neutral gel in fluoride media often exhibited sizes in the range of 10-15 µm and high hydrophobicity, due to a limited number of structural defects.24 Despite all these attractive properties, very few studies have been led concerning the synthesis of zeolite films in fluoride media.27-30 Tosheva et al. have hence compared the syntheses of BEA-type films on seeded silicon wafers in alkaline and fluoride media.27 The latter involved a near-neutral gel, prepared from tetraethylammonium fluoride (TEAF) as both the structuredirecting agent and the mineralizer. It has been shown that the film thickness depends strongly on the type of mineralizer, ranging from 200 nm in alkaline media to over 100 µm with the fluoride route. The films prepared with TEAF were firmly bonded to the substrates, but were fairly rough and had nonuniform thickness. Mitra et al. prepared zeolite beta films on stainless steel substrates, using hydrofluoric acid (HF) as the mineralizer.28 The obtained films were polycrystalline, continuous, 15 µm thick, and exhibited a low dielectric constant. These studies confirmed the applicability of the fluoride route for the synthesis of thick zeolite films, without providing details of the key parameters. The goal of the present work is to investigate the factors controlling the formation of BEA-type films in fluoride media using the seed-induced growth technique. Glass substrates were (11) Bonaccorsi, L.; Proverbio, E. Microporous Mesoporous Mater. 2004, 74, 221. (12) Wee, L. H.; Tosheva, L.; Vasilev, C.; Doyle, A. M. Microporous Mesoporous Mater. 2007, 103, 296. (13) Mintova, S.; Bein, T. AdV. Mater. 2001, 13, 1880. (14) Li, Z.; Johnson, M. C.; Sun, M.; Ryan, E. T.; Earl, D. J.; Maichen, W.; Martin, J. I.; Li, S.; Lew, C. M.; Wang, J.; Deem, M. W.; Davis, M. E.; Yan, Y. Angew. Chem., Int. Ed. 2006, 45, 6329. (15) Li, S.; Li, Z.; Medina, D.; Lew, C.; Yan, Y. Chem. Mater. 2005, 17, 1851. (16) Wang, Z.; Wang, H.; Mitra, A.; Huang, L.; Yan, Y. AdV. Mater. 2001, 13, 746. (17) Vilaseca, M.; Coronas, J.; Cirera, A.; Cornet, A.; Morante, J. R.; Santamaria, J. Sens. Actuators, B 2007, 124, 99. (18) Rauch, W. L.; Liu, M. J. AdV. Mater. 2003, 38, 4307. (19) Mintova, S.; Shangyi, M.; Bein, T. Chem. Mater. 2001, 13, 901. (20) Urbiztondo, M. A.; Valera, E.; Trifonov, T.; Alcubilla, R.; Irusta, S.; Pina, M. P.; Rodriguez, A.; Santamaria, J. J. Catal. 2007, 250, 190. ¨ hrman, O.; Hedlund, J.; Msimang, V.; Moeller, K. Microporous (21) O Mesoporous Mater. 2005, 78, 199. (22) Okumura, K.; Yoshino, K.; Kato, K.; Niwa, M. J. Phys. Chem. B 2005, 109, 12380. (23) Blasco, T.; Camblor, M. A.; Corma, A.; Esteve, P.; Guil, J. M.; Martinez, A.; Perdigon-Melon, J. A.; Valencia, S J. Phys. Chem. B 1998, 102, 75. (24) Caullet, P.; Paillaud, J. L.; Simon-Masseron, A.; Soulard, M.; Patarin, J. C. R. Chim. 2005, 8, 245. (25) Mies, M. J. M.; Rebrov, E. V.; Jansen, J. C.; de Croon, M.H.J.M.; Schouten, J. C. J. Catal. 2007, 247, 328. (26) Mies, M. J. M.; Rebrov, E. V.; Jansen, J. C.; de Croon, M.H.J.M.; Schouten, J. C. Microporous Mesoporous Mater. 2007, 106, 95. (27) Tosheva, L.; Hoelzl, M.; Metzger, T. H.; Valtchev, V.; Mintova, S.; Bein, T. Mater. Sci. Eng., C 2005, C25, 570. (28) Mitra, A.; Cao, T.; Wang, H.; Wang, Z.; Huang, L.; Li, S.; Li, Z.; Yan, Y. Ind. Eng. Chem. Res. 2004, 43, 2946. (29) Gualtieri, M. L.; Gualtieri, A. F.; Prudenziati, M. Microporous Mesoporous Mater. 2008, 111, 604. (30) Muraza, O.; Rebrov, E. V.; Chen, J.; Putkonen, M.; Niinisto¨, L.; de Croon, M. H. J. M.; Schouten, J. C. Chem. Eng. J. 2008, 135, S117.

Jakob et al.

chosen because of their high chemical affinity for the considered synthesis gels and light weight that favors their use in spacecraft devices. Moreover, since satellites comprise glass parts, it would be of great interest to use them as substrates rather than to incorporate new materials that would present additional weight. A comparative study of the sorption performances of films and powders in respect to different molecules such as aromatics and linear alkanes is also among the goals of the investigation.

2. Experiment 2.1. Materials. The substrates used in this study were commercial silica glass microscope slides (Marienfeld, 69.87 wt % SiO2, 12.71 wt % Na2O, 8.91 wt % CaO, 3.98 wt % MgO, 2.43 wt % N2O5, 0.82 wt % SnO2, 0.58 wt % Al2O3, 0.39 wt % K2O, 0.22 wt % SO3, and 0.12 wt % Fe2O3), cut into pieces measuring 20 mm × 10 mm × 1 mm. Monocrystalline (1 0 0) silicon wafers (Wacker) were also employed. The dissolution rate of glass substrates was determined by immersing them in a solution of molar composition 0.6 TEAOH/ 0.6 HF/10.5 H2O at 140 °C for 1, 5, and 9 days. The mass of substrates was measured before and after synthesis, hence leading to the dissolution rate. Dimensions of the substrates were 20 mm × 10 mm × 0.725 mm. The substrates were cleaned in an ultrasonic bath by successive 10 min immersions into acetone, ethanol, and distilled water. They were finally removed from the ultrasonic bath, rinsed with distilled water, and air-dried. 2.2. Seed Preparation. The seed crystals were prepared from a clear solution of molar composition 0.35 Na2O/9 TEAOH/0.5 Al2O3/ 25 SiO2/295 H2O.31 Two solutions (A and B) were first prepared. Solution A was obtained by dissolving freeze-dried colloidal silica (Ludox AS-40, Aldrich) into one-half of a tetraethylammonium hydroxide solution (TEAOH, 35% in H2O, Aldrich). Solution B was formed by hydrolyzing aluminum isopropoxide (98%, Aldrich) into the remainder of TEAOH solution. The two solutions were then mixed together and stirred at 500 rpm for 24 h, before being sealed in a polypropylene bottle and hydrothermally treated at 95 °C for 9 days. The resulting milky solution was purified by centrifugation at 20 000 rpm for 30 min and washed with deionized water. The operation was repeated four times. The pH of the colloidal suspension was then adjusted to 9 with 0.1 M NH3 solution. 2.3. Film Synthesis and Post-treatment. After the substrates had been cleaned, they were placed in a 1 wt % aqueous solution that was obtained by diluting a solution of cationic polymer (poly(diallyldimethylammonium choride), 20 wt % in H2O, Aldrich) so as to reverse their surface charge. Then the substrates were immersed in a zeolite beta colloidal suspension. Once the seeds had been adsorbed, the substrates were calcined at 600 °C for 6 h. The seeded substrates were treated in a gel of molar composition 0.6 TEAOH/1 SiO2/0.6 HF/10.5 H2O. A typical gel preparation started with the addition of fumed silica (99.9%, Cab-O-Sil) to TEAOH (35% in H2O, Aldrich) until complete dissolution of the silica. HF (40 wt %, Riedel de Hae¨n) was then added, and a viscous gel was formed. Finally, a vertically fixed seeded substrate was immersed into the gel and placed into a 20 mL autoclave, and hydrothermally treated for different periods of time at 100, 120, 140, 150, 160, 170, and 180 °C. After the synthesis, the so-obtained films were cleaned in an ultrasonic bath for 10 min to remove the loosely attached crystals, rinsed with distilled water, and air-dried. In order to remove the structure directing agent occluded in the zeolite pores, the films were calcined at 550 °C for 6 h under air atmosphere. 2.4. Characterization. The substrate composition was determined by X-ray fluorescence (Philips, MagiX). The size of the zeolite seeds was measured by dynamic light scattering (DLS, Malvern HPPS-ET). Morphological features of zeolite crystals and films prepared in fluoride media were investigated with a scanning electron microscope (SEM, Philips XL30 FEG). Powder samples were studied using a conventional X-ray diffractometer (PANanalytical, X’Pert (31) Mintova, S.; Valtchev, V.; Onfroy, T.; Marichal, C.; Knoezinger, H.; Bein, T. Microporous Mesoporous Mater. 2006, 90, 237.

Sorption Properties of Zeolite Beta Films in Fluoride

Langmuir, Vol. 25, No. 6, 2009 3551

Pro MPD), while films were characterized by grazing incidence X-ray diffraction (GID). The structural properties of the latter, i.e., the orientation of the crystallites in the film, were investigated by measuring the Bragg reflections with the reciprocal lattice vector parallel to the samples’ surface. Depth sensitivity in the growth direction is obtained by changing the incident angle (Ri) at the order of the critical angle (Rc) for total external reflection, typically some tenths of a degree. The incident angle was varied from 0.05° to 0.3° in order to change the scattering and the depth from about 4 to 700 nm. The accuracy in setting the incident angle was ∆Ri ) 0.005°. The wavelength used (k ) 0.154 nm) was selected by a double crystal monochromator from the undulator spectrum of beamline ID1 at the ESRF, in Grenoble. The footprint on the sample’s surface was about 0.5 mm × 0.5 mm, given by the cross-section of the parallel incident beam and the collimated beam behind the sample. Nitrogen sorption measurements were realized with an ASAP 2040 MP instrument. The specific surface was calculated with the Brunauer-Emmett-Teller (BET) equation while the microporous volume was separated from the mesoporous volume by the t-plot method. Sorption properties of organic molecules on the zeolite films and powders were studied using a SETARAM TGA 92 thermobalance under dynamic conditions. Placed in a platinum crucible, the zeolite samples were first heated up to 400 °C at a rate of 10 °C/min and maintained at this temperature for 2 h to achieve a complete dehydration. After cooling down to 25 °C, sorption of organic molecules was isothermally carried out under two nitrogen flows. The first was a reactive one circulating through a cell containing the studied organic liquid. The cell was maintained at 25 °C in a thermostatic bath, which allowed us to determine the equilibrium vapor pressure by the “cold point process”. On the other hand, the second nitrogen flow was kept inert and used as a carrier gas to dilute the reactive one. The P/P0 value, where P is the adsorbate pressure and P0 is the equilibrium vapor pressure, was hence set up. Thus, it was possible to isothermally study the quantity of organics that would adsorb on the given zeolite material as a function of time. In these experiments, the sorption properties of n-hexane (99.0 wt %, Fluka), cyclohexane (99.5 wt %, Aldrich), p-xylene (98 wt %, Fluka), trimethylbenzene (TMB, 98 wt %, Aldrich) and tetramethoxysilane (TMOS, 98 wt %, Fluka) were investigated on zeolite beta powder and films at a relative pressure of P/P0 ) 0.5, as this relative pressure involved similar nitrogen flows for carrier and reactive gases. The quantity of supported zeolite beta was estimated to be 40 mg by considering the density (d ) 1.53 g cm-3) of the zeolite and the quantity of nitrogen adsorbed at P/P0 ) 0.5. The density of the BEA-type material was calculated considering its framework density, which is 15.3 T/1000 Å3, with T representing a silica tetrahedron. The amount of supported BEA-type zeolite was calculated using a specific procedure. First of all, the quantity adsorbed by a given amount of BEA-type powder for a relative pressure of P/P0 ) 0.5 was determined using the t-plot method. The adsorbed volume was calculated using the following equation:

V)q×m

(1)

where V is the volume adsorbed by the BEA-type powder at a relative pressure of P/P0 ) 0.5 (cm3), q is the quantity adsorbed by the BEA-type powder at P/P0 ) 0.5 (cm3 · g-1), m is the weight of the tested material (g). Then, the quantity adsorbed by the corresponding zeolite films for a relative pressure of P/P0 ) 0.5 was determined using the t-plot method. The quantity adsorbed by the glass substrates was regarded as negligible, since they are not porous materials. Using eq 1, it was possible to calculate the volume adsorbed by the supported BEA-type material and, hence, to deduce its weight, considering the volume adsorbed by a given amount of the counterpart powder. A similar quantity of zeolite beta was used for corresponding sorption measurements on powder material.

3. Results and Discussion 3.1. Design of the Synthesis Gel. The advantages of fluoride media synthesis were listed in the introduction section. On the

Figure 1. XRD pattern (a) and particle size distribution measured by DLS (b) of BEA-type nanocrystals employed for seeding of the substrates.

Figure 2. Low (a) and high (b) magnification views of seeded glass substrate.

other hand, serious disadvantages of F--rich gels are high viscosity and low wetting ability that result in nonuniform film formation. These drawbacks in the application of F-containing gels in zeolite film growth require the design of an initial composition with relatively low viscosity able to provide large zeolite beta crystals. In order to obtain such a gel, preliminary experiments with variation of the silica source and water content were performed. First, tetraethoxysilane (TEOS) and fumed silica were tested as silica sources. All the gels prepared with TEOS were too dry to allow a uniform film formation, whereas fumed silica led to a gel that enabled the synthesis of extended zeolite layers. Further, the water content in the latter was adjusted in order to obtain pure silica zeolite beta, as larger amounts of water promote the formation of zeolite MFI. The final composition of the employed gel was set up to 0.6 TEAOH/1.0 SiO2/0.6 HF/10.5 H2O. 3.2. Effect of Seed Crystals. X-ray diffraction (XRD) was used to verify the cristallinity of seed crystals (Figure 1a) while their size was measured by DLS (Figure 1b). As can be seen, the seed crystals presented a standard BEA-type pattern with a slight peak broadening due to the small crystal size. The average size of the particles was estimated to be 105 nm, which make them suitable for seeding purposes. In order to avoid polymer decomposition at temperatures higher than 140 °C and to promote strong interactions with the substrate, the polymer agent was eliminated at 600 °C for 6 h. The SEM inspection showed that the glass substrate was uniformly covered with nanosized BEA-type crystals (Figure 2). Films obtained at 140 °C for 9 days on seeded and unseeded substrates were subjected to an XRD study (Figure 3). In addition to highly crystalline BEA-type material, fluorite CaF2 was detected in the films prepared on unseeded substrates (2θ ) 28.4° and 47.4°). The presence of fluorite is most probably due to partial dissolution in fluoride media of the glass substrate that contains non-negligible quantities of calcium oxide (ca. 8.9 wt %). Small quantities of zeolite MFI were also revealed by the weak peak at 2θ ) 23.2°. In contrast, seeded substrate provided a pure zeolite beta film. This result revealed the protective role of zeolite beta seeds, which induced fast film formation, thus limiting the dissolution of the substrate and formation of undesired phases, which was estimated to approximately 3.7 mg h-1 on unseeded

3552 Langmuir, Vol. 25, No. 6, 2009

Jakob et al.

Figure 3. XRD patterns of zeolite beta films synthesized on seeded (a) and unseeded (b) substrates after 9 days of hydrothermal treatment at 140 °C.

Figure 4. SEM micrographs of zeolite films prepared on seeded (a) and unseeded (b) substrates after hydrothermal treatment at 140 °C for 9 days.

substrates immersed in solution of molar composition 0.6 TEAOH/0.6 HF/10.5 H2O at 140 °C for 9 days. The peak profile at 7.1° 2θ differs from the one observed in the XRD pattern of nanoseeds (Figure 1a). In particular, the ratio between the two polymorphs in the intergrown film is different from the one of employed zeolite beta nanoparticles. Therefore, the film growth was not influenced by seed crystals, or, if such an effect exists, it is limited to the beginning of the crystallization process. The SEM study of the unseeded zeolite films showed that they were up to 20 µm thick and not uniform, comprising several layers with different morphological features (Figure 4a). Prismatic crystals and random aggregates prevail in the top layer, while traces of etching of the glass can be seen at the substrate film interface. In contrast to the latter, smooth and homogeneous films were observed on seeded substrates after hydrothermal treatment under similar conditions (Figure 4b). The thickness was reduced to 4 µm and the film was built of an intergrown layer, covered with randomly distributed well-shaped bipyramidal zeolite beta crystals. SEM inspection of cross-sections of vertically positioned plates was performed along their length, and no etching could be observed for seeded substrates. 3.3. Kinetics of Growth. The growth kinetics of seeded zeolite beta films in fluoride media was investigated on silicon wafers. These substrates were preferred to silica glass because of smoothness and the possibility to perform GID study on the synthesized films. Top and side views of the obtained films are shown in Figure 5. After 1 day of hydrothermal treatment, the substrates were covered with a thin (0.5 µm) layer of nanosized particles that did not exhibit distinct crystalline features (Figure 5a,b). Upon prolongation of the synthesis time, the particles continue to grow, filling the interparticle space. When the substrate was hydrothermally treated for 4 days, the layer was 2 µm thick and built of bipyramidal well-faceted crystals typical of zeolite beta morphology (Figure 5c,d). The prolongation of the synthesis time to 7 days increased the film thickness up to 3.5 µm. The film was continuous, built of intergrown crystals, and presented

Figure 5. SEM top (a,c,e) and side (b,d,f) views of zeolite beta films obtained after hydrothermal treatment at 140 °C for 1 day (a,b), 4 days (c,d) and 7 days (e,f).

Figure 6. Film thickness as a function of the hydrothermal treatment time for seeded silicon wafers treated at 140 °C.

a good adhesion to the substrate, as the layer was not affected by a 30 min ultrasonic treatment. The uniform monolayer was built of 3-4 µm zeolite beta crystals (Figure 5e,f), while the crystals obtained in the bulk were larger (4-6 µm). This size difference may be attributed to the steric hindrance during the growth steps that prevented the film crystals from growing further. Figure 6 presents the evolution of film thickness as a function of synthesis time for hydrothermal treatment at 140 °C. The thickness assessment was based on the SEM study of the crosssectional views of at least 10 measurements along film length. 3.4. GID Study of Zeolite Films. The orientation of the zeolite beta crystals in the zeolite films was investigated by Bragg reflections in the regions 2θ ) 5-10° and 2θ ) 20-25°. Figure 7a shows the XRD pattern recorded at different incident angles for a film obtained after a 1-day hydrothermal treatment at 140 °C. With an incident angle of Ri ) 0.05°, the zeolite beta film only exhibited three reflections with (h k l) equal to (1 0 1), (3 0 2), and (4 0 1). The observed orientation along the (h 0 l) planes was attributed to the steric hindrance in the growing layers. It should be mentioned that the intensity of the recorded peaks was

Sorption Properties of Zeolite Beta Films in Fluoride

Langmuir, Vol. 25, No. 6, 2009 3553

Figure 8. XRD pattern of zeolite films obtained after hydrothermal treatment for 9 days at 140 °C (a), 160 °C (b), and 180 °C (c).

Figure 7. Radial 2θ scans at Ri ) 0.05° (a), 0.1° (b), and 0.3° (c) of zeolite beta films grown in fluoride media after 1 day of hydrothermal treatment at 140 °C.

very low, indicating that the upper layer was not well crystallized. This low intensity of the XRD pattern suggests that the top layer is not completed, and the observed diffraction emanates from a few crystals grown on the top of the film. In contrast, the XRD patterns of the deeper layers (Figure 7b,c) shows that this orientation was gradually lost (Ri ) 0.1°) and, for an incident angle of Ri ) 0.3°, the BEA-type crystals were randomly oriented. On the basis of this data, we anticipate that the growth of deeper layers was controlled by the seeds, which are complex aggregates of randomly intergrown zeolite beta crystallites, and thus a preferred orientation is not expected to be observed. Thus, the GID steady revealed that the effect of the seeds is limited to the first 10-20 h of synthesis. After this stage, the film growth is governed by the growth rate of different faces, which, in this case, is the pinacoidal one. This result is in agreement with the SEM study, where well-developed pyramidal and small pinacoidal faces were observed. As known, the face with the fastest growth rate is absent or less presented in the ultimate crystal morphology. 3.5. Influence of Temperature on the Film Growth. Seeded silica glass substrates were hydrothermally treated for 9 days at temperatures varying between 140 and 180 °C. XRD patterns (Figure 8) showed that pure and highly crystalline pure zeolite beta films were obtained in the temperature range 140-170 °C. At higher temperature (180 °C), zeolite MFI dominated in the intergrown film. At this temperature CaF2 was also detected, revealing that considerable amounts of calcium (Ca2+) were extracted from the substrate. The experiments at 100 and 120 °C did not provide a zeolite beta film even after 2 months of hydrothermal treatment. Besides on the type of crystallizing material, the crystallization temperature had a pronounced effect on the morphological characteristics of obtained layers. The synthesis at 140 °C provided a 4 µm well-intergrown monolayer of zeolite beta (Figure 9a,b).

Figure 9. SEM top (a,c,e) and side (b,d,f) views of zeolite beta films obtained at 140 °C (a,b), 150 °C (c,d), and 160 °C (e,f) after 9 days of hydrothermal treatment.

At 150 °C much thicker (20 µm) film was formed, where truncated square bipyramids were observed on the top part of the film (Figure 9c,d). These relatively large crystals were not organized in a compact layer due to the partial intergrowth between them. In contrast, the bottom part of the film was strongly influenced by seed crystals, and a well intergrown layer was formed. The film thickness increased along with temperature, reaching up to 40 µm for films treated at 160 °C (Figure 9e,f). Figure 10 presents the evolution of film thickness as a function of temperature of hydrothermal treatment. 3.6. Sorption Properties of Zeolite Beta Films and Powders. 3.6.1. Nitrogen Sorption Properties. Sorption properties of zeolite beta films synthesized at 160 °C for different times of hydrothermal treatment were investigated. The specific surface area of the nontreated silica glass substrates was 0.1 m2 g-1, while the one of zeolite beta powder obtained during zeolite film synthesis was 577 m2 g-1. Supported films showed an increasing specific surface area for hydrothermal treatment time up to 5 days (SBET ) 28.4 m2 g-1). The prolongation of the synthesis time to 9 days did not lead to a significant increase of the specific

3554 Langmuir, Vol. 25, No. 6, 2009

Jakob et al.

Figure 10. Film thickness as a function of the crystallization temperature for the samples obtained after 9 days of hydrothermal treatment.

Figure 11. Nitrogen adsorption (filled symbols)/desorption (open symbols) isotherms at -196 °C for zeolite beta films obtained after 1 day (squares), 5 days (triangles), and 9 (circles) days of hydrothermal treatment at 160 °C. Table 1. Specific Surface Areas for Zeolite Beta Films Prepared at 160 °C for Different Periods of Time synthesis time (days)

specific surface area (m2 g-1)

surface loading (g m-2 of substrate)

1 3 5 9

1.8 2.0 28.4 30.3

4.43 4.66 58.86 62.80

surface area (SBET ) 30.3 m2 g-1), which revealed that the films were not growing any more in this time range (Table 1). Small amounts of zeolite were deposited on the glass slides after 1 day of hydrothermal treatment, which results in very low nitrogen adsorption. The type I adsorption/desorption isotherm of the sample synthesized for 5 days revealed the microporous character of the synthesized film. A steep rise in the uptake at low relative pressures corresponds to the filling of micropores with N2 (Figure 11). A second uptake at high relative pressure comprising a hysteresis loop was observed in the isotherm of the film obtained after 9 days hydrothermal treatment, which is indicative of the generated intercrystalline mesoporosity (Figure 11). The latter is a consequence of the relatively large size of the crystals in the top layer. As a consequence, the surface loading increased up to 62.8 g m-2, showing that a thick and homogeneous layer was obtained after 9 days of hydrothermal treatment. In comparison, zeolite beta films synthesized at 140 °C for 9 days presented a BET specific surface area of 6.4 m2 g-1, due to the low amounts of zeolite deposited at this temperature. So as to maximize sorption properties of organic molecules, films and powders prepared at 160 °C for 9 days were employed for sorption experiments.

3.6.2. Sorption Capacity and Kinetics of BEA-Type Films. Molecules with substantial difference in kinetic diameter32-35 (see Table 2) were chosen in order to study the influence of this parameter on the kinetics and sorption capacity of the zeolite films and powders. The hydrophobicity of the BEA-type films and powder was investigated by water sorption tests. With a kinetic diameter inferior to that of nitrogen, water fills only 31.2% and 32.8% of the microporous capacity for films and powder, respectively, confirming the hydrophobic character of the zeolitic material. It was found that the adsorbed quantities depend not only on the kinetic diameter but also on the shape of the molecule. Thus, cyclohexane presented a higher sorption rate than p-xylene, while these two adsorbates are of similar kinetic diameter. This might be due to the branching of p-xylene, which is packed less efficiently in zeolite beta channels than cyclohexane. The adsorbed quantities of larger molecules (TMOS and TMB) are significantly lower than those measured for smaller ones, which is most probably due to the difficulties in the diffusion and lower compacting ability of these species in zeolite channels. Another topic of interest deals with the sorption capacity of the zeolite film with respect to the corresponding powder sample. Sorption experiments showed that the adsorbed quantity on the zeolite film is about 90% of the theoretical amount for cyclic molecules (aromatics and alcane), while it is significantly lower for TMOS (Table 3). With a kinetic diameter close to the one of zeolite beta pores (7.1 Å × 7.7 Å), TMOS could not diffuse efficiently in zeolite film, decreasing its sorption efficiency compared to the corresponding zeolite beta powder. In other words, lower adsorption capacity of zeolite beta films in respect to their powder counterparts is most probably due to the abundant intergrowth in the film which presents additional diffusive barriers. Moreover, the geometry of the film, which is deposited on a flat and dense substrate, only allows access to the top zeolite layer, while all the crystals are equally reachable in the powder material. In other words, the growth of the bottom crystals was strongly influenced by the seeding step, resulting in a strongly intergrown layer that hindered the organic molecule diffusion. Table 2 presents additional information concerning the sorption properties of zeolite beta films and powders. The last two columns of this table present the adsorbed amount of organic to the adsorbed amount of nitrogen in percent, which allows the filling efficiency to be deduced. Owing to their large kinetic diameter, TMB and TMOS showed the lowest sorption capacities, respectively filling 13.0% and 8.1% of the zeolite beta film capacity. On the other hand, cyclohexane was proven to be the best adsorbate, as its cyclic geometry allowed an efficient diffusion and compacting in the zeolite beta porosity. Adsorption kinetics of the organics is represented in Figure 12. Besides better compacting, cyclohexane was found to present the fastest kinetics on both films and powders, which supports the previous observations concerning its high diffusion efficiency through zeolite channels. For all adsorbates, the adsorption kinetics were faster on zeolite beta powder (Table 3) as a result of the greater pore accessibility of the isolated crystals. However, in the case of TMOS and cyclohexane, the adsorption equilibrium was reached first on the zeolite films, because of the smaller amount adsorbed on the films in respect to the corresponding powders. (32) Van Bavel, E.; Meynen, V.; Cool, P.; Lebeau, K.; Vansant, E. F. Langmuir 2005, 21, 2447. (33) Xomeritakis, G.; Naik, S.; Braunbarth, C. M.; Cornelius, C. J.; Pardey, R.; Brinker, C. J. J. Membr. Sci. 2003, 215, 225. (34) O’Connor, C. T.; Moller, K. P.; Manstein, H. KONA 2007, 25, 230. (35) Funke, H. H.; Frender, K. R.; Green, K. M.; Wilwerding, J. L.; Sweitzer, B. A.; Falconer, J. L.; Noble, R. D. J. Membr. Sci. 1997, 129, 77.

Sorption Properties of Zeolite Beta Films in Fluoride

Langmuir, Vol. 25, No. 6, 2009 3555

Table 2. Adsorbed Amounts of Zeolite Beta Films and Powders at P/P0 ) 0.5 adsorbate

kinetic diameter (Å)

BEA-type film nads (mmol g-1)

BEA-type powder nads (mmol g-1)

nads(film)/ nads(powder)

% nadsa(film)

% nadsa(powder)

nitrogen water n-hexane cyclohexane p-xylene TMB TMOS

3.5 2.75 4.3 6.0 5.8 7.3 8.9

7.14 2.23 1.08 1.38 1.07 0.93 0.58

7.14 2.34 1.39 1.51 1.19 1.01 0.94

1 0.95 0.78 0.91 0.90 0.92 0.62

31.2 15.1 19.3 15.0 13.0 8.1

32.8 19.5 21.1 16.7 14.1 13.2

a

% nads ) nads(organic)/nads(nitrogen)

Table 3. Comparison of the Average Adsorption Kinetics Towards N2 and Organics for Zeolite Beta Film and Zeolite Beta Powder adsorbate

BEA-type film Vads (10-5 mmol g-1 s-1)

BEA-type powder Vads (10-5 mmol g-1 s-1)

Vads(film)/ Vads(powder)

nitrogen n-hexane cyclohexane p-xylene TMB TMOS

33.48 6.74 7.36 4.53 2.04 3.58

31.14 10.20 13.12 6.00 5.87 4.71

1.08 0.66 0.56 0.76 0.35 0.76

Figure 12. Adsorption kinetics of cyclohexane (diamonds), n-hexane (squares), p-xylene (triangles), TMB (crosses), and TMOS (circles) on zeolite beta powder (a) and film (b).

The sorption kinetics appeared to be 1.3 to 1.8 times slower on zeolite films than on zeolite powders toward cyclohexane, n-hexane, p-xylene, and TMOS. On the other hand, the sorption kinetics of TMB appeared to be dramatically slower on the films than on the powders, which seems to be related with the additional diffusion barriers in intergrown crystals and some of the particularities of this molecule (Table 3). In other words, the limited accessibility of some parts of the film due to the abundant intergrowth of zeolite crystals substantially decrease the adsorption kinetics of large branched molecules.

Another topic of interest deals with the performances of zeolite films in comparison to their counterpart pellets that are shaped with plasticizers. The advantage of intergrown zeolite coating in respect to the corresponding pellets is that surfaces that are integrated as part of the satellite can be used as substrates. Thus, additional weight can be avoided. Moreover, the intergrown films do not comprise a binder, in contrast to the zeolite-containing pellets.

4. Conclusions The synthesis of zeolite beta films from a near-neutral gel was studied along with their sorption properties for space applications. A gel with appropriate viscosity and wetting properties was designed so as to grow uniform films on different types of substrates. The conditions of synthesis were adapted for flat glass slides and silicon wafers, proving that there was no substrate etching in the presence of seed crystals, and both substrates were successfully covered with intergrown zeolite beta films. Film formation growth and kinetics were studied at different temperatures. Thus, the fluoride route was adapted to the synthesis of pure and highly crystalline BEA-type films with controlled thicknesses ranging between 0.7 and 40 µm. The obtained films were uniform and presented a good adhesion to the substrate, as 30 min ultrasonic treatment did not result in film pealing or any other changes in the films’ morphology. Sorption measurements toward several organic molecules showed that the adsorbed quantities depend on both the molecule shape and kinetic diameter. Inferior quantities of TMOS and TMB adsorbed in BEA-type films in respect to corresponding powders suggest a lesser accessibility of intergrown layers for molecules of kinetic diameter close to zeolite pores. Nevertheless, the obtained values are still compatible with the target application, that is, control of the molecular contamination level in low earth orbit. Acknowledgment. The financial support from the CNRS and the CNES was greatly appreciated. Dr T. Metzger from ERSF (ID-01) Grenoble is gratefully acknowledged for GID analysis. Supporting Information Available: Details of the dissolution of glass substrate in the initial solution. This material is available free of charge via the Internet at http://pubs.acs.org. LA8033963