Patent Grant
RE40299
The present invention relates generally to a mesoporous silica material made by an evaporative method. More specifically, the present invention relates to films, fibers and powders having mesoporous structure made from a silica precursor solution by layer thinning, for example spin casting; drawing; and spraying respectively. As used herein, the term “silica” means the presence of silicon (Si), without precluding additional metals.
Porous silica powders, with ordered porosity in the nanometer scale, have utility for catalysis, gas separation and high surface area supports for self-assembled monolayer films. Mesoporous micro-bubbles in particular, have applications in separations, thermal barriers and micro-encapsulation for drug delivery.
Micron-sized bubbles composed of solid silica walls are commercially available and are used as fillers and within reflective paint for highway signs. U.S. Pat. No. 2,797,201 (Standard Oil Co., Ohio) describes hollow glass spheres with solid walls, by spray drying liquid alkali metal silicates containing a blowing agent. Sizes range from 50-300 μm. Because these products are not porous, they are not useful for catalysis, and gas separation.
Other formation techniques for mesoporous powders and films, discussed in the literature, involve slow growth from supersaturated solutions for several hours to one week. The previous methods are based on a precipitation processes in which dissolved silica co-precipitates with the surfactant micelles to form a mesoporous structure and typically involve heating the reactants in an autoclave for several hours to a week. A disadvantage of these methods is that there is no control over particle size and/or shape. Filtration, often tedious because of small particle size, is required to separate the solution from the mesoporous particles.
Work described in U.S. Pat. Nos. 5,264,203, 5,098,684, 5,102,643, and 5,238,676 shows mesoporous powder formation by in-situ solution-phase precipitation, which again requires substantial time from a minimum of about 1.25 hour to about 168 hour to obtain precipitated powders.
Huo et al., Chem. Mater. 1994, 6, 1176, discussed a method for producing mesoporous silica by an acid route. Tetraethoxysilane (TEOS) was added to a dilute aqueous solution of cetyltrimethyl ammonium chloride (CTAC) and HCl. The solution composition on a mole basis was: TEOS 1.0; CTAC 0.12; HCl 9.2; water 130. After ˜30 min of stirring at room temperature, particles precipitated and were filtered from the remaining solution. Again, a significant amount of time (30 min) is needed to obtain precipitation of particles from the solution phase.
Tanev, P. T.; Pinnavaia, T. J.; Science, 1996, 271, 1267 used surfactant vesicles to template silica vesicles in a reaction mixture. The reaction mixture was vigorously stirred at ambient temperature for 18 hours to obtain the templated lamellar product with vesicular morphology, denoted MSU-V.
The method discussed by both Kresge, C. T., et al., Nature 1992, 359, 710; and Beck, J. S., et al., J. Am. Chem. Soc. 1992, 114, 10834 involves a slow growth, or co-precipitation, of silica and surfactant micelles over a period of 4 hours to 144 hours (5 days). Beck, J. S.; Hellring, S. D.; Vartuli, J. C. Abstract # COLL-311, ACS National Meeting, April 13-17, San Francisco, Calif., 1997, further indicate that 1700 m2/g is presently an upper limit of surface area.
Porous silica films have applications in catalysis, environmental remediation, energy storage, thermal barriers and energy storage. Porous silica films, in particular, are potentially useful as low dielectric constant interlayers in semiconductor devices, as low dielectric constant coatings on fibers and other structures, and in structured catalytic supports. Porous silica films produced by previous methods can be divided between random, gel-like silica films, and surfactant-templated films in which the pores are within a hexagonal lattice, with the characteristic pore diameter defined by the surfactant micelle.
Previous work resulting in mesoporous membranes from surfactant-templated powders and structures by in-situ solution-phase precipitation has been described in co-pending U.S. patent application Ser. No. 08/344,330. In-situ solution-phase precipitation requires substantial time from about 4 hours to 1 week to form a mesoporous membrane or film.
Hrubesh, L. W.; Poco, J. F., J. of Non-Cryst. Solids 1995, vol. 188, p. 46 applied “aerogel” technology to produce high-porosity films with random porosity. In the aerogel synthesis route, a hydrolyzed silicon-alkoxide solution is metered onto a spinning substrate. To avoid drying, the spinning apparatus is in an atmosphere saturated with solvent vapor. The spinner is stopped with a brake, and the retained spinning solution gels within a few minutes. The gel-coated substrate is immersed in solvent and subsequently dried under supercritical conditions.
Smith et al. (Smith, D. M.; Anderson, J.; Cho, C. C.; Gnade, B. E., Mat. Res. Soc. Symp. Proc. 1995, 371, 261, and Smith, D. M.; Anderson, J.; Cho, C. C.; Johnston, G. P.; Jeng, S. P., Mat. Res. Soc. Symp. Proc. 1995, 381, 261) applied “xerogel” technology as an alternative to aerogels. Here, the spin-cast silica sol-gel film is aged, washed and solvent exchanged, silated with a trimethylchlorosilane solution in heptane, and dried. In contrast to the aerogel process, the film is dried at ambient pressure. The aging and chemical treatment minimizes pore shrinkage during drying and makes the film hydrophobic, but the film becomes hydrophilic on heat-treatment, unless done in a forming gas environment.
Both techniques for spin-casting (1) aerogel and (2) xerogel films are complicated by the fact that spinning must be performed in solvent-saturated atmospheres (requiring explosion proofing) to avoid premature drying of the film.
In other work on mesoporous silica films, Ogawa (Ogawa, M., J. Am. Chem. Soc. 1994, 116, 7941) fabricated spin-cast silica-CTAB films. Ogawa used a CTAB/TMOS mole ratio of 0.40 in a solution that avoided gelation or precipitation and produced films that were lamellar, containing alternating layers of silica and bilayers of CTAB, and therefore not calcinable; surfactant can not be removed without degradation of the film structure. Accordingly, Ogawa did not calcine his silica films. Although Ogawa noted that rapid evaporation was essential for the formation of highly-ordered, lamellar CTAB-silica composites, those composites would not be expected to be stable to calcination, and would also not contain useful pore structures.
Further work by Ogawa (M. Ogawa, A SIMPLE SOL-GEL ROUTE FOR THE PREPARATION OF SILICA-SURFACTANT MESOSTRUCTURED MATERIALS, Chem. Commun., 1996, 1149-1150) was with a CTAC/TMOS ratio of 0.25. However, he used a substoichiometric ratio of water to silica (TMOS) of 2 (stoichiometric ratio of water to silica is 4). Ogawa's product, before calcination, has the 100, 110 and 200 reflections in the XRD pattern corresponding to a hexagonal structure. However, no information is given on calcined films in which the surfactant has been removed. It is inferred that Ogawa's product is unstable against calcination.
Porous silica fibers, with ordered porosity in the nanometer scale, have potential applications in catalysis, environmental remediation, thermal insulation and chemical sensors. Nanoporous or mesoporous fibers using the previously described sol-gel methods and stable against calcination have not been reported.
In the previous methods in the literature, there is no direct means for controlling particle size or pore volume fraction in powder, films or fibers.
Accordingly, there remains a need for mesoporous products having well defined morphology on both the nanometer scale (1-20 nm) (solid silica and pores) and the micrometer scale (0.1 μm-100 μm) (the characteristic dimension of the mesoporous product), and a method for making them in less time and without the need for filtration. Where spin-casting is done, there remains a need for a straight-forward method for producing mesoporous film(s) without supercritical drying, aging, silation of the film(s), or controlled gas environments.
This invention pertains to the development of surfactant-templated, nanometer-scale porosity of a silica precursor solution and forming a mesoporous material by first forming the silica precursor solution into a preform having a high surface area to volume ratio, then rapid drying or evaporating a solvent from the silica precursor solution. The mesoporous material may be in any geometric form, but is preferably in the form of a film, fiber, powder or combinations thereof. The rapid drying or evaporation of solvent from the solution is accomplished by forming a preform by any of layer thinning, for example spin casting; drawing; or spraying respectively.
It is critical to the present invention that the silica precursor solution avoid gelation or precipitation in order to permit formation of the mesoporous material by templating and evaporation of solvent(s). In a precipitation process, the composition of the mesoporous material is governed by a partitioning between the aqueous silica precursor solution and solid phases. In the evaporative process of the present invention, mesoporous material composition is directly related to the composition of the silica precursor because all of the non-volatile components of the precursor solution (namely the silica and the surfactant) are incorporated into the mesoporous structure. Therefore, in the present invention, the pore volume per gram of silica is controllable by varying the surfactant to silica precursor mole ratio in the silica precursor solution.
It is further critical to the present invention that the silica precursor be an alkoxide silica precursor, and that the mole ratio of an amount of a surfactant to the alkoxide silica precursor for templating be great enough to avoid producing a dense, non-porous film yet low enough to avoid producing a lamellar structure that is not calcineable, or any other non-calcineable structure. Left to themselves, some alkoxide silica precursor solutions will gel or precipitate over time if left alone from about 10 seconds to about 5 days or a week depending upon the solution. Thus, preforming must be done within a time before gelation or precipitation occurs. Finally, the rate of evaporation is critical to the formation of the mesoporous product. The slower the evaporation, the less ordered the mesopores. Accordingly, it is preferred that the solvent be evaporated or removed from the templated mesoporous structure in a time less than about 5 minutes, preferably less than about 1 minute, and most preferably less than about 10 seconds.
Production of a film is by layer thinning, wherein a layer of the silica precursor solution is formed on a surface followed by removal of an amount of the silica precursor solution and leaving a geometrically thinner layer of the silica precursor solution from which the solvent quickly escapes via evaporation. Layer thinning may be by any method including but not limited to squeegeeing and/or spin casting. Spin-cast mesoporous films are formed on the order of a minute or even seconds.
Advantages for the layer thinning method of the present invention include (1) films are formed within a minute (apart from time required for post-treatment and calcination), (2) no special atmospheres, pressures or supercritical drying equipment are required as in the case of aerogel film fabrication, and (3) the porosity is ordered, and of a controlled pore size rather than a random, gel-like structure in the case of aerogels and xerogels; the volume fraction of porosity and the structural order within the film are controllable by the silica to surfactant content or mole ratio. Further advantages are realized from thinning with a spin-coater, which is standard equipment in the microelectronics industry. Advantages of using a spin-coater include (1) films have uniform interference colors, indicating uniform film thicknesses (2) film thicknesses are repeatable from sample to sample (for example ±0.006 μm thick film, or a 1% variation) and controllable by varying the ethanol and water dilution and the spinning speed, and (3) the spin-casting technique does not require the use of large solution batches in which only a small fraction of the solution is used for film growth, as in the case of the earlier film growth technique for which the solution is depleted with film growth and must either be replaced or somehow regenerated. In the interfacial growth technique, bulk solutions of silicate and surfactant are used.
In powder formation by spray drying, the same conditions of fast drying exists as in spin-casting (as well as in fiber spinning) because of the high surface-area to volume ratio of the product. When a powder is produced by liquid spraying, the particles or micro-bubbles within the powder are hollow spheres with walls composed of mesoporous silica.
The volume within the interior of the mesoporous micro-bubble is undesirable for some applications including catalytic processes in which coking occurs. In these cases, the micro-bubbles may be broken by crushing or grinding. Mesoporous silica powders impregnated with catalytically active metals have applications in catalysis. The pore size, ˜25-40 Å, allows access of large molecules to catalysis sites. The high surface area of the powders allows high catalytic activity. The surface area of mesoporous powders was determined to be ˜900 m2/g by nitrogen absorption. The powders may be pressed or mixed with binders and extruded to produce pellets, tubes and other shapes for structured catalyst supports. Thus, the particle size in spray-drying may be controlled for a particular application. Because the micro-bubble walls are permeable, many applications such as micro-encapsulation is possible. Silica is ingestible. Containing a drug within the micro-bubble to allows passage through the stomach where it would normally degrade. The drug is released through the porous walls into the intestinal tract. Materials, including surfactants and polymers, adsorbed to either the outside of the bubble or within the pores can acts as pH-sensitive gates for the release of the drug.
Encapsulation may be done wherein a non-drug substance may be permanently caged within the bubble by closing off the pores with silane treatment, silica precipitation, or surfactant absorption.
Mesoporous fiber formation starts with a similar silica precursor solution but with an added pre-polymer making a pituitous mixture. The pituitous mixture is drawn into a thin strand from which solvent is evaporated leaving the mesoporous fiber(s). Mesoporous silica fibers may be impregnated with catalytically active metals for applications in structured catalytic packing. The small thickness, on the order of 10 to 100 μm, minimizes the diffusion distance from the bulk to the catalytic sites on the internal surface of the silica. The high aspect ratio of the fibers give the advantages of high throughput, combined with high reactive areas. The fibers may be wound or assembled in reactor modules. Hollow mesoporous fibers fabricated by rapid drying with heated gas (e.g. air) may be bundled into a module to form a catalytic membrane reactor. Reactants can flow through the hollow fiber and diffuse radially outward, through the mesoporous wall, past catalytically active sites. The reactor is especially useful for reactions where short contact times and good temperature control are required (e.g. partial oxidation). The temperature is uniform because reactions occur along the length of hollow fibers. The high surface area of the fibers allows high catalytic activity. The surface area of mesoporous fibers was determined to be ˜1100 m2/g by nitrogen absorption. Mesoporous fibers have further use in high-performance thermal insulation. The pore size, ˜25-40 Å, is such that transport of phonons of specific wavelength is reduced, limiting conductive heat transfer within the fiber.
It is an object of the present invention to provide a method of making mesoporous materials having high surface area to volume ratios.
It is a further object of the present invention to provide a method of making the mesoporous materials by solvent evaporation.
The subject matter of the present invention is particularly pointed out and distinctly claimed in the concluding portion of this specification. However, both the organization and method of operation, together with further advantages and objects thereof, may best be understood by reference to the following description taken in connection with accompanying drawings wherein like reference characters refer to like elements.
The method of the present invention relies upon a silica precursor mixed with a surfactant in an aqueous solution for templating the silica precursor together with a catalyst (acid) for hydrolysis of the silica precursor. The silica precursor is then made into a preform that has a high surface area to volume ratio and the aqueous solution quickly evaporated to form the mesoporous material. The evaporative process for mesoporous material has the following steps: (1) the silica precursor in aqueous solvent is formed into a preform of high surface area to volume ratio (by spinning, drawing, atomizing) and (2) the solvent is quickly evaporated, leaving mesoporous silica in a similar shape (film, fiber, sphere). High surface area is necessary for fast evaporation of the solvent. A third step of heating may be used to remove any residual solvent and to further condense the silica, followed by calcining which further removes any residual surfactant.
The evaporation is fast in comparison to precipitation. With the preform having a high surface area to volume ratio, and with heated air, solvent is rapidly evaporated from the preform. Mesoporous materials are formed in less than five minutes, preferably less than one minute. For particles, a particle may be formed in less than one second (<<1 s). In a few minutes of spray drying, several grams of powder are produced. Dry powders are formed directly in the spray-dryer and no filtration step is required.
In spray drying, the precursor solution is atomized into fine droplets. Solvent evaporation leaves behind a shell of mesoporous silica. In spraying, the droplet size, and thus the mesoporous particle size, is controlled by modifying the rheological properties of the spraying solution (through solvent dilution or addition of polymeric thickeners) and by changing spray conditions (by the use of different nozzle geometries and varying solution pressure).
The silica precursor may be an alkoxide silica precursor or tetrachlorosilane. A preferred alkoxide silica precursor is tetraethyl ortho-silicate (TEOS). Other alkoxide silica precursors include orthosilicates, including but not limited to tetramethyl orthosilicate (TMOS), tetrapropyl orthosilicate, and tetrabutyl orthosilicate. Iso-propyl, sec-butyl and tert-butyl orthosilicates are included as well but may have limited commercial availability. In addition to the alkoxide, a metal halide salt may be added, especially a metal chloride as well as a metal nitrate. Metal halide salt(s) and/or metal nitrate(s) combine with the alkoxide. More specifically, iron chloride or nitrate, aluminum chloride or nitrate combines with the alkoxide. Additional metal(s) may be incorporated into the mesoporous silica structure. These additional metal (s) result in reducing solubility of the mesoporous silica structure and may impart a negative charge to the mesoporous silica structure.
A preferred surfactant contains an ammonium cation, either a quaternary ammonium cation, for example cetyltrimethylammonium chloride (CTAC), or a tertiary ammonium cation. Variations of CTAC as described by Huo (See Background) include substitution of ethyl and propyl groups for the methyl group that may also be used. In addition, it is possible to produce mesoporous materials using alkyl trimethylammonium chloride or bromide surfactants with different alkyl chain lengths. Variation in alkyl chain length (e.g. C12, C14, C16, C18) permits control of the pore diameter wherein shorter alkyl chain lengths produce smaller diameter pores.
An alternative method of varying pore size is by adding a swelling agent to the silica precursor solution. For example, addition of 1,3,5-trimethylbenzene produces pore diameters about 2-5 times greater than pores made without the swelling agent.
In the present invention, the mole ratio of templating surfactant (or surfactant) to alkoxide silica precursor is preferably from about 0.05 to about 0.3. Below about 0.05, a dense, non-porous silica phase is produced and above about 0.3, a lamellar phase is produced that is not calcineable. The lamellar phase is not calcineable because the lamellar structure collapses upon removal of the surfactant that occurs during calcination. Specifically for CTAC/TEOS, the mole ratio is preferably less than about 0.24. More specifically, with a mole ratio of between about 0.1 to about 0.15, the pores are well ordered. However, at high ratios, above 0.24, specifically from about 0.26 to 0.28, give high pore volume fraction and high specific surface area (e.g. 1770 m2/g), surprisingly higher than reported in the prior art.
In the present invention, the mole ratio of water to silica precursor is preferably greater than or equal to a stoichiometric ratio. More preferably the ratio of water to silica precursor is about 7. Use of stoichiometric or super stoichiometric amounts of water is believed to help preserve the hexagonal structure of the product upon calcination.
Control of the morphology on the micrometer scale is also unique to the evaporative process. In spin casting, a flat substrate flooded with the precursor solution is accelerated to high rpm. Excess solution flows off during spinning, leaving a thin film of the solution which forms a solid mesoporous film by evaporative concentration. Films may be deposited on no-flat surfaces by spraying, painting or dip coating. It should be noted that according to the present invention it may be advantageous to insure that the substrate surface is hydrophilic. For silicon wafer substrate, a hydrophilic surface may be obtained by sonication in deionized (d.i.) water, followed by soaking in a solution of sulfuric acid and finally rinsing with d.i. water and drying.
In fiber spinning, the precursor solution (mixed with high molecular-weight polymer) is drawn into a strand. The solution evaporates leaving the mesoporous fiber. Drawing may be either by contacting an object to the precursor solution and moving the object away and forming a strand of precursor solution, or by permitting the precursor solution to flow from a vessel under pressure or by gravity. Flow under pressure may include extrusion.
In the hydrolysis of the silica precursor, tetrahydroxysilane is produced which undergoes condensation reactions to form silica oligomers. With the alkoxide silica precursor, an alcohol is a byproduct of hydrolysis. With the tetrachlorosilane precursor, hydrochloric acid is a byproduct.
An experiment was conducted to demonstrate making mesoporous films by the method of the present invention.
Silicon wafers were obtained from Silicon Source and cut into 2.5×2.5 cm2 squares. The silicon wafers were pretreated by sonication in deionized (d.i.) water, followed by soaking overnight in a solution of sulfuric acid and Nochromix™ (Godax Labs) and finally rinsing with d.i. water and drying by aspiration.
The silica precursor solution had the mole ratios of TEOS (Aldrich) 1.0; deionized water 7.2 (18 MΩ resistance); ethanol (punctilious; Quantum Chemicals) 5.7; HCl (Mallinckrodt) 0.10. The (CTAC) (T.C.I. America) was added after hydrolysis. CTAC/TEOS mole ratio was varied from 0 to 0.30 to determine its influence on film properties.
Spin-castings were performed with a Specialty Coating System Model P-6204A spin coater. With the silicon wafer at rest, the entire top surface of the silicon wafer was covered with the hydrolyzed TEOS-surfactant solution. The covered silicon wafer was spun at 4000 rpm for 60s with maximum acceleration (spin-up time <1s). A substantial amount of the hydrolyzed TEOS-surfactant solution flowed off the covered silicon wafer during rotation. The remaining solution was geometrically at a high surface area to volume ratio. Flowing ceases as the viscosity of the hydrolyzed TEOS-surfactant solution increases because of the increased concentration of silica precursor and surfactant from loss of solvent through evaporation.
The precalcined mesoporous silica film was post treated with ammonia vapors which improved the stability of the pore structure after calcination. A few drops of ammonium hydroxide were put in a closed petri dish containing the spin coated silicon wafer. The spin coated wafer was ammonia treated for about 15 minutes.
The spin coated silicon wafer was removed from the ammonia treatment then heated to 105° C. for several hours to complete drying of the solvent and increase condensation of the silica. Finally, the dried spin coated silicon wafer was calcined at 550° C. for 5-10 minutes.
The mesoporous silica film was characterized by X-ray diffraction and ellipsometry.
Although the pores were hexagonally ordered within a narrow composition range, within a broader range the films were thermally stable and the porosity may be fine-tuned by adjusting the CTAC/TEOS ratio.
Calcined films were characterized by X-ray photoelectron spectroscopy with analyzed volume on the surface of approximately 1×1 mm2 in area and 20 to 40 Å in depth. Silicon, oxygen and a small amount of adventitious carbon were identified. Within the resolution of the equipment, no chlorine nor nitrogen were found, demonstrating that calcination yields relatively pure silica without contamination from other chemicals used in the process. Cross-polarized optical microscopy of films deposited on glass slides, before and after calcination, did not reveal liquid crystalline-like optical anisotropy. Atomic force microscopy of the mesoporous film (not shown for brevity) revealed a surface morphology dominated by undulations ˜1 μm in diameter and raised rims ˜0.1 μm in width. Height variations were within 60 nm. The film was continuous and not the result of the deposition of pre-existing particles.
An experiment was conducted to demonstrate that quick drying was important to the film structure.
Silicon wafers were pre-treated in the manner described in Example 1.
The silica precursor solution had mole ratios of deionized water 7.1; ethanol 5.4; HCl 0.1; TEOS 1.0; and CTAC 0.11.
Spin castings were performed in the manner described in Example 1. The remaining silica precursor solution was evaporated by natural convection in an open glass bottle. The spin coated wafer and the evaporated silica precursor solution were not post treated with ammonia vapors. The spin coated silicon wafer and evaporated silica precursor solution were heated to 105° C. for several hours.
The mesoporous silica film and the evaporated silica precursor solution were characterized by XRD.
An experiment was conducted to demonstrate making mesoporous material powder. The silica precursor solution was prepared with the following composition by mass: TEOS 51.80 g; water 26.264 g; 38 wt % hydrochloric acid 1.756 g; and CTAC 10.333 g. TEOS, deionized water and HCl were first combined together, following by addition of CTAC. Normally immiscible, TEOS combines with water in the presence of the surfactant, allowing the hydrolysis reaction to occur. The solution became hot upon mixing from the exothermic hydrolysis; the sample bottle was cooled under running water.
In Example 1, ethanol was used to dilute the precursor solution for spin casting. However, in the spray-drying process, potential explosion or flammability hazards from either added flammable solvents or the ethanol reaction by-product must be minimized. Hence, no ethanol dilution was used because of the potential hazards. Instead of ethanol, water was used for dilution, hydrolysis, and solvent for the surfactant.
The precursor solution was spray-dried in a Büchi-190 Mini Spray Dryer operating with an inlet air temperature of 174° C., an outlet temperature of 76° C., a pump speed setting of “5,” a heat setting of “13,” an airflow setting of 300 and a gate valve setting between ¼ and ½. The precursor solution was pumped through a water-cooled nozzle into a flow of heated air and down the length of a ˜30 cm drying tube. The solvent in the droplets of the spray quickly evaporate, leaving behind the nonvolatile material in the particulates which are collected at the bottom of a venturi separator. It takes on the order of a second for material to pass through the spray nozzle and be collected.
The powder was heated at 105° C. for ˜2 h, followed by calcination at 600° C. for ˜30 min. The absence of ammonia treatment for this powder demonstrates that ammonia is not required for producing a calcineable material. After calcination, the powder lost 40.6% of its mass, compared with a theoretical mass loss of 40.9% based on the precursor solution composition (assuming silicon exists in the form of SiO2 and each surfactant molecule has a chloride counter-ion). The close agreement between the experimental and theoretical values supports the assertion that, in the rapid drying process, the mesoporous solid composition is directly determined by the solution composition, and all of the nonvolatile species are incorporated into the solid.
Scanning electron micrographs of the spray-dried mesoporous silica after calcination showed that the particles were in the form of hollow spheres or shells with diameters ranging from approximately 4 to 40 μm (micron). Potentially, hollow spheres may range in size from about 1 μm to about 300 μm. The wide range of particle size was probably due to non-uniform atomization from the spray nozzle. The hollow nature of the particles was evident by the fraction of collapsed particles which have the appearance of deflated balls. From the width of the folds in the collapsed particles, the thickness of the particle bubble shell was estimated to be less than 0.5 μm. Electron dispersive X-ray analysis confirmed the calcined particle micro-bubbles were composed only of silicon and oxygen (see FIG. 6).
A surfactant-silica solution has been spray dried into a powder. X-ray diffraction data of calcined powder clearly showed the existence of mesoporosity by a low-angle primary peak corresponding to a d-spacing of ˜31 Å.
In the PXRD pattern of the powder after calcination, a second peak of low intensity at 5.1°, corresponding to a d-spacing of 17.3 Å, corresponding to a d-spacing of 17.3 A, is apparent. The (110) and the (200) reflections were not clearly resolved.
An experiment was conducted to demonstrate mesoporous silica fibers. Poly(ethyleneoxide) (PEO) with a MW of 5×106 was mixed with 18 MΩ deionized water to form a 3.7 wt % stock solution and was allowed to dissolve overnight. The pituitous mixture was prepared with the following composition by mass: TEOS 8.02 g; water 3.60 g; 38 wt % hydrochloric acid 0.39 g; CTAC 1.71 g, and 3.7 wt % PEO solution 1.26 g. TEOS, deionized water and HCl were first combined together, followed by CTAC. The solution became warm upon mixing from the exothermic hydrolysis; the sample bottle was cooled under running water. The PEO solution was then added to modify the rheology of the silica/CTAC solution to allow drawing of fibers.
Fibers were drawn onto a spindle with four wooden dowels. The dowels were covered with parafilm™, on which fiber samples were collected. The spindle was driven either by hand or by an electric hand drill attached by the drill chuck to the end of the metal rod. A thin, stream of solution, was drawn up from the solution with a pipet tip, and wound onto the spindle. The rotating spindle served to collect, pull and rapidly dry the fibers. After drawing, the fibers were peeled away from the dowels and collected in a crucible.
The fibers were ammonia treated for ˜1h by dripping a few drops of ammonium hydroxide solution onto a tissue taped to underside of a cover and placing the cover over the crucible containing the fiber sample. The ammonia vapor raises the pH in the fibers, and increases condensation of the silica and improves the stability of the porous silica structure during subsequent high-temperature calcination. The fibers were placed in an oven at 105 to 180° C. for several hours, followed by heat treatment at 600° C. for ˜1 h. The fully calcined fibers were white, presumably due to opacity arising from large defects incorporated during the hand-drawing procedure. The mass loss after calcination was 55% compared with an expected value of 57% calculated based on the amount of SiO2, CTAC, and polymer that would be obtained from the pituitous mixture.
The diameter of the fibers varied from 5 to 100 μm with a range of 35 to 60 μm being more typical. Fibers with small diameter of ˜10-30 μm are desirable because of better bending endurance and weavability. Small diameters are easily achieved with state-of-the-art fiber drawing (extrusion) equipment.
Fibers were ground with a mortar and pestle and characterized by powder X-ray diffraction (PXRD).
A second peak at low intensity at 3.9°, corresponding to a d-spacing of 23 Å, (4.4° or 20 Å after calcination). The (110) and (200) reflections are not clearly resolved similar to the mesoporous powder XRD pattern from example 3. The pore ordering in the fibers is improved by varying the CTAC/TEOS mole ratio and by drawing smaller diameter fibers.
TEM photos confirm a well ordered mesoporous structure within the fibers. Cross-polarized optical microscopy of the calcined fibers revealed liquid crystalline-like optical anisotropy.
An experiment was performed to demonstrate incorporation of aluminum into the mesoporous silica film.
Silicon wafers were pre-treated in the manner described in Example 1.
The aluminosilicate precursor solutions had mole ratios of deionized water 7.3; ethanol 5.3; HCl 0.09; CTAC 0.14; TEOS 1.0. The mole ratio of aluminum nitrate (Al(NO3)3 9H2O) to TEOS mole ratio was varied between 0.035 to 0.25. The solutions were prepared by combining deionized water, ethanol, hydrochloric acid, CTAC and aluminum nitrate together, followed by the addition of TEOS.
Spin castings were performed in the manner described in Example 1.
The spin coated wafers were not post treated with ammonia vapors.
The spin coated wafers were heated and calcined in the manner described in Example 1.
The calcined mesoporous aluminosilicate films were characterized by XRD in a 2-theta range of 1-10°. XRD results for the primary reflections are summarized in Table E6-1. With increasing Al/TEOS mole ratio, the d-spacing of the primary reflection decreased. The calcined mesoporous aluminosilicate film with a Al/TEOS mole ratio of 0.064 had the greatest primary reflection peak intensity. However, an Al/TEOS mole ratio of 0.064 should not be considered an optimum value for a well-ordered film structure because the pore ordering is also dependent on the surfactant content in the aluminosilicate precursor solutions. The dependence of the pore ordering on the CTAC/TEOS mole ratio was demonstrated in Example 1 for mesoporous silica films without aluminum. In the XRD patterns of the calcined mesoporous aluminosilicate films with Al/TEOS mole ratios of 0.035, 0.064 and 0.13, reflections of low intensity were observed at half the d-spacing of the primary reflection.
To identify possible crystalline phases distinct from the amorphous pore walls, the calcined mesoporous aluminosilicate film with an Al/TEOS mole ratio of 0.25 was characterized by XRD in a 2-theta range of 5-30° using a slow scan rate (0.04°/75 s). A wider 2-theta range was not used because of the strong reflections from the silicon wafer substrate above 30°. A low intensity peak with a d-spacing of 4.02 Å was observed. The only possible matching reference data for the Si—Al—O—H system was for cristobolite (SiO2). Therefore, the XRD pattern did not show a separate aluminum-oxide crystalline phase.
The calcined mesoporous aluminosilicate film with a TEOS/Al mole ratio of 0.25 was characterized by SEM. The calcined mesoporous aluminosilicate film was homogeneous; no crystal gains were observed. A small amount of surface roughness was observed which had the same appearance of the AFM image discussed in Example 1. EDS characterization of the calcined mesoporous aluminosilicate film showed the presents of aluminum. The EDS characterization was not quantitative because of significant penetration of the electron beam through the calcined mesoporous aluminosilicate film and into the silicon wafer substrate.
Dry-Spun Mesoporous Fibers
An experiment was conducted to demonstrate making well-ordered mesoporous fibers by the method of the present invention.
The spinning solution was formed by combining deionized water, hydrochloric acid (Mallinckrodt), 5×106 MW poly(ethylene oxide) (PEO) (Polysciences) from a 4 wt % aqueous stock solution, ethanol (punctilious, Quantum Chemicals), followed by TEOS (Aldrich). The solution was mixed to promote the hydrolysis reaction. Finally, CTAC was added to obtain final mole ratios of: 7.0 H2O, 0.050 HCl, 0.10 PEO (repeat unit), 4.0 ethanol, 1.0 TEOS, 0.24 CTAC. A thin strand of the pituitous solution was drawn from a pipette tip, and wound at a rate of 300 m/min onto a spool consisting of six dowels. Fibers were air dried at 105° C. overnight, and calcined by heating at 350° C. for 1 h and 600° C. for 3 h.
Samples were analyzed by powder X-ray diffraction using a Philips diffractometer with Cu Ka radiation. Pore-size distributions and BET surface areas were determined from nitrogen adsorption/desorption isotherms with the Quantachrome Autosorb 6-B gas sorption system, using the BJH and multi-point BET methods, respectively.
Mesoporous fibers were dry spun by drawing the precursor solution into continuous filaments and collecting on a spool. Fibers crossing on the spool during spinning tend to fuse together, creating a gauze-like product at the end of spinning which was cut away in sections from between the dowels. The as-spun fibers were pliable and pressable into pellets or rolled into tubes. With drying and calcination the fibers become brittle. Low temperature oven drying promotes condensation between silica oligomers and increases calcination stability of the silica phase. During calcination, silica undergoes further condensation; surfactant and polymer are removed, leaving the porous structure. The mass loss of 59% after calcination compared to a value of 57% calculated from the spinning solution composition, assuming the dried fibers contain SiO2, PEO, and surfactant with chloride bridging-ions (for the powders, mass losses were within 1% of the calculated values). The excess loss was attributed to incomplete drying/silica condensation prior to calcination.
A scanning electron micrograph of these calcined fibers showed fiber diameters are on the order of 40 μm and where varied by modifying solution composition and spinning conditions. The distribution of fiber diameters was due to the hand spinning technique presently used; more uniform fibers are achievable with state-of-the-art spinning equipment. The fiber cross sections typically had a kidney-shape, characteristic of dry-spun fibers where high evaporation rate at the air-fiber interface and comparatively slow solvent diffusion rates through the fiber caused the skin to collapse around the soft cores. Self-assembly of silica and surfactant occurred first at the air-fiber interface, followed by progressive conversion of the entire fiber to a mesophase structure. The mesoporous products do not form by the aggregation of preexisting mesoporous particles. Precursor solutions are clear, typically stable for several days, and eventually gel rather than form particles, as in the acid-route synthesis of Huo et al. Chem. Mater. 1994, 6, 1176-1191.
Powder X-ray diffraction (PXRD) patterns of the dried and calcined fibers are shown in FIG. 10. The (100), (110), (200), and (210) reflections corresponding to a hexagonal structure are visible in the PXRD pattern for the dried fibers, although only the first three reflections are visible for the calcined fibers. The increase in peak intensity after calcination (note scale indicated on
By nitrogen-adsorption analysis, the mesoporous fibers have a surface area of 1100 m2/g and a 20 Å pore diameter (see FIGS. 11,12). Though the pore size was smaller than that of MCM-41 materials, because of calcination shrinkage, the total surface area is comparable. The adsorption/desorption isotherms showed no hysteresis within the resolution of the equipment, indicating that the pores were unconstricted. The hydrophilic polymer is presumably dispersed within the silica phase (no crystalline XRD peaks are observed in the dried fibers). However, there appears to be no residual porosity from pyrolysis of the polymer; the adsorption data showed no evidence of a micropore contribution. The volume fraction of mesoporosity was 54%, which correlates well with the 57 vol % porosity calculated from the volume contributions of the surfactant and silica phases, after taking into account the volumetric shrinkage measured by the shift in the d100 peak. We have shown previously that the pore volume fraction in mesoporous films can be controlled by varying the CTAC/TEOS mole ratio in the precursor solution [8]. Raman spectroscopy of calcined fibers indicate that the fibers consist of meta-silicate, consist for silica with hydroxide terminated surfaces.
The as-spun, dried and calcined fibers all showed birefringence between cross polarizers in an optical microscope. Consistent with pore orientation along the fiber axis, maximum light transmission occurred with the fiber 45° to the polarizers and nearly complete extinction occurs when parallel and perpendicular to the analyzer. Transmission electron microscopy of microtome sections showed pore alignment over a length scale of at least 10 μm. However, because of difficulties in the microtome technique, the absolute pore orientation with respect to the fiber axis could not be established.
Spray-Dried Mesoporous Powders
An experiment was conducted to demonstrate making well-ordered mesoporous powders of high surface areas by the method of the present invention.
The spray-drying solution was formed by combining deionized water, HCl, CTAC, followed by TEOS to obtain final mole ratios of: 10.0 H2O, 0.050 HCl, 0.12 to 0.28 CTAC, 1.0 TEOS. The solution was mixed to promote the hydrolysis where the surfactant acts as a emulsifying agent to combine the aqueous and alkoxide phase. In the solution formulation, water rather than alcohol dilution is used to avoid possible explosion hazards. Solutions were spray-dried in a Buchi 190 Mini Spray Dryer operating with an outlet temperature of 120° C. Powders were collected under a cyclone and calcined under the same conditions as the fibers.
Samples were analyzed by powder X-ray diffraction and by nitrogen sorption as described in Example 7 above.
In spray drying the particle morphology was dependent on the precursor solution composition and drying conditions. The surfactant to silica mole ratio was varied between 0.12 and 0.28. SEM analysis showed particle morphology was similar to that of the hollow spheres (see Example 3), except the walls had collapsed during drying. Depending on the spray drying conditions, a range of particle morphologies were possible from solid spherical particles to collapse particles to hollow particles.
Pore volume fraction and the surface area as a function of surfactant concentration are shown in FIG. 13. The pore-size distribution plots for the highest surfactant ratio sample are shown in FIGS. 14,15, respectively. A multi-point BET, adsorption/desorption analysis was consistent with the pore size analysis. A maximum pore volume fraction of 63% (37 vol % silica) and a surface area up to 1770 m2/g was achieved at the highest surfactant concentration. Nitrogen adsorption/desorption curves had no hysteresis and indicate a constant pore size of 25 Å for all surfactant concentrations. In the PXRD patterns for the as-synthesized powders in this series, the (100), (110), (200) and (210) reflections corresponding to a hexagonal array were evident. After calcination the (210) reflections were absent (see FIG. 16). Interestingly, the d100 values were relatively constant with surfactant concentration (˜38 Å as synthesized and −32 Å after calcination).
Spray-Dried Mesoporous Powders
Aluminum was incorporated into the spray dried powders by the addition of aluminum chloride to the precursor solutions.
Aluminum chloride (hexa-hydrated form, Fischer) was combined with d.i. water, hydrochloric acid, CTAC followed by CTAC in the following mole ratios 11.3 H2O/TEOS; 0.10 HCl/TEOS; 0.106 CTAC/(AlCl3+TEOS); the AlCl3/TEOS ratios were 0.00, 0.031, 0.063, 0.125 and 0.25. Solutions were spray-dried in a Buchi 190 Mini Spray Dryer operating with an outlet temperature of 120° C. Powders were dried overnight at 105° C. and calcined in air at 350° C. for one hour and 600° C. for 1.5 h.
For an aluminum to silica mole ratio of 0.25 (FIG. 17a), the aluminum in the synthesized powders was a mixture of tetrahedral (framework) and octahedral. Octahedral aluminum would not impart a negative charge to the aluminosilicate. For an aluminum to silica mole ratio of 0.063, the aluminum in the as-synthesized powders was predominately framework, as determined by 27Al-NMR (see FIG. 17c). However, the framework substitution was not stable with calcination (see FIG. 17d). For an aluminum to silica mole ratio of 0.031, the aluminum in the calcined powders was roughly two-thirds tetrahedral (framework) and one-third in an octahedral coordination (see FIG. 17b).
The x-ray diffraction patterns for the mesoporous powders with aluminum addition are shown in
Loading of a Catalytically Active Metal into Mesoporous Fibers
An experiment was conducted to demonstrate loading of a catalytically active metal into the mesoporous fibers which were produced by the method of the present invention.
A mesoporous fiber supported rhodium catalyst was prepared by incipient wetness impregnation of the mesoporous silica fibers with a rhodium (III) nitrate solution. Rhodium is a good catalyst for reactions including methanol decomposition, alkane partial oxidation and fuel combustion. The loading for rhodium metal was 5% by weight. A solution of rhodium (III) nitrate (10 wt % assay, in nitric acid, Engelhard) was diluted with d.i. water in a volumetric cylinder until the 1.5 ml index was reached. A mass of 2.07 g or the mesoporous silica fibers was used in the catalyst loading. The mesoporous silica fibers were tumbled and the rhodium (III) nitrate solution was added drop wise. Once the rhodium (III) nitrate was impregnated on the mesoporous silica fibers, the fibers were dried at 100° C. in a vacuum overnight, followed by calcination at 350° C. for at least one hour. Prior to catalyst testing mesoporous silica fiber supported rhodium catalyst were activated (reduced to metallic rhodium) with a mixed gas of 10% hydrogen and 90% helium (by volume) at 120° C. for at least one hour.
Fibers with 5 wt % rhodium were used as a supported catalyst to convert methane and air to hydrogen and carbon monoxide. The amount of methane was 29.5 vol % and the amount of air was 70.5 vol %. A small amount of rhodium/fiber supported catalyst (0.041 cm3) was used. Residence time was 8 milli-seconds and the reaction was carried out separately at two temperatures 360 and 445° C. On day 1, the supported catalyst showed activity for both temperatures. The supported catalyst was permitted to cool overnight. However, on day 2, the supported catalyst showed no activity. Because the amount of catalyst was too small to perform an analysis of the failure, a second series of tests were done with an increased quantity of fiber supported catalyst (0.442 cm3).
The residence time for the second series was 50 milliseconds. The supported catalyst was run at 400° C. Again, the supported catalyst was permitted to cool overnight. No loss of activity was observed on the second day.
Because the failure was not duplicated, the reason for it was not determined, nor was the reason for the subsequent success determined.
An experiment was performed to demonstrate the coating of mesoporous silica onto glass cover slips. The precursor solutions were modified from Example 1 by reducing the amount of acid by half. The calcination temperature was also lowered to 450° C.
Microscope-slide-cover-slipe substrates (22×22 mm2, 0.13 to 0.16 mm thick) were soaked in a solution of sulfuric acid and Nochromix (Godax Labs) and rinsed with deionized water. The precursor solutions were prepared in a 30 ml glass bottle. The bottle was rinsed and dried to remove particulates. Reagents were added by mass using disposable transfer pipettes. Spin-coating precursor solutions were prepared by combining cetyltrimethylammonium chloride (CTAC) (T.C.I. America), deionized water, ethanol (punctilious; Quantum Chemicals), hydrochloric acid (Mallinckrodt) and tetraethyl orthosilicate (TEOS) (Aldrich). The surfactant, water, ethanol and acid were mixed together to allow the surfactant to completely dissolve before TEOS was added. Mass amounts of each reagent in the preparation are shown in Tables E-10a and E-10b for two separate formulations. In the second formulation (# 56483-5), the amount of ethanol was reduced approximately by half.
After TEOS hydrolysis (indicated by the exothermic reaction), the solutions were aged for 1 h prior to coating. The silica species within the precursor solutions would be expected to change with aging. Though these have not be characterized, it was found that the XRD peak intensities improved after 1 h aging.
The substrates were flooded with spin-coating solutions and spun 3000 rpm with a Specialty Coating System Model P-6204A, using the maximum acceleration setting (spin-up time <1 s). To increase silica condensation, coated substrates were post-treated by exposing films to the vapors from drops of concentrated ammonia under an inverted beaker for about 15 min, followed heating at 105° C. overnight in air and calcination at 450° C. To prevent cracking of glass substrates, samples were placed in the box furnace prior to bringing up to temperature.
In the XRD patterns (see FIGS. 20 and 21), the (100) and (200) reflections are apparent. The absence of the (110) reflection (or other higher order peaks) suggests the (100) family of planes of the hexagonal array are parallel to the substrate surface. The width of the (100) reflection does not change significantly with calcination, indicating the good stability of the mesoporous structure. Because of the difficulty in located the sample in the same position in the X-ray diffractometer before and after calcination, no interpretation can be made on the change in the peak height (i.e., the decrease of the peak height in
The top surface of the spin-coater chuck, on which the covers slips sits during spin coating, consists of a series of radial groves approximately 1 mm apart. This pattern of groves also appears in the interference colors of the films spun from both precursor solutions. It is believed that the chuck acts as a heat sink for the thin cover slip. As the solvent evaporates during spin coating, spatial variations in heat transfer through the cover slip results in temperature variations which in turn results in variations in the film thickness.
While a preferred embodiment of the present invention has been shown and described, it will be apparent to those skilled in the art that many changes and modifications may be made without departing from the invention in its broader aspects. The appended claims are therefore intended to cover all such changes and modifications as fall within the true spirit and scope of the invention.
This application is a continuation-in-part of U.S. application Ser. No. 08/753,573 filed Nov. 26, 1996, now abandoned.
This invention was made with Government support under Contract DE-AC06 76RLO 1830 awarded by the U.S. Department of Energy. The Government has certain rights in the invention.
| Number | Name | Date | Kind |
|---|---|---|---|
| 4913966 | Garvey | Apr 1990 | A |
| 5057296 | Beck | Oct 1991 | A |
| 5098684 | Kresge | Mar 1992 | A |
| 5102643 | Kresge | Apr 1992 | A |
| 5104515 | Chu | Apr 1992 | A |
| 5108725 | Beck | Apr 1992 | A |
| 5112589 | Johnson | May 1992 | A |
| 5145816 | Beck | Sep 1992 | A |
| 5156829 | McCullen | Oct 1992 | A |
| 5198203 | Kresge | Mar 1993 | A |
| 5211934 | Kresge | May 1993 | A |
| 5215737 | Chu | Jun 1993 | A |
| 5238676 | Roth | Aug 1993 | A |
| 5250282 | Kresge | Oct 1993 | A |
| 5256277 | Del Rossi | Oct 1993 | A |
| 5264203 | Beck | Nov 1993 | A |
| 5300277 | Kresge | Apr 1994 | A |
| 5321102 | Loy | Jun 1994 | A |
| 5364797 | Olson | Nov 1994 | A |
| 5470802 | Gnade | Nov 1995 | A |
| 5472913 | Havemann | Dec 1995 | A |
| 5494858 | Gnade | Feb 1996 | A |
| 5504042 | Cho | Apr 1996 | A |
| 5523615 | Cho | Jun 1996 | A |
| 5561318 | Gnade | Oct 1996 | A |
| 5565142 | Deshpande | Oct 1996 | A |
| 5622684 | Pinnavaia | Apr 1997 | A |
| 5625108 | Perego | Apr 1997 | A |
| 5647962 | Jansen | Jul 1997 | A |
| 5661344 | Havemann | Aug 1997 | A |
| 5723368 | Cho | Mar 1998 | A |
| 5736425 | Smith | Apr 1998 | A |
| 5753305 | Smith | May 1998 | A |
| 5789819 | Gnade | Aug 1998 | A |
| 5795556 | Jansen | Aug 1998 | A |
| 5795559 | Pinnavaia | Aug 1998 | A |
| 5800799 | Pinnavaia | Sep 1998 | A |
| 5804508 | Gnade | Sep 1998 | A |
| 5807607 | Smith | Sep 1998 | A |
| 5840271 | Carrazza | Nov 1998 | A |
| 5847443 | Cho | Dec 1998 | A |
| 5858457 | Brinker et al. | Jan 1999 | A |
| 5922299 | Bruinsma | Jul 1999 | A |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 08753573 | Nov 1996 | US |
| Child | 08921754 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 08921754 | Aug 1997 | US |
| Child | 09481988 | US |
