Membranes made from nanomaterials

In some instances, membranes may be made (nearly) entirely of products of nanochemistry. The advantages of these materials may include improved processing, due for example to the lower sintering temperature required when nanoparticles are used as precursors to ceramic membranes. Or they may include include improved performance linked directly with the properties of the materials.

Ceramic membranes derived from mineral nanoparticles. Mineral membranes have been made from a variety of mineral nanoparticle precursors. Commercially available ceramic membranes are typically made from metal oxides such as Al2O3, ZrO2, and TiO2 [15]. However, membranes can be made from many other nanomaterials ranging from gold [16, 17] to SiO2 [18]. In most cases nanoparticles are deposited on a support surface and then calcined to create the membrane. Processes differ in the manner in which nanoparticle precursors are prepared. One common procedure for producing nanoparticle precursors to these membranes is to precipitate particles under controlled conditions creating a suspension or sol of nanoparticles that are deposited on a surface and dried to form a gel. This procedure is known as sol-gel.

Sol-gel involves a four-stage process: dispersion, gelation, drying, firing. A stable liquid dispersion or sol of the colloidal ceramic precursor is initially formed in a solvent with appropriate additives. In the case of alumina membranes, this first step may be carried out with 2-butanol or iso-propanol. By removing the alcohol, the polymerization of aluminum monomers occurs leading to a precipitate. This material is acidified, typically using nitric acid, to produce a colloidal suspension. By controlling the extent of aggregation in the colloidal sol, a gel of desired properties can be produced. The aggregation of colloidal particles in the sol may be controlled by adjusting the solution chemistry to influence the diffuse layer interactions between particles, by adding stabilizing agents such as surfactants, or through ultrasonification. Knowing that the properties of the gel will influence the permeability of the future membrane, it is clear that the gelation step is extremely important. This gel is then deposited, typically by a slip-cast procedure, on an underlying porous support. In variations on this procedure, functionalized surfaces may also be used to achieve a more ordered deposit or micelles can be used to direct film formation in specific geometries through self-assembly. In conventional sol-gel, the excess liquid is removed by drying and the final ceramic is formed by firing the resulting gel at higher temperatures. Drying and firing conditions have been shown to be very important in the structural development of the membranes, with higher drying rates resulting in more dense membrane films.

The sol-gel approach of reacting small inorganic molecules to form oligomeric and polymeric nanoparticles has several limitations such as difficulties in controlling the reaction conditions, and the stoichiometries, solubility, and processability of the resulting gel. It would thus be desirable to prepare nanoparticles in a one-pot bench-top synthesis from readily available, and commercially viable, starting materials, which would provide control over the products. One strategy for producing nanostructured membranes involves an environmentally benign alternative to the sol-gel process for ceramic membrane formation. Metal nanoparticles such as alumoxanes [19] and ferroxanes [20] can be produced based upon the reaction of boehmite, [Al(O)(OH)]n, (or lepidi-crocite in the case of the ferroxanes) with carboxylic acids [21]. The physical properties of such metal-oxanes are highly dependent on the identity of the alkyl substituents, R, and range from insoluble crystalline powders to powders that readily form solutions or gels in hydrocarbon solvents and/or water. Thus, a high degree of control over the nanoparticle precursors is possible. Metal-oxanes have been found to be stable over periods of at least years. Whereas the choice of solvents in sol-gel synthesis is limited, the solubility of the carboxylate metal-oxanes is dependent on the identity of the carboxylic acid residue, which is almost unrestricted. The solubility of the metal-oxanes may therefore be readily controlled so as to make them compatible with a coreactant. Furthermore, the incorporation of metals into the metal-oxane core structure allows for atomic scale mixing of metals and formation of metastable phases. In the case of the aluminum-based alumoxanes, the low price of boehmite ($ 0.5 per kilogram) and the availability of an almost infinite range of carboxylic acids make these species ideal as precursors for ternary and doped aluminum oxides.

Application of a metal-oxane-based approach to creating ceramic membranes reduces the use of toxic solvents and energy consumption. By-products formed from the combustion of plasticizers and binders are minimized, and the use of strong acids eliminated. Moreover, the use of tailored nanoparticles and their deposition on a suitable substrate presents an extremely high degree of control over the nanostructure of the resulting sintered film. The versatility of the process can be used to tightly control pore-size distributions. The MWCO of the first generation of alumoxane-derived membranes is approximately 40,000 daltons [22], which is in the ultrafiltration range. Table 9.4 shows a comparison of the ceramic and sol-gel methods with that of the carboxylate alumoxanes for the synthesis of alumina and ternary aluminum oxides. The ease of modification of the alumoxanes suggests that a single basic coating system can be modified and optimized for use with a range of substrates.

There has been interest in using ceramics as electrolyte materials for proton exchange membrane fuel cells because of their thermal, chemical, and mechanical stability and their lower material costs [23]. However, traditionally ceramic membranes have exhibited comparatively small proton conductivities. The conductivities of silica glasses fired at 400 to 800°C is in the order of 10~6 to 10~3 S/cm [24]. The conductivities of silica, alumina, and titania sintered at 300 and 400°C are in the range of 10~7 to 10~3 S/cm [25].

However, recent work suggests that membranes derived from fer-roxane nanoparticles may be attractive alternatives for such proton exchange membranes. With a conductivity of approximately 10 2 S/cm the ferroxane-derived membrane represents a large improvement over other ceramic materials prepared by the traditional sol-gel method, with conductivities close to that of Nafion (Table 9.5).

The protonic conductivity of these membranes varies as a function of the temperature at which they are sintered. For example, when fer-roxane films are sintered at 300°C the resulting membranes display a

TABLE 9.4 Comparison of the Alumoxane and Sol-Gel Synthesis Methods

Alumoxane

Sol-gel

Methodology

Atomic mixing

Metastable phases

Stability

Solubility

Processability

Time

Cost simple yes yes excellent readily controlled good <8 h low complex yes yes fair difficult to control good >20 h med.-high

TABLE 9.5 Representative Conductivities of Oxide Membranes and Nafion® Compared with Preliminary Results for Ferroxane-Derived Membrane. Conductivity is Reported at 100% Humidity and at 20°C

Material/Study

Conductivity

0 0

Post a comment