Nanostructured Silica and Silica-Derived Materials (Nanotechnology)

INTRODUCTION

Silicates are the most commonly found minerals as they are compounds of silicon and oxygen, which together make up almost 75% of the earth’s crust. Consequently, silicates are a basic building block of various natural or artificial structures. With emerging interests in nanoscale science and technology, silicate materials containing nanometer-sized structures have been studied and their new and interesting physical, chemical, and biological phenomena are begun to be understood. The control of silica material structures at a nanoscopic level attracts a great deal of interest among the scientific community in the last decade and remains an ongoing challenge. While various kinds of silica polymeric materials have been known for years, new morphologies have recently appeared, such as spherical hollow spheres, transparent films, meso-porous materials with narrow pore size distribution, and nanoporous thin films. In this topic, we discuss some old and new aspects of the synthesis of silica polymeric nanostructures, as well as their use in emerging technologies.

SILICA NANOPARTICLES

In 1968, Stober and Fink[1] hydrolyzed a dilute solution of tetraethylorthosilicate (TEOS) in ethanol at high pH and obtained uniform spheres of amorphous silica whose sizes could be varied from 10 nm to 2 mm simply by changing the concentrations of the reactants. Examples of dense silica nanoparticles synthesized according to the Stober process are depicted in Fig. 1.


This method was later improved by many others and appears to be the simplest and most effective route to monodispersed silica spheres.[2] Two types of reactions occur in the formation of silica particles: silanol groups are formed by hydrolysis of alkoxysilane groups and siloxane bridges are formed by a condensation polymerization reaction.

Recent mechanistic studies suggest that several types of micrometer-sized colloidal particles prepared by the precipitation method are in fact, the result of aggregation of much smaller subunits, or nanometer-sized primary particles, rather than continuous growth by diffusion of species from the solution to the nucleating surfaces.[3] In some cases, a broad range of size distribution was observed during the period of growth, indicating the occurrence of multiple nucleation events.[4] The uniformity of size in the final product may be achieved through a self-sharpening growth process during which small particles grew more rapidly than larger ones.

The surfaces of these silica colloids are typically terminated with three silanol types: free or isolated silanols, hydrogen-bonded or vicinal silanols, and geminal silanols (Scheme 1).

Silica nanoparticle surfaces can be rendered non-polar by post-treatment with hexamethyldisilazane (commonly used in industry for the manufacture of hydrophobic fumed silica), alkoxysilanes, or alcohols (esterification).[5-7] The treated silica nanoparticles can then be dispersed in various nonalcoholic solvents, such as methyl isobutyl ketone, cyclohexane, n-octane, or diethyl ether. Protected and unprotected mono-dispersed silica colloidal spheres are commercially available in large quantities from Nissan Chemical Industries, a company that has produced silica colloids for many years and whose monodispersed products are marketed under the trade name Snowtex®. Those readers with interest in the applications of monodispersed colloidal silica spheres can find a recent survey by Xia et al.[8]

Whereas dense silica particles have been known since 1956, spherical silica materials with hollow interiors have drawn interest in recent years because of their potential uses as low-density capsules for controlled-release drugs, dyes and inks, development of artificial cells, protection of proteins, or enzymes, and catalysis.[9] Many procedures for the generation of hollow silica spheres exist, including layer-by-layer deposition on (sub)micrometer-sized latex particles,[10] interfacial synthesis,[11] the use of hybrid polystyrene/siloxane latex particles,[12,13] and more recently, the controlled precipitation of silicic acid on functionalized polystyrene latexes[14] or the liquid phase deposition of silica onto fullerenol surfaces.[15] Two major requirements of any of these synthetic methods are to ensure a homogeneous deposition of silica and to prevent multiparticle aggregation. Hereafter we describe two of the aforementioned strategies as examples for the synthesis of spherical silica particles with hollow interiors. The first approach is divided in two steps as shown in Scheme 2. First, 3-(trimethoxysilyl) pro-pyl methacrylate (MPS) is covalently attached to the surface of polystyrene latex particles leading to silanol-functionalized particles. Then, the silanols are used to nucleate the deposition of a silica layer onto the hybrid particles’ surface by reaction with TEOS in aqueous basic alcoholic suspensions. Using this process, stable nonaggregated suspensions of hollow silica nanoparticles with 20 nm thick shells are produced as illustrated by the transmission electron microscope (TEM) images in Fig. 2. It is notable that dense silica particles are not detected, indicating that the polycondensation reaction only takes place at the seed surface.

Scanning electron microscope pictures of dense silica nanoparticles (left: 50 nm diameter; right: 150 and 200 nm diameter).

Fig. 1 Scanning electron microscope pictures of dense silica nanoparticles (left: 50 nm diameter; right: 150 and 200 nm diameter).

The second approach involves the slow controlled precipitation of silica from a sodium silicate solution onto polystyrene latexes bearing amine functionalities or amine and carboxylate groups (zwitterionic surface). Control of the sodium silicate solution pH is critical. A pH of 9.7 ensures a homogeneous coating of the latex particles in 24 hr and minimizes the non-templated precipitation of silica particles (Scheme 3). Interestingly for this approach, deposition of the coated particles on a surface from aqueous solution by slow evaporation leads to highly organized hexagonal lattices. The voids between adjacent spheres are still open, indicating that the organized array is indeed formed by the stacking of silica coated spheres (Fig. 3).

Scheme 1 Surface silanol types.

Scheme 1 Surface silanol types.

For both strategies, the latex core is removed by calcination in a final step leading to the formation of hollow silica spheres with theoretically closed central pores, the pore size being dictated by the size of the starting polymeric core. While closed cell porosity is advantageous in the case of hollow silica nanoparticles for the applications mentioned previously, other applications benefit more from open porosity and high surface area. These include ion exchange, gas adsorption, heterogeneous catalysis, hydrogen and or methane storage, and many others.

NANOPOROUS MEDIA

Porous media are very useful in numerous applications for every day life. When the pore size reaches several nanometers, they can find potential applications in state-of-the-art technologies including electronics, optics, biotechnology, and so on. Porous materials may be fully crystalline (zeolites), ordered on a meso-scopic length scale but amorphous on a atomic length scale (surfactant-templated materials), or fully disordered (silica gels).[16'17] While the engineering of micro structure has been achieved in the case of zeolites for more than 50 years, precise control of the pore size distribution and the shape and volume of the void spaces for mesoporous silicates only emerged in the early 1990s with the synthesis of the so-called M41S periodic mesoporous silica.[18] In the synthesis of this material (Scheme 4), the hydrolysis and polycondensation of the silica precursor was templated by micelles preformed from a surfactant, an amphiphilic molecule that self-organizes in aqueous media into supramole-cular arrays due to polarity differences between the lyophobic head group and lyophilic tail. While a structure-directing agent (surfactant) is used for the synthesis of both zeolites and M41S materials, only in the latter case is a true template mechanism operative as a direct correlation of the surfactant array size and shape to final pore size and geometry is observed. The pores of these solids are classified according to size: pores size below 2 nm are called micropores, those in the range 2-50 nm are denoted mesopores, and those above 50 nm are macropores.[19,20]

Scheme 2 Formation of hollow silica beads.

Scheme 2 Formation of hollow silica beads.

Transmission electron microscope micrographs of (A) SiOH-functionalized latex particles, (B) silica-coated latex particles, and (C) hollow silica nanoparticles.

Fig. 2 Transmission electron microscope micrographs of (A) SiOH-functionalized latex particles, (B) silica-coated latex particles, and (C) hollow silica nanoparticles.

Scheme 3 Silica coating of amino-functionalized latex particles.

Scheme 3 Silica coating of amino-functionalized latex particles.

Scanning electron microscope image of an organized multilayer of 200-nm-sized coated spheres after calcinations. With the chosen acceleration voltage, the complete coating is not visible in this case.

Fig. 3 Scanning electron microscope image of an organized multilayer of 200-nm-sized coated spheres after calcinations. With the chosen acceleration voltage, the complete coating is not visible in this case. 

Depending upon the nature of the surfactant (e.g., cationic, anionic, neutral, zwitterionic, bolaamphiphile, Gemini, divalent, and commercially available polymers), different mesophases are obtained: MCM-41 (hexagonal), MCM-48 (cubic), MCM-50 (lamellar), SBA-1, SBA-6, HMS, MSU-n, MSU-V, and many others.[21] When a charged surfactant is used, calcination of the as-synthesized material is mandatory to fully remove the template, whereas in the case of a neutral surfactant washing is sufficient. This neutral templating route allows the direct synthesis of hybrid organic-inorganic mesoporous materials where an organic molecule can be incorporated either in the channels or inside the walls.[22,23] These organically functionalized silicas present a very high surface area and tightly controlled porosity with a loading of organic groups that can be varied over a large range.

Scheme 4 General mechanism for the formation of M41S materials.

Scheme 4 General mechanism for the formation of M41S materials.

A typical structure of multilayered BEOL.

Fig. 4 A typical structure of multilayered BEOL.

In the semiconductor industry, scientists have recently drawn inspiration from the molecular tem-plated synthesis of mesoporous solids to develop materials with low dielectric constant (low k) for advanced microelectronics. As electronic devices get smaller and smaller, improvements in back end of the line (BEOL) interconnect performance require the reduction of resistance and capacitance, hence, new insulating materials having lower dielectric constant are needed.[24] Fig. 4 represents a typical BEOL structure containing multilayers of hard mask, dielectrics, etch stop, and cap coating. For over 30 years, silicon dioxide (SiO2) has been the dielectric insulator of choice for the semiconductor industry due to its excellent dielectric breakdown strength, a high modulus, good thermal conductivity, and excellent adhesion to other materials. However, SiO2 is being replaced with materials possessing lower permittivity to achieve reduced capacitance for next technology generations. For example, at the 65 nm technology generation, the target effective dielectric constant (keff) according to the 2001 International Technology Roadmap for Semiconductors is 2.3-2.6.[25] The keff is a composite value comprised of the dielectric, hard mask, barrier cap layer, and etch stop layer as schematically shown in Fig. 4. To achieve this ultra low dielectric constant, materials with k < 2.2 will be required. Porous structure is unavoidable choice for dielectric material in order to attain sufficiently low dielectric constants.

The incorporation of nanosized pores into a matrix structure leads to a significant decrease in the dielectric constant of the bulk material since the dielectric constant of air is about one. The dielectric constant of the material can therefore be adjusted by simply varying the level of nanoporosity. Among a variety of method to generate nanoporosity in organic or inorganic films, an approach developed in IBM, known as the so-called sacrificial porogen approach, has been recognized as a simple and effective route to thin films containing nanometer sized pores.[26] It utilizes phase separation of two component systems where one component (e.g., organosilicates matrix) crosslinks into a network effectively limiting domain growth and coarsening of the porogen phase (an organic, labile polymeric component) that is ultimately expelled from the film by thermal decomposition. Fig. 5 depicts the schematics of pore generating process. The morphology of nanohybrid, where the phase separated porogen domain is entrapped within crosslinked organosilicate matrix, and hence the pore morphology is strongly dependent on the interaction between porogen and matrix material, and molecular weight, molecular architecture, and loading level of the porogen.

Schematics of the sacrificial porogen approach to nanoporous organosilicates. Two distinct microstructures of pores are observed with different molecular architectures of porogens as shown in the right side.

Fig. 5 Schematics of the sacrificial porogen approach to nanoporous organosilicates. Two distinct microstructures of pores are observed with different molecular architectures of porogens as shown in the right side.

Transmission electron microscope images of porous thin films generated by two different routes. Left: NG; Right: TP.

Fig. 6 Transmission electron microscope images of porous thin films generated by two different routes. Left: NG; Right: TP.

This process is able to generate various morphology and dimensions of the pores. Two distinctive pore morphologies could be observed depending on the phase separation kinetics. As shown in the right side of Fig. 5, when the porogen and matrix undergo the liquid-solid phase separation caused by the crosslink-ing of the matrix material, which is a typical nucleation and growth (NG) process, random shapes of pores with broad size distribution are obtained. Porogens of homopolymers or copolymers with linear or branched molecular architectures show NG type phase behavior. In contrast, well-defined, quite-ordered pore structure with narrow size distribution is obtained when preassembled porogens template the porous structure.[27] By the definition of templating (TP), the procedure involves a preassembled mould with the specific morphology that can be transferred to the final porous structure. As shown by the cross-sectional TEM images in Fig. 6, microstructures of pores generated by two routes are very different. TP process (right-side image of Fig. 6) provides well-defined regular pore structures while NG process (left-side image of Fig. 6) gives random structures of pores.

In addition to the low dielectric constant, the nanoporous organosilicate thin films have a number of interesting properties which make them very attractive for many potential applications including optical components, catalyst supports, separation media, high density biosubstrates, and so on. Due to the nanoscopic dimensions of pores, these materials are optically homogeneous within visible wavelength of light. As shown in Fig. 7, for example, the refractive index of thin films of organosilicates also can be easily controlled over broad ranges (1.16-1.35 for this example) by simply varying the amount of porgen loadings. One simple application of this controllable refractive index is optical coating on glass to enhance transmission of light (i.e., antireflection effect).[28,29] With a single layer of homogeneous dielectric material coating on a substrate, antireflection effect can be achieved if n22 = n1n3 and h = 1/4, where h is coating thickness, 1 is the wavelength of incident light, n1, n2, and n3 are the refractive indices of substrate, dielectric material, and air, respectively. A dielectric material of n2 = 1.22 is required for glass substrate (n1 ~ 1.5, n3 ~ 1), which is not attainable with pure solid dielectric materials as the lowest n2 of dielectric materials is about 1.35 (cryolite, n2 = 1.38 for MgF2).[30] Fig. 8 shows the optical transmission spectra of glass slides coated with ~60 nm of nanoporous poly(methylsilsesquioxane) (PMSSQ) generated from 50wt% porogen loading (n ~ 1.22). Compared to the uncoated glass, single and double sided nanoporous coatings increase the optical transmission through the glass from ~91.7% to ^94.4% and ^98.2%, respectively. In addition, thin films of the nanoporous film with optical thickness ranging within visible wavelength are found to be robust.

Refractive index of porous organosilicate thin films as a function of porosity.

Fig. 7 Refractive index of porous organosilicate thin films as a function of porosity.

 Antireflection effect of nanoporous thin films coated on a glass slide.

Fig. 8 Antireflection effect of nanoporous thin films coated on a glass slide.

Due to the nanoscopic dimension of pores and intrinsic hydrophobicity of PMSSQ, the porous PMSSQ films show very selective sorption behavior. A quartz crystal microbalance (QCM) combined with reflectance infrared spectroscopy (IR) study revealed this selectivity.[31] With organic liquids having surface tension below the critical value (38-48 dyne/cm), the extent of sorption increased with porosity of PMSSQ films. Extremely low amounts of sorption were measured with the liquids having a surface tension higher than the critical value. The selective sorption behavior can be interpreted by capillary condensation resulting from the lowered vapor pressure in the nanoscopic pores and the experimental data are in good agreement with calculations using the Kelvin equation.

Water contact angles of dense PMSSQ surface as a function of UV/ozone treatment time at 30°C.

Fig. 9 Water contact angles of dense PMSSQ surface as a function of UV/ozone treatment time at 30°C.

Nanoporous PMSSQ surface with hydrophilicty-contrasted patterns.

Fig. 10 Nanoporous PMSSQ surface with hydrophilicty-contrasted patterns.

The nanoporous PMSSQ thin films show potentials as biosubstrates when the hydrophilicity is controlled. A wide range of surface hydrophilicity can be obtained by a simple UV/ozone treatment on nanoporous PMSSQ films. Although the precise mechanism of UV/ozone treatment has remained unclear in literature, it has been widely used for a variety of etching/ cleaning applications in the microelectronics industry. It is known that ozone is dissociated by absorption of 253.7 nm radiation or thermal heating into atomic oxygen which is postulated to be the predominant etchant species. Over the temperature range from room temperature to —300° C, organic materials are broken down into simple volatile oxidation products such as carbon dioxide, water, etc.[32,33] Fig. 9 shows water contact angles on dense PMSSQ surface as a function of UV/ozone treatment time at 30°C. The treatment gives surfaces having water contact angles ranging from 105° to 25° with increasing treatment time.

Since biosystems are based on aqueous environment, controlled hydrophilicity of solid surface is very useful for numerous applications in biotechnology including microarrays, affinity separation channels, etc. The ability to control surface hydrophilicity is applicable for pattern generation by limiting the area of UV/ozone treatment using a mask. Hydrophilic patterns in hydrophobic matrix can be obtained since only the area exposed both to UV and ozone becomes hydrophilic while masked areas remain as hydro-phobic. Fig. 10 shows the patterning process and an optical micrograph of a patterned surface where hydrophilic area (diameter = 250 mm) is decorated with a fluorescent dye (6-FAM amidite). Due to the porous, three-dimensional structure of the matrix, the patterned substrate has higher number density of functional groups than flat surface, which results in ten times higher fluorescence intensity than the corresponding flat silicon wafer surface. The arrays of patterns containing high surface area are desirable to reduce the overall size of an array while maximizing the number of reaction sites within the pattern and minimizing the required reagent and sample volume.[34]

CONCLUSIONS

Silicates with controlled nanoscopic structures are very attractive for a variety of potential applications due to their novel and interesting physical, chemical, and biological properties. Effort continues to explore new nanostructures, to optimize properties, and to utilize them for numerous end-uses, including optical components, electronics, separation media, catalysis, and biotechnology.

Next post:

Previous post: