J. Phys. Chem. B 2008, 112, 11881–11886
11881
ARTICLES Transparent, Hydrophobic Composite Aerogels with High Mechanical Strength and Low High-Temperature Thermal Conductivities Te-Yu Wei,†,‡ Shih-Yuan Lu,*,† and Yu-Cheng Chang‡ Department of Chemical Engineering, National Tsing-Hua UniVersity, HsinChu, Taiwan 30013, Republic of China, and Industrial Energy-SaVing Technology DiVision, Energy & EnVironment Laboratories, Industrial Technology Research Institute, HsinChu, Taiwan 310, Republic of China ReceiVed: June 2, 2008; ReVised Manuscript ReceiVed: July 21, 2008
A facile one-step polymer-incorporation sol-gel process, together with a surface modification and an ambient pressure drying processes, was developed to prepare silica-poly(vinylpyrrolidone) composite aerogels. These composite aerogels are with high hydrophobicity (static contact angle >120°), good mechanical strength (Young’s modulus of bending >30 MPa), and low high-temperature thermal conductivity (0.063 W/m-K at 300 °C), which are critical characteristics for practical applications of aerogels, particularly in energy saving areas, for long-term usage and large scale production. Introduction The ever worsening problems of fossil energy depletion and global warming have become a grand challenge and survival threat to human being, and have thus greatly boosted the research efforts in development of clean alternative energy and emission control of global warming gases. These efforts however would prove insufficient if not accompanied with energy saving practices. One of the major energy saving practices is thermal insulation, e.g., reducing fuel and electricity usages through better insulating boilers and buildings, respectively. Silica aerogels, a class of mesoporous materials with extremely high porosities and specific surface areas, have found potential applications in high temperature insulation, low dielectric constant thin layer, low refractive index glass, catalysts, electrodes, and absorbents.1-6 Their applications as thermal insulators are particularly promising because of the existence of high percentage of highly interconnected pores, with apparent thermal conductivities commonly lower than 0.1W/m-K at room temperature.6,7 Native silica aerogels however have found difficulties in commercialization because of their fragility, hydrophilicity, and demand of supercritical drying in production. Fragility makes it very difficult to apply silica aerogels in monolithic form while hydrophilicity restricts silica aerogels from long-term usage since cracking can readily occurs through moisture adsorption at ambient conditions. Furthermore, supercritical drying limits silica aerogels from large scale production. Because of these drawbacks, silica aerogels have found only very limited use in highly specific situations such as in NASA’s Stardust program for capturing particles in space8 and in nuclear reactors as radiation detectors.9,10 Consequently, researchers in this area devote efforts to improve the mechanical strength of, introduce hydrophobicity to, and avoid * To whom correspondence should be addressed. E-mail: sylu@ mx.nthu.edu.tw. † Department of Chemical Engineering, National Tsing-Hua University. ‡ Industrial Energy-Saving Technology Division, Energy & Environment Laboratories, Industrial Technology Research Institute.
use of supercritical drying in production of silica aerogels, while trying to maintain their low thermal conductivities, particularly at high temperatures. Researchers have tried to improve the mechanical strength of silica aergoels by aging11 and by incorporating fibers as reinforcement reagents.12 These efforts however received only limited success since the Young’s moduli of bending of the resulting aerogels were only 1.6 MPa and 23 MPa, respectively, and the products were all hydrophilic. Hydrophobic methylsilsesquioxane-based (Me-SiO1.5) aerogels and xerogels of good mechanical strength however have been prepared with a onestep sol-gel route by using a single trifunctional precursorsmethyltrimethoxysilane.13 The production of such aerogels however requires use of surfactants and supercritical drying, and the values of thermal conductivity were not reported. A more promising approach is hybridization of polymers into silica aerogels, taking advantages of the mechanical strength and possibly hydrophobicity of the polymer and the thermal stability of the silica backbone. Two processes are particularly worthy mentioning. The first process involves surface modification of the silica backbone with amino groups through cohydrolysis of silica precursors with 3-aminopropyltriethoxysilane (APTES), followed by cross-linking of desired polymer precursors to form a conformal polymer coating, such as polyurethane, polyurea, and polystyrene, on the silica backbone, from which the mechanical strength and possibly hydrophobicity (for the polystyrene case) of the aerogel can be greatly improved.14-18 These wet silica-polymer composite gels were then supercritically dried to obtain the final products. Young’s moduli of bending as high as above 700 MPa have been reported for such composite aerogels. Nevertheless, in the process, toxic reagents, such as acetontrile, a carcinogenic chemical, and supercritical drying were often required and the whole process was rather complicated, often involving amino group decoration of silica backbone surfaces and cross-linking of polymer precursors for conformal polymer coating. In addition, these products were at
10.1021/jp804855v CCC: $40.75 2008 American Chemical Society Published on Web 08/27/2008
11882 J. Phys. Chem. B, Vol. 112, No. 38, 2008 most translucent and their high-temperature thermal conductivities were not reported. The second process introduces polymer into the silica network through cohydrolyzing surface functionalized core-shell polymer nanoparticles with the silica precursor. The incorporated polymers, both core and shell, possess low glass transition temperatures, from which the improvement in mechanical strength of the resulting composite aerogel is realized. This process, although succeeded in introducing mechanical strength and hydrophobicity into the otherwise brittle and hydrophilic silica aerogel, is again rather complicated, requiring preparation and surface modification of core-shell polymer nanoparticles and subsequent cohydrolysis of the polymer nanoparticles with the silica precursor.18,19 In this work, a facile one-step polymer-incorporation sol-gel process, together with a surface modification and an ambient pressure drying processes, was developed to prepare silica-poly(vinylpyrrolidone) (SiO2-PVP) composite aerogels. The polymerincorporation step was achieved by simply mixing solutions of desired polymers with the silica precursor solution at the beginning of the sol-gel process, avoiding the more complicated multistep preparations of the polymer additives as reported in the literature. 14-19 These composite aerogels were characterized to possess high hydrophobicity (static contact angle >120°), good mechanical strength (Young’s modulus of bending >30 MPa), and low high-temperature thermal conductivity (0.063 W/m-K at 300 °C). Here, PVP, a hydrophilic and water soluble polymer, was chosen as the polymeric additive for its relatively high thermal stability, up to 300 °C,20 and for enabling an ethanol based sol-gel process without use of hazardous organic solvents. The hydrophilicity of PVP was later shielded with a surface modification process for introduction of hydrophobicity to make possible the final ambient pressure drying operation.21-23 Because of the relatively high thermal stability of the polymeric additive, the resulting composite aerogels can be applied at temperatures up to 300 °C in areas such as solar tank24,25 and insulation boxes for accommodation of batteries and electronic devices.2 The hydrophobicity of the final product enables longterm applications, while the ambient pressure drying warrants potential large scale production. The thermal conductivity of 0.063W/m-K of the composite aerogels at 300 °C was lower than that of glass fibers (0.08 W/m-K), which are the traditional high-temperature thermal insulation materials.26 Experiment The silica-PVP composite aeorgels were synthesized with the hydrolysis and condensation of tetraethoxysilane (TEOS) in the presence of ethanolic PVP solution of different concentrations. The wet gel was dried at ambient pressure after the hydrophobicity-acquiring surface modification process. The molecular weight (M.W.) of the PVP used was 58,000. The procedures for the surface modification and ambient pressure drying followed from our previous work.21-23 In a typical run, TEOS (20.8 g) was mixed with EtOH (13.8 g) and was then added HCl (1.8 g of 0.14 wt % HCl/H2O solution) as the acidic catalyst. The resulting mixture was stirred for 1.5 h and then added to it an ethanolic PVP solution (0.03 g PVP in 23 g of EtOH for a final PVP weight percentage of 0.5, 0.045 g for 0.75 wt %, 0.06 g for 1 wt %) of desired concentration. Some amount of NH4OH (4.685 g of 0.15 wt % NH4OH/H2O solution) was then added as the basic catalyst. The molar ratio of TEOS: EtOH:H2O:HCl:NH4OH was set at 1:8:3.6:7 × 10-4:2 × 10-3. After the sample gelled, it was aged for 2 days, and then washed with EtOH to remove the nonreacted chemicals. The washing
Wei et al. process was repeated every 24 h for 3 times. After the washing, the ethanol trapped inside the pore structure of the wet gel was exchanged with n-hexane to avoid reaction with the surface modification reagent, trimethylchlorosilane (TMCS), later applied to the gel. The exchanging process was repeated every 24 h for 4 times to ensure a complete exchange. The wet gel was then made hydrophobic with repeating TMCS treatments, every 24 h for 4 times. With this treatment, the -OH groups exposed on the silica backbone surface were replaced with -OSi(CH3)3 groups from TMCS to make the aerogels hydrophobic.21 The acids, produced from the surface modification step and/or carried over from the unreacted chemicals, were removed by 4 times of n-hexane washing. The resulting hydrophobic wet gel was then dried at ambient condition for several days to obtain the desired silica-PVP composite aerogel. The apparent density was obtained by weight divided by the volume of the samples. The BET specific surface area, specific surface area of the micropores, BJH pore volume, micropore volume, average pore size, and the pore size distribution were obtained from N2 adsorption/desorption analyses at 77 K (Quantachrome, NOVA e1000). The morphology of the composite aerogel was characterized by high-resolution TEM (HRTEM, JEOL JEM-3000) operated at 300 kV. The chemical bondings of the pure PVP and composite aerogel were determined from FT-IR (Perkin-Elmer Spectrum RX-I) and the static contact angle was determined from the water drop shape placed on the sample surface (DSA100, Kruss, Hamburg, Germany). The thermal stability of the PVP and composite aerogel was characterized by TG analysis from room temperature to 700 °C (Dupont instruments, TGA951). The thermal conductivity of the samples, thin disks with a thickness of about 0.6 cm and diameter of 2.1-2.3 cm, at different temperature were determined with the transient plane heat source method (Hot-Disk, Gothenburg, Sweden).33 Young’s moduli of bending of the samples were measured by three point flexural bending method (Lloyd, LFPlus series). Result and Discussion Composite aergels of three different PVP concentrations (0.5, 0.75, 1wt %) were prepared and compared. Here, the weight percentage is based on the final dried product. Also included in the comparison was the native silica aerogel prepared with the same sol-gel conditions but dried supercritically with CO2. These aerogel samples were prepared in the form of monolithic thin disk with a thickness of about 0.6 cm and diameter of 2.1-2.3 cm. As shown in Figure 1, the native silica aerogel and the three silica-PVP composite aerogels all appeared transparent and carried a blue tone. The native silica aerogel appeared bluer and more transparent than the three composite aerogels. This observation can be supported by the typical transmittance spectra of the native and composite aerogels shown in Figure 2. The lower transmittances in the shorter wavelength regime accounted for the blue tone. As a comparison, the native silica aerogel gave significantly lower transmittance in the wavelength range of 400-520 nm, but higher transmittance at wavelengths longer than 520 nm, consistent with the appearance observation. The high enough transparency and not much blue tone of these composite aerogels are good for them to serve as exterior insulation glass for buildings. Table 1 lists the relevant structural parameters, thermal conductivities, and Young’s moduli of bending of these samples. There can be observed several points from this table. First, expectedly, with incorporation of PVP, the density of the aerogels increased while the specific surface area and pore
Composite Aerogels
J. Phys. Chem. B, Vol. 112, No. 38, 2008 11883
Figure 1. Photographs of monolithic thin disks of the native silica aerogel and silica-PVP composite aerogels of three different PVP concentrations (0.5, 0.75, 1.0wt %), showing good transparency.
Figure 2. Transmittance spectra of the native silica aerogel and composite aerogel of 1 wt % PVP.
volume dropped, as compared with those of the native silica aerogels. Here, the increase in density resulted not only from the incorporation of PVP, but also from the more intensive volume shrinkage encountered in ambient pressure drying. The volume shrinkage, defined as (1 - Vm/Vo) × 100% with Vo and Vm representing the apparent volumes of the as-prepared sample before and after drying, respectively,21 of the composite aerogels and native silica aerogels from supercritical drying were 78-80% and 45%, respectively. The more pronounced volume shrinkages of the composite aerogels were a direct result of the surface tension force developed when the solvent left the
Figure 3. Nitrogen adsorption/desorption isotherms of native silica aerogel and silica-PVP composite aerogels of three different PVP concentrations (0.5, 0.75, 1.0wt %). The left inset shows the pore size distributions of the four aerogel samples and the right inset illustrates a proposed scheme for the composite aerogel structure.
aerogels in the ambient pressure drying process. Other direct results of the more intensive volume shrinkage are the drastic decreases in pore volume, more than 100% reduction, and pore size (discussed later). Although the density increased quite significantly, there was only a slight increase in room temperature thermal conductivity. The mechanical strength, manifested as the magnitude of Young’s modulus of bending, however improved drastically, from below the detection limit to over 20 MPa. Second, with increasing incorporation of PVP, the density, thermal conductivity, and Young’s modulus of bending of the aerogel continued to increase, while the pore volume continued to drop. Interestingly, the specific surface area of the aerogel increased and micropores started to appear with increasing PVP concentration. This phenomenon implies that the incorporated PVP was present in the form of aggregates within the mesopores of the silica network instead of as a coating layer for the silica backbone, as illustrated in the right inset of Figure 3. The aggregates offered extra surface area and emerging micropores. At the PVP loading of 0.5 wt %, the incorporated PVP nanoparticles were mainly distributed on the surface of the silica backbone. There were no micropores created from the distribution of the PVP nanoparticles at this stage. Further increasing of the PVP loading resulted in the packing and aggregation of the PVP nanoparticles for generation of micropores. The micropore volume thus did not scale linearly with the PVP loading. As to the improvement in mechanical strength, the increasing density with increasing PVP loading played a positive
TABLE 1: Properties of Silica-PVP Composite Aerogels
aerogela 0.5 wt %b 0.75 wt %b 1 wt %b
density,c g/cm3
BET specific surface area,d m2/g (specific surface area of micropores)e
BJH pore volume,f cm3/g (micropore volume, cm3/g)g
thermal conductivity,h W/m-K
Young’s modulus of bending,i MPa
0.218 ( 0.002 0.368 ( 0.002 0.377 ( 0.003 0.430 ( 0.003
1199 ( 14 (0) 823 ( 4 (0) 896 ( 8 (85) 1011 ( 11 (225)
3.43 (0) 1.53 (0) 1.43 (0.05) 1.39 (0.13)
0.041 ( 0.0003 0.045 ( 0.0006 0.048 ( 0.0004 0.049 ( 0.0005
below detection limit 24.3 ( 1.8 30.7 ( 2.0 39.3 ( 2.0
a Native silica aerogels. b PVP concentration. c Average of six samples. d Specific surface area obtained with the Brunauer-Emmett-Teller model based on data of P/P0 of 0.1 to 0.3. e Specific surface area from micropores estimated with the t-plot method. f BJH pore volume obtained with the Barrett-Joyner-Halenda model based on N2 desorption data. g Micropore volume obtained with the t-plot method. h Thermal conductivity at room temperature (averages of six samples). i Averages of six samples.
11884 J. Phys. Chem. B, Vol. 112, No. 38, 2008
Wei et al.
Figure 4. TEM images of (a) the native silica aerogel and (b) the composite aerogel of 1 wt % PVP, revealing the hierarchical highly porous structure in both samples.
role as expected. But the drastic increase in Young’s moduli of bending came mainly from the incorporation of PVP. The pore structure of the samples was further analyzed with N2 adsorption/desorption at 77K. As shown in Figure 3, all samples exhibited type IV isotherms typical for mesoporous materials.27-29 The characteristics of the hysteresis loops were however different for the composite and native silica aerogels, with the former being type H3 and the latter type H1. Type H3 hysteresis loops are indicative of slit-shaped pores while type H1 hysteresis loops indicated cylindrical-like pores.27-29 The limited N2 uptake obtained at low relative pressures (
