21582
J. Phys. Chem. B 2006, 110, 21582-21587
Structural Characteristics of Silica Sonogels Prepared with Additions of Isopropyl Alcohol D. A. Donatti, A. Iban˜ ez Ruiz, M. M. Kumakawa, and D. R. Vollet* UNESP - UniVersidade Estadual Paulista, IGCE, Departamento de Fı´sica, Cx.P. 178, 13500-970 Rio Claro (SP), Brazil ReceiVed: July 11, 2006; In Final Form: August 17, 2006
Wet silica gels with ∼1.4 × 10-3 mol SiO2/cm3 and ∼92 vol % liquid phase were obtained from sonohydrolysis of tetraethoxysilane (TEOS) with different additions of isopropyl alcohol (IPA). The IPA/TEOS molar ratio R was changed from 0 to 4. Aerogels were obtained by supercritical CO2 extraction. The samples were analyzed by small-angle X-ray scattering (SAXS) and nitrogen adsorption. The wet gels exhibit mass fractal structure with fractal dimension increasing from D ∼ 2.10 to D ∼ 2.22, characteristic length ξ decreasing from ∼9.5 to ∼6.9 nm, as R increases from 0 to 4, and an estimated characteristic length for the primary silica particles lower than ∼0.3 nm. The supercritical process apparently eliminates a fraction of the porosity, increasing the mass fractal dimension and shortening the fractality domain in the mesopore region. The fundamental role of isopropyl alcohol on the structure of the resulting aerogels is to decrease the porosity and the pore mean size as R changes from pure TEOS to R ) 4. A secondary structure appearing in the micropore region of the aerogels can be described as a mass/surface fractal structure, with correlated mass fractal dimension Dm ∼ 2.7 and surface fractal dimension Ds ∼ 2.3, as inferred from SAXS and nitrogen adsorption data.
Introduction A large variety of glass and glass ceramics has been obtained by a sol-gel process from the hydrolysis of tetraethoxysilane (TEOS).1 Since TEOS and water are immiscible, a mutual cosolvent such as ethanol is usually employed as a homogenizing medium in the conventional sol-gel method. Ultrasound is an alternative method to promote the TEOS hydrolysis in acidified solventless heterogeneous TEOS-water mixtures.2 The overall process involves hydrolysis and polycondensation reactions leading to the growth of clusters that eventually collide together to form a gel. Wet gels frequently exhibit a mass fractal structure consisting of a continuous solid network imbibed in a large-volume-fraction liquid phase. The aggregation process of very small silica particles to form mass fractal structures in wet sonogels has been studied at different temperatures.3 Supercritical drying (aerogel), freeze-drying (cryogels), and evaporation drying (xerogels) are usual methods in dried gel production. Most of the porosity is eliminated on the obtaining of xerogels due to capillary stresses imposed by the liquid phase on the drying process. The structure of aerogels is often similar to that of the wet gels except by the contrast variation resulting from the supercritical extraction of the liquid phase and some structural modification at a high-resolution level due to an apparent defining of a secondary mass/surface fractal particle.4 The characteristics of the structure of the gels are important for a wide variety of applications. Silica gels have been considered as appropriate matrixes for the preparation of complex-centerdoped materials for metallic ions5,6 and for encapsulation of organic7,8 and inorganic compounds,9 with interesting optical and/or electronic properties. The mesoporous structure has been considered as an important transport medium for a variety of applications such as controlled-release carrier-implantable materials for low-weight drugs in biological systems10,11 and as substitute materials for membrane processes in fuel cells.12 * Corresponding author: Tel/Fax: +55-19 526 2237. E-mail: vollet@ rc.unesp.br.
Beside the conditions of drying of the gels, the structural properties of the final product depend primarily on the starting materials and on the conditions for hydrolysis and aging of the gels as pH, alkoxide/water proportion, type of catalyst, temperature, solvent and structure-modifier additions, and the method employed for hydrolysis (ultrasound or conventional). Isopropyl alcohol (IPA) has been considered as an additive in the silica sol-gel process.13-15 The interchange reaction between metal alkoxide and isopropyl alcohol has been studied by NMR15 in the case of tetramethoxysilane. Isopropyl alcohol additions during the hydrolysis step of TEOS could affect the structure of the resulting gels due to alcoholysis, steric, and capillary effects. In this work, we probe the structural properties of the wet gels and aerogels prepared from sonohydrolysis of TEOS with different additions of IPA. The silica concentration in the wet gels was kept the same for all the IPA additions. The structural properties of the sonogels were studied by smallangle X-ray scattering (SAXS) and nitrogen adsorption as a function of the IPA/TEOS molar ratio. Experimental Section The samples were prepared from the sonohydrolysis of mixtures of TEOS, distilled and dionized water, isopropyl alcohol (IPA), and 0.1 N HCl as a catalyst. The IPA/TEOS molar ratio (R) was changed according to the nominal values R ) 0, 0.5, 1, 2, and 4. The proportion of the components in the mixtures was chosen according to Table 1 in order to obtain mixtures with the same water/TEOS molar ratio (∼6.45) and the same total volume (∼38 cm3) during the hydrolysis step of the process. The hydrolysis was promoted for 10 min under a constant power (∼0.7 W/cm3) of ultrasonic radiation. Each of the sols was diluted in water according to Table 1 in order to obtain sols with the same final silica concentration (∼1.4 × 10-3 mol SiO2/cm3). The sonication was then continued for 2 min for complete homogenization. A volume of 3.5 cm3 of 0.1
10.1021/jp0643552 CCC: $33.50 © 2006 American Chemical Society Published on Web 09/22/2006
Silica Sonogels Prepared with Isopropyl Alcohol
J. Phys. Chem. B, Vol. 110, No. 43, 2006 21583
TABLE 1: Preparation of the Silica Sols with Different Additions of Isopropyl Alcohol (IPA)a,b hydrolysis step
sol dilution
R
TEOS
H2O
IPA
0.1 N HCl
H2O
0.1 N NH4OH
0 0.5 1 2 4
25.0 22.5 20.4 17.2 13.1
8.0 6.7 5.5 3.9 1.8
0 3.8 7.1 11.9 18.1
5.00 ( 0.05 5.00 ( 0.05 5.00 ( 0.05 5.00 ( 0.05 5.00 ( 0.05
37.9 30.0 23.3 13.1 0
3.50 ( 0.05 3.50 ( 0.05 3.50 ( 0.05 3.50 ( 0.05 3.50 ( 0.05
a The quantities are given in cm3. b Nonspecified errors are in (0.1 cm3.
N NH4OH was yet added to the sols in order to accelerate the gelation process by increasing the pH. The resulting sols were cast in sealed containers and kept under saturated conditions for 30 days at 40 °C for gelation and aging. Monolithic pieces of wet gels were then obtained. The wet gels were characterized by density measurements, thermal gravimetric (TG) analysis, and small-angle X-ray scattering (SAXS). To obtain aerogels, the liquid phase of the wet gels was exchanged by ethanol at room temperature. The ethanol was then exchanged by liquid CO2 in an autoclave followed by supercritical CO2 extraction. Monolithic, 1-cm-diameter, 2-cmheight, cylindrical-shaped samples of aerogels were obtained after the supercritical CO2 extraction. The aerogels were studied by nitrogen adsorption and small-angle X-ray scattering. The SAXS experiments were carried out using synchrotron radiation with a wavelength λ ) 0.1608 nm. The beam was monochromatized by a curved silicon monochromator and collimated by a set of slits defining a pinhole geometry. A 1D position-sensitive X-ray detector was used to record the SAXS intensity as a function of the modulus of the scattering vector q ) (4π/λ)sin(θ/2), where θ is the scattering angle. The cross section of the beam at the sample was about 0.5 mm × 2.0 mm. The convergence system and the set of slits give a beam cross section of about 0.03 nm-1 × 0.03 nm-1 at the detector window. The experimental setup allowed us to get SAXS data from q0 ) 0.13 nm-1 to qm ) 3.4 nm-1 with a resolution of ∆q ) 4.87 × 10-3 nm-1. The data were corrected by the sample attenuation and the parasitic scattering and were normalized with respect to the beam intensity and the logarithm of the attenuation, which is proportional to the thickness of the sample. Nitrogen adsorption isotherms were obtained at liquid nitrogen temperature using an ASAP 2010 Micromeritics apparatus. The samples were degassed at 200 °C for no less than 24 h before the adsorption tests. The adsorption data were analyzed for the BET specific surface SBET, the total pore volume per mass unit Vp, as the total adsorbed volume of nitrogen at a single point close to the saturation pressure, the pore mean size lBET ) 4Vp/SBET, and the pore size distribution (PSD), as determined by the classical Kelvin equation and the Harkins & Jura model for cylindrical pores.16 Results Wet Gels. Table 2 shows the values for the density of the wet gels Fwg as a function of the IPA/TEOS molar ratio R. Figure 1 (bottom) shows the mass loss of the wet gels as determined by thermal gravimetry (TG) carried out at a heating rate of 5 °C/min and (top) the derivative of the TG signal. The great mass loss (∼83 wt %) up to about 300 °C as inferred from the peak in the derivative of the TG test is mainly associated with the evaporation of the liquid phase of the wet gel. The values of mass fraction of the liquid phase in the wet gels (∆m/m0)L was
thus estimated from the mass loss peaks and are shown in Table 2 as a function of R. The evaporation of the liquid phase is more delayed in the samples with lower R as suggested by the mass loss rates in the TG tests. This is likely due to the lower boiling point of the isopropyl alcohol compared to that of water. It could yet be due to some structural difference in the “mesopore” structure of the wet gels as a function of R, which would make the liquid-phase release more difficult in the samples with lower IPA content. Very few residual organic compounds remain in the silica structure, accounting for signifying mass loss due to oxidation reactions after the liquidphase evaporation. Thus, hydrolysis and polycondensation reactions are practically 100% complete for this set of lowdensity wet gels after the long period of aging at 40 °C under sealed conditions. The density of the liquid phase FL can be estimated for the wet gels from the measured Fwg and (∆m/m0)L by the equation
(1/FL) ) (1/FS) + [1/(∆m/m0)L][(1/Fwg) - (1/FS)]
(1)
where FS is the density of the solid particles (silica) of the gel, assumed as 2.2 g/cm3 as frequently quoted for fused silica. In addition, the volume fraction of the liquid phase φL can be obtained by φL ) (FS - Fwg)/(FS - FL). Table 2 shows the estimated values FL and φL for the wet gels. Figure 2 shows the SAXS intensity I(q) as a function of the modulus of scattering vector q of the wet gels. Since the volume fraction of liquid phase (φL) was found to be about 92% in the wet gels, we are more properly concerned with the scattering from the solid particles of the system. The SAXS intensity in Figure 2 shows characteristics of the scattering from a mass fractal system. The mass fractal characteristics can be recognized by typical power-law decrease on q of the SAXS intensity in a q-range given by ξ . q-1 . a, where ξ is the characteristic length of the fractal structure, and a is the characteristic length of the primary particles composing the structure.17 The SAXS intensity departs from the power-law regime at low values of q due to the finite correlation length ξ of the mass fractal structure. An approach accounting for this effect can be written as18
I(q) ) A sin[(D - 1) arctan(qξ)]/(1 + q2ξ2)(D-1)/2(D - 1)qξ (2) where D is the mass fractal dimension of the structure, a physically acceptable value between 1 < D < 3, and A is a constant for a given D and ξ. Figure 2 shows that the curves of the SAXS intensity of the wet gels are well fitted by eq 2. Table 2 shows the mass fractal parameters D and ξ as obtained from the fitting process. The mass fractal dimension D increases from D ∼ 2.10 up to D ∼ 2.22, while the characteristic length ξ decreases from ξ ∼ 9.5 nm to ξ ∼ 6.9 nm as R increases from pure TEOS (R ) 0) up to R ) 4. The values for ξ could be effectively greater than those in Table 2 since a clear crossover to Guinier’s law was not found in the SAXS intensity at low q. A slight positive deviation from eq 2 is apparent in the SAXS intensity curves at high q, particularly for the samples with isopropyl alcohol additions. This contribution may be associated with a slightly different secondary mass fractal structure at a higher-resolution level. Equation 2 cannot account for the fractality limit at high q associated with the finite characteristic length a of the primary particles composing the fractal structure. The lack of a crossover to Porod’s law at high q [I(q) ∼ q-4] suggests that a < ∼1/qm
21584 J. Phys. Chem. B, Vol. 110, No. 43, 2006
Donatti et al.
TABLE 2: Structural Properties of the Wet Gelsa R
Fwg (g/cm3)
(∆m/m0)L (%)
FL (g/cm3)
φL
D
ξ (nm)
Fξ (g/cm3)
0 0.5 1 2 4
1.093 ( 0.005 1.079 ( 0.005 1.061 ( 0.005 1.032 ( 0.005 0.994 ( 0.005
83.5 ( 0.5 85.5 ( 0.5 83.3 ( 0.5 83.4 ( 0.5 78.1 ( 0.5
∼0.994 ∼0.993 ∼0.961 ∼0.933 ∼0.862
∼0.92 ∼0.93 ∼0.92 ∼0.92 ∼0.90
2.10 ( 0.01 2.18 ( 0.01 2.18 ( 0.01 2.22 ( 0.01 2.22 ( 0.01
9.5 ( 0.1 9.3 ( 0.1 8.9 ( 0.1 7.4 ( 0.1 6.9 ( 0.1
∼1.05 ∼1.07 ∼1.04 ∼1.04 ∼0.98
a
The values for FL, φL, and Fξ are shown preceded by the symbol ∼ to emphasize the speculative character of the evaluation.
Figure 3. SAXS intensity curves of the aerogels compared with those of the corresponding wet gels. Figure 1. (bottom) Mass loss of the wet gels prepared with different IPA/TEOS molar ratio (R) as determined by thermal gravimetric (TG) analysis carried out under a heating rate of 5 °C/min. (top) The derivative of the TG signal.
Figure 2. SAXS intensity as a function of the modulus of scattering vector q for wet gels prepared with different IPA/TEOS molar ratio R. For the sake of clarity, the curves are shown once more after shifting vertically by different factors. The dash lines are fittings of eq 2 to the experimental data.
∼ 0.3 nm. Such an order of magnitude could be checked from an independent estimation of the density Fξ of the mass fractal structure of the wet gels. The density Fξ can be evaluated from the mass fractal dimension D and the fractality range ξ/a through the following equation, as an analogy to another first presented for aerogels:18
∆Fξ ) ∆FS(ξ/a)D-3 where ∆Fξ ) Fξ - FL and ∆FS ) FS
-
(3)
FL. Table 2 shows the
estimated Fξ of the wet gels from the measured D and ξ (Table 2), assuming in all cases that a ) 0.3 nm and FS ) 2.2 g/cm3. The estimated Fξ values were found to be in reasonable agreement with the bulk-density Fwg. The agreement would be expected since the structure for correlation distances greater than ξ appears to be essentially homogenous to SAXS. Aerogels. Figure 3 shows the SAXS intensity curves for the aerogels compared with those from the corresponding wet gels. The scattering from the aerogel is more intense than that from the wet gel due to the increase in the electronic density difference between the silica particles and the “pores”, resulting from the supercritical extraction of the liquid phase. At lowand intermediate q-regions, the curves of the aerogels are moreor-less parallel to those of the corresponding wet gels, so they could be associated fairly with the original mass fractal structure. Apparently, the characteristic length ξ in the aerogels is diminished with respect to that in the wet gels; however, the determination of D and ξ by fitting eq 2 is not very precise since the fractality range was found to be too shortened and not very well defined in the aerogels. The SAXS curves of the aerogels are significantly different from those of the wet gels at high q-values. Apparently the curves cross over at qms ∼ 0.85 nm-1 to a distinct power-law decaying I(q) ∼ q-R, with the exponent R essentially equal to 3. The exponent R ∼ 3 is high enough to be associated with a mass fractal structure. It could be associated with a space-filling rough surface or, as we will see, a secondary mass/surface fractal structure developed with the supercritical treatment. The nitrogen adsorption isotherms shown in Figure 4 were obtained for the aerogels after degassing the samples at 200 °C for 24 h. The isotherms were found to be more properly classified as type IV, according to the classification by the International Union of Pure and Applied Chemistry (IUPAC).19 Table 3 shows the values for the BET specific surface area SBET, the total pore volume per mass unit Vp, and the pore mean size lBET, as determined from the nitrogen adsorption isotherms. Table 3 also shows the values estimated for the particle mean
Silica Sonogels Prepared with Isopropyl Alcohol
J. Phys. Chem. B, Vol. 110, No. 43, 2006 21585
Figure 4. Nitrogen adsorption isotherms of the aerogels carried out at the nitrogen liquid temperature.
Figure 5. Pore size distribution (PSD) for the aerogels as obtained from the nitrogen adsorption isotherms through the Kelvin equation and the classical Harkins & Jura model for cylindrical pores.
size lS ) 4VS/SBET (as an analogy to the pore mean size, so VS ) 1/FS), the bulk density of the aerogels F, through the relation (1/F) ) (1/FS) + Vp, and the pore volume fraction φ ) (FS F)/FS, assuming in all cases the value 2.2 g/cm3 for the silica density FS. Figure 5 shows the pore size distribution (PSD) as a function of the pore width (2r) for the aerogels. The pore volume fractions φ of the aerogels (Table 3) were found to be between 14 and 23% lower than the typical φL ) 0.92 found for the wet gels (Table 2). This means that some porosity apparently is eliminated with the supercritical process and degassing the samples at 200 °C. Discussion According to SAXS data, the structure of the wet gels is built up by primary silica particles with characteristic length lower than ∼0.3 nm, forming a mass fractal structure with fractal dimension D and characteristic length ξ given in Table 2. In the case of aerogels, the mass fractal characteristics are modified with the supercritical process, mainly at a high level of resolution, or at high values of q. The parameters D and ξ associated with the aerogels at low and intermediary q-values could not be precisely obtained by SAXS since the mass fractal range was found too shortened and not very well defined in the aerogels.
According to adsorption data, SBET diminishes slightly while lS increases correspondingly as R increases. The fundamental role of the increase of isopropyl alcohol additions on the structure of the resulting aerogels is to diminish the porosity φ and the pore mean size lBET. The diminution of lBET and the increase of lS are compatible with a pore elimination accompanied by a slight particle coarsening process that could be actuating with the overall supercritical process and degassing of the sample at 200 °C. The values between 2.1 and 2.3 nm found for the silica particles lS in the aerogels were compatible with some coarsening of the original silica particles of the wet gels, since the minimum a ∼ 0.3 nm there corresponds to a Bragg distance of 1.9 nm. However, the adsorption method could be applying stress on the sample during measurement, leading to an underestimation of pore volume, as pointed out by Scherer et al.20 in the case of low-density aerogels (typically for density values smaller than 0.35 g/cm3). But the effect could be negligible for this set of not very low density samples, which have presented densities between 0.46 and 0.66 g/cm3 (Table 3). On the other hand, the macroporosity could even be underestimated because nitrogen should not necessarily fill all the macropores in nitrogen adsorption experiments in the case of aerogels, since most of the solid surface of the aerogel has positive curvature, making the nitrogen condensation more difficult21,22 and thus underestimating the macroporosity. The modification of the mass fractal characteristics with the supercritical process was probed from the PSD curves as determined from nitrogen adsorption isotherms. For this purpose, we employed an earlier proposed method23 for the generation of a mass fractal structure from the PSD curve by a sequence of approximations. We start from a homogeneous solid with a density equal to that of the fused silica FS and follow probing the resulting bulk density F(r) as each incremental pore volume per mass unit δVp(ri) is incorporated to the structure. The process can be represented as
1 F(r)
)
ri ) r
1 FS
+
∑ δVp(ri)
(4)
ri ) 0
For aerogels, the bulk density Fξ of the mass fractal structure can be calculated as18
Fξ ) FS(ξ/a)D-3
(5)
which is a particular case of eq 3 with the liquid phase substituted by pores. Then, the evaluation of F(r) in the range a e r e ξ for such a system should yield
F(r) ) FS(r/a)D-3
(6)
Figure 6 shows the function F(r) for the aerogels, in a loglog scale with the structure length scale (as probed by the porewidth 2r), as evaluated through eq 4 from the PSD curves shown in Figure 5. Accordingly, the density F(r) tends to the constant value equal to the aerogel bulk density F (Table 3) as the porewidth 2r goes to the maximum value in each case, let us say 2rξ. The values for 2rξ in the aerogels (Table 4) diminish with
TABLE 3: Structural Properties of the Aerogels As Determined from Nitrogen Adsorption Isotherms R
SBET (102 m2/g)
Vp (cm3/g)
lBET (nm)
lS (nm)
F (g/cm3)
φ
0 1 2 4
8.45 ( 0.02 8.01 ( 0.02 7.79 ( 0.02 7.74 ( 0.02
1.73 ( 0.01 1.34 ( 0.01 1.21 ( 0.01 1.06 ( 0.01
8.2 ( 0.1 6.7 ( 0.1 6.2 ( 0.1 5.5 ( 0.1
2.15 ( 0.05 2.27 ( 0.05 2.33 ( 0.05 2.35 ( 0.05
0.458 ( 0.005 0.557 ( 0.005 0.602 ( 0.005 0.659 ( 0.005
0.79 ( 0.01 0.75 ( 0.01 0.73 ( 0.01 0.70 ( 0.01
21586 J. Phys. Chem. B, Vol. 110, No. 43, 2006
Donatti et al. Dm shown in Table 4. Dm was found all around ∼2.7 for the aerogels, fairly independent of IPA additions. The apparent crossover at ∼2rms in the plot of Figure 6 approximately matches the Bragg distance 2π/qms associated with the characteristic length ∼1/qms of the crossover observed by SAXS in Figure 3. Thus the mass fractal structure associated with the micropore region as inferred from adsorption was associated with the power-law decaying I(q) ∼ q-R in the SAXS intensity at high q in Figure 3. The values of Dm (between 2.67 and 2.70) would be in agreement with the value of the exponent R ∼ 3 (Table 4) of the SAXS intensity decaying if we consider that such a secondary structure behaves as a mass/surface fractal. The intensity from a mass/surface fractal obeys a power law decaying with the scattering exponent R given by25
R ) 2Dm - Ds Figure 6. The bulk density F(r) as a function of the structure length scale (as probed by the pore-width 2r) evaluated through eq 4 from the PSD curves for the aerogels. The structure length scale as probed by adsorption (2rξ and 2rms) in the aerogels and the Bragg distance probed by SAXS in the corresponding wet gels (2πξ) and in the aerogels (2π/qms) are indicated in each case.
TABLE 4: Properties of the Fractal Structure of the Aerogels As Deduced from Nitrogen Adsorption and SAXS Data R
D
0 1 2 4
2.38 ( 0.02 2.43 ( 0.02 2.43 ( 0.02 2.52 ( 0.02
2rξ (nm) 2rms (nm) ∼39 ∼29 ∼24 ∼22
∼9.3 ∼6.7 ∼6.5 ∼5.8
Dm
R
Ds
2.67 ( 0.02 2.70 ( 0.02 2.69 ( 0.02 2.68 ( 0.02
3.05 ( 0.01 3.08 ( 0.01 3.00 ( 0.01 3.06 ( 0.01
2.29 ( 0.03 2.32 ( 0.03 2.38 ( 0.03 2.30 ( 0.03
the IPA additions and, in all cases, 2rξ values are smaller than the corresponding 2πξ (the Bragg distance associated with the characteristic length ξ probed by SAXS) in the wet gels. It is a common practice in SAXS literature24 to define the resolution associated with the scattering vector q using the Bragg relation dBragg ) 2π/q. We have found that the pore structure of several silica aerogels as determined by nitrogen adsorption is in good agreement with the determination from SAXS, if we assign the Bragg distance probed by SAXS to the pore width probed by nitrogen adsorption.23 The reason for this could be associated with the low density of the aerogels, which allows for the structure length scale to be probed practically by the pore width. The difference which was found between 2rξ and 2πξ in the present work is compatible with the elimination of some porosity in the supercritical process and/or underestimation of the macroporosity in the adsorption method. According to eq 6, the plot in a log-log scale of the function F(r) with the structure scale length 2r should be a straight line with slope D-3. Figure 6 shows that eq 6 fits reasonably well to the experimental F(r) of the aerogels in a first not too great interval of 2r in the mesopore region, between the maximum 2rξ and an intermediary value, let us say 2rms. Table 4 shows the values for D obtained from the fitting process. The values for D were found in the interval between 2.38 for R ) 0 and 2.52 for R ) 4. These values are greater than those corresponding to the wet gels (2.10 and 2.22, respectively) as determined from SAXS. This suggests that the supercritical process and degassing the samples at 200 °C also increase the mass fractal dimension together with the shortening of the fractality range in the mesopore region. Figure 6 also shows a second mass fractal regime associated with the micropore region in the range below ∼2rms. From the slopes Dm-3 of the straight lines fitting eq 6 in the micropore region we obtained the values of the mass fractal dimension
(7)
where Dm and Ds are the dimensionality of the correlated mass and surface fractal, respectively. The surface fractal character could come from the surface roughness possibly evolved in the supercritical process. Ds was evaluated from eq 7 through the measured Dm and R, and it was found all essentially equal to ∼2.3 (Table 4), fairly independent of R. Finally, we think that the basic reason for the IPA additions to affect the characteristics of the mass fractal structure of the wet gels could be related to the alcohol exchange reactions. Metal alkoxides tend to react with alcohols, resulting in the replacement of their alkoxy groups. It has been shown that alcoholysis occurs between TMOS and IPA when water is added to the mixture.15 A two-step mechanism beginning with fast hydrolysis of TMOS followed by either exchange between silanol and IPA or exchange between partially hydrolyzed alkoxide group and IPA yielding alkoxy derivatives of silicon has been proposed.15 It could carry some modification in the hydrolysis and polycondensation process in the gel synthesis, proportioning small variations in the structural parameters D and ξ in the wet gels. Small modifications in the mass fractal structure in the wet gels could result in greater structural modifications in the resulting aerogels. Conclusions The structure of wet gels with ∼1.4 × 10-3 mol SiO2/cm3 and ∼92 vol % liquid phase prepared from TEOS sonohydrolysis with different additions of isopropyl alcohol (IPA) can be described as a mass fractal structure built up by primary silica particles of characteristic length a lower than ∼0.3 nm. The dimensionality D and the characteristic length ξ of the mass fractal structure were found changing between D ∼ 2.1 and D ∼ 2.2 and ξ ∼9.5 nm and ξ ∼6.9 nm, respectively, as the IPA/ TEOS molar ratio was changed from 0 to 4. The supercritical process and degassing the samples at 200 °C apparently increase the mass fractal dimension D, diminish the characteristic length ξ, and shorten the fractality range in the mesopore region. The fundamental role of the increase of isopropyl alcohol additions on the structure of the resulting aerogels is to diminish the porosity φ and the pore mean size lBET. This is compatible with elimination of some porosity in the supercritical process and/or underestimation of the macroporosity in the adsorption method. A secondary mass/surface fractal structure is evolved in the micropore region of the aerogels. Acknowledgment. This research was partially supported by LNLS - National Synchrotron Light Laboratory, FAPESP, and CNPq, Brazil.
Silica Sonogels Prepared with Isopropyl Alcohol References and Notes (1) Brinker, C. J.; Scherer, G. W. Sol-Gel Science: The Physics and Chemistry of Sol-Gel Processing; Academic Press: San Diego, 1990. (2) Blanco, E.; Esquivias, L.; Litra´n, R.; Pin˜ero, M.; Ramı´rez-del-Solar, M.; de la Rosa-Fox, N. Appl. Organomet. Chem. 1999, 13, 399. (3) Vollet, D. R.; Donatti, D. A.; Iban˜ez Ruiz, A. J. Non-Cryst. Solids 2001, 288, 81. (4) Donatti, D. A.; Vollet, D. R.; Iban˜ez Ruiz, A.; Mesquita, A.; Silva, T. F. P. Phys. ReV. B 2005, 71, art. no. 014203. (5) Morita, M.; Kajiyama, S.; Rau, D.; Sakurai, T.; Iwamura, M. J. Lumin. 2003, 102-103, 608. (6) Sujatha Devi, P.; Ganguli, D. J. Non-Cryst. Solids 2004, 336, 128. (7) Litra´n, R.; Blanco, E.; Ramı´rez-del-Solar, M. J. Non-Cryst. Solids 2004, 333, 327. (8) Parvathy Rao, A.; Venkateswara Rao, A. Sci. Technol. AdV. Mater. 2003, 4, 121. (9) Feng, Y.; Yao, R.; Zhang, L. Physica B 2004, 350, 348. (10) Ahola, M.; Kortesuo, P.; Kangasnieme, I.; Kiesvaara, J.; Yli-Urpo, A. Int. J. Pharm. 2000, 195, 219. (11) Radin, S.; El-Bassyouni, G.; Vresilovic, E. J.; Schepers, E.; Ducheyne, P. Biomaterials 2005, 26, 1043. (12) Colomer, M. T.; Anderson, M. A. J. Non-Cryst. Solids 2001, 290, 93. (13) Mehrotra, R. C. Struct. Bonding 1992, 77, 1.
J. Phys. Chem. B, Vol. 110, No. 43, 2006 21587 (14) Kim, S. I.; Kim, G. Y.; Lim, H. M.; Lee, B. W.; Nah, J. W. Bull. Korean Chem. Soc. 2000, 21, 817. (15) Sivade, A. Bourret, D.; Sempe´re´, R.; Bouaziz, J. J. Non-Cryst. Solids 1992, 147-148, 167. (16) Gregg J.; Sing, K. S. W. Adsorption, Surface Area and Porosity; Academic Press: New York, 1967. (17) Schaefer, D. W.; Keefer, K. D. Phys. ReV. Lett. 1984, 53, 1383. (18) Vacher, R.; Woignier, T.; Pelous, J.; Courtens, E. Phys. ReV. B 1988, 37, 6500. (19) Leofanti, G.; Padovan, M.; Tozzola, G.; Venturelli, B. Catal. Today 1990, 41, 207. (20) Scherer, G. W.; Smith, D. M.; Stein, D. J. Non-Cryst. Solids 1995, 186, 309. (21) Scherer, G. W. J. Colloid Interface Sci. 1998, 202, 399. (22) Reichenauer, G.; Scherer, G. W. J. Colloid Interface Sci. 2001, 236, 385. (23) Vollet, D. R.; Donatti, D. A.; Iban˜ez Ruiz, A. Phys. ReV. B 2004, 69, art. no. 064202. (24) Glatter, O. In Small-Angle X-ray Scattering; Glatter, O., Kratky, O., Eds.; Academic Press: London, 1982; p 136. (25) Pfeifer, P.; Ehrburger-Dolle, F.; Rieker, T. P.; Gonza´lez, M. T.; Hoffman, W. P.; Molina-Sabio, M.; Rodrı´guez-Reinoso, F.; Schmidt, P. W.; Voss, D. J. Phys. ReV. Lett. 2002, 88, art. no. 115502.
